What's new
Van's Air Force

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

SDS CPI2

The alternator control was super fast and the battery seemed to soak up anything that we'd consider remotely dangerous.

Let's remember the whole point of the standards is demonstrating immunity to a difficult case.
 
Let's remember the whole point of the standards is demonstrating immunity to a difficult case.

Agreed. We'll do some further testing with the complete PCBs which are starting 2nd proto stage now. RF testing to follow that.
 
Good Progress, Finally

Taking Dan's advice, we did some additional testing today with the battery disconnected, just running on the alternator with about 10-20 amps constant load and adding about 40-50 more with our load tester which allows very quick application of load (or reduction (edit)).

We saw spikes up to 27 volts and much longer duration now- to about 100ms.

The voltage is of no worry but we'll adjust the crowbar hardware and software to reflect this new data.

Other progress today- the keypad design will be ready to go out for quoting tonight and the colors have been toned down to dark and light grey with white outlines- no more green.;)

2nd proto controller boards may be done for Friday and we'll begin populating them with components over the weekend. That will take some time by hand as there are a lot of very small ones on this design.

Next week, the CAD will go to the machine shop for the proto programmer housing and the week after, the controller box CAD files will go to the other machine shop to run a proto enclosure.
 
Last edited:
Wait a minute. You mean to tell me after I?ve waited all this time your not going to offer the CPI2 in ?green?.

I appreciate the extra testing and transparency Ross.
 
Taking Dan's advice, we did some additional testing today with the battery disconnected, just running on the alternator with about 10-20 amps constant load and adding about 40-50 more with our load tester which allows very quick application of load .

We saw spikes up to 27 volts and much longer duration now- to about 100ms.


Set it up with alternator output about 40 amps into that load tester, and then quickly disconnect the tester load.
 
Yes, that's what we were doing today. The load tester uses a contactor and switch. We can dial in any load and fairly instantly apply or release it.

Didn't see anything much different from similar tests carried out over the years even though we had much higher loads applied today. Amplitude and duration were increased with higher alternator outputs somewhat, as expected.

BTW, the test vehicle was running on an SDS EM-5. Did around 100 switching events, engine never skipped a beat.

We did observe that the wipers activated from some of these spikes though. Interesting.
 
Last edited:
Wait a minute. You mean to tell me after I?ve waited all this time your not going to offer the CPI2 in ?green?.

I appreciate the extra testing and transparency Ross.

I know many of you will be sorry to see the green go...:rolleyes:

We are about to run another batch of original CPIs, green and all, if you MUST have green.:D
 
We saw spikes up to 27 volts and much longer duration now- to about 100ms.

Well, ok. As noted previously, the current test standard for ordinary automotive applications is >75V. Recent unfortunate events have shown it is possible for some aircraft installations to reach spike and sustained voltages at that level. For example, consider the note below. EarthX thought a 60V max was adequate when designing the BMS. As it turned out, it wasn't enough out in the real world, where customers make random alternator choices. That led to the recent requirement for an overvoltage system when combining an EarthX with a alternator of more than 20A capacity.

I'm entirely in support of the EarthX OV system requirement. That said, some customers will ignore it...theirs and yours.

During his flight, his regulator failed and his voltages climbed up and remained at +29V and 40+ amps for many minutes. A couple of times the voltage spiked above 60V per the data from the EFIS.

http://www.vansairforce.com/community/showthread.php?t=154137
 
Well, ok. As noted previously, the current test standard for ordinary automotive applications is >75V. Recent unfortunate events have shown it is possible for some aircraft installations to reach spike and sustained voltages at that level. For example, consider the note below. EarthX thought a 60V max was adequate when designing the BMS. As it turned out, it wasn't enough out in the real world, where customers make random alternator choices. That led to the recent requirement for an overvoltage system when combining an EarthX with a alternator of more than 20A capacity.

I'm entirely in support of the EarthX OV system requirement. That said, some customers will ignore it...theirs and yours.



http://www.vansairforce.com/community/showthread.php?t=154137

In our case, even if it our OV system activated, the controller is isolated and switches over to the backup battery. As previously stated, the critical components on the input power are rated to 125V. We don't rely on other OV systems in the aircraft which is why we went to so much time and trouble here. As you say, not everyone will install an alternator crowbar system.

We were loading the circuit to around 60 amps (max alternator output in this case) with no battery connected at all- an unlikely but possible scenario. The contactor release spike was almost 56V sometimes but that spike was microseconds, which won't damage anything.
 
Hi Ross,

To clarify (at least for me), are you actually running the engine (controller operating) on just alternator power when you do the 'load dump'? That, or more accurately, dumping both battery and the load while operating, would be the 'real world' test.
 
