Simplified Approach

KatanaPilot · Feb 22, 2023

DerekS said:
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. .

Assuming you have the dual ECU system - I’m not sure why you can’t wire for separate power sources. You have one ECU, fuel pump and coil pack on a respective (hot battery) bus. Injector power is determined by the PRI/BAK switch on a 4 cylinder. On the 6 cylinder power is separated 1,2,3 to one bus and 4,5,6 on the other or all on one bus - as determined by the PRI/NOR/BAK switch.

That’s how we did our RV-10. We have a cross tie that’s used only to tie both ETX-900 batteries together for engine start. No diodes required. Some intervention required if a bus is lost or an ECU fails.

keitht · Feb 22, 2023

Analysis versus lab testing

Steve, I don’t disagree with your views on experimental results versus analysis especially when there are a number of unknown variables and the experiment is to test a hypothesis, gather data to refine the model and better predict future performance. Both Dan’s excellent work on firewall forward fire protection and Ross’s ECU work fall into that category I believe.

Another good example is airframe fatigue testing. There are expectations, always surprises and the big area of debate is always the pressure and stress profile that is used to do accelerated testing and the locations and extent of damage in the test article compared to what shows up on the in-service fleet.

In fact, I took great interest in the firewall flammability work, learned a lot and made changes to my own RV-7 firewall penetration design as a result.

My issue with the fuse test was not so much with doing the test but that the test conditions were extreme limit conditions. If you hit a 5 amp fuse in series with a 20 amp fuse with 100 amps what will happen is pretty obvious and you don't learn much - which is usually the case with a quick and dirty test that doesn’t represent an in service fault condition.

What happens when you change the test set up to be more representative of why you have a 20 amp and a 5 amp fuse in series and the likely in service conditions.

The 20 amp fuse is in a distribution path upstream with multiple other feeds coming off it and is passing a current of say 75% of rating (15) amps. The 5 amp fuse is in one of those downstream feeds and passing a current of say 70% of rating (3.5 amps) as the starting condition.

The 5 amp feed then goes short to ground with a resistance of 0.5 ohms in circuit. The answer is neither obvious nor predictable with a simple model - it depends on the type of fuses, ambient temperature of the two fuses, actual voltage and source impedance of the supply and how the other feeder supply lines are likely to react to a short on one of the feeder lines especially if they are feeding constant power demand loads that react to a reduction in supply voltage by increasing their current draw.

I am sure you get my point.

The fuse test done in a more dynamic representative configuration would yield considerable usable data that would be difficult to obtain by simulation or modeling. The test as performed convinced me of the exact opposite.

Keith Turner

DerekS · Feb 22, 2023

KatanaPilot said:
Assuming you have the dual ECU system - I’m not sure why you can’t wire for separate power sources. You have one ECU, fuel pump and coil pack on a respective (hot battery) bus. Injector power is determined by the PRI/BAK switch on a 4 cylinder. On the 6 cylinder power is separated 1,2,3 to one bus and 4,5,6 on the other or all on one bus - as determined by the PRI/NOR/BAK switch.

That’s how we did our RV-10. We have a cross tie that’s used only to tie both ETX-900 batteries together for engine start. No diodes required. Some intervention required if a bus is lost or an ECU fails.

Nothing wrong with that approach (or the variation in Dan's picture) and I'm always a fan of fewest components. - more isolation should reduce the chances of bad interactions in that part of the system. You'll see bus tie in lots of my earlier posts, but was working on a solution without a bus tie to avoid the temptation to link the batteries together, which means either a bus a/b switch or diodes. Defaulted to diodes with no moving parts and no switches to switch (bus failure only). I also defaulted to keeping Ross's instructions in place for standardization and ease of communication since I didn't see a compelling reason to change. (my standard only, I see the merit to your approach), and we underline why each two bus system is a little different, and add to the considerations for folks building their own.

DerekS · Feb 22, 2023

Freemasm said:
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.

Those are the ones. They are not power redundancy, they function basically as surge suppression. They are a diode that you connect backwards, but they have the property that as they hit their reverse break down voltage they basically short (in this case to ground, and they do it quickly), so as soon as the hot goes above the breakdown voltage they clamp like a surge suppressor, ensuring the voltage stays on the hot stays at the breakdown voltage. If the voltage stays high too long, of course the fuse will fail and the voltage will spike, and you have different issues.

