DIY Electric Car Forums banner

Experience w/Paul & Sabrina's Controller

45K views 193 replies 24 participants last post by  GE11 
#1 ·
Hi,
I'm looking to begin a discussion about peoples experience with Paul and Sabrina's controllers.
I'm working on my first EV conversion and am very interested in their controller. But as they're not a "manufacturer" with the accompanying documentation, I am hoping to find some experienced users info.
Anyone?
Thanks!
 
#2 ·
Hi
I built two of Paul's 500 amp units and I'm currently using his 1600 amp unit

500 amp unit
The kit and instructions are superb - I did manage to blow my first one up but as this was the first electronic thing I had ever built my learning curve was very steep - and I should not have left it open in a dirty garage for 9 months!

The second one was much easier and worked fine

Then I got one of Paul's 1600 amp 400 volt units
It's been working well - Paul has made a few software changes for me

Very pleased with both of them - work well, very cheap, I learned a ton - and Paul is very helpful
 
#3 ·
Hi Duncan,

Thank you for your reply.

I've been reading through the DIY for years now and have finally found the right project to "start" my EV adventures.
I have a 38 volt DC forklift pump motor that should work for an ATV conversion.
I've been advised that I need a controller rated at 150 to 175 Amps for an hour and 400 to 500 Amps for 1 minute.
Fine - and I've been looking at various off the shelf options.
BUT, when I discovered the P&S 's controllers it seemed like they were more versatile for voltage input. As in, I can use it for a larger motor project at some later date.
I'm also trying to figure out what other parts I'll need to go with the P&S controller.

What was your first conversion project?

Thanks again.
 
#5 ·
... It did die on me at the drag races back in March - but I have acquired a replacement! which is now in the car - final re-assembly and back on the road today (I hope)
Just to clarify for gabtrac - it is the motor in Duncan's car which died, not the controller. And the car is custom-built (with Subaru suspension parts), rather than a conversion of a production car.
 
#6 ·
Hi,

This is a great thread for me to keep in touch as I have recently got two P&S DC Control Boards, and I'm slowly doing a Chevy S-10 Conversion and I will be happy to share my P&S related stuff here, and I'd love to see what others are doing... so thanks and please do keep it coming. During my dialog with Paul he was very helpful and he did offer lots of relevant info and recommendations on compatible parts such as the throttle, links to IGBT's etc. which I would be happy to share if anyone is interested.

cheers!
-Ocean
 
#7 ·
Paul is still active on ecomodder.com. http://ecomodder.com/forum/showthre...erter-ac-controller-10839-318.html#post560622

Through the years his controller(s) evolved from DC, to AC 3 phase, to universal DC/AC, all with power stages.
The controllers are a good representation of the progress in the ereas of EVs, electronics and microcontrollers.
His latest version is a "logic board" only for the Leaf inverter+motor.
As can be seen in the posts following the post in the link above.

Really great stuff. I have read a lot of the posts on ecomodder (not all).
He did get a lot of support from "expert engineers" and other contributors.
 
#8 ·
The difficulty I'm having is that there's never an overview of these projects telling you what's going on. If you have followed all 7000 posts of the discussion over the years you'd know, but there's never a summary and "where we are now".

Like, this started as the OpenRevolt project, and there's still a wiki for that, but it hasn't been updated in ages and there's no files available. There's no one that tells you that it kinda switched over to the P&S site and that it's not as open anymore. Or that there are old versions with errors in them, that you'd never get told unless you asked the right person. There was never a wrapup posted that, this is X and X works this way.

Paul's said stuff like "Oh we lost those files 8 years ago, there are no more boards like that" etc. Which is fine, it's just, unless you stumble across those specific posts, you don't even know what you don't know.

I would like to build a simple DC controller, but I can't find decent documentation on any controller that was at any stage of 'This now works." Only "Here was a revision as we were working on things".

Just my two cents. I'm not saying anyone owes me anything (they don't), I'm just lamenting that, from someone who wants to take advantage of this open source project (kinda the point of having one?), it doesn't feel accessible to me. Unless, it seems, I was capable of engineering it myself, which defeats the purpose.
 
#9 ·
Ok!
Thanks to all of you.

First - Ocean - I would be very interested in anything you could share. My next project (if I ever get through this one) will be an S-10.

DUNCAN!!!
Amazing - you're out of my league, but I'll be giving it the old college try.
I too am building a house. Seems like I'm just a little behind you in my timing.

Ok, so my first question of many to come - how difficult is it to program the P&S and what hardware do I require?