We were loading the circuit to around 60 amps (max alternator output in this case) with no battery connected at all- an unlikely but possible scenario. The contactor release spike was almost 56V sometimes but that spike was microseconds, which won't damage anything.

The alternator manufacturer did a nice job with the regulator design. Whose is it?
 
The alternator manufacturer did a nice job with the regulator design. Whose is it?

This was Hitachi. I expect the same performance from IR Denso designs as we've done similar, though less severe testing with those in the past.
 
Last edited:
Working!

The first 2nd prototype board is now running. We're very happy with that and will start running all the tests over again to verify performance and operation.
 
Have both CPI2 boards loaded now and should be talking to each other tomorrow.

Spent some time on software to detect if either of 2 coil pack fuses are blown and makes the check light come on.
This won't allow backup battery to send power to bad coil pack so this way the backup battery will still run the good coil pack if required.
 
Last edited:
Testing continues. This is a long process, making sure it all works seamlessly and repeatedly with no glitches.
 
Update

Things are working pretty well now after a couple more mind benders.

Just tweaking some software on the charging and aural warning circuits now. Prototype enclosure designs will go to CAD around August 6th we think and hopefully have those in aluminum around the 13th.

It's been a long, painful journey, hopefully happy times will come soon now!
 
Barry was cleaning up some code today and testing the serial communication today. Making rapid progress now.

Most hardware and software issues have now been addressed and a lot of testing on unusual battery conditions, switchover and power on/off events has been done.

Light at the end of the tunnel. It will be good to freeze the design soon and get production PCB in process.
 
Considering a Larger Backup Battery

Hello Ross,

I am impressed with your progress and game plan for the CPI2 and have pretty much made my mind up to purchase one this winter. I have a question regarding a backup battery. I will be adding a Skyview flat panel at the same time and it also needs a backup battery. I am thinking about using a single lithium iron phosphate (10-12 Ah) backup battery to power the CPI2 and the avionics bus in case a voltage problem occurs. This should give me plenty of time to sort out the problem and get to an airport. Will the CPI2 keep this larger backup battery charged and be able to switch over and have the capacity to also supply current to the Avionics bus, or will I need to manually charge and switch on the backup battery? What are your concerns with this idea?
 
Last edited:
Hello Ross,

I am impressed with your progress and game plan for the CPI2 and have pretty much made my mind up to purchase one this winter. I have a question regarding a backup battery. I will be adding a Skyview flat panel at the same time and it also needs a backup battery. I am thinking about using a single lithium iron phosphate (10-12 Ah) backup battery to power the CPI2 and the avionics bus in case a voltage problem occurs. This should give me plenty of time to sort out the problem and get to an airport. Will the CPI2 keep this larger backup battery charged and be able to switch over and have the capacity to also supply current to the Avionics bus, or will I need to manually charge and switch on the backup battery? What are your concerns with this idea?

Mike, we appreciate you waiting for the CPI2.

We charge the backup battery at a very low rate to just keep it topped up. There is no provision to charge this battery if the alternator takes a dump but with 10-12 Ah, you can run the CPI2 for several hours. It will automatically switch over to the backup when voltage falls below a preset value. We also have selectable thresholds for different battery chemistries.

Not sure what kind of current the Skyview draws or how long you want to run after the alternator dies but all that is pretty easy to calculate if current draw is known.
 
Want to Keep My Key Switch

Hello Ross,

I'm sure many folks, myself included, will be converting from traditional magnetos or PMags to the CPI2, and we all need to decide what to do with the key switch and existing p-leads. I don’t like the idea of removing the key switch and possibly replacing it with two ignition on-off switches and a start button. I have looked at the wiring diagram on your website to use the existing p-leads to activate a relay to power the ignition modules. I would much prefer to use the p-lead to momentarily shut off the ignition for a mag check than to depend on a p-lead powered relay to provide full time 12 volts to an ignition module.

I know that it is late in the CPI2 introduction process, but I have a request. Is there anywhere in the new CPI2 module to add a p-lead to momentarily ground out a component that powers the coil pack to enable a mag check? This will make the conversion to the CPI2 more conventional, easier to install and, in my reasoning, more reliable. And we can keep our fully functioning key switch.

Thanks for your time,


Mike
 
Last edited:
Completing some RF interference testing on the boards today and happy how that went. There will be some more thermal imaging done on the board soon before we freeze the PCB design and send it off for production before month end hopefully.

Keypad layout is frozen now. That will go out for quotes soon too.
 
Hello Ross,



I know that it is late in the CPI2 introduction process, but I have a request. Is there anywhere in the new CPI2 module to add a p-lead to momentarily ground out a component that powers the coil pack to enable a mag check? This will make the conversion to the CPI2 more conventional, easier to install and, in my reasoning, more reliable. And we can keep our fully functioning key switch.

