What's new
Van's Air Force

Don't miss anything! Register now for full access to the definitive RV support community.

Simplified Approach

Not sure what you're talking about. Its a very simple concept. Put things on one bus, put things on another. If power supply goes offline from one, the busses get tied to keep everything powered.

Don’t disagree - it is a very simple concept. But what if the original failure could be put in the category of “malicious” - for example transient overvoltage from a generator or diode failure in an alternator rectifier circuit that puts AC on to the DC supply bus. Without adequate monitoring circuitry to automate the bus tie connect/disconnect or monitor crowbar circuitry to protect against particular failure modes connecting the two system strings together could result in complete loss of the system rather than partial loss of function. So we get to the concept of accepting an increase in overall system failure rate but gaining system availability by adding an automated busss tie with a high integrity monitor circuit. For vehicles and systems that require high levels of availability then system redundancy and complex high integrity monitoring may be the only answer.
For example Cat 3B autoland requires 3 autopilot channels all to be up and running before committing to land. The period of exposure is only 10 minutes but modern technology cannot make the availability or integrity number required with less than 3 independent very complex systems with redundant monitoring on each channel. The tentacles of complexity travel a long way into other systems for the auto-land case - I know its an extreme example but it makes the point about what it takes when the mission demands the capability. We dont have that demand (most of us) and are prepared to accept loss of function or degraded capability following a failure because without complex high integrity monitoring we could be making the situation much worse by tying the two strings together so goes the thinking. In the end analysis it depends on your level of scar tissue, experience and paranoia how you define your system architecture and develop your flight procedures following a failure. Balancing all of our systems design around the number one single point failure ( being single engine) and the associated failure rate which is likely some number in the 10-E4 range per hour puts us in the category of keeping it as simple as possible, living with either loss of function on a first failure or severely degraded capability following a first failure. Adding complex monitoring and system reconfiguration capability just gets us to a higher overall failure rate, more complexity in maintenance and troubleshooting without compensating benefits in availability.
Fly the first leg of a flight IFR with a full up system, fly the second leg of the flight or get back home in good VFR after the first failure could be one way to describe the philosophy.
Keith Turner
 
What you're suggesting is more or less old school certified practice. Current EAB practice (Nuckolls et al) piles avionics on an "essential bus" and gives it a battery direct backup feed via a diode. Think about it in the context of your question.

No argument I have an old-school mindset. I still have all round gauges, except for the EMS. Heck, I still code in FORTRAN too! So this is all a very useful, helpful exercise for me. Thanks!

Now consider the architecture as presented...or again, the essential bus feed above. If an avionics bus or the wire connecting it to power was actually shorted, the circuit protection device (fuse, fusible link, ANL, breaker) feeding that wire would blow. If it has a problem, it turns itself off. If it doesn't have a problem, why turn it off?

Feed%20Protection.jpg

There is a real key aspect here that changes my thinking. You have every single device on its own fuse. So the idea you describe here makes total sense ("If it has a problem, it turns itself off...")
This makes all the difference. It not only makes the event of cabin smoke far less likely, but also means that whatever the source will go off-line on its own pretty quickly. Not sure if that makes the old-school emergency response still relevant or not (all four rockers down, in your case).

In my "old-school" system, I have a breaker for groups of major items, but several items may draw from that breaker. I do have an avionics fuse block so individual fuses for radio, transponder, encoder, GPS, Trutrak ADI-Autopilot.

It does beg the question, where ever do you put the fuse block for all those fuses? (especially in an RV-8!!!) I was tempted to count them up, but there must be dozens of fuses. FAR 91.205 (c) (6) says we have to be able to access and replace fuses in flight, and carry three sets of spare fuses. My ops limits say if I am going to fly at night, I have to comply with FAR 91.205 (b), (c). Admittedly this is old-school thinking. If there is a problem that blows a fuse, why would you want to put another fuse in there, to re-energize the problem device so it can blow the new fuse too? I can sort of actually access my avionics fuse block in flight, but I don't think I could change a fuse and maintain control of the airplane.

Ahhh, then you'll love the new approach. Light up the whole panel on one battery, and crank on the other. All your worry parts are isolated from cranking gremlins.

Ah! And here is the very best reason to keep your two avionics master switches separate. Turn on the left master, and the right avionics master, and start (or shut down). Brilliant!


Exactly what I drew first, but switching ECU power when running makes Ross nervous. As drawn now, the ECU power goes on before cranking and stays on for the duration.
Oh, right. It's basically a computer. No point, and probably not good for it, to be re-booting it from a brief power interruption. So, two switches for the pumps and coils, and two switches for the ECUs. I forgot what the fifth switch was for...


I'll assume the Garmin gang would have done the same.

So the Garmin devices that take two separate power inputs, we are assuming/trusting that they would have put diodes or other isolation in there between the two supplies with enough 'capacity' to tolerate a short on one of the supply leads. It makes sense - they would surely be smart enough to do that.
 
Last edited:
It does beg the question, where ever do you put the fuse block for all those fuses? (especially in an RV-8!!!)

Good question, at least regarding an -8 I have a swing down fuse block panel in mine (old photo below), and it can be accessed in flight. That said, I don't carry spare fuses, as I'm not replacing any in flight. Nuff said. I think a decent FAA guy is aware the circuit protection popped for a reason.

It's basically a computer. No point, and probably not good for it, to be re-booting it from a brief power interruption.

Not sure of the reason. I kill and resume power to both the Ford EDIS module and the Autosport Labs timing control computer at every runup. Beat my dog too. (I kid, I kid!)

So, two switches for the pumps and coils, and two switches for the ECUs. I forgot what the fifth switch was for...

Actually two for the pumps/coils and two best thought of as "bus power" for each SDS ECU/pump/coil bus. Fifth selects which ECU has control of the injectors.

So the Garmin devices that take two separate power inputs, we are assuming/trusting that they would have put diodes or other isolation in there between the two supplies with enough 'capacity' to tolerate a short on one of the supply leads.

Ok, I'll ask ;)

EDIT...Thinking about it. Ya'll correct me if wrong, but assume a supply lead is shorted. The fuse on that lead will pop of course. The stated concern boils down to "Can power from the other bus reach that short through the diodes?" Answer would be no. The issue is how much reverse voltage the diode will withstand. Here the reverse voltage is simply bus voltage. Situation normal.
.
 

