What's new
Van's Air Force

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

Couple SDS CPI2 install questions

N546RV

Well Known Member
Planning, planning, planning...such is my life these days. This must be that last 90% everyone talks about. This has surfaced a couple questions about installing/packaging the CPI2 components.

#1: How accessible should the ECU and backup battery units be? Both these units have fuses that might require service at some point. I have no expectation of making them accessible in-flight, but I'm also not sure that burying them behind the panel is smart either.

My first inclination was to package the ECU alongside my fuse blocks, which I intend to place on a hinged mount behind the panel for on-ground accessibility, but getting the ECU on there alongside the fuse blocks and associated stuff is pretty tight. Moving the ECU to my upper avionics shelf - which will only be accessible by removing the center of the panel - would make things a good bit roomier.

I figure in a general sense if these fuses are blowing on any regular basis, I've got obvious issues, so I suppose this mostly comes down to possibly dealing with teething problems early on. I still feel like overall, having these units somewhat buried wouldn't be a bad thing.

#2: I'm interested in the "switchology" for the CPI2. I'll be full electronic, and thus running a dual ECU unit. I intend to use toggle switches, not a key switch, for ignition control. My current schematic iteration has two DPST switches, one for each ECU. Each switch controls power to the ECU and coil (purple and red wires) for that ECU.

What I'm wondering is whether I should also provide a way to kill the coils with a dedicated switch or switches, either for the purpose of killing the engine quickly, or for killing the coil packs individual (as with a traditional mag check). I know all the preflight stuff can be done from the controller - is there any benefit to using the controller in lieu of external switches for the runup procedure?

One thing I thought about was to replace those DPST switches with ON-OFF-(ON) DPDT switches for this purpose. The center position would of course have everything powered off, up would be the normal "on" position in-flight, and down would be the momentary position, and could activate the external kill function for the connected ECU. This would enable both traditional "mag checks" as well as the ability for a quick engine kill.

The other alternative would be to simply have a third switch that killed both ECUs, and was used only for an emergency shutdown, and all runup operations could just run through the controller.

This is definitely the more fun question, and I'm very interested in people's thoughts here. Thanks!
 
My Installation

I have been running the dual CPI-2 for almost 100 hours. I have had to access the main control box exactly one time. That was to add an SDI recommended capacitor to quiet false low current indications in my particular system. No fuses have ever blown. I have power for one side fed through my VPX-pro system and one side fed through a direct battery buss (bypassing the master). I would suggest that having access for occasional maintenance is critical, but you will not need regular access to the fuses.

My battery is low on the forward firewall. I would have put it inside the cockpit if I could have made a spot that was easy to service, but completely out of the way. I mounted it low and outboard to keep it out of the high heat after shutdown, but high enough to be away from the exhaust. The back of the engine is a pretty busy environment with access needed for regular oil filter access. I am sure in the cabin would be a kinder environment for the battery life, but nowhere to put it.

Switchology is another matter. You will need two on off switches to power the system up. You can do the run-up mag check on the head, but it is not very straightforward with a lot of button pushing. Having dedicated switches to tell the system to suspend ignition 1 and 2 for an ignition check is faster and more straight forward in my opinion. With its own internal back up power you will also need a switch to tell the system to shut down. Simply turning off the power will only switch it to back up power.

I found it simpler to use a traditional key switch to accomplish all of those tasks. Starting the engine is simply to turn on the two CPI systems power and key the starter. Run up ignition check is key position 1 and 2 ‘mag’ check. Shutting down is mixture, system power switches off, key off. All accomplished with two power switches and 1 key switch.

Once the false ‘low volts’ issue was solved with the addition of the capacitor, the system has been flawless. I have an IO540 thunderbolt. I just flew a 1310 nm trip from SLC to PTK at 11,500 alt, WOT fuel flow of 10.5 to 11 gph (lean of peak), 2450 rpm with an average of 160-165 kts true airspeed. I was in loose formation with another RV10 (1 mile lateral separation). When I switched on the LOP advance I picked up 2-3 kts and slowly pulled away from the other 10. I am a happy customer! :D
 
Last edited:
#1: How accessible should the ECU and backup battery units be? Both these units have fuses that might require service at some point.

I'm all for service access, but anything which can be reached in ten minutes with simple tools is plenty good here.

I'm interested in the "switchology" for the CPI2.

It's in the manual.

For each ECU and coil, one DPST for the purple and red power supply wires, and one SPST switch for runup check, gray wire for left IGN, yellow for right IGN. When closed, the SPST switch grounds the circuit and shuts down the respective ignition.

The runup switch can be a toggle wired to be normally open, or a normally open push button, or a rotary key switch. I'd pick the toggle, mounted upside down so up is open. I'd rank the rotary switch last for reliability, but operation is innate for a pilot transitioning from most traditional GA birds. That is a big plus. Always good if your switchology conforms with convention.

One thing I thought about was to replace those DPST switches with ON-OFF-(ON) DPDT switches for this purpose. The center position would of course have everything powered off, up would be the normal "on" position in-flight, and down would be the momentary position, and could activate the external kill function for the connected ECU. This would enable both traditional "mag checks" as well as the ability for a quick engine kill.

If the engine was running, the center OFF would make the system switch to the backup battery, not quick kill.

That's not necessarily a bad thing. Up would be normal run, middle would check operation of the backup system, and the spring-loaded down would kill ignition for runup checks. It would be a one switch runup pad function check for each ignition.

Note that to shut down the engine using the ignition switch rather than the mixture knob, you would need to hold both switches in the down position at the same time until engine rotation ceased. Only then would releasing the switches put the systems in a fully OFF mode.

The other alternative would be to simply have a third switch that killed both ECUs, and was used only for an emergency shutdown, and all runup operations could just run through the controller.