Failure mode is typically a short, but open is an option which is why we have the fuse in series, and smaller - e.g. sized to match the typical load on the device and quite a bit smaller than the one protecting the circuit so that a failure won't blow the power supply.

Its purpose is to sink any short duration, high voltages that might upset the ECU. Common feature in electrically dependent part 23 world, so I figure worth considering for our world.

DanH · Feb 23, 2023

keitht said:
My issue with the fuse test was not so much with doing the test but that the test conditions were extreme limit conditions.

You're reading a awful lot into a simple demonstration.

Ivan's question was...

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

I responded with "I dunno, give it try", but later, in the shop, I got thinking that smacked of being dismissive. So I walked over to the junk drawer, spent five minutes on bench assembly, and took one cell phone picture.

The question specifically addressed different fuse ratings in series, thus ATO fuses for both feeder and load so rating would be the only variable. Other forms of circuit protection can be faster or slower to open. If I had put something else in the feed, the demonstration would not have been as clear.

No good deed goes unpunished.

If you hit a 5 amp fuse in series with a 20 amp fuse with 100 amps what will happen is pretty obvious...

Given the question, apparently not to everyone.

The answer is neither obvious nor predictable with a simple model....I am sure you get my point.

That a detailed analysis is more complex. Yep.

idubrov · Feb 23, 2023

DanH said:
That a detailed analysis is more complex. Yep.

Thank you for doing doing your demo! It surely matches the intuition (and the theory, as per below).

I looked for a theory of fuses, and found this document.

It gives a pretty specific rule how fuses should be selected when they are connected in series (page 5, "Selective Coordination"). It also should allow comparing fuses of different types as well.

Not sure how do I estimate this "arcing I2t" (to calculate the "total clearing I2t"). On a glance, 20A and 5A have a big difference in their melting I2t (according to Littelfuse ATOF datasheet, it's 520 and 26), so maybe that's the extent I am willing to dip in the theory.

Freemasm · Feb 23, 2023

DerekS said:
Those are the ones. They are not power redundancy, they function basically as surge suppression. They are a diode that you connect backwards, but they have the property that as they hit their reverse break down voltage they basically short (in this case to ground, and they do it quickly), so as soon as the hot goes above the breakdown voltage they clamp like a surge suppressor, ensuring the voltage stays on the hot stays at the breakdown voltage. If the voltage stays high too long, of course the fuse will fail and the voltage will spike, and you have different issues.

Failure mode is typically a short, but open is an option which is why we have the fuse in series, and smaller - e.g. sized to match the typical load on the device and quite a bit smaller than the one protecting the circuit so that a failure won't blow the power supply.

Its purpose is to sink any short duration, high voltages that might upset the ECU. Common feature in electrically dependent part 23 world, so I figure worth considering for our world.

Thanks. I realized those weren't for rectifying. Your post got me thinking since there is no redundancy for fuel injectors, only redundant power from separate buses; how do i best protect them? Seems like cheap insurance if applied correctly and if it doesn't introduce new failure modes. Still a little conflicted but will think more about it.

keitht · Feb 23, 2023

DanH said:
Given the question, apparently not to everyone.

That a detailed analysis is more complex. Yep.

That a detailed analysis is more complex ……….and there are variables where you don’t know the numbers so doing a practical test or experiment will be the best way forward, give the highest fidelity answer.

I’ll PM you…… feel the need to get the wheels back on the wagon and focus on what is essential for the Vans community. On reflection, I made some incorrect assumptions and over reacted.

Keith Turner

DerekS · Feb 23, 2023

A few pictures

Just thought I would add sketches of what I see in my mind when I say diode-or, A/B switch, Bus Tie, and cirrus approach for a dual bus system

I don't have an avionics bus (or 4), in these drawings because I didn't build one into my planes, but each one could be attached to any of the busses (or switched from the hot for that matter as per the builders preference without affecting the base).

All of these approaches presuppose you have made the decision to have two batteries and two alternators. If not the answers of course look different.

Depending on what devices you have on each of the busses, all three are valid approaches in my mind and as simple as I can come up with for a dual bus.