I'm not sure if mentioned that I've been lurking all over this site learning as much as I can for a number of years now. The problem is, I can't remember half of what I've learned...
Please bear with me.
Thanks.
 
#10 ·
Hi
Re-House
Have you designed it yet?
If not I can give you my tuppence worth

Reprogramming the P & S
For me that is the most scary and awkward part!
So I kept mine as simple as possible!

You will have an issue if you build one of the earlier units -
(1) There is the base level programing - I skipped this by buying a chip that Paul had already done
(2) the final programing - can't be too hard because I managed - BUT you may need an old machine with XP on it

Overall the most frightening part of making your own controller is programing it - but Paul has been incredibly helpful
 
#11 ·
Thanks Duncan,

The reason I asked about the programming is that I'm a Mac guy. Regular PC stuff is way off my radar so even the simplest of maneuvers is complicated...
I have an old PC laptop that I used to program my first drone, but it wasn't pretty.

I designed the house. Plans and permits all done. Just now coming out of the ground.
The house is why my entries will be slow to come.
 
#12 ·
Programming issues with microcontrollers for a DC traction motor?

Well, in that case I've got good news for you.

DC traction motors have been used for propulsion for decades without the aid of microcontrollers.

Although Paul has done a splendid job with his controller board, it is still basically a buck regulator.

There are a lot of good ASIC buck regulators available, for instance the LM2694 that I'm using in the buck stage of my charger design.
It has hysteretic control, which is ideal for DIY No programming needed.

But it does run at a frequency that is probably too high for a 144V/500A power stage.
But who knows, I haven't done the math for modern IXYS mosfets and schottky diodes and it might very well be possible to lower the frequency enough.
 
#13 ·
I have been using my open revolt controller since Sept. 09. My conversion is an 87 Ford ranger. There has been no problems with the controller.

Programming was done on a parallel port programmer I made. Changing the program can be done with RTD Explorer with a USB to serial cable adapter.
Even with windows10.

I still have schematic and BOM.

Alvin
 
#17 ·
MattsAwesomeStuff is right that the documentation hasn't been great. I never knew how to make changes to the Wiki, and then the person who was storing the files quit storing them. I have been so busy the last couple years with 2 full time jobs, but now as of today I am unemployed (in greenland!) , so I hope in the next couple months to be really improving the documentation (in between looking for a new job haha). There is a new DC control/driver board that I had made that will be here in about 4 days. If anybody wants the software and whatever else for it, they are welcome to it. the same goes for the AC control/driver board. You are welcome to the most recent files. Since I'm a little "website design" challenged, I just email it to people when they ask for it. My email is pandspowerelectronics@gmail.com.

Here's a video of someone who got my new DC board (I ordered a couple and assembled them to make sure it was all good before getting a batch done):
https://blackfamily-my.sharepoint.c...ct&slrid=78ef629e-d0e2-5000-8f80-7375769e162b
 
#18 ·
MattsAwesomeStuff is right that the documentation hasn't been great. I never knew how to make changes to the Wiki, and then the person who was storing the files quit storing them. I have been so busy the last couple years with 2 full time jobs, but now as of today I am unemployed (in greenland!) , so I hope in the next couple months to be really improving the documentation (in
between looking for a new job haha). There is a new DC control/driver board that I had made that will be here in about 4 days. If anybody wants the software and whatever else for it, they are welcome to it. the same goes for the AC control/driver board. You are welcome to the most recent files. Since I'm a little "website design" challenged, I just email it to people when they ask for it. My email is pandspowerelectronics@gmail.com.

Here's a video of someone who got my new DC board (I ordered a couple and assembled them to make sure it was all good before getting a batch done):
https://blackfamily-my.sharepoint.c...ct&slrid=78ef629e-d0e2-5000-8f80-7375769e162b


Hey Paul I hope the unemployment was not something that a bad thing happened to you! Hope things are ok otherwise.

Hey would you happen to have the latest documents on your synchronized dc controller design that you did way back?

Also what is this new DC board for? What are the new improvements?
 
#19 · (Edited)
It just required frequent traveling, but my wife has M.S., and can't drive herself, so I need to be near home, and can't take long trips. Otherwise she and my little kids get stuck at home for extended periods of time.

Oh no, the SR controller. Hmm... I don't have those. I made the boards several years ago, the hard drive went bad, and I never re-did it. This is just a surface mount version of the DC control/driver board that plugs into up to three 600v 600amp IGBTs. Fairly similar to the one duncan has, but also has a USB plug for conversing with a computer using realterm.
 