Not just no. Four letter no ;)
 
Last edited:
I?m in the process of gathering up all the components to convert my Rocket to a dual CPI-2 set up. I have two double pole switches that will control CPU/coil power to ignition A and B. I also have two momentary push buttons that will be wired to shut down either ignition. I thought I had everything handled with the this setup but after reading the above posts I?m asking myself: ?What prevents the engine from running off the backup battery once the two main power switches are turn off and the engine is shut down?? In other words, will the back up battery fire the ignition if the prop is turned on the ground? Do I need a separate switch to shut off the back up power?
 
Ok...RTFM:
? Power down of the CPi2.
Due to the ecu having its own backup battery, power down is unique on the CPi2. When you turn off Mainbus 12 volt
power to the CPi-2 it automatically switches over to its backup battery. If there is no engine rpms, and no keypad keys are pressed, it will begin a power off countdown within 5 seconds, and after counting down from 10 to 0 it will shut off unless Mainbus power is switched back on. So total power off time is approximately 15 seconds.?

So stay clear of the prop within 15 sec of shut down.
 
CPI-2

I do my run up check using the controller. It’s very easy and efficient. The system from the control head, momentarily shuts off a coil for a few seconds and then it fires back up automatically. It allows you to check both coils independently.

I have the main system wired to my master buss. When the master is on it powers up an ignition bus ( made by Blue Seas) with 4 circuits ( 2- 3amp fuses, 2 -10 amp fuses).
I also have two dipole switches which can cut off the power individually to each coil. These are switched after the ignition bus.

I use a key start, the two dipole switches, located just above my ignition switch( up for on), obviously have to be turned on to start the engine. When I shut down, I pull the mix, shut off both switches, turn off the key, turn off the master. The dipole switches are needed. I first had the system wired without them just using the master. The problem was there was some apparent left over voltage in my panel wiring. (back up batteries) that kept the control head from completely shutting off. The system works well.
 
Last edited:
If the engine was running, the center OFF would make the system switch to the backup battery, not quick kill.

That's not necessarily a bad thing. Up would be normal run, middle would check operation of the backup system, and the spring-loaded down would kill ignition for runup checks. It would be a one switch runup pad function check for each ignition.

Note that to shut down the engine using the ignition switch rather than the mixture knob, you would need to hold both switches in the down position at the same time until engine rotation ceased. Only then would releasing the switches put the systems in a fully OFF mode.

Yup, this is pretty much exactly the procedure I had in mind; I kind of feel like the switch setup has a certain elegance to it.
 
I?m in the process of gathering up all the components to convert my Rocket to a dual CPI-2 set up. I have two double pole switches that will control CPU/coil power to ignition A and B. I also have two momentary push buttons that will be wired to shut down either ignition. I thought I had everything handled with the this setup but after reading the above posts I?m asking myself: ?What prevents the engine from running off the backup battery once the two main power switches are turn off and the engine is shut down?? In other words, will the back up battery fire the ignition if the prop is turned on the ground? Do I need a separate switch to shut off the back up power?

From the installaton manual:

If you don?t have a typical aircraft key/mag type switch, you can wire up yellow and gray to one or two normally open momentary push switches or toggle switches to kill ignition which will then shut down the engine. Remember turning off Mainbus power to the CPi2 just forces it over to backup battery power so your Mainbus or ecu power switch will not shut off the engine.

So, use your key switch to turn the ignition off. That grounds the ignition just like it would if you had traditional mags. Otherwise the engine is going to keep happily running until the CPi-2 backup battery runs out of power.

SDS also recommends having a separate switch from the master switch to prevent back feed as was mentioned previously, and separate breakers for the coilpack and cpu for each of the CPi-2 units. So, if you have a dual unit you'll need a total of 4 breakers and two double pole switches. Each switch controls power to one cpu and one coilpack via the main CPi-2 control unit.
 
You can do this with only two switches

I've been running a dual CPI-2 setup for just over a year. Great unit. Like an earlier poster, I've only needed access once to install a capacitor into each of two wiring connectors.

I'm finalizing a new panel design with SteinAir, and based on feedback from SDS, I will be using a pair of locking toggle switches to provide "off", "run", and "mag check" positions for each ECU.

That portion of the panel design looks like this:

CPI2+Switchology.jpg


This is a wiring diagram that I got from SDS. I understand that this was created by another SDS customer, and reviewed by SDS.

CPI2+Switch+Wiring.jpg


I'm very tight for panel space, so this makes the most of a limited amount of real estate.
 
I've been running a dual CPI-2 setup for just over a year. Great unit. Like an earlier poster, I've only needed access once to install a capacitor into each of two wiring connectors.

I'm finalizing a new panel design with SteinAir, and based on feedback from SDS, I will be using a pair of locking toggle switches to provide "off", "run", and "mag check" positions for each ECU.

That portion of the panel design looks like this:

CPI2+Switchology.jpg


This is a wiring diagram that I got from SDS. I understand that this was created by another SDS customer, and reviewed by SDS.

CPI2+Switch+Wiring.jpg


I'm very tight for panel space, so this makes the most of a limited amount of real estate.

It appears you have both CPUs and Coilpacks wired directly to the main bus through the switch. SDS recommends individual breakers or fuses for each CPU and each Coilpack.

You may want to double check this with Ross or Barry.

From the installation manual:

"The Purple 20 ga on the 14pin main harness plug, needs a 2 to 5 amp fuse or breaker."

"The Red 18 ga wire on the 12pin coil harness plug. Needs a 10 to 15 amp fuse or breaker."
 
It appears you have both CPUs and Coilpacks wired directly to the main bus through the switch. SDS recommends individual breakers or fuses for each CPU and each Coilpack.

You may want to double check this with Ross or Barry.