Attachments

  • P7020007.JPG
    P7020007.JPG
    260.7 KB · Views: 100
Last edited:
Ok, I'll ask ;)

EDIT...Thinking about it. Ya'll correct me if wrong, but assume a supply lead is shorted. The fuse on that lead will pop of course. The stated concern boils down to "Can power from the other bus reach that short through the diodes?" Answer would be no. The issue is how much reverse voltage the diode will withstand. Here the reverse voltage is simply bus voltage. Situation normal.
.[/QUOTE]

The safe answer is to put a suitably rated diode in line in the wiring close to the Garmin connector in each power input line then you can be sure you are protected and isolated. Garmin and Dynon seem to have a strict policy of not divulging any circuit information regarding what is inside the box although you may get an answer to the question about what would happen with one power pin grounded and power on the other. Or you could do the test ………………………………��
There is a distinct possibility that the double pin power input is provided for redundancy to cover the case of pushed back connector pins or a broken wire and the expectation is that both pins are fed from the same source and go into the same printed circuit board layer ,at the unit connector. Most Garmin equipment is designed to TSO’d standards and this would be standard practice on transport category equipment to eliminate single point connector failures.
Keith Turner
 
Last edited:
This one for Ivan.

Previously...

The bus feeds get circuit protection (fuse, fusible link, ANL, CB, etc) with a rating much higher than the individual device feeds. In your example, the 5 amp servo fuse would melt before the 20 amp feed fuse got warm.

What guarantees this property if there is a 100A short current? Technically, both should trip. Why 5A would necessarily trip faster?

I suggested he give it a try. In the shop this afternoon, it occurred to me that I probably have a more extensive junk drawer. So, I dug out a bag of fuses, a fuse block, an inline fuse holder, a battery, and some jumpers:

Feeder%20Fuse.jpg


That's a 20 amp in the inline "bus feed", and 15, 10, 7.5, 5, and 3 in the fuse block. Be assured each of the smaller fuses sacrificed itself in a little flash of light when I touched its respective output terminal with the jumper, while the 20 remained ready to rock.

To further illustrate, I loaded a 7.5 into the bus feed and burned a 5. Then I swapped for a different 7.5 and burned another 5.

Could be a pattern here.
 
I would love to see the following experiment;

-load the bus to a constant 15A draw for a few minutes
-then short the 5A breaker

Now the result should theoretically depend on the fuse characteristics: slow blow vs. fast blow
 
Last edited:
One more sneaky one...

Dan,

The ETXs (or any low internal resistance battery) have failure mode when used in pairs. If one is at a low state of charge and the second is fully charged. When connected in parallel the fully charged one will dump amps into the lower voltage one with enough current to do bad things. If memory servers a V delta of more than 0.3, but it is likely battery dependent. Certainly a 900 into a 640 is a bad idea, don't know how picky two 900s are.

If I've been keeping up, I think you have that scenario if you run one battery done through the bus, and then hit both masters.

Derek

https://earthxbatteries.com/lithium-battery-series-parallel-operation-2/
 

Attachments

  • 2023-02-19_18-39-36.jpg
    2023-02-19_18-39-36.jpg
    104.4 KB · Views: 70
Last edited:
The ETXs (or any low internal resistance battery) have failure mode when used in pairs. If one is at a low state of charge and the second is fully charged. When connected in parallel the fully charged one will dump amps into the lower voltage one with enough current to do bad things. If memory servers a V delta of more than 0.3, but it is likely battery dependent. Certainly a 900 into a 640 is a bad idea, don't know how picky two 900s are.

If I've been keeping up, I think you have that scenario if you run one battery done through the bus, and then hit both masters.

Thanks Derek, excellent comment. I had not considered high current between the two due to unequal charge. I'll contact EarthX for comment. The webpage asks that they be within 0.3V of each other at connection, but doesn't attempt to define amps to be expected vs voltage difference. Worth knowing, because the caution would apply to any sort of dual bus system.
 
How dependent is this whole architecture on the cool "enable/standby" mode of the MZ-30L?

What, if anything would need to be changed if the second alternator was a more traditional pad-mounted back-up alternator?
 
Hey Dan, thanks for doing the fuse experiment.

I can relate a relevant story. In my previous life looking after a cable TV headend (think data centre), plugging in one new but faulty piece of gear took out power for the entire floor, killing cable TV and internet for an entire city...
It turns out that the REALLY BIG breakers feeding power to the floor had a faster trip response time (aka curve) than the individual 20 amp breakers feeding each rack.

To bring this back to yours and other reader's planes, if the scenario was your 20 amp or higher automotive blade fuse feeding a bunch of typical (say) 5 amp thermal breakers, I would imagine the 20 amp blade fuse will be the hero long before the breaker.

I don't have the bits to repeat your experiment Dan...


 
There is a distinct possibility that the double pin power input is provided for redundancy to cover the case of pushed back connector pins or a broken wire and the expectation is that both pins are fed from the same source and go into the same printed circuit board layer ,at the unit connector. Most Garmin equipment is designed to TSO’d standards and this would be standard practice on transport category equipment to eliminate single point connector failures.

This is what I learned from the G3x AEA class a couple weeks ago: if it’s labeled PWR 1 and PWR 2 or BACKUP PWR, like on the G5, those are 2 inputs and are internally dioded, in case power is lost for whatever reason on either input.

However, on the back of the GNX for example, there are 3 pins labeled AIRCRAFT POWER. This is because 22awg wire/pins are used and can’t carry the load on a single pin, so it’s spread across multiple pins. Internally these all connect to the same place and therefore WILL back feed to each other.

This specific problem was addressed in the class, answer is to run 2 discrete wires in the case of the G5 (presumably from 2 different power sources, one as a backup).

And one larger wire with the split-out down to 22awg as close as possible to the connector in the case of the GTxx/GNxx/etc, since it’s unlikely your short would happen in the last 3 inches. And if it did, either the small 22awg pig tail would fry, or the breaker for the larger wire would pop.

I don’t see much documentation in the G3x manual for the internal specifics, but the class is taught by Levi, a former Garmin engineer, and owner of Midwest Avionics, so seems authoritative.
 
I don’t see much documentation in the G3x manual for the internal specifics

The G3X installation manual makes this pretty clear... for each LRU that features multiple diode-isolated power inputs, this statement appears:

AIRCRAFT POWER 1 and AIRCRAFT POWER 2 are “diode ORed” to provide power redundancy.
 
How dependent is this whole architecture on the cool "enable/standby" mode of the MZ-30L?
What, if anything would need to be changed if the second alternator was a more traditional pad-mounted back-up alternator?

The MZ is a generator; just spin it for output. The likely alternator choice, a B&C 462-3H, requires an outside power source for initial excitation of the field.

Details revolve around regulator function. The prescribed installation for a B&C SB1B-14 backup regulator connects both voltage sense and field power to the main bus, with separate connections. A Nuckolls Z101 connects the aux alternator field supply to a fuse on a battery bus, but lists a generic Ford regulator. I don't know the reason, and given he drew a B&C LR3 on the primary alternator, it would be best to ask. Maybe an SB1B-14 can be wired as attached, or maybe not. Again, B&C's instructions have always been to install separate sense and field connections.