Disclaimer: Don't blindly use these to build a plane, and note that some of the smaller well understood pieces (flyback diodes, ANLs etc) were omitted for clarity, and these are not the only way to skin this cat.

mstrauss · Feb 24, 2023

First, thank you all for this great thread. I am just starting to research EFI/EI solutions, and plan to reread most of this thread again to glean what I can. That said, I did not see mention of FlyEFII's Bus Manager. It looks like an interesting solution for bus management, with some additional support for auto failover of the fuel pump.

KatanaPilot · Feb 24, 2023

mstrauss said:
First, thank you all for this great thread. I am just starting to research EFI/EI solutions, and plan to reread most of this thread again to glean what I can. That said, I did not see mention of FlyEFII's Bus Manager. It looks like an interesting solution for bus management, with some additional support for auto failover of the fuel pump.

I expect you will get some feedback on this question, particularly around the "single point of failure" aspect. Some folks will tell you this magic box is the greatest thing ever and others will tell you that it's too many eggs in one basket - especially for an electrically dependent airplane. It's up to you to decide your risk tolerance.

It's not for me, although I do have an electrically dependent airplane.

Suggest you use Google advanced search for the bus manager with the site "vansairforce.net" versus the VAF search feature.

mstrauss · Feb 24, 2023

KatanaPilot said:
I expect you will get some feedback on this question, particularly around the "single point of failure" aspect. Some folks will tell you this magic box is the greatest thing ever and others will tell you that it's too many eggs in one basket - especially for an electrically dependent airplane. It's up to you to decide your risk tolerance.

It's not for me, although I do have an electrically dependent airplane.

Suggest you use Google advanced search for the bus manager with the site "vansairforce.net" versus the VAF search feature.

Krea, Thanks for the response. "The more I learn, the..."

As a newbie, I'm really starting to appreciate this forum, and all it offers. For now, I'm going to just keep swimming. My engine order will not be fulfilled until late this year. Hopefully, by then I will have enough education to be confident in choosing a direction.

rocketman1988 · Feb 24, 2023

bus manager

I am using some of the functions of the bus manager, namely the diode bridge and auto fuel pump switch. I have had zero issues with it.

Realize that the auto fuel pump switch can easily be duplicated with a fuel pressure switch and a relay. Likewise, the diode bridge can also fashioned with readily available parts.

There is no "magic" to the bus manager, it just consolidates the features into one box.

DanH · Feb 24, 2023

rocketman1988 said:
I am using some of the functions of the bus manager, namely the diode bridge and auto fuel pump switch. I have had zero issues with it.

Bob, curiosity question please. Have you ever noted the temperature rise of the diode box under full electrical load? Does Robert want it hot on purpose to reduce forward voltage? Does the fan run all the time, or part time to hold a temperature range?

DanH · Mar 1, 2023

DerekS said:
If you do decide to keep the busses tied to together and haven't already take a look at the cirrus SR22 diagram below. ....note the zener diodes to catch OV spikes...

Another good contribution. I'm a terrible electronics guy, but the symbols suggest they are TVS diodes, a zener subset intended for this kind of application. Very interesting.

The applications as drawn raise a few questions. Want to take a shot?

To clarify, I stripped and simplified. Here are the individual child busses:

Note the arrangement of the TVS diodes and companion fuses. Different fuse ratings, 20's and 8's, with the 20's between the TVS and ground, while the 8's are on the bus side of the TVS. What is going on here?

rocketman1988 · Mar 1, 2023

DanH said:
Bob, curiosity question please. Have you ever noted the temperature rise of the diode box under full electrical load? Does Robert want it hot on purpose to reduce forward voltage? Does the fan run all the time, or part time to hold a temperature range?

Are your questions out of genuine curiosity or a dig at EFII?

DanH · Mar 1, 2023

rocketman1988 said:
Are your questions out of genuine curiosity or a dig at EFII?

Curiosity. The heat is a power loss, but forward voltage is less when hot. For a given amperage, a designer could reduce voltage drop by increasing thermal resistance, i.e. less heat sink. Example attached.
.

rocketman1988 · Mar 1, 2023

DanH said:
Curiosity. The heat is a power loss, but forward voltage is less when hot. For a given amperage, a designer could reduce voltage drop by increasing thermal resistance, i.e. less heat sink. Example attached.
.

Thanks for that info. I have not noticed an appreciable temperature rise, though I haven't specifically looked for it. I have a bunch of temp sensors that I made for the airplane, I will attach one to the bus manager and put one in the fan stream, as well.