From the installation manual:

"The Purple 20 ga on the 14pin main harness plug, needs a 2 to 5 amp fuse or breaker."

"The Red 18 ga wire on the 12pin coil harness plug. Needs a 10 to 15 amp fuse or breaker."

Hmm...I didn't think of it earlier, but I'd have this same problem with my proposed switch setup. There's no way, in one switch, to provide two separate switched current paths and a switched ground path for the kill function.
 
Re-read my post. I got this diagram from Barry. He has already vetted it. It originally came from another owner, but Barry sent it to me.
 
Re-read my post. I got this diagram from Barry. He has already vetted it. It originally came from another owner, but Barry sent it to me.

Interesting. I take it this setup only has two bus feeds (one for each switch) rather than the four described in the manual? What are your fuse/breaker values? Naively one might go with 20A by summing the recommendations from the manual, though it also seems like 15A might be fine.

It surprises me a bit to see this approved by Barry, because the manual is quite explicit about providing separate bus feeds for the ECO and coils, and using a DPST switch for them. The way it's written implies to me that it's important that those feeds be separate, but it seems like that's not the case - perhaps I'm reading too much into the documentation.

I wouldn't mind getting back two slots in my planned fuse blocks. :)
 
Yup, this (three position switch)is pretty much exactly the procedure I had in mind; I kind of feel like the switch setup has a certain elegance to it.

It would need to be a three position, three pole, or if unavailable in your preferred brand, three position, four pole with one unused.

In terms of serviceability, using an ON-OFF double pole for the red and purple, and a separate ON-OFF single pole for the kill means you could get field replacements almost anywhere, unlike the fancy switch.

It surprises me a bit to see this approved by Barry, because the manual is quite explicit about providing separate bus feeds for the ECO and coils, and using a DPST switch for them. The way it's written implies to me that it's important that those feeds be separate, but it seems like that's not the case - perhaps I'm reading too much into the documentation.

I too am surprised. Private guidance from SDS insisted on separate red and purple power supply all the way from the source bus or battery.
 
I have the SDS dual CPI2 in my RV-8. I installed the provided filter supplied by SDS. My ECU?s are mounted to the bottom of the cover that is next to the right gear tower. My backup battery is also mounted there. To gain access, remove the screws and pull the cover. There is a sufficient service loop in the wiring harness to sit in the seat and rest the ECU and cover it?s mounted to on my right knee. Satisfied with this location so far.

I wired the ECU in accordance to the SDS diagram. The two hot leads go directly to my primary battery.

My display/controller is in my panel. I like it there, easy to monitor the ignition and check battery health.

Really easy to operate. The SDS support has been fantastic. Glad I waited for CPI2. Works as advertised.

Run ups and mag/system checks are easy.
 
SDS CPi-2 Wiring Diagram - VPX, Garmin G3X

...I too am surprised. Private guidance from SDS insisted on separate red and purple power supply all the way from the source bus or battery.
Same here. The documentation is very clear.... power source -> breaker -> switch -> cpu feed AND source -> breaker -> switch -> Coilpack feed. (x2 if you have a dual unit). I also verified the need for the separate feed and switch for each with Ross via phone as it didn't make sense to me why you'd have to have redundant switching (master switch + individual switch).

Also SDS highly recommends not using the VPX breakers as your source of power for the Coilpack.

My system consists of:

  • VPX and G3X (panel and all avionics by SteinAir)
  • CPi-2 Dual System
  • Backup alternator and backup SDS CPi-2 battery
  • Traditional On-Off-L-R switch for my ignition. Two ground wires go from the ignition switch to the CPi-2 units that ground them out and allow testing and engine shutdown just like traditional mags would work.
  • I ran a separate feed to a separate small bus that has 4 breakers - 2-10a (Coilpack feeds) and 2-5a (cpu feeds) as SDS highly recommends NOT using the VPX breakers to feed the Coilpacks.
  • 2 DPST switches (a Coilpack and a cpu feed to each) between the breakers and the CPi-2 units.

Since my CPi-2 panel controller is on the other side of the cockpit I also have a third "LOP MODE" switch that turns the LOP mode on and off remotely and also causes a LOP Mode On annunciation on the G3X PFD.

My wiring diagram is in PDF format and is too large to load here as an image file, but if you'd like a copy you can download it here from Dropbox. Note that there are multiple pages, and the third page is the pinout drawing that's key to understanding the wiring diagram on the first page.

Here is the notice that I have in my wiring diagram that will go in my POH regarding the SDS switches in the panel:

Shutdown-XL.jpg


Here are the pin numbers in the connectors. I gave the connectors my own "C" numbers to make them easier to identify for me:

Pin%20Numbering-XL.jpg


Here are my switches on the left side of the panel. We decided late in the build to add the LOP MODE switch so I used the Spare switch shown here and just ordered a new switch cover. Thank goodness for spare switches!

Switches-XL.jpg


And finally here's the panel controller and the breakers on the co-pilot's side of the panel:

CPi%20Panel%20Controller-XL.jpg
 
Last edited:
Barry had a look at the posts here and made these comments:

Nobody has reported a fuse blown yet. Not critical to mount with fuses accessible, but I would do it if it's possible. Might be good to mount the gold ecu such that you can get it out easily, since it's likely there could be a software upgrade in the future and the gold ecu needs to be removed for this.

"Engine shutdown via keypad is really only meant for installs where there is no more room for kill switches. Likely nobody would use this feature, but it was easy to add in software.

Best use the diagram posted using the honeywell switches or install dedicated kill switches. If you use pushbutton kill switches you should mount them close together so you can put your thumb across both for shutdown.