Obviously you could use the Ford regulator, but I assume there would be no OV protection, not a good idea with Gen1 EarthX batteries.

Speaking of which, remember, they internally disconnect in the event of over-discharge or excessive amp output. Potential fail modes here would be (1) running the batteries down too far before bringing the backup alternator online, and (2) shorting a fat wire. Mr. Nuckolls says those kinds of shorts tend to clear themselves by burning away the offending structure. Sounds scary, but makes sense. Here however, the battery BMS may disconnect...meaning no output to fire up the alternator for somewhere between 1 and 3 minutes. I think it's low risk.

The major advantage to the Monkworkz choice is cost, size, and weight. The B&C on an RV tends to require a spacer and a new oil filter adapter in addition to the cost of the alternator. It's no secret that I despise blast tubes on anything, but it's all about considered compromise. Bill says they are needed to get 30 amps from such a tidy package, which does make sense.
.
 

Attachments

  • Alternator Backup.jpg
    Alternator Backup.jpg
    29.5 KB · Views: 92
Last edited:
Thanks Derek, excellent comment. I had not considered high current between the two due to unequal charge. I'll contact EarthX for comment. The webpage asks that they be within 0.3V of each other at connection, but doesn't attempt to define amps to be expected vs voltage difference. Worth knowing, because the caution would apply to any sort of dual bus system.

A few numbers to play with

According to google, in rough numbers the internal resistance of the LIPO cells is 0.01 Ohm (or a little higher). There are 4 in series to get to our voltages so you can figure 0.04 Ohm for a single series. There are a bunch in parallel to get to the total mAh for the battery. 1C for a LIPO cell is 6A, so V=IR gives us 0.24V before things get interesting using those rough numbers. Earthx will have better ones for the cells they use.

Since we have a bunch in parallel the amps go up. If the max charge rate for the battery is 60 and the numbers jive, we have 10 in parallel, and the (rough) math says 0.5V is then 120A

The numbers aren't perfect, but the message is that the current will grow quickly but more or less linearly as the delta volts increase.

The worst case is the same scenario, but with the big alternator turning as most of those 20-30 additional amps would additively flow to the discharged battery.

Its an issue for any dual battery system that can be connected and disconnected. In my plane it earned a placard. A well placed shunt and an over amp alert would likely also do the trick, since they can take quite a bit of abuse for a short time (another EarthX question).

Derek
 
Last edited:
So the Garmin devices that take two separate power inputs, we are assuming/trusting that they would have put diodes or other isolation in there between the two supplies with enough 'capacity' to tolerate a short on one of the supply leads. It makes sense - they would surely be smart enough to do that.

Confirmed...they are definitely smart enough.
 
Last edited:
Don’t disagree - it is a very simple concept. But what if the original failure could be put in the category of “malicious” - for example transient overvoltage from a generator or diode failure in an alternator rectifier circuit that puts AC on to the DC supply bus. Without adequate monitoring circuitry to automate the bus tie connect/disconnect or monitor crowbar circuitry to protect against particular failure modes connecting the two system strings together could result in complete loss of the system rather than partial loss of function. So we get to the concept of accepting an increase in overall system failure rate but gaining system availability by adding an automated busss tie with a high integrity monitor circuit. For vehicles and systems that require high levels of availability then system redundancy and complex high integrity monitoring may be the only answer.
For example Cat 3B autoland requires 3 autopilot channels all to be up and running before committing to land. The period of exposure is only 10 minutes but modern technology cannot make the availability or integrity number required with less than 3 independent very complex systems with redundant monitoring on each channel. The tentacles of complexity travel a long way into other systems for the auto-land case - I know its an extreme example but it makes the point about what it takes when the mission demands the capability. We dont have that demand (most of us) and are prepared to accept loss of function or degraded capability following a failure because without complex high integrity monitoring we could be making the situation much worse by tying the two strings together so goes the thinking. In the end analysis it depends on your level of scar tissue, experience and paranoia how you define your system architecture and develop your flight procedures following a failure. Balancing all of our systems design around the number one single point failure ( being single engine) and the associated failure rate which is likely some number in the 10-E4 range per hour puts us in the category of keeping it as simple as possible, living with either loss of function on a first failure or severely degraded capability following a first failure. Adding complex monitoring and system reconfiguration capability just gets us to a higher overall failure rate, more complexity in maintenance and troubleshooting without compensating benefits in availability.
Fly the first leg of a flight IFR with a full up system, fly the second leg of the flight or get back home in good VFR after the first failure could be one way to describe the philosophy.
Keith Turner

https://www.dailywritingtips.com/why-we-need-paragraphs/ ......:(
 
Oh yeah. Look him up if you head south. Warren is the sort who will poke you in the ribs, then grin big and buy you a drink ;)

Of course, you are right - I get the point and he is right - I am remiss in not going back and breaking it up into paragraphs, correcting grammar and spelling and adding punctuation.

Must admit I find it distracting when I see there for their, to for too and a few others of the same ilk.

Function distracted by form……… Need to stay away from the dark humor and focus on sharing our knowledge.
KeithTurner

Writing it properly wasn’t difficult at all even though it was one finger typing on an Ipad mini.
 
Thank you for this. It matches my configuration, so maybe I can label all my electrical layout "recommended by Dan Horton".

One question. What is the rationale for having the switch and pullable circuit breaker inline? Wouldn't just the switch be sufficient?
 
Thank you for this. It matches my configuration, so maybe I can label all my electrical layout "recommended by Dan Horton".

One question. What is the rationale for having the switch and pullable circuit breaker inline? Wouldn't just the switch be sufficient?

I assume you're referring to the little alternator drawing. Please note there are a few questions to be asked at B&C before anyone can call it recommended.

The pull breaker was just me being sloppy while drawing too fast. Please substitute a standard breaker if used with a switch, or a keep the pull breaker and delete the switch.
 
B&C regulators LR3 and SB1 present a passive load

Post 19

… My plan also has the Stby Alt (B&C 410-14) tied to the hot side of the main contactor. I wondered about the battery back-feed, and have been meaning to talk to the folks at B&C…
Hi Shawn,

B&C regulators LR3 and SB1 do present a passive load in the V sense connection, about 30 mA.

However your schematic “ Primary Power System” rev Feb 17 2023 as posted on VAF and Aeroelectric List has the B&C 410 field and sense connections, correctly IMO, on your Avionics/Essential Bus which is not tied directly to the battery but thru the Essential Bus relay so all is well.

BTW Bob Nuckolls says about the SB1: "The stand-by versions of the alternator controller are not recommended (by Bob) for simple dual alternator installations where the auto-switch feature is not necessary... this was a 'bell-n-whistle' aimed primarily at the STC market for TC aircraft that were receiving the pad driven, aux alternators.". Nothing wrong with the SB1 if you already have it but an LR3 set to say 13.5V can be used. What the SB1 doesn’t have is the OV test button that most folks don’t use. BTW2 the SB1 comes set to 13V, 13.5 is better per Bob.
.
 