I won't be able to do anything, though, until the airplane comes back from the paint shop...currently down at Evoke...

DerekS · Mar 1, 2023

I can guess....

DanH said:
Note the arrangement of the TVS diodes and companion fuses. Different fuse ratings, 20's and 8's, with the 20's between the TVS and ground, while the 8's are on the bus side of the TVS. What is going on here?

Might not have been the best example since it opens up a whole other topic..

I'll open with no relation to the program, so guessing and don't have the the parts manual so don't know the size of the diodes, but those are some pretty big fuses so very likely also some beefy diodes. The order doesn't matter.

Best guess, came from the lightning impact analysis based on the size and placement. Nav/Strobes are lighting zone 1, and lots of energy potential there.

The 2nd big one (ESS1) does either side of the essential bus, and since close to the PFD1 etc, at the same time serves as protection to the screens as the 8A ones do elsewhere.

The other two 8A ones are protecting the busses with the screens which they've been doing for a while. If you look at some of their other layouts you will see small ones just associated with the PFD/MFDs. Size increase from previous likely grew out of the same lightning analysis. I'm not aware of anyone else who does this for the screens for small airplanes, but Cirrus has had it from the early glass days, so could be baked in somewhere.

I'd be more worried about the ECUs than the screens which is where I see the application in our space.

Lots of conjecture here, but seems like it would fit.

Derek

rocketbob · Mar 1, 2023

DanH said:
Note the arrangement of the TVS diodes and companion fuses. Different fuse ratings, 20's and 8's, with the 20's between the TVS and ground, while the 8's are on the bus side of the TVS. What is going on here?

Terrible scheme to do OV protection. When the OV condition occurs, the TVS closes and causes the fuse to blow. The problem is, in the amount of time it takes a thermal device such as a fuse to work, the damage has been done.

And...the fuse should be on the supply lead, the TVS anode connected to ground. So its drawn wrong. Otherwise if the circuits are wired exactly as the drawing then it simply wont work for anything.

DerekS · Mar 1, 2023

rocketbob said:
Terrible scheme to do OV protection. When the OV condition occurs, the TVS closes and causes the fuse to blow. The problem is, in the amount of time it takes a thermal device such as a fuse to work, the damage has been done.

It's not OV protection in that sense but rather for transients. e.g lightning, spikes, noise. The TVS won't pull the voltage to zero it will only provide a path to ground to while the voltage is above the design voltage - and that is does MUCH faster than a fuse, basically holding the voltage at its design limit until it melts or the small fuse opens

Part of the design is that if short high voltage transient that exceeds the capability of the diode / fuse combo to route the excess power to ground it absorbs all it can and then bows out, which is what happens when the fuse blows. To your point, at that point the device is subject to whatever is on the wire. The diode becomes a non factor.

A sustained OV with enough amps behind it would of course exceed the capability of the protection, and the fuse would blow. The slower acting normal OV protection that is (relatively slowly) warming up the fuse or CB on the field line takes care of that case.

You don't want the scenario where you have a big transient zap that melts the diode saving the main device, only to lose the main device because the diode fails as a short to ground, hence the fuse.

Derek

Mich48041 · Mar 5, 2023

Really Simplified Approach

Below is a really simple electrical system. If the alternator fails, the battery is the backup.
Mount switches in order of importance with least important on far right.
If alternator fails, shut off unneeded loads on the right side.
The pilot won't have to remember what loads are on what bus.

Mich48041 · Mar 5, 2023

With Brownout protection

Same as schematic above but with brownout protection added.

rocketbob · Mar 5, 2023

DerekS said:
It's not OV protection in that sense but rather for transients. e.g lightning, spikes, noise.
Derek

Cannot work for that purpose. Fuses are too slow.

Freemasm · Mar 6, 2023

Mich48041 said:
Same as schematic above but with brownout protection added.

Unprotected leads off of a battery? There's a lot more here that could be scrutinized.

Mich48041 · Mar 6, 2023

What do you mean by "Unprotected leads off of a battery?"
The only thing connected to the battery is the battery contactor.
If you are referring to the feeder for the main power bus, every diagram in
Bob Nuckolls' book is wired that way.

PaulvS · Mar 6, 2023

Freemasm said:
Unprotected leads off of a battery? There's a lot more here that could be scrutinized.