#21 ·
Hi,
I had built this controller about 2 years back. Since then the car have been working great @ 96v.
Recently one of the mosfets blew so I replaced it. Now I got a funny problem. Early in the morning the motor just wont run though the green and yellow and the contactor comes on.
But as soon as the car gets little heated in the sun, the car works as normal as ever .But in the morning the motor just wont move. Need help to please find the problem...
Regards
Arnold
 
#25 ·
Hi Everyone,

I'm such a busy guy that I'm kinda a lurker here although I enjoy getting involved whenever I can.... and of course now that I need something.... here I am. I just hope my story helps someone else too....

A while back I got two DC Control boards from Paul over at P&S Power Electronics. Very nice transaction... one was for a Chevy S-10 conversion I'm doing (slowly) and the other for.. well, as a backup mainly, or something else.

Meanwhile driving the Bradley on the way home, in the 105degF summer heat, after driving 35 miles, pulling only about 150 - 200 amps, my Logisystems controller burned out. Check out my post in the Controller forum for pics!

So that put a fire under my ass to complete one of these P&S Controllers 'cause now my main ride is down!!!

So here's some pics of my progress. I found this case with two fans and amazing heatsink long ago in a scrapyard... in fact it had 5 IGBT's mounted to it... but alas those are not compatible with the P&S controller, so I got the ones Paul recommended. I built up the plexiglass walls... and I ordered the three current sensors... they should be here soon. I have a nice TYCO 500 amp contactor.... and a pre-charge resistor.... So as I put this together, I'd love feedback from anyone with experience with these things.

Regarding the Main Capacitors...

I know the caps are not what Paul suggested... I'm on a budget here and I showed him what I have, and he thinks they might serve... given they are rated 95degC (although they are older) but he had concerns about their ability to manage the Ripple current. I don't know a lot about Ripple Current but I think it has something to do with the rise and fall of the current as the IGBT is switched on and off.... anyone??? And this impacts the heat output of the IGBT, which is the critical factor... yes???

I would be interested if it were possible to add additional CAP's - not necessarily big ones, but certain ones that would minimize this "Ripple"... any ideas or suggestions here would be greatly appreciated.

The system is supplied by a bank of 4 parallel Chevy Volt Lithium Ion packs @150volts. Each pack is 47 amp hours... and I believe they have no problem supplying ample current... so maybe the Ripple handling is more important than the extra Farads... so again, anything on handling the Ripple and keeping the heat of the IGBT's down would be greatly appreciated.

Moving on to the IGBT's.

If anyone here with IGBT experience can confirm the contact assignments I would be grateful...

For the IGBT INPUT: Starting from the Main Breaker @ Battery POS.... to 500amp SHUNT... to TYCO Contactor... to CAP POS / IGBT contact #1.

(Noting the pre-charge resistor which will be controlled by an additional contactor)

IGBT OUTPUT: IGBT contact #3 to Motor POS (A1) .

Does this make sense, or is the proper flow of current reversed through the IGBT?

The three current sensors recommended by Paul would go on the barstock leading to IGBT Contacts #1.

Battery NEG to CAP NEG / Motor S1.

The motor is the older 8" ADC.

Interesting to note that the Logisystem controller had three main connections - Motor NEG (S1), Battery NEG and Battery POS (which was also connected to Motor A1)... and the motor had A2 connected to S2.

So it seems like this P&S Controller setup would have slightly different arrangement from the Logisystems.... almost as if the P&S IGBT's are controlling the current coming from Battery POS whereas the Logisystems controller was controlling current returning to the Battery NEG... although it's funny because the blowout occurred at a MOSFET near the Battery POS terminal of the Logisystems controller....

Anyways, thanks in advance everyone.... and best wishes on your projects!
-Ocean
 

Attachments

#26 · (Edited)
...
Moving on to the IGBT's.

If anyone here with IGBT experience can confirm the contact assignments I would be grateful...

For the IGBT INPUT: Starting from the Main Breaker @ Battery POS.... to 500amp SHUNT... to TYCO Contactor... to CAP POS / IGBT contact #1.

(Noting the pre-charge resistor which will be controlled by an additional contactor)

IGBT OUTPUT: IGBT contact #3 to Motor POS (A1) .

Does this make sense, or is the proper flow of current reversed through the IGBT?
It looks like you have it backwards.