Thanks for your time,


Mike

Mike, I?ve sent you a PM
 
2nd thermal imaging test

Shot the 2nd proto board today to confirm what we already knew. The SMT voltage regulator ran hot despite tons of via holes nearby it to try to sink the heat away. Lower left. While within specs, we never like hot running components on our boards as they may see 50C ambient conditions in some southern states.



Barry devised a fix today to mount a through hole, old school 7805 to the board with a cutout to the allow it to sink directly to the chassis like the EM-5 does. The mod is already done in CAD which should be the last of the changes before we submit the files for production. Lots of components and traces to move to fit the much larger device into the space where the small SMT component was.
 
Last edited:
Enclosure Design

Now that the PCB outline has been frozen, Barry has been able to get the enclosure design into CAD and check the dimensions with the PCB CAD. This will be sent on to the machine shop this week for prototyping.

 
Update

The machine shop has been unable to keep up with other parts demand and is only now getting to machining the proto enclosure this week.

Keypad design has been sent out for quoting to 3 vendors.

Programmer head enclosure CAD is complete, files sent off to machine shop, might see that by the end of the month.

Revised the main PCB and are eliminating some unneeded components. That will go out for production next week.

This stuff takes forever when you are busy with other orders. September has been crazy for us. Good in one way, bad for getting this design in production.
 
CAD Snippet



This render shows part of the dual CPI2 enclosure with the MAP ports and part of the main connector package plus PCB outline for the top board.
 
Update

We had to make a slight change to the CPI-2 keypad for production considerations but the design is now in process for manufacture.



This rendering shows the face layout. The outside square dimensions of the keypad are 2.8 X 2.8 inches.

I should have some photos of the CNC'd enclosure for this later today.
 
Great to see some progress. Any info you are willing to share about price points & buttonology?

How?s the aux battery tray coming along? How soon will it be ready for the market place?

Missed you at Reno.

Thanks Ross
 
Great to see some progress. Any info you are willing to share about price points & buttonology?

How’s the aux battery tray coming along? How soon will it be ready for the market place?

Missed you at Reno.

Thanks Ross

No pricing yet until all the bills come in. As for buttonology, pretty similar to the EFI stuff or the CPI-1. The LOP is self evident. The PROG button allows access to either controller when a dual board version is used. The round holes are all for indicator LEDs to show what's active and which power source the unit is running on.

The CNC shops have been so busy churning out other parts for EFI systems, the battery boxes just kept getting pushed back. I expect I may have some next week finally.

CNC'd parts for the CPI-2 are in process now and I'll be picking up prototypes ones today and next week. Keypads are in process. PCBs are mostly finalized and will go to production soon. This is a complex thing to test and that's been ongoing for months now. We won't release the thing for sale until it passes all of our tests. That could be Nov. if things go smoothly or later if things require more software revisions and more re-testing.
 
Last edited:


This is the prototype box for the LCD/ keypad/ PCB. We'll test fit the PCB in this as soon as it's done before doing a production run.
 
Sorry for overstating the obvious, but...

Wow! Even the box is machined from billet rather than folded aluminum. This is why SDS products are on my airplane upgrade wish list!
 
Sorry for overstating the obvious, but...

Wow! Even the box is machined from billet rather than folded aluminum. This is why SDS products are on my airplane upgrade wish list!

Build it in billet as we say. You wouldn't want to know what this first box costs after CAD, tool paths were established, fixtures made and it's actually machined though. Fortunately most of those costs are not recurring...
 
Ross,

Referencing the nomenclature on the two LED?s depicted on the top of the CPI2, are those LEDS meant to identify the main power source vs the auxiliary power source? Main battery/aux battery?
 
Progress



Photo of production CPI2 keypad. Happy to have these in hand now so we can test fit to the enclosure and get those in production as well
 
We have been receiving components from our suppliers all week. Production PCBs should arrive tomorrow, stencils next week, prototype enclosures and battery boxes early next week.

Things are coming together now.
 
Great news. I hope I?m near the front of the line for the CPI2 and related battery components. I already have the FF components. Any idea what the price will be yet?
 
You're near the front of the list. I'll be getting most of the bills in next week as we are just finishing the final component orders tonight and Friday. Will also be getting the rest of the CNC'd parts then too so I can do most of the costing. Still have to get a quote for loading and cooking the PCBs but I have a rough idea of what that will be from many previous jobs.

I hope to have prices figured by the 3rd week in Nov.
 
I know many of you will be sorry to see the green go...:rolleyes: We are about to run another batch of original CPIs, green and all, if you MUST have green.:D
I read the whole thread... whoooa. Ross you are a patient man. Take all the armchair designers input with grain of salt, and don't rush nothing. Take all the time you need or want.