It's essentially the same as the RV Aircraft Wiring Schematic that was provided in the Vans wiring kit, apart from an ammeter shunt between the master relay and the (single) bus. Do you think something is not correct?

Mich48041 · Mar 6, 2023

If it is desired to measure current, a hall effect sensor has advantages over a shunt.
The wire of interest does not have to be cut or terminals crimped or shunt mounted.
https://grtavionics.com/media/AMP_HE_RevD.pdf

DanH · Mar 6, 2023

Joe's diagrams are both fundamental and traditional. No concerns.

If I may, the difference between the "simplified approach" in post #1 and Joe's really simple system is the equipment they support, and the mission implied by that equipment. Once upon a time, we wired 90% of the EAB fleet just like Joe describes. Now we're moving to three screens and EFI...

DerekS · Mar 6, 2023

rocketbob said:
Cannot work for that purpose. Fuses are too slow.

Another way of looking at it:

Try to picture it without the fuses. The fuses do not play into the protection for transients, the reverse biased diode does that work. The fuse is there in case the diode fails short.

Derek

Freemasm · Mar 6, 2023

Mich48041 said:
What do you mean by "Unprotected leads off of a battery?"
The only thing connected to the battery is the battery contactor.
If you are referring to the feeder for the main power bus, every diagram in
Bob Nuckolls' book is wired that way.

Sorry. My bad, big time. The comments made me go back and look. From the phone screen, thought you had added an unprotected "brown out battery".

Now hopefully a better question. Assuming you manage the load during start, why wouldn't you diode into the buss versus individual radios?

Thanks

Mich48041 · Mar 6, 2023

why wouldn't you diode into the buss versus individual radios?

The voltage booster that I suggested does not have the capability to power a whole bus. Even if it did, the main power bus would then have to be isolated from the start circuit with a diode capable of carrying the full aircraft load. Without that additional diode, the starter would pull the bus voltage down regardless of any voltage booster. The dual diodes that I suggested only cost a dollar or so each. They are electrically isolated so that they can be bolted directly to sheet metal.

PaulvS · Mar 6, 2023

Mich48041 said:
If it is desired to measure current, a hall effect sensor has advantages over a shunt.
The wire of interest does not have to be cut or terminals crimped or shunt mounted.
https://grtavionics.com/media/AMP_HE_RevD.pdf

I was puzzled by freemasm's comment, since it is my understanding that the master contactor is what "protects" the feed to the bus in the Nuckols diagrams, hence my question.

I already installed the Dynon shunt on the firewall. The connections between the two contactors and to the shunt are copper bar. I just went with the shunt that Dynon provided in the installation kit and followed pictures on VAF and Vans drawings for the firewall mounting.

I don't mean to hijack Dan's thread but the traditional approach is simple and appropriate for my simple RV.

Mich48041 · Mar 6, 2023

the traditional approach is simple and appropriate for my simple RV.

The advantage of building our own aircraft is that we can personalize it, including the electrical system.
A builder might want a complicated system with lots of features.
Others might want a simpler system with less to go wrong.

DanH · Mar 6, 2023

PaulvS said:
I don't mean to hijack Dan's thread but the traditional approach is simple and appropriate for my simple RV.

No problem here Paul. The core system for mine is also stone simple...but again, it was conceived almost 15 years ago to fly with a lead battery, mags, and a single VFR EFIS. LiFePO, EFI, and multiple LRU's change the requirements. Add-ons and workarounds have their limits. And we're running it all with the same meat servo C and C.

scsmith · Mar 6, 2023

Mich48041 said:
If it is desired to measure current, a hall effect sensor has advantages over a shunt.
The wire of interest does not have to be cut or terminals crimped or shunt mounted.
https://grtavionics.com/media/AMP_HE_RevD.pdf

I'm guessing by the name in this URL that the GRT EFIS will support a hall-effect current measurement as described. Will other brands of EFIS support hall-effect current sensing?

Will most EFIS nowadays support two current sensors so that I could monitor BOTH the alternator output and the battery charge/discharge? Wouldn't that be nifty?

johnbright · Mar 6, 2023

Mich48041 said:
Same as schematic above but with brownout protection added.

Is the brownout booster capable of responding quickly enough? Have you tested it in situ?