Last edited:
More questions

Reading Bob Nuckoll's comments about the Z101 system prompted additional questions.

"The B-lead is hot all the time, which has post-crash implications, and you should be careful to disconnect the (-) battery terminal before wrenching."

So, Dan, in your design, the B-lead to the standby generator is hot all the time.

Assuming this is an RV-8 with an angle-valve engine, we want the batteries in the back. Even if they are LiPo4, there are two, and I would still prefer that weight in the back. So, (A) not very convenient to disconnect the (-) terminals, and (B) pretty long wires. The lead between the battery and the MZ-30L is probably a 8 AWG?

It probably makes sense to put those battery leads, and any other 'always hot' wires, such as the avionics bus and SDS bus power, inside a protective conduit? That helps at least up to close by the connections at the generator and regulator. Still wouldn't want to touch a wrench to that terminal while working on something else. I guess it would blow the ANL before it would weld the wrench in place.

As an aside, years ago I knew a guy who got his wedding ring caught between the starter solenoid terminal and something that was at chassis ground while working on a car. He lost the finger.
 
Last edited:
So, Dan, in your design, the B-lead to the standby generator is hot all the time.

Well, it's fused 40 amps at the battery, so it's not going to burn off any fingers. However, if it's a worry, move the MZ's contactor to the back of the RV-8, with the batteries. Still worried? Keep the 40 amp fuse.
.
 

Attachments

  • MZ w aft battery.jpg
    MZ w aft battery.jpg
    34.1 KB · Views: 50
B&C regulators LR3 and SB1 do present a passive load in the V sense connection, about 30 mA.

Thank you John. The passive load is parasitic, fundamentally a battery drain when parked, if connected to the hot side of the master?
 
The lead between the battery and the MZ-30L is probably a 8 AWG?

The power connections are as follows:
3x 14 AWG wires from the generator to the regulator, Dress, cut to length and terminate with provided ring terminals. Keep away from exhaust plumbing and route such that there is no strain on any of the leads. The regulator has onboard fuses so no external current limiting is required.
2 x 10 AWG wires from the regulator are used for ~14.4 VDC output and ground. An onboard fuse is provided so no additional current limiting is necessary. If there is no main generator breaker, the positive 14.4 volt lead can be wired directly to the main bus or the load side of the master relay. If there is an existing circuit breaker for the old generator, the output can be wired to that where the previous power source was connected. The breaker maximum rating can exceed 30 amps as the MZ-30L regulator will current limit at 30 amps. This breaker will provide the following benefits:
protect the wiring between the breaker and regulator output
provide a means for isolating the MZ-30L for testing
provide a means of isolating the regulator output from the bus for any reason.

monkworkz.jpg

Appears that since the regulator is internally restricted to 30amps, they are specifying a 10awg?

I think these small accessory driven generator/alternator devices provide for some new and interesting electrical system designs. Perhaps 2 of them carrying the entire airplane load with a failover to an essential bus that is less than 30amps? (Not sure if all accessory cases have 2 vacuum pads normally).
 
Well, it's fused 40 amps at the battery, so it's not going to burn off any fingers. However, if it's a worry, move the MZ's contactor to the back of the RV-8, with the batteries. Still worried? Keep the 40 amp fuse.
.

Ah yes! I had not noticed that there was a contactor as part of the MZ-30 circuit. That totally solves it.

This still leaves the SDS bus and avionics bus as "always hot" but that is kinda the point, isn't it. I guess I would run those in a conduit up to the panel from the batteries, just as an extra layer of protection for them.
 
This still leaves the SDS bus and avionics bus as "always hot" but that is kinda the point, isn't it.

Yep. Times have changed, The loads have changed.

Think about it. The original "essential bus" proposal was merely an alt path to bare minimum avionics. OFF with the master, ON with an essential bus switch feeding a navcom and a map light. Very sensible given a lot of those airplanes had a unprotected feed to a 60 amp breaker in the panel. The only way to kill those sparks was to open the main contactor. No problem if you needed to do it, as mags and a vacuum pump took care of the rest.

Two decades later, the concept is almost unrecognizable. "Essential" is now a G3X suite of modules, an autopilot and a navigator box. EI/EFI requires full time power. They are not minor add-ons, but rather, fundamental to continued flight.

We design like we think. I'm suggesting we think of those busses as primary, and consider the main bus as the auxiliary. Think of it as a place to park big loads and a big generator, with big contactors and big wires to feed a big starter, none of which must absolutely, positively work to continue flight.

I guess I would run those in a conduit up to the panel from the batteries, just as an extra layer of protection for them.

Conduit is good. Alt plan...my own RV-8 has two hot leads running from the aft batteries to the panel for dual EI. They're not in conduit, but they are carefully routed and secured, using shielded two conductor wire. The shield braid has no electrical purpose. Here it's just armor.
 
Great synopsis. Helps with perspective.

Yep. Times have changed, The loads have changed.

Think about it. The original "essential bus" proposal was merely an alt path to bare minimum avionics. OFF with the master, ON with an essential bus switch feeding a navcom and a map light. Very sensible given a lot of those airplanes had a unprotected feed to a 60 amp breaker in the panel. The only way to kill those sparks was to open the main contactor. No problem if you needed to do it, as mags and a vacuum pump took care of the rest.

Two decades later, the concept is almost unrecognizable. "Essential" is now a G3X suite of modules, an autopilot and a navigator box. EI/EFI requires full time power. They are not minor add-ons, but rather, fundamental to continued flight.

We design like we think. I'm suggesting we think of those busses as primary, and consider the main bus as the auxiliary. Think of it as a place to park big loads and a big generator, with big contactors and big wires to feed a big starter, none of which must absolutely, positively work to continue flight.



Conduit is good. Alt plan...my own RV-8 has two hot leads running from the aft batteries to the panel for dual EI. They're not in conduit, but they are carefully routed and secured, using shielded two conductor wire. The shield braid has no electrical purpose. Here it's just armor.
 
True story Steve. About 20 years ago I installed the full essential bus treatment in a steam gauge RV. It was loaded...certified S-Tec AP, Garmin 430, Dual Lightspeed EI, even a Stormscope. Needed the essential bus because the amp requirement with everything running was significant...and it was all hung on a single alternator, with a single Concorde lead acid battery.

Here's the punchline. I did the work in return for a set of keys. The two owners, Beech and Piper guys, could not grasp the concept no matter how much I explained it. "Huh?' they said. "You want us to turn off the master because the alternator died? And we flip which switches to bring back what? And then only part of it works?"