For a DC motor H-bridge drive with forward and reverse, 3 goes to the DC buss (+), 2 goes to DC buss (-), and 1 goes to the motor. And you would only be using 2 devices to form the H-bridge.

i haven't found the schematic but would guess that the third IGBT is being used as an overall ON/OFF switch to control power to the DC buss.

[edit: i found some discussion where he is using the same power devices for either AC or DC, where the "three phases" are wired together...? Got no clue what that means]

For a 3-phase AC motor you would use all 3 devices with the same wiring but each device output (contact 1) would go to a separate phase.

Capacitors:
4 parallel capacitors at 2,100uF gives you 8,400uF total. i don't know what frequency it is running, but i would guess it is a PWM controller at somewhere between 2 and 20 kHz. At these frequencies the capacitor looks like a short circuit to ground and lots of current will flow thru it and cause heating. The ESR of the capacitors will determine how hot they will get due to the ripple current from the PWM switching.

Does the kit include a large inductor somewhere? for example, https://www.diyelectriccar.com/forums/showpost.php?p=800450&postcount=22
notice that the caps here are much smaller than yours.

Really need to see a schematic to know what was intended.
 
#27 ·
Thanks for your response Kenny,

But let me be more clear about this setup. As I understand it, these are three half-bridge IGBT's, and they are all running in parallel. There is no electrical reverse, just a gearbox with reverse.

I need to study up on what "half bridge" actually means, because I don't really know.... however....

The GATE (labelled) pin of the circuit board is attached to PIN#4 of the IGBT, which appears to control the Gate between contact #1 and #3.... looking closely at the IGBT diagram and the circuit board layout.

The DC control board has three identical Gate drivers, and I understand that they all do exactly the same thing in parallel....

Also, it is clear to me that there will be three current sensors - one for each IGBT, and that the programming allows each IGBT to run up to 300+ amps, for a total of 900+ amps.

So I think the setup has to be very straight forward...

And the CAPS at high frequency acting like a short to ground... this is because??? I can imagine that current moves into and out from them very fast and this causes the heating... perhaps that is what you mean?

thanks!
-Ocean
 

Attachments

#28 ·
For AC currents the caps are low impedance. Not for DC.
And they're not completely capacitive.
There's some real power involved in charging/discharging. High ripple current means big caps are needed to dissipate this power.

Judging by the size of these things I guess they can handle about 20Arms each at max temp and 10 kHz.

OK, the IGBTs run synchronously in parallel. So Paul didn't come up with some clever control scheme to lower the ripple.
In that case 80Arrms isn't enough for a 900A DC controller. It is enough for 3phase AC though. Farads are probably OK.

Don't know any details about the setup, but I would add at least five 77Arms foil caps.
 
#29 ·
Thank you Tony,

So I think you are saying that the caps need to absorb the ripple current from the IGBT's....

I will look more closely into this ASAP... got to go to work now, but thank you.

Can you say something about the flow of current through the IGBT's? Thinking of what Kenny said, and looking at the IGBT schematic, I realized maybe there are bypass diodes on those bridges, and so perhaps the flow does indeed need to be POS at Contact#3 and Output to motor at Contact #1...

thanks, I'll be back tonight.
-Ocean
 
#30 · (Edited)
Okay now i see how he is using the power devices.

Each device has 2 igbt circuits, one between 1 and 2, and another between 1 and 3. This device is sometimes called a "totem pole" or half of an H-bridge. Since you have a gearbox for reverse then you don't need the full H-bridge functionality, so he is just using the "top" or upper part of the totem as the power switch, then running all three in parallel to increase the total output current.

Gate pin 4 is used to switch the upper igbt which is between 1 and 3. Terminal 3 is the power supply (+) input and terminal 1 is the output of the igbt.

The ripple in the current is due to the switching on and off of the igbt at the pwm frequency. Those big capacitors can't stop these high frequency pulses and they shoot right thru from + to - as if they weren't there.

A "snubber" circuit is often used in order to filter or smooth out the ripple and requires a tuned set of high-voltage components (lower value caps, resistor, diode, etc) in combination with the big capacitors (used to hold up the DC buss voltage) and the inductance and resistance of the motor windings.

See for example, http://www.cde.com/resources/catalogs/SCD.pdf , for snubber caps.
 
#32 ·
Thank you Kenny... now I think we are on the same page. I did have the current reversed in my original thoughts. Now it makes sense how you describe it, as I see it in the pins and on the IGBT diagram. I think I need to look into the "snubber" circuit to see how that works. I really want to understand these things as best as I can before putting them together.

So I will check out your example...
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top