Mich48041 · Mar 6, 2023

I have not tested a DC - DC voltage booster. Bob Nuckolls included one in
his Z-101B electrical architecture. Once the start switch is activated, it will
take time for the start contactor to energize. The DC-DC booster should
start working immediately since it has no moving parts and is all electronic.
By the time that the pilot realizes that the engine has started and releases
the start switch, the bus voltage will have recovered.

DanH · Mar 11, 2023

Lots of excellent contributions and comments.

Referring back to the wiring in post 1, we had two comments of particular interest. The first was the possibility of an OV event taking out some or all components.

There are several things in play here, notably (1) how quickly the OV crowbar kicks in, (2) how much voltage the "victim" components are designed to withstand, and (3) interplay with the EarthX battery BMS.

In this case, the installed Garmin avionics are designed to meet a DO-160 specification which requires they withstand 32.2V for five minutes, as well as 60V for 100ms, then 40V for one full second, then nominal voltage, repeated 3 times at 10 second intervals. It's probably true of all the major manufacturers.

60V for 100ms tells us much about how much time is available to deal with OV.

Dave at B&C was kind enough to quantify the performance of an LR3D. The trip time is based on an RC circuit, which sets a time constant. The actual time delay is thus based on beginning voltage and the peak voltage seen on the bus during the OV event. For a nominal 14.4V bus subjected to 30V, the time to crowbar trip (a dump to ground) is 6ms. 32.2V (the 5 minutes DO-160 value) is 5 ms. A 60V spike makes it 2ms. The point here is the LR3D intervenes far, far faster than the DO-160 time limits.

In this application, the most sensitive component may be the EFI ECU. I'm told it has been tested at 30V for 15 seconds. It doesn't mean the ECU won't meet the DO-160 standards, but rather, only that it has not been tested or certified to the standard. I'm not aware of any burned ECUs from either of our vendors...and only the good Lord knows how many SDS units are in cars without OV protection.

The EarthX BMS interaction with the regulator also appears to be a non-issue. It allows a full two seconds before it removes itself from the system, eons of time even with a very marginal OV problem. At 17V, the LR3D crowbars the field in 55ms.

It can be argued a lot depends on the LR3D being reliable. True enough. In my opinion, it's as much a matter of installation quality as device quality, maybe more.

The other item was interconnection of two EarthX batteries when one of them was at a higher state of charge than the other. The website cautions about a 0.3V difference. EarthX engineering was kind enough to plot the actual expected amperage vs deltaV. Turns out the caution doesn't realistically apply for two batteries connected with #2 cables and master contactors. For a pair of ETX900's, a full volt difference results in 70 amps for about 10 seconds, falling to 10 amps or less. Please note the caution is entirely valid if the system offers a battery-to-battery current path via small wire, which may well become a fuse.

JLuckey · Mar 24, 2023

Dan or anyone else,

Do you have .dxf's of the relays/contactors you use in your drawings that you're willing to share?

DanH · Mar 24, 2023

JLuckey said:
Dan or anyone else,
Do you have .dxf's of the relays/contactors you use in your drawings that you're willing to share?

Try this...

http://www.aeroelectric.com/PPS/

http://www.aeroelectric.com/PPS/ACAD_Symbols_Library/

TASEsq · Apr 27, 2023

DanH said:
I'm going to post recent work, and risk being branded a heretic.

Heretic can also be pronounced as 'innovator'

DanH said:
On a bad day, both main contactors can be opened, which immediately dumps the large loads and brings the MZ-30L on line. Avionics load is about 10A, maybe 13 running autopilot servos, and can be less. The pilot has access to both batteries and the MZ-30L, in any combination.

On a really, really bad day, the MZ-30L carries the SDS, and the G5 is on internal battery.

Ok, i have read over all the posts in this thread, and all your explanations many times!

Is the purpose of the avionics switches for isolating the avionics in the event of a smoke in the cockpit event? i.e. you would close the battery masters (dumping the non-essential / high current devices). If the smoke persisted, you would close both avionics switches?

If so, then why not put the avionics loads on the switched side of the battery contactors and remove the avionics switches?

Why are there 2 switches? Given all the units are dual power fed (or the single power users are dual diode fed), turning off one of the 2 avionics switches would not remove power from the device? (you have to turn off both).

Also, i don't see a field switch for the main alternator? is this needed?