In truth, it worked exactly as intended, but it made 'em nervous, and as time went on I was the only one flying it. Not a bad deal for me, but really drove home a realization. It's easy to build a system which works. It's hard to build one everybody understands. The best requires the pilot to know or do very little.
.
 

Attachments

  • circuit_diagram.png
    circuit_diagram.png
    133.7 KB · Views: 69
Last edited:
Conduit is good. Alt plan...my own RV-8 has two hot leads running from the aft batteries to the panel for dual EI. They're not in conduit, but they are carefully routed and secured, using shielded two conductor wire. The shield braid has no electrical purpose. Here it's just armor.
Using shielded wire for power sounds good however: I once had a short on the unprotected shielded wire from a Lightspeed Ign, when the inner wired melted it normally would have just opened up, but the shield prevented that from happening and the short continued until the shield finally burned thru. The result was eye opening as the smoke filled the airplane and the shielded wire continued to burn for what seemed like eternity. It also burnt all the wires in the bundle and burnt up the LSI box as well as it was the ground.
Lesson learned: protect all wires (despite the recommendation from LSI) at the battery and using shielded wire to protect power wires is not a great idea.

And just to add my 2c to this tread, mags/simple bus/battery backup for primary instruments is all you need.
All this complexity for an RV is crazy IMO, risk/vs gain so you get easier starts or a couple knots at most, save a gal/hr LOP? :D
Is it really worth all this fancy design work/added complexity??
Is it more reliable than the millions of hrs accumulated with simple carb/FI/mags??
If it fails away from home base can it be repaired easily with help from local A&P??
 
Last edited:
Thank you John. The passive load is parasitic, fundamentally a battery drain when parked, if connected to the hot side of the master?

Hi Dan,

Yes, the ~30 mA passive load of a B&C LR3 or SB1 regulator V sense connection is parasitic to the battery if it is connected directly to the battery, will drain it over time if no charging system is active.

In Z101B the Ford regulator sense wire is jumpered to the power connection at the regulator so if the sense connection presents a passive load it doesn't matter because power is off.

Bob Nuckolls and B&C connect the B&C LR3 or SB1 sense wire to the bus so it does not see voltage fluctuations due to variations in field current into the regulator. These voltage fluctuations can make the regulator unstable. Some on VAF report jumpering V sense to power (terminals 3 and 6) with good results; the fewer contacts (switch and CB), connections, and wires along the way, the better.

If using LR3 or SB1 for the battery alternator in Z101B, one could jumper 3 and 6 and see what happens or one could add a relay for the V sense lead. I put the regulator on the engine bus in Z101 so no additional relay with the curious side effect that if the engine bus relay fails, the battery alternator setpoint goes up by the diode voltage drop which in the case of a Schottky diode is ~0.5V.

BTW I hope everyone with a backup alternator will do periodic stress tests in flight by turning off the main alternator.
.
 
yep

Using shielded wire for power sounds good however: I once had a short on the unprotected shielded wire from a Lightspeed Ign, when the inner wired melted it normally would have just opened up, but the shield prevented that from happening and the short continued until the shield finally burned thru. The result was eye opening as the smoke filled the airplane and the shielded wire continued to burn for what seemed like eternity. It also burnt all the wires in the bundle and burnt up the LSI box as well as it was the ground.
Lesson learned: protect all wires (despite the recommendation from LSI) at the battery and using shielded wire to protect power wires is not a great idea.

And just to add my 2c to this tread, mags/simple bus/battery backup for primary instruments is all you need.
All this complexity for an RV is crazy IMO, risk/vs gain so you get easier starts or a couple knots at most, save a gal/hr LOP? :D
Is it really worth all this fancy design work/added complexity??
Is it more reliable than the millions of hrs accumulated with simple carb/FI/mags??
If it fails away from home base can it be repaired easily with help from local A&P??

That's the beauty of our hobby; everyone can built what they want based on their own opinions, and that doesn't mean it's "crazy" to them...
 
Originally Posted by Mich48041 in post #6

Dan, what is the purpose of the 1N4933G diode in parallel with the Monkworkz contactor?
If the contactor contacts develop resistance, wouldn't the diode smoke?

Bill Judge of Monkworkz suggested it so the regulator's OV protection would not trip into lockout if the contactor was opened under load three times in the same power cycle. I suppose burned contacts could smoke the diode, but it would merely set up the possibility of a lockout.

BTW MZ-30L OV protection is set to 15.3V.

In one of my brainstorming schematics:
  • A Schottky 1N5822 is used to clamp any spike at a little lower voltage (maybe 525 versus 1,000 mV over battery voltage) compared to a standard diode although I'm not certain there's a scenario where it matters.
  • A progressive switch is used, if it's moved slowly / not slammed from up to down the generator will be disabled before the relay is opened.
 
Last edited:
Using shielded wire for power sounds good however: I once had a short on the unprotected shielded wire from a Lightspeed Ign...

The operative word being "unprotected". The armor example has a fusible link at the battery end. The idea is to make it harder to short...but it does, the burned wire will be a link enclosed in a fiberglass sleeve. Note it doesn't even hurt the tube.
.
 

Attachments

  • Fusible Link 2.jpg
    Fusible Link 2.jpg
    80.3 KB · Views: 112
The operative word being "unprotected". The armor example has a fusible link at the battery end. The idea is to make it harder to short...but it does, the burned wire will be a link enclosed in a fiberglass sleeve. Note it doesn't even hurt the tube.
.

Hi Dan, I am sure you meant "but if it does".

Could you point me to the thread that you talked about making your own fusible links.

Thanks.
 
More heresy...

Dan,

At the risk of being branded a heretic or of beating a dead horse….

A fear my last comments/thinking got lost in the bus-tie discussion, but with your stated goal of simple, and minimal pilot interactions so I thought I’d put the following thought experiment out there that I stumbled into while doing my panel.

I see two purposes of an essential bus. One is load shedding, and the other is bypassing the failed contactor. Since you have two batteries, and therefore two contactors the second one goes away. If you separate the busses, then so does load shedding. If we cross the bridge that two contactors replace the backup function of the essential bus, (here comes the heresy) then you can delete both Avionics/ESS switches, and the backup power generator switch (and deletions are always good if we still meet the requirements)

When I was doing my panel, since I’m more preoccupied with OV than most, I didn’t feed the G3X from both sides since 60V on either bus would then be a very bad day. After I did that, I did the split of equipment for safety since I could easily loose half the equipment.