On a serious note, I LOVE GREEN :D Glad the original CPI is still available since that meets my needs. I'm standing by on the CPI2 price, which is a cool design and might change my mind, but at this time my plan is provide my own OV protection with dual original CPI's:

Ignition A) powered from main bus battery (which is crowbar protected from alternator going crazy OV).
Ignition B) powered from E-bus battery charged from main bus through isolation diode, with independent over-volt protection relay (automatic).
* Under-voltage (alternator off line) will trigger alarms and pilot action (land).
 
Last edited:
Warning: CNC Porn

I read the whole thread... whoooa. Ross you are a patient man. Take all the armchair designers input with grain of salt, and don't rush nothing. Take all the time you need or want.

On a serious note, I LOVE GREEN :D Glad the original CPI is still available since that meets my needs. I'm standing by on the CPI2 price, which is a cool design and might change my mind, but at this time my plan is provide my own OV protection with dual original CPI's:

Ignition A) powered from main bus battery (which is crowbar protected from alternator going crazy OV).
Ignition B) powered from E-bus battery charged from main bus through isolation diode, with independent over-volt protection relay (automatic).
* Under-voltage (alternator off line) will trigger alarms and pilot action (land).

Thanks George, it's been a long haul. Our designer will be happy to hear you like the green! No green available on the CPI2 though...

Today we finally got our boards and prototype dual box together. Just 2 minor changes to bring the boxes to full production spec. Here's some PCB and CNC eye candy for those patient souls awaiting release:



Boards are panelized for machine loading and soldering. We had the quote back today and it was fair so this will be a go early next week.



Trevor turns out his usual impeccable work on the Haas



Some folks might be tempted just to buy CPI2 for the box it comes in...



Notice the heat sink platforms for the big through hole type 5V regulators we had to fit after the earlier thermal imaging showed the SMT regs were getting too hot.

Anyway this is one reason why we love our jobs. It's very satisfying to see all the numerous individual pieces start coming together after over a year of hard work and multiple challenges along the way. Inching ever closer to the finished project.
 
Hi Ross

I am trying to follow the thread from the beginning but getting a bit confused on the marketing\branding of the various components. Can you help clarify a bit. At times, part of this thread seems to show CPI2 as the programmer type device that would go into the panel and other times CPI2 seems to be the replacement for the previous gold colored EM-5 box?

Is the CPI2 systems composed of the programmer + a EM-5 type box? I may be the only one, but a picture\block diagram of the main components and naming would help me greatly.

Thx

Steve
 
Hi Ross

I am trying to follow the thread from the beginning but getting a bit confused on the marketing\branding of the various components. Can you help clarify a bit. At times, part of this thread seems to show CPI2 as the programmer type device that would go into the panel and other times CPI2 seems to be the replacement for the previous gold colored EM-5 box?

Is the CPI2 systems composed of the programmer + a EM-5 type box? I may be the only one, but a picture\block diagram of the main components and naming would help me greatly.

Thx

Steve

The EM-5 is a totally separate product from the CPI and CPI2. The EM-5 does fuel and spark control and uses a 3 1/8 panel mount programmer and blind mounted ECU.

The CPI is an ignition only, single module, with the ignition controller and programmer built into one box which can be blind or panel mounted.

CPI2 is 2 part, ignition only with a separate blind mount controller and separate 2.8 X 2.8 inch panel mount programmer.
 
All electronic components in hand now along with the solder stencil for production loading and soldering. Off to the circuit board house in a day or so to get produced.:)
 
Boards and bits were off to the loading/ soldering facility today. We've been slammed with EM-5 orders this week but will be working on CPI2 software cleanups, documentation and pricing over the next couple weeks as time permits (that means evenings). We've also given the go ahead on CNC production of single and dual enclosures, backup battery boxes and panel display enclosures.

Hopefully in about 3-4 weeks all this stuff will start flowing into our hands and we can start delivering to those folks on our list who've been patiently waiting in early 2019 after the holidays are over.
 
Update

Completed PCBs should arrive today or Wednesday. Single and dual enclosures plus battery boxes are off the mill and will go to anodizing today. Programmer boxes and faces are machined and done anodizing. LCDs and keypads have been in hand for a few weeks now.

It's all coming together now. Production software is done and bench tested. Documentation will hopefully be completed by mid January, about the same time we finish engine running tests if the weather permits.
 
Christmas Comes Early

Received a bunch of CNC'd parts today from our hard working anodizer, Christa before they closed down for the holidays.



CPI2 battery box parts



Dual and single CPI2 enclosures

 
Back
Top