Your design has really gotten me thinking on how to implement a system for EFII/SDS in my -14 build at the moment. Thank you for posting!

DanH · Apr 27, 2023

TASEsq said:
Is the purpose of the avionics switches for isolating the avionics in the event of a smoke in the cockpit event? i.e. you would close the battery masters (dumping the non-essential / high current devices). If the smoke persisted, you would close both avionics switches?

They do bring that capability, as well as the possibility of re-engaging one side at a time to bring the big screens back on line. However, chances are none of that ever becomes necessary. A major design goal here is to require no pilot action. Note bus feed protection at the battery connection. Short one of the bus feeds and the fuse burns open. Very limited smoke, and the displays, fed by the other bus, don't even blink.

When the main contactors open, this is essentially a dual bus system. The dormant MZ30 automatically wakes to extend endurance on one bus.

If so, then why not put the avionics loads on the switched side of the battery contactors and remove the avionics switches?

Because they would have no access to the batteries or the MZ30.

Why are there 2 switches? Given all the units are dual power fed (or the single power users are dual diode fed), turning off one of the 2 avionics switches would not remove power from the device? (you have to turn off both).

We don't shut down a bus because of a device failure. An individual fuse removes power from the shorted device it feeds, leaving the rest of the bus intact.

On a daily basis, two avionics masters allow lighting up the entire G3X system prior to engine start on either battery, while reserving the other (clearance delivery or flight plan entry, for example). If preferred, they can both be off for engine cranking, jump cart hookup, etc. With two feeds, the "single point failure" argument against an avionics master is moot.

Also, i don't see a field switch for the main alternator? is this needed?

Nope.

Your design has really gotten me thinking on how to implement a system for EFII/SDS in my -14 build at the moment. Thank you for posting!

You're welcome.

TASEsq · Apr 27, 2023

DanH said:
Note bus feed protection at the battery connection. Short one of the bus feeds and the fuse burns open. Very limited smoke, and the displays, fed by the other bus, don't even blink.

We don't shut down a bus because of a device failure. An individual fuse removes power from the shorted device it feeds, leaving the rest of the bus intact.

Thanks for the explanation. It’s interesting because my mind was in a different place regarding failures. You’ve considered two scenariosI think:
1. shorted feed to the bus, or shorted bus itself - protected automatically by a fuseable link on the feed line.
2. Shorted device - protected by the CB or Fuse feeding the device.

It’s funny because I don’t really think a short is a major issue and very unlikely - you can protect wires and busses etc as well as you like. In my mind smoke in the cockpit would come from an internal failure inside one of the LRU’s. This may or may not trip the circuit protection to the device (As you mentioned these devices are designed to take 30 or 60v for x Ms etc). But I can imagine something going wrong inside a box and smoking is more likely than a bus or device power feed short?

In this case the only action is to turn off all power, pull all the cb’s, and put them back on one by one.

In your case, this would mean both masters off removing power from the white bus. (And I guess also shutting down the alternator field wire fed from that bus?). Both avionics off removing power from both avionics busses. If the busses were the other side of the contractors then masters off would kill everything at once.

On the other side of the coin the advantage is being able to turn on avionics from one battery and crank on the other battery? I think with dual feeds to devices and “keep alive” circuits from devices like the gad27 this might not be needed for all architectures.

Is this the only reason to have the avionics on the hot side of the battery?

It’s got me thinking on why we even need a battery contractor at all? Why not just have everything on the hot side and switched or relayed based on the load on the bus? Couldn’t you do away with the contractors and use a smaller lighter relay to turn off power to the white bus? Just use 15a switches for the avionics bus feeds as you have done? (No relay). And 15a engine bus switches for the engine busses?

DanH · Apr 27, 2023

TASEsq said:
It’s funny because I don’t really think a short is a major issue and very unlikely - you can protect wires and busses etc as well as you like. In my mind smoke in the cockpit would come from an internal failure inside one of the LRU’s. This may or may not trip the circuit protection to the device (As you mentioned these devices are designed to take 30 or 60v for x Ms etc). But I can imagine something going wrong inside a box and smoking is more likely than a bus or device power feed short?

Designed max voltage is not relevant to your concern. Our interest is amperage. Garmin is wonderfully specific about actual current draw and a recommended fuse rating. Nothing in those LRU's is gonna make much smoke before popping the recommended fuse.