Also for proper shutdown its best to shut off the CPi2 power switches first, then press or flip kill switches, then shut off your main bus power last. This insures the CPi2 will do it's automatic power-off routine properly. Cpi2 may have trouble turning off if it's still connected to the mainbus, due to capacitors and or residual voltage from other devices in the airplane."
 
4 circuit bus

I posted earlier. Here is the four circuit bus I used. Made by Blue Seas.
 

Attachments

  • 9748F8EC-6A59-4C36-AEE1-4B22831406F6.jpeg
    9748F8EC-6A59-4C36-AEE1-4B22831406F6.jpeg
    129.3 KB · Views: 249
I've been running a dual CPI-2 setup for just over a year. Great unit. Like an earlier poster, I've only needed access once to install a capacitor into each of two wiring connectors.

I'm finalizing a new panel design with SteinAir, and based on feedback from SDS, I will be using a pair of locking toggle switches to provide "off", "run", and "mag check" positions for each ECU.

That portion of the panel design looks like this:

CPI2+Switchology.jpg


This is a wiring diagram that I got from SDS. I understand that this was created by another SDS customer, and reviewed by SDS.

CPI2+Switch+Wiring.jpg


I'm very tight for panel space, so this makes the most of a limited amount of real estate.

Breakers are to protect wires, not devices. If you power both the ECU and the coilpack from one breaker, all the wires to them need to be large enough that the breaker pops before any of the wires burn. So, if it's a 15A breaker, all the wires need to be able to carry 15A. That purple wire probably needs to be upsized. Other than that, I don't know why this schematic wouldn't work.

Ed Holyoke
 
Breakers are to protect wires, not devices. If you power both the ECU and the coilpack from one breaker, all the wires to them need to be large enough that the breaker pops before any of the wires burn. So, if it's a 15A breaker, all the wires need to be able to carry 15A. That purple wire probably needs to be upsized. Other than that, I don't know why this schematic wouldn't work.

Ed Holyoke

That switch is listed in the diagram as DPDT, but it is really a DP3T for those looking to use something like that. However, it only acts at a 3T when the jumpers are installed. It is needed in this case, as it continues to "make" the exsitiing connection (via a wiper) on one pole as the other pole "makes" a new connection in the top or bottom position.
 
Last edited:
Breakers are to protect wires, not devices. If you power both the ECU and the coilpack from one breaker, all the wires to them need to be large enough that the breaker pops before any of the wires burn. So, if it's a 15A breaker, all the wires need to be able to carry 15A.

Ed Holyoke

Good point, Ed. I had not caught that. The purple is 20 gauge & the CPI doc calls for a 2 to 5A breaker or fuse. The red is 18 gauge & calls for a 10A breaker or fuse. Looks like you would need to upgrade both wires to 14 gauge with a 15A breaker.
 
The purple wire only powers the CPU section. 2-5amp breaker, 20 ga wire which is already part of the harness.
 
The purple wire only powers the CPU section. 2-5amp breaker, 20 ga wire which is already part of the harness.

Right, but if the purple and red wires are tied together on the same circuit - as depicted in the aforementioned diagrams that demonstrate using a single switch to both remove power and/or activate the killswitch function - then that purple wire will be protected by a 15A breaker, rather than 2-5A.

This seems to be the concern being expressed here - it may be OK operationally to combine ECU and coil power feeds into a single circuit, but it raises issues with proper protection of the purple wire in the harness.

Of course, none of this applies if the system is set up as depicted in the installation manual, with two separate breakers for each ECU (one for the purple wire and one for red).

Put another way, David isn't calling out a deficiency with the harness as constructed, but a deficiency with the proposed single-switch setup.
 
Layout

Here is how I wired my system. Click on the image
 

Attachments

  • EFC20B0D-6A8B-4089-914A-0D5B72D7CC02.jpg
    EFC20B0D-6A8B-4089-914A-0D5B72D7CC02.jpg
    152.3 KB · Views: 274
Last edited:
Right, but if the purple and red wires are tied together on the same circuit - as depicted in the aforementioned diagrams that demonstrate using a single switch to both remove power and/or activate the killswitch function - then that purple wire will be protected by a 15A breaker, rather than 2-5A.

This seems to be the concern being expressed here - it may be OK operationally to combine ECU and coil power feeds into a single circuit, but it raises issues with proper protection of the purple wire in the harness.

Of course, none of this applies if the system is set up as depicted in the installation manual, with two separate breakers for each ECU (one for the purple wire and one for red).

Put another way, David isn't calling out a deficiency with the harness as constructed, but a deficiency with the proposed single-switch setup.

From Barry:

"Best to keep purple and red on separate circuits & breakers and use a double pole switch. The CPi2 should work if purple and red are tied together on one circuit but if the breaker or switch was not in good condition this could introduce some voltage fluctuation into the purple wire input and since this powers the cpu circuits, extra noise here is not a good thing to have. This is why separate breakers and a double pole switch are best."

Since coil power goes through the CPI2 red wire and there are large current surges as the coil charges, the decision was made in the design stage to separate CPU and coil power to reduce the likelihood of noise. Coils can be noisy devices. By combining these wires, you negate the design intention here. Might work ok, might not.
 
From Barry:

"Best to keep purple and red on separate circuits & breakers and use a double pole switch. The CPi2 should work if purple and red are tied together on one circuit but if the breaker or switch was not in good condition this could introduce some voltage fluctuation into the purple wire input and since this powers the cpu circuits, extra noise here is not a good thing to have. This is why separate breakers and a double pole switch are best."

Since coil power goes through the CPI2 red wire and there are large current surges as the coil charges, the decision was made in the design stage to separate CPU and coil power to reduce the likelihood of noise. Coils can be noisy devices. By combining these wires, you negate the design intention here. Might work ok, might not.