Then suddenly my power supply diagram got a lot simpler (adjusted for your equipment): I created two completely conventional power systems that are fully independent – and no pilot interaction needed:

Small Bus / Alternator / Contactor: (Always powered by the backup alternator):
G5 (which is right next to my PFD)
GAD29
Right GDU460 (If Amps allow at low RPM – Will revert to G5 for ADHRS)
GMC507
GNX375
GSA28 Pitch
GSA28 Roll
GTR20 #1


Turn off the big contactor (or any other big failure on the big bus) – Engine still turns, you can talk on com1 (audio panel fail safe), have the G5 to keep upright, and the autopilot to compensate for only having the small screen, and a GNX to fly an approach – better than most steam gauge planes.


Big Bus / Alternator / Contactor (Always powered by the main alternator):
GEA24
GAD27
GDU Left
GEA24
GSU25
GAD27 E Trim
GTR 20 #2
GMA 245R
GDL51R
LEMO 1 &2
And the rest of the conventional bus.

Turn off the small contactor (or any other big failure on the small bus) – Engine still turns, you can talk on com2, Big screen to navigate and to keep myself upright. you are hand flying, and have lost the IFR navigator but still have the VFR G3X big screen gps to fly on.

Two alt/bat switches is all I need to control it Load shedding happens by default. No new switches to teach the pilot. Just a list of which equipment is on bus 1 or bus 2 somewhere on the panel. Really easy to understand what is powering what when.

Keeping them separate all of the time solves the battery imbalance issue, and weird lightning, HIRF or OV issues crossing over.

Although not strictly needed I still added a bus-tie and CBs to allow me to share across busses, since you are a fuse guy – that could be replaced with a bus1/2 for the 375 or anything else deemed essential (engine) to the same effect, but for the most part the idea is that it would have to be a really bad day before you would look at sharing across the busses. If the split is good I should never have to tie the busses, but I like to hedge my bets…

it’s a big pivot but feels (to me) like a better fit to your stated goals… it goes without saying that you are free to use or ignore at your preference.

Derek
 
Hi Dan, I am sure you meant "but if it does".

Could you point me to the thread that you talked about making your own fusible links.

Thanks.

Warren, I don't recall ever doing one.

Easy enough, good inline splices, fiberglass/silicone tube, adhesive heat shrink. Ordinary aviation wire 4 AWG less than the protected wire. Or buy some automotive fuse link wire.

B&C has kits.
https://bandc.com/product/fusible-link-kit-20-16-awg/
https://bandc.com/product/fusible-link-kit-24-20-awg/
 
Using shielded wire for power sounds good however: I once had a short on the unprotected shielded wire from a Lightspeed Ign, when the inner wired melted it normally would have just opened up, but the shield prevented that from happening and the short continued until the shield finally burned thru. The result was eye opening as the smoke filled the airplane and the shielded wire continued to burn for what seemed like eternity. It also burnt all the wires in the bundle and burnt up the LSI box as well as it was the ground.
Lesson learned: protect all wires (despite the recommendation from LSI) at the battery and using shielded wire to protect power wires is not a great idea.

And just to add my 2c to this tread, mags/simple bus/battery backup for primary instruments is all you need.
All this complexity for an RV is crazy IMO, risk/vs gain so you get easier starts or a couple knots at most, save a gal/hr LOP? :D
Is it really worth all this fancy design work/added complexity??
Is it more reliable than the millions of hrs accumulated with simple carb/FI/mags??
If it fails away from home base can it be repaired easily with help from local A&P??

Its good to see a voice of experience providing sound advice.

Adding an extra layer of insulation on power cables can be good - may not be necessary if the power cable is appropriately routed and secured, but it won’t hurt. Using screened or shielded cable when there is no need to protect against RF/EMI increases the risk of an uncontrollable short to ground/ screen/shield, adds unnecessary weight and isn’t part of any best practices that I am aware of.

Using a proven architecture from a respected and experienced source has much lower risk, even with minor modification than going with any alternative.

As I once heard the chief engineer for a large airplane program say during a design review “ I don’t see too many ponies but I do see what they left behind”.

What I think I see in parts of this thread is a lot of beer pitchers but they don’t contain beer. Y’all need to be careful what you drink.


Keith Turner
 
At the risk of being branded a heretic or of beating a dead horse….

Not by me. A pair of independent busses is a reasonable approach. I've already expressed why I would leave out the crosstie contactor.

I also very much support a KIS approach in every case where equipment requirements allow it. Mostly VFR, mags or P-mags or a combination, a carb or constant flow FI? Yeah, put a Z11 in it, or a TCW IBBS. My own is pretty much like that, conventional, with a small IBBS for the single EFIS, a big lead-acid main, and a little backup for the second EI. I've punted an alternator over the Smokies and flown home without concern.

Thing is, not everyone is me, or Walt, or Keith.

Using a proven architecture from a respected and experienced source has much lower risk, even with minor modification than going with any alternative.

As I once heard the chief engineer for a large airplane program say during a design review “ I don’t see too many ponies but I do see what they left behind”.

What I think I see in parts of this thread is a lot of beer pitchers but they don’t contain beer. Y’all need to be careful what you drink.

Ahhh, the ghost of Friar Torquemada. Insistence on dogma as the only true path, based on faith. No factual illustration, but plenty of inflammation. Fire up the crowd, light the torches!
 
Consider adding some zeners into to the mix.

Thanks Dan,

If you do decide to keep the busses tied to together and haven't already take a look at the cirrus SR22 diagram below. Also very low pilot interaction needed. For me two points of interest - keeping alternator two live all the time allows you to detect a failure of it before you need it (can be mitigated by a preflight test), and also note at all the zener diodes to catch OV spikes, and the diode between the small and large bus to auto shed load.

They were one of the 1st single engine part 23 certs after the hiatus, and one of the 1st where keeping the electrons flowing was critical (for the avionics), so I feel it got a little complex, but there are some really good ideas.

Diamond was more keeping the engine running focused, and learned a lesson about HIRF and electrically dependent engines near a particularly powerful antenna in Austria - good news when you get far enough away the engine starts again... . Their design is quite a bit different (they really like batteries), but again note the zeners protecting the (Da50) ECUs (they don't bother else where), that I'm told solved the problem - though certainly not a guarantee, they helped. They only use them on the engine which at least for me feels like a good fit here.

The zeners can be done with small fuses on them, so they won't protect against long duration overvoltage, but they do stop quick transients without risking blowing the circuit, or big fuses so they work as a ov cut out, like the B&C OV circuit. Since you already have the big fuse version in the B&C, I'd suggest you don't need one elsewhere, unless part of a fail to battery back up solution.

Also, note the fuses in series, which give a nice certified example in support of your small fuse/big fuse thinking and testing.

On the topic of 4 master switches, I had a conversation with the small airplane directorate on that very topic. For part 23 their position (20 years ago) is that even though the 23.13something rule says "one hand with a single movement" to turn off all masters their limit was two unless you did a bar like a twin cessna. Apparently our fingers are different lengths. Other than that it would be certifiable :D

Those that trust certified airplanes, and enjoy sausage should not watch either being made...