It’s got me thinking....(snip)

Draw a diagram and start a thread.

tims88 · May 3, 2023

I have also read over this entire thread many times because I'm very interested in going with this architecture in my RV-10. There's a couple things that I've recently noticed that I wanted to bring up since I haven't seen them discussed yet:

1 - The Monkworkz B lead and the wire to the engine bus are both connected to the battery with a single fuse so a short on the B lead would blow the 40 amp circuit protection and remove both the generator and the battery from one side of the engine bus. There would still be the other battery and the primary alternator so this probably wouldn't be a major issue but it seems like two separate connections would be better so that a short on the B lead wouldn't remove both power sources from one side of the engine bus.

2 - The primary alternator and the master bus wire are both connected to the same side of the ANL so a short on either removes both of them from the electrical system. I know some load shedding might be necessary when the primary alternator is lost but it might also be nice to turn be able to use some of those items on that bus when they are needed.

DanH · May 3, 2023

tims88 said:
I have also read over this entire thread many times because I'm very interested in going with this architecture in my RV-10. There's a couple things that I've recently noticed that I wanted to bring up since I haven't seen them discussed yet:

1 - The Monkworkz B lead and the wire to the engine bus are both connected to the battery with a single fuse so a short on the B lead would blow the 40 amp circuit protection and remove both the generator and the battery from one side of the engine bus. There would still be the other battery and the primary alternator so this probably wouldn't be a major issue but it seems like two separate connections would be better so that a short on the B lead wouldn't remove both power sources from one side of the engine bus.

Good comments Tim.

Easy to do separate connections as above with no downside, so sure, go for it. I drew it in below.

Since the original post I've been kicking around a change to the SDS engine bus arrangement, using one of the dual diode blocks and a single bus. See below. Mostly I'm just easing toward "single bus, dual feed" vendor guidance for consistency, but FWIW, with this arrangement losing one feed would not take out one of the ECU's and its coil.

2 - The primary alternator and the master bus wire are both connected to the same side of the ANL so a short on either removes both of them from the electrical system. I know some load shedding might be necessary when the primary alternator is lost but it might also be nice to turn be able to use some of those items on that bus when they are needed.

There's nothing on the main bus I absolutely need. I have seen folks use two ANL's (one for the B lead and one for the main bus feed), and that's fine too. Third approach is no circuit protection on the main bus feed. Not my favorite, but as Mr. Nuckolls explains, shorting a main bus feed to structure tends to burn away the light gauge material and stop the short.

I'm gonna stick with the ANL's. As with the fusible link previously discussed, I prefer melting things be inside a containment

.

Walt · May 3, 2023

Dan's going to the dark side

You might look at the midi fuses, smaller than ANL, same basic function but smaller package and fuse blocks available with multi midi's inside.

DanH · May 3, 2023

Walt said:
Dan's going to the dark side

Aww heck Walt, I invented the dark side. I just let Darth use it.
.

TASEsq · May 3, 2023

DanH said:
Since the original post I've been kicking around a change to the SDS engine bus arrangement, using one of the dual diode blocks and a single bus. See below. Mostly I'm just easing toward "single bus, dual feed" vendor guidance for consistency, but FWIW, with this arrangement losing one feed would not take out one of the ECU's and its
.

Is there any concern about a bus short or fault in this arrangement? (A bunch of water lands on the fuse block etc)

Simplified Approach

Well Known Member

Well Known Member

Well Known Member

Well Known Member

Attachments

Legacy Member

Well Known Member

Well Known Member

Well Known Member

Well Known Member

Attachments

Well Known Member

Well Known Member

Well Known Member

Well Known Member

Legacy Member

Legacy Member

Well Known Member

Legacy Member

Attachments

Well Known Member

Well Known Member

Well Known Member

Well Known Member

Well Known Member

Attachments

Well Known Member

Attachments

Well Known Member

Well Known Member

Well Known Member

Well Known Member

Well Known Member

Legacy Member

Well Known Member

Well Known Member

Well Known Member

Well Known Member

Well Known Member

Legacy Member

Well Known Member

Well Known Member

Well Known Member

Legacy Member

Member

Legacy Member

Well Known Member

Legacy Member

Well Known Member

Legacy Member

Well Known Member

Legacy Member

Attachments

Well Known Member

Legacy Member

Attachments

Well Known Member