And now we've kind of come full circle. Barry's words here reflect what I get out of the installation manual, that separating the power leads is best practice. This, however, seems to contradict this proposed wiring diagram which is said to have been vetted by Barry.

To be clear, I'm not trying to belabor a point here, just wanting to clarify what looks to me (and apparently, to a few others) as mixed messages. I'm not sure where the disconnect might be happening here - maybe the "vetting" was not the "two thumbs up" endorsement I imagine, but something more akin to "well the plane won't explode if you do it that way..."

So I'm really just seeking to resolve this disconnect - and not even really for my own sake. For me, best-practice and separating the feeds seems like the right decision. But I'd also like to be sure that anyone stumbling across this discussion down the road comes away with the right idea.

I appreciate your responses. I'm pretty sure I've said it before, but I'll say it again: your involvement in the community here is a huge factor in my decision to go with the CPI2 for my build.
 
I think you make a good point here. Most of our recommendations for systems and wiring lean towards best practice. Often other ways of doing things also work fine but sometimes they don't and we end up on the phone trying to sort out how to fix issues that crop up.

We've seen some vexing problems over the years, many caused by worst practice on installations that were only revealed when we reviewed photos or video of the installations. Tie wrapping sensor cables to spark plug wires was one (done by an A&P and avionics tech!). Injectors in backwards was another memorable one.

Best practice gives you the best chance of avoiding any weird issues. :)
 
As a CPI2 owner/operator I’ve been following this thread with interest. As said previously I wired my dual channel system per plans. I do however appreciate the points made about having the purple and red wires tied together.

I currently have two 16 gauge red wires coming off the + side of my primary battery. One wire for each channel. The wires are Each individually protected immediately after the connection to the battery with Fuses. Then it’s wired per SDS plans.

If I’m getting this right, the suggestion would be to run two additional wires from the power source, in my case the positive battery terminal, to supply power for the purple wires?

I’ve had no issues with current scenario but prefer to use the “best practice” approach if I can discern what that actually is.

A revised wiring diagram might help.
 
I'd say if your installation is working well, no need to change. The proof is in the pudding as they say.

We've seen a large variation in electrical noise between airplanes and installations over the years. Where one is wired not with best practices, we may see no issues while another wired with best practices may have multiple noise issues. There are no guarantees...
 
I'd say if your installation is working well, no need to change. The proof is in the pudding as they say.

We've seen a large variation in electrical noise between airplanes and installations over the years. Where one is wired not with best practices, we may see no issues while another wired with best practices may have multiple noise issues. There are no guarantees...

Thanks for the quick response Ross.

My CPI2 works great as wired per diagram.

On another note, never had a fuse blow on CPI2.
 
I now have 80 hours on my no issues CPI2 and love it.
Power is simply wired through main switch then to the controller then inline 2 10amp and 2 5amp fuses mounted on a swingdown easily accessible panel, all functions work easily/correctly through the controller with no issues.
Fully tested (ground&inflight) by turning master power off, I get 2 red lights for power loss and backup battery active and the engine keeps purring without a hiccup on the backup battery.
For normal shutdown the procedure is: Fuel lean/cutoff then master off (takes 15 seconds)...if immediate shutdown is needed there are 2 momentary pushbuttons on the panel that instantly kill both coil packs. (also tested many times)
The result is a simple compact but powerful ignition system that works well in a small aircraft.
Thanks go to Ross for this great ignition. (and for patience with me as well)
Mike Kellems
RV3B N931M
 
Last edited:
Best Practice

There has been a lot of back and forth about how some people have wired their SDS CPi-2 systems and what works vs. what are the actual requirements / recommendations / best practices.

I agree that the documentation could (and should) be clear, definitive guidelines should be given for the recommended setup. Then, if you decide to deviate from that you at least know that you're deviating! ;)

To me the only confusing part was regarding the switching. The power requirements / recommendations are very clearly spelled out - a separate 2-5A breaker or fuse is required for each CPU (purple wire) and a separate 10-15A breaker or fuse is required for each coilpack. See the excerpt from the manual below.

What is confusing is the switching recommendations. For example, for the purple 20 gauge cpu conductor the manual says very clearly "you must have a dedicated switch for this wire" however two paragraphs later it says

"Four +12V sources to the CPi2 ecu need to be switched. If you have an existing Mainbus switch you could wire all 4 of these to the switch but you will need to have 4 fuses or breakers between the Mainbus switch and each wire.

Dedicated switches are not necessary but some installers may desire this anyway. If installing dedicated switches then you want your breaker or fuse before your switch."

First off, on a dual CPi-2 system there aren't 4 - 12V sources to the CPi ecu, there are two to the cpus, and two to the coilpacks. Second, you can't say "you must have a dedicated switch for this wire" then go on to say "dedicated switches are not necessary".

Then it goes on to say

"Switching could be done using two DPST switches. One DPST switch for both Purple wires and a 2nd switch for Coil power sources +12V. Four breakers/fuses are needed here.

You could also do two SPST switches for each coil power source. Switch rating of 8 amps minimum.

Note, the red 18ga Coil power wire mentioned above needs a switch since there may be a small amount of current flow if it is not switched off, likely 3 to 10mA only but enough to run down your main battery after a month or two."

I realize that SDS is trying to give us a lot of options, but seriously, it would be nice to just have a clear "this is our recommended setup" diagram" and perhaps notes regarding why that setup is recommended.

Here's what I did that meets the strictest requirements / recommendations per the manual:

Unit 1
Power bus > 5a breaker > DPST Switch > Purple CPU-1 wire
Power bus > 15a breaker > DPST Switch > Red Coilpack-1 wire

Unit 2
Power bus > 5a breaker > DPST Switch > Purple CPU-2 wire
Power bus > 15a breaker > DPST Switch > Red Coilpack-2 wire

SDS backup battery, traditional key switch L-R-Off-On.