Derek
 

Attachments

  • SR22.jpg
    SR22.jpg
    285.4 KB · Views: 110
  • DA50.jpg
    DA50.jpg
    49.8 KB · Views: 77
  • 310.jpg
    310.jpg
    36.8 KB · Views: 48
Not by me. A pair of independent busses is a reasonable approach. I've already expressed why I would leave out the crosstie contactor.

Thing is, not everyone is me, or Walt, or Keith.

Ahhh, the ghost of Friar Torquemada. Insistence on dogma as the only true path, based on faith. No factual illustration, but plenty of inflammation. Fire up the crowd, light the torches!

Dan,
And you could have added "burn the heretics".

But you would be wrong - I am all for a simple, easy to analyze electrical distribution system that provides the maximum capability following a first failure and has no pathways for cascading failures or nasty surprises.

In a past life (not the one you refer to) I worked on a number of civil and military aircraft programs as an engineer and manager including the one that got in the news for shooting down a balloon. Having collaborated and led some of those activities, I cannot think of one of those folks that I worked with that would have even considered doing the fuse blowing experiment. They would have done the analysis plugged in the variables and tabulated the results.

So you could say that was the first clue that I was, maybe, observing the ghost of the Pied Piper (we need to stay away from politics and religion).

My observation (from over 40 years experience) is that doing this kind of work (designing control systems and power distribution electrical systems) requires a good understanding of electrical engineering, batteries, rotating electrical machines (alternators, permanent magnet generators) and sufficient electronic circuit design engineering skills to be able to analyze the control circuits used for these devices.

We could have a more detailed discussion about battery characteristics or the reason why most modern electronics present a negative impedance load to the power distribution system. How to tie high performance LIPO batteries together without having a fire is a subject all on its own.

We could talk about how the source impedance of the power input to those units affects the stability of the individual unit switched mode power converter and how to compensate for those effects.

We could have a detailed discussion on fuses, circuit breakers and home made weak links. BTW the home made weak links was the second clue to the ghost from the past.

A few years back we (you and I) had an on line discussion about MIL-HDBK-217B and failure rates applicable to our aircraft, how COTS had changed the landscape and what numbers to use. We could have a comprehensive discussion on FMEA, failure rates and failure analysis.

For many builders the electrical system, selection of avionics and wiring is a big unknown grey area. Its an area to take very seriously - not saying that you don't but I think you need to move from, what appears to me, to be the "suck it and see", "patch it until it works" approach to the professional, analytical, theoretical proven by test methods that starts off with a set of written requirements and objectives.

Its your turn to fill the pitcher - beer would be good, I am allergic to cool aid.

Keith turner


Friar Tomas de Torquemada - Spanish Inquisition - Not so nice guy - you couldn't possibly be comparing me to him.

Archbishop Juan de Torquemada - uncle to Tomas - intelligent, articulate, well educated - not so well known and not a bad person as far as one can tell. Maybe you were thinking of this guy?

Friar Juan de Torquemada - Franciscan monk famous for Monarquía indiana Engineer, philosopher, Architect, Public Servant- all around good guy - sure you must have been thinking of comparing me to this guy in your reply.
 
Last edited:
... I cannot think of one of those folks that I worked with that would have even considered doing the fuse blowing experiment. They would have done the analysis plugged in the variables and tabulated the results.


...... Its an area to take very seriously - not saying that you don't but I think you need to move from, what appears to me, to be the "suck it and see", "patch it until it works" approach to the professional, analytical, theoretical proven by test methods that starts off with a set of written requirements and objectives.

Another example of how different folks approach things differently.

Dan is a "I believe it when I see it" kind of guy, not shy to build his own test rig to experimentally explore an issue. Take for example the fire protection work he has done.

Another that is strongly in the experimental testing approach is Ross Farnham. In a conversation with him about liquid cooling system design for aircraft engines, he essentially said, "that 1st law of thermo stuff is fine, but I find that you never really know until you build it and test it" - I'm paraphrasing from a long-ago conversation, so I hope I don't mis-characterize Ross' point of view.

Throughout my career, I was an unusual guy because I did a lot of analytical and computational work, and also a lot of experimental work. Use the approach that most easily gets at what you want to know. I suppose if you had a detailed analytical model of a series set of fuses including a heat-transfer model that would predict time to blow, then you might turn to that tool to answer the question. But it was pretty quick and easy for Dan to wire up and smoke a few fuses to demonstrate the conclusion. Similarly, I could imagine building a computational heat transfer model of a firewall with various thermal protection devices on it. In fact folks in my group did this to design thermal protection systems for re-entry vehicles. Everything from ablative layers like Apollo and Orion, to reusable like Shuttle tiles. Honestly I doubt my computational model would be very accurate for the firewall insulation problem, and I would much more trust the experimental approach there.

Both approaches have merit.
 
Keeping them separate all of the time solves the battery imbalance issue, and weird lightning, HIRF or OV issues crossing over.

Derek, I am inclined towards this line of thinking (at least as a thought experiment, this whole thread has really got me to thinking overall). In the hypothetical airplane this thread is about, where would your SDS-style injector/ECU/fuel pump be located bus wise? You would at least need enough switches to verify failsafes on runup, no?

Also, it wasn't completely clear if your scenario had 2 batteries, one on each bus, or 2 tied into both busses?

I don't know if it helps to think of an "engine-keep-running bus" and a "everything else bus", with more protection/redundancies built into the first, and willingness to allow the 2nd to die completely (assuming at least a G5 with battery backup). This would keep the prop turning and head for the ground, a good fighting chance even in IMC.
 
For many builders the electrical system, selection of avionics and wiring is a big unknown grey area.

Whether we are the "magneto's from my cold dead hands" type or not, people ARE building electrically dependent airplanes, many times because of Dan's initial reasoning: I need to keep this other avionics stuff running to fly this thing, and therefore need redundancy/failsafe/etc, might as well go all in and make the engine that way too.

I think this discussion is a great way to leading us in time toward a somewhat best practices approach to these new realities.

It kind of feels like the kit airplane and mostly the power source are solved problems. Most aren't experimenting with new wing shapes and many aren't experimenting with new power sources, sticking with the known, proven drop-ins (Vans kit and lycoming).

Many builders are searching for the equivalent in the electrical/avionics arena, and arrive at the top 2 (3?) vendors, which are actually multi-node complex systems. Is most of it overkill for a weekend warrior VFR burger-run plane? Probably. But the zeitgeist seems to be moving towards very capable machines with lots of complexity, even after simplification has been applied. Having a Knuckles 2.0-style reference geared toward total electrical dependency would be a great starting place for many builders, without 40 years of electrical design experience being a prerequisite or rediscovering the wheel each time.
 