SDS-XL.jpg
 
Unfortunately many people want to wire stuff their own way on the EFI as well as the ignition systems so we get many email and phone calls- I want to do it this way, will it work? Can I use this type of switch? etc. With people building their own planes, you have many different preferences. No two airplanes will be built or wired the same way.

We publish simple schematics for our recommendations on wiring the EM-5 but almost nobody follows that advice and we see similar things on the CPI-2.

The CPI-2 is more complicated in some ways because it has a backup power source in most cases and voltage sensing and switching circuits. It can be hard to shut down because it will just revert over to the backup battery and stay running. Any tiny amount of voltage leaking from other devices onto the power bus will keep the ECU active or partly so (often unpredictably) and we've seen a number of these issues from backup glass power sources.

There has been customer feedback and resultant changes in wiring the switches over the last year or so. Be sure to read the latest versions of the manuals. April 21/20 was the latest revision.

Here is what we recommend and we think that it's pretty straightforward (but maybe it doesn't suit everyone):



Are there other ways to do it? Certainly and some work just fine as we've heard on this thread.
 
Thanks Ross. I understand what you're saying, however I still contend the documentation does not convey that clearly at all. I recommend that you re-read this section and at a minimum make the "must haves" (i.e. required) and the recommendations not conflict with each other.

After reading, re-reading, and re-re-reading this area many times I feel that it could be highly condensed and simplified to read something like this:

"Each ECU requires two power conductors, one for the the cpu (purple) and one for the coilpack (red). We highly recommend a separate fuse or circuit breaker for each (3-5a for the cpu and 10-15a for the coilpack), and that they be able to be isolated from the main bus by a switch. The reasons for these recommendations are shown below, as is a recommended wiring diagram.

1. We recommend that a separate fuse or breaker be used because....

2. We recommend that a switch be installed to isolate cpu and coilpack from the main bus even though there is a master switch for the following reasons:
a) The cpu should be isolated because....
b) The coilpack should be isolated because...

3. We do NOT recommend using VPX breakers for the coilpacks because....

4. We also HIGHLY recommend that you install the backup battery because....

While there are many different ways to wire this system that very well may work, the above is our highly recommended configuration based on our experience and the design of the system."

These simple, non-conflicting instructions and explanations, combined with the wiring diagram that you posted in the link above would be FAR less confusing and would convey what you're trying to convey much clearer IMO.
 
I now have 80 hours on my no issues CPI2 and love it.
Power is simply wired through main switch then to the controller then inline 2 10amp and 2 5amp fuses mounted on a swingdown easily accessible panel, all functions work easily/correctly through the controller with no issues.
Fully tested (ground&inflight) by turning master power off, I get 2 red lights for power loss and backup battery active and the engine keeps purring without a hiccup on the backup battery.
For normal shutdown the procedure is: Fuel lean/cutoff then master off (takes 15 seconds)...if immediate shutdown is needed there are 2 momentary pushbuttons on the panel that instantly kill both coil packs. (also tested many times)
The result is a simple compact but powerful ignition system that works well in a small aircraft.
Thanks go to Ross for this great ignition. (and for patience with me as well)
Mike Kellems
RV3B N931M

So, you're running power unfused through a "main" switch and then putting a fuse after the controller? This confuses me. I would hope you're not running all your power through a switch. If you are running power from a "main" buss, isolated by a battery contactor, one would expect the fuses to be at that buss and not after a device or switch. If you have a separate dropdown panel for an "ignition buss", one would wonder how the wire to that bus is protected and why power is running through a switch and a device without protection.

I don't dispute that this will all work normally. My concern is what happens if any of the unprotected wires short. Depending on the location of the short, it could take down both ignitions as well as smoking and possibly starting a fire in the cabin. I guess the battery backup would probably keep the ignitions alive, but the smoke and fire are still a problem.

Ed Holyoke
 
Thanks Ross. I understand what you're saying, however I still contend the documentation does not convey that clearly at all. I recommend that you re-read this section and at a minimum make the "must haves" (i.e. required) and the recommendations not conflict with each other.

After reading, re-reading, and re-re-reading this area many times I feel that it could be highly condensed and simplified to read something like this:

"Each ECU requires two power conductors, one for the the cpu (purple) and one for the coilpack (red). We highly recommend a separate fuse or circuit breaker for each (3-5a for the cpu and 10-15a for the coilpack), and that they be able to be isolated from the main bus by a switch. The reasons for these recommendations are shown below, as is a recommended wiring diagram.

1. We recommend that a separate fuse or breaker be used because....

2. We recommend that a switch be installed to isolate cpu and coilpack from the main bus even though there is a master switch for the following reasons:
a) The cpu should be isolated because....
b) The coilpack should be isolated because...

3. We do NOT recommend using VPX breakers for the coilpacks because....

4. We also HIGHLY recommend that you install the backup battery because....

While there are many different ways to wire this system that very well may work, the above is our highly recommended configuration based on our experience and the design of the system."

These simple, non-conflicting instructions and explanations, combined with the wiring diagram that you posted in the link above would be FAR less confusing and would convey what you're trying to convey much clearer IMO.

I'll have Barry read this over again when he has some time and look at your recommendations. Perhaps it can be clarified and simplified. Does anyone find the schematic confusing?

People can cut power to the red and purple wires any way they wish. In the end, that's what's important. In our view, the easiest way is depicted in the published schematic.

Regarding the VPX, our caution is to set the breaker current for the coil high if you decide to run the red wire through it (VPX). As I've discussed many times on VAF, coil current is far from constant and a breaker trip here will be bad news. The VPX doesn't trip the same way a conventional breaker or fuse does- rather via peak current over a very short time period.

Thanks for the feedback in this thread guys. It helps us improve the devices and documentation.
 