Humor an example. Note the dual bus block diagram I posted above has no crosstie contactor. Having one requires the PIC to sort out why the dead bus went dead, before he/she connects it to the live bus.

Agree and disagree. You are going to have enough battery reserve to get somewhere close assuming no sudden failure of one of the batteries where it isnt shorted, opened, or disconnected by its own BMS. Being prudent you leave everything alone, land, do the diagnosis on the ground and determine if tying the busses can get you home safely, as would be in the case where one alternator is dead. The next common failure item is a dead battery contactor. In that case the buss tie would get you home. If one battery is moderately discharged, the likely case with LifePO4 batteries running with no charge for lets say 30 min, as others mentioned the inrush current can be a problem. I have doubts about this as they are of low impedance and charge quickly. Easy enough to test.
 
Another example of how different folks approach things differently.

Dan is a "I believe it when I see it" kind of guy, not shy to build his own test rig to experimentally explore an issue. Take for example the fire protection work he has done.

Another that is strongly in the experimental testing approach is Ross Farnham. In a conversation with him about liquid cooling system design for aircraft engines, he essentially said, "that 1st law of thermo stuff is fine, but I find that you never really know until you build it and test it" - I'm paraphrasing from a long-ago conversation, so I hope I don't mis-characterize Ross' point of view.

Throughout my career, I was an unusual guy because I did a lot of analytical and computational work, and also a lot of experimental work. Use the approach that most easily gets at what you want to know. I suppose if you had a detailed analytical model of a series set of fuses including a heat-transfer model that would predict time to blow, then you might turn to that tool to answer the question. But it was pretty quick and easy for Dan to wire up and smoke a few fuses to demonstrate the conclusion. Similarly, I could imagine building a computational heat transfer model of a firewall with various thermal protection devices on it. In fact folks in my group did this to design thermal protection systems for re-entry vehicles. Everything from ablative layers like Apollo and Orion, to reusable like Shuttle tiles. Honestly I doubt my computational model would be very accurate for the firewall insulation problem, and I would much more trust the experimental approach there.

Both approaches have merit.

Well stated as always. “Degrees of merit”, maybe better? Comes down to costs and consequences as with most things. Lots of early Rocket failures; then, some short decades later the first Space Shuttle flew with a live crew (wisdom of which is very debatable). Later still, SoaceX is somewhere in the middle and has/will sacrificed some expensive launch vehicles. Being a non-gov entity, I’m sure they found where they thought the development costs and risk curves crossed.

Neither approach is 100% right all of the time. We get to see a decent observation of both here and which I enjoy (and counts for nothing).
 
Also, it wasn't completely clear if your scenario had 2 batteries, one on each bus, or 2 tied into both busses?

I don't know if it helps to think of an "engine-keep-running bus" and a "everything else bus", with more protection/redundancies built into the first, and willingness to allow the 2nd to die completely (assuming at least a G5 with battery backup). This would keep the prop turning and head for the ground, a good fighting chance even in IMC.

The base line is two standard battery / alternator / buses, where you can live with either side dying. One of each on each bus

There are a few options you get to pick from when you get to the SDS since you can't split it across the bus in a way that either bus can fail nicely. The SR22 diagram solves and basically does what you ask flowing one way from the main to the essential bus. I'd delete all of the little busses down to just main and essential if were going that path, but I also dropped my avionics masters when my avionics stopped being "tuneable", so its a preference thing. With that design have two busses always live, if the big ones dies it's dead, but if the small bus dies you still have the large bus picking up the slack, and they are for the most part fully independent.


One other way (my preferred) would be to diode-or just the engine essential bus to both of your electrical systems. The diodes only allow current to flow into the bus, but don't allow anything to go from one bus to the other. The main reason I prefer it, is that it minimizes stuff on a shared bus and means I can use smaller diodes, again its a preference thing. Its probably also slightly fewer components.

You start with two full battery / alternator /regulator / contactor systems splitting the avionics / devices so you can live on either and then for the SDS, you wire it up just as Ross draws it, i.e. take power from the hot side of
each contactor but add a diode in each line to only allow the flow from the respective bus to the engine bus.

Diode-or has one disadvantage in that the highest voltage will pass through so you don't have OV protection transfer protection on that bus, but as Dan points out that's what your external OV protection is for, but something to think about. One mitigant would be the zener's that I mention above and shown in the DA50 pic. (and connecting them without the diodes of course also has that issue)

Zooming out on the DA50 as an example of a diode or for ECUs - looks something like this - crudely edited one is the SR22, where main Dist Bus 2 is the always up bus. The colourful picture a few posts back is SR22 in all its glory.
 

Attachments

  • da50-2.jpg
    da50-2.jpg
    25.9 KB · Views: 54
  • SimpleCirrus.jpg
    SimpleCirrus.jpg
    83.9 KB · Views: 56
Last edited:
The base line is two standard battery / alternator / buses, where you can live with either side dying. One of each on each bus

There are a few options you get to pick from when you get to the SDS since you can't split it across the bus in a way that either bus can fail nicely. The SR22 diagram solves and basically does what you ask flowing one way from the main to the essential bus. I'd delete all of the little busses down to just main and essential if were going that path, but I also dropped my avionics masters when my avionics stopped being "tuneable", so its a preference thing. With that design have two busses always live, if the big ones dies it's dead, but if the small bus dies you still have the large bus picking up the slack, and they are for the most part fully independent.


One other way (my preferred) would be to diode-or just the engine essential bus to both of your electrical systems. The diodes only allow current to flow into the bus, but don't allow anything to go from one bus to the other. The main reason I prefer it, is that it minimizes stuff on a shared bus and means I can use smaller diodes, again its a preference thing. Its probably also slightly fewer components.

You start with two full battery / alternator /regulator / contactor systems splitting the avionics / devices so you can live on either and then for the SDS, you wire it up just as Ross draws it, i.e. take power from the hot side of
each contactor but add a diode in each line to only allow the flow from the respective bus to the engine bus.

Diode-or has one disadvantage in that the highest voltage will pass through so you don't have OV protection transfer protection on that bus, but as Dan points out that's what your external OV protection is for, but something to think about. One mitigant would be the zener's that I mention above and shown in the DA50 pic. (and connecting them without the diodes of course also has that issue)

Zooming out on the DA50 as an example of a diode or for ECUs - looks something like this - the colorful one is the SR22 which I should likely simplify down to make my point clearer....

Can you help an electrically challenged ME out? I assume the zeners you mention are those depicted between the hot and ground of the ECUs. I will (have to) diode redundant power to the fuel injectors if I want any power redundancy for them.

How would you go about sizing those components?

What are these diodes (most likely) failure mode(s) i.e. will any associated risk outweigh the intended purpose?

Any efforts to get me smarter here would be greatly appreciated. Thanks.
 
Back
Top