Those Red and Purple Wires

I converted to a CPI-2 ignition on my RV-9a and I now have over 50 trouble free hours. I'm very satisfied with this new ignition and described this conversion in a previous post. Last year I had a few phone conversations with Barry and Ross before I started the wiring. My questions were mainly focused on the supply current to the red and purple wires. I am reading this thread and couldn't help but mention my thoughts and experiences relating to those red and purple wires.

Maybe I missed something or I'm just a knucklehead or both, but I could not see the reason why two circuit breakers and one switch were required for each module when only 1 circuit breaker would do the job. I used a single 15A breaker for each module and connected it directly to the red and purple wires. My thinking is that the red and purple wires are both supplied by 14v from the main buss and it was not clear to me why it was required, what function do they serve, or if it would make any difference for current to pass thru those extra components. I understand it's Best Practice but I am a firm believer in the KISS principle. So any time I can eliminate two breakers and two switches from a wiring diagram, those are four components that won't fail in the future and will not require valuable real estate on my panel now.

Both ignition breakers are on the Master buss, and the backup battery has one switch and two fuses as outlined in the Installation Manual. The backup battery is isolated from the main buss thru a large schottke diode and the Avionics buss is isolated from the Main buss thru two large schottke diodes. So when I turn off the Master and Backup battery switches, the ignition modules are completely isolated from any other backup batteries or capacitors. If I need to turn off one module I can pull one of the breakers. So, until I have a good reason to rewire the modules and add four more components to the panel, I will leave the wiring as it is. Please let me know if I am making a big mistake or missed something important.
 
Last edited:
I converted to a CPI-2 ignition to my RV-9a and I now have over 50 trouble free hours on this system. I'm very satisfied with this new ignition and described this conversion in a previous post. Last year I had a few phone conversations with Barry and Ross before I started the wiring mainly focusing on the supply current to the red and purple wires. I am reading this thread and couldn't help but mention my thoughts and experiences relating to those red and purple wires.

Maybe I missed something or I'm just a knucklehead or both, but I could not see the reason why two circuit breakers and one switch were required for each module when only 1 circuit breaker would do the job. I used a single 15A breaker for each module and connected it directly to the red and purple wires. My thinking is that the red and purple wires are both supplied by 14v from the main buss and it was not clear to me why it was required, what function do they serve, or if it would make any difference for current to pass thru those extra components. I understand it's Best Practice but I am a firm believer in the KISS principle. So any time I can eliminate two breakers and two switches from a wiring diagram, those are four components that won't fail in the future and will not require valuable real estate on my panel.

Both ignition breakers are on the Master buss, and the backup battery has one switch and two fuses as outlined in the Installation Manual. The backup battery is isolated from the main buss thru a large schottke diode and the Avionics buss is isolated from the Main buss thru two large schottke diodes. So when I turn off the Master and Backup battery switches, the ignition modules are completely isolated from any other backup batteries or capacitors. If I need to turn off one module I can pull one of the breakers. So, until I have a good reason to rewire the modules and add four more components, I will leave the wiring as is. Please let me know if I am making a big mistake or missed something important.

If a purple wire shorts, will it burn up before the 15A breaker blows? Seems I read somewhere that it is 20awg. Don't think that'll carry 15A. Any and all wires downstream of a breaker need to be able to carry the full capacity of the breaker or you might as well not have one. I think the red wire is 18awg. Will that carry 15A? Another question would be "is the switch rated for 15A?"

Ed Holyoke
 
Last edited:
I converted to a CPI-2 ignition on my RV-9a and I now have over 50 trouble free hours. I'm very satisfied with this new ignition and described this conversion in a previous post. Last year I had a few phone conversations with Barry and Ross before I started the wiring. My questions were mainly focused on the supply current to the red and purple wires. I am reading this thread and couldn't help but mention my thoughts and experiences relating to those red and purple wires.

Maybe I missed something or I'm just a knucklehead or both, but I could not see the reason why two circuit breakers and one switch were required for each module when only 1 circuit breaker would do the job. I used a single 15A breaker for each module and connected it directly to the red and purple wires. My thinking is that the red and purple wires are both supplied by 14v from the main buss and it was not clear to me why it was required, what function do they serve, or if it would make any difference for current to pass thru those extra components. I understand it's Best Practice but I am a firm believer in the KISS principle. So any time I can eliminate two breakers and two switches from a wiring diagram, those are four components that won't fail in the future and will not require valuable real estate on my panel now.

Both ignition breakers are on the Master buss, and the backup battery has one switch and two fuses as outlined in the Installation Manual. The backup battery is isolated from the main buss thru a large schottke diode and the Avionics buss is isolated from the Main buss thru two large schottke diodes. So when I turn off the Master and Backup battery switches, the ignition modules are completely isolated from any other backup batteries or capacitors. If I need to turn off one module I can pull one of the breakers. So, until I have a good reason to rewire the modules and add four more components to the panel, I will leave the wiring as it is. Please let me know if I am making a big mistake or missed something important.

The main reason for separating the CPU and coil power is noise isolation. If it works well as is, noise doesn't seem to be a factor in your installation but it has been in others. It's a case of applying best practice again.
 
Best practice

Best practice would be to protect the 20g purple wire with the appropriate size fuse. The size that is recommended in the install instructions.
 
Since my CPi-2 panel controller is on the other side of the cockpit I also have a third "LOP MODE" switch that turns the LOP mode on and off remotely and also causes a LOP Mode On annunciation on the G3X PFD.

I'm running the CPI-2 also, and have it pretty far over on the right side of the panel. I can reach it, but it's not convenient. I don't ever recall seeing anything about the CPI-2 supporting a remote LOP mode switch. Is that an undocumented feature? If it's available, I'd like to use it.

Thanks!
 
I'm running the CPI-2 also, and have it pretty far over on the right side of the panel. I can reach it, but it's not convenient. I don't ever recall seeing anything about the CPI-2 supporting a remote LOP mode switch. Is that an undocumented feature? If it's available, I'd like to use it.

Thanks!
Give SDS a call and talk with Ross or Barry. It is possible.
 
Yes, you can use an external switch for LOP instead of the keypad. We'd need to ship out a couple parts to do that.

Can only use one or the other, not both.
 
sds cpi-2

I replaced an unreliable left mag with the CPI-2 last December in the RV4 and it has been working great. Installation for me was straight forward and the few questions I had were answered quickly. Service could not have been better.
Nice to have a quicker starting and smoother running engine plus I really like having the display with RPM, MP and advance and the ability to scroll through other data and change settings if I want to. Also great to have no moving parts or seals to worry about.
 
Just my experience...

You won't find a company that's more responsive, and competent, than SDS. You have direct contact with the ultimate decision makers that have "skin in the game". I don't compliment people or companies like this often, but Ross and Barry are among the best I've ever dealt with. They're totally committed to their products and their customers, and they're extremely competent in what they do.

I'm proud to be a SDS customer.
 
Thanks Steve. We're always burning the midnight oil trying to make things better.

We think our customers are also the best!
 
Another compliment for SDS

Just my experience...

You won't find a company that's more responsive, and competent, than SDS. You have direct contact with the ultimate decision makers that have "skin in the game". I don't compliment people or companies like this often, but Ross and Barry are among the best I've ever dealt with. They're totally committed to their products and their customers, and they're extremely competent in what they do.

I'm proud to be a SDS customer.

I have to second this endorsement. Ross & Barry are the absolute best. They've helped me through some tricky troubleshooting and have always been eager to help & very generous with their time and advice.

I'm also proud to be an SDS customer.
 
Thought my install was all good, but - issues

Cross posted from another site, so some of you may see this twice.

I've had to let my chase plane buddy know that I'm postponing my first flight, which was to be tomorrow morning.

Turns out that on my first actual "mag check" with my dual SDS CPi2 setup, it's only running on one side. I can't believe I got this many engine runs into the project before using the A-B/L-R test function, but it's due in large part to my lack of understanding how completely the backup battery function takes over on these things - the power switches they say are required to turn it all the way off do NOT kill backup power to it and thus DO NOT substitute for a mag check done by grounding each side of the ignition by key-switch or by keypad menu entry. After I figured this out and actually did a true "mag-check" yesterday, the engine died with the key switch in the R or OFF positions but ran in L or BOTH. Same when tried with keypad. Left a voicemail at the shop 4:30 yesterday and called Barry's number later Friday evening - calls have not been returned yet. Given their reputation for customer service and my own prior experience, I know I'll hear from them at some point.

My other issue is that the ignition often lights off a charge when I first power it up. :eek: Typically it just kicks a blade back 60 degrees or so (still a hazard to anyone near the prop at that point!) but last night it fired one that sounded like a .44 mag without ear pro. Blew soot and fuel all over the floor of the hangar. Okay, I _might_ have had a hot start flooding issue just before that... This behavior is now no longer something I can ignore. A pilot should not expect an untouched prop to move by itself before cranking is begun. It's too easy to leave an ignition toggle on at shutdown and have the prop kick back the moment the Master is turned on - has already happened to me. Barry said it's a reported issue and he had a software type of fix for it but I'd have to send the CPU's back to him. Given the difficulty of access to my units and the approaching first flight target date, I decided at first to live with it and mitigate by powering up the CPU's first and then the coil packs (my left-to-right toggle sequence is LEFT COIL, LEFT CPU, RIGHT COIL, RIGHT CPU). Next time I tried this the fuel-air charge lit off as soon as I powered up the LEFT CPU with the coil pack switch still OFF!

It's obvious the unit is capable of discharging all plugs at once on power-up regardless of power/no power on the red wire to the coil pack, and one of these spark events will be on a cylinder with some residual fuel-air and sometimes an open valve or two.

What I've learned for certain is that mounting the CPU's way up under the forward skin between firewall and sub-panel was a big mistake. There were not any other attractive options I could see with the real estate taken up by the Advanced QuickPanel remote components, but I mounted them over the rudder pedal cross tubes thinking I'd never have to touch them again for the life of the airframe - and honestly, I shouldn't - but now I do. As it is, the seat and joystick must come out just to allow the tortuous yoga position needed for unbolting the ECU's from the stringers I put across the ribs from firewall to sub-panel.

I'm thankful to have found this issue before I naively launched on one ignition. I feel divinely protected from my own ignorance and folly once again.
 
Next time I tried this the fuel-air charge lit off as soon as I powered up the LEFT CPU with the coil pack switch still OFF!

When they want to fire a coil, a CPU like yours sends a command to a coil module (sometimes called an ignitor) and that module creates a path from the coils ground wire to a frame ground for a few milliseconds, then opens that connection to force the spark. Each coil (sometimes many coils in a coil pack) gets a 12V feed and an output/ground (in packs, the 12V feed is common, with individual grounds for each coil). If you haven't supplied 12 volts to a coil, it should be impossible for the CPU to generate a spark from that coil.

If your CPU can generate a spark with the 12V feed switch off, I feel that you have a wiring issue somewhere, allowing 12V to the coil with the switch in the off position. You should lay out a schematic and follow each of your actual connections against that schematic.

Once you address the above, you should be able to cure your initial, power on firing issue by powering up the CPU's prior to applying coil power.

I don't know anything about the SDS system and how it manages power or B/U power for the coils, so don't mean to answer for them or step on toes.

Larry
 
Last edited:
Back
Top