Plug-In Supply DIY

So i heard back from rob he said he could do a cash on delivery on a 4kwh pack. This makes me alot less nervous especially hearing the stories on hear

 

That's good to hear gunner. If you schedule it, I'd still be prepared for the appointment to get pushed back a few times while they wait for components to arrive from China, but at least your cash isn't tied up while you're waiting. Its always possible the supply chain problems may be sorted, so you could be pleasantly surprised

Look forward to hearing how it goes.

Rob

 

Thanks for the in depth explanation. My question is how is PIS system connected to GEN 3 Prius, there are no threads on this forum about converting GEN 3 with high voltage (no DC-DC converter) battery? Is it the same approach as gen 2, but without Can View or BMS+, what are the disadvantages?

 

They do "something" to encourage the stock controller to use more power, but it's not SOC spoofing. Whether they have found some way to force the controller to recalibrate the calculated SOC (and because the voltage is held high by the lihtium pack it recalibrates to a high SOC) or stop the controller from counting the current used we don't know. Be interesting to find out though.

 

Not sure if they do this for GEN 3, but on my GEN 2 the system seems to be 'flooding' the CAN bus with high SOC level messages. I know this because I've set my Scangauge to display both passive and active SOC request values. The passive SOC value typically displays 74 or 79% SOC. The active SOC value varies, but typically shows values from 58% to as high as 65% and tries to stay around 62%. I believe those lower values are the true SOC and the high SOC values are bogus. I've also noticed that the PIS system will "shut itself down" (i.e. stop feeding current to the OEM pack) if the 'true' SOC drops below ~58%. That condition can occur if I try to accelerate too quickly.

 

I put in a request for a quote for our lexus RX ( Similar in Design to the Ford Escape) but still haven't heard back yet. Even charging 1 car at home, we still have a lot of solar surplus - so it might as well be used to charge our Lexus, as a PHEV .

 

If anyone is in the SF Bay Area this weekend, you can come talk to Robb about his system at Green Drive Expo. Someone who's going could gather up all the questions and ask them all at once and then report back. Might save some effort.

 

usnavystgc

Nah, never got em. But it kinda evens out because I haven't paid the rest of my shipping dues. I'm excited to get the airbag shocks though. This should make my ride a lot more comfortable.

 

That's interesting Frog, that sounds very much like the classic SOC spoofing approach. Spoofing was based on the observation that the hybrid ECU would essentially ignore a low SOC value from the stock battery ECU if it received a higher SOC value from the add on controller soon after. Its interesting because I also remember Kiet mentioning that they are not just spoofing. That may have been in the context of the high speed EV mode though, rather than just the basic SOC management strategy. If the PIS controller is just plugged into the OBDII port as I seem to recall, then I guess it would kind of have to be spoofing as the stock battery ECU is still out there talking. The alternate approach that I believe the BMS+ uses is to actually isolate the battery ECU on its own CAN bus, and then dynamically decide which CAN messages to pass on to the hybrid ECU via the main CAN bus, and which messages to block or modify before passing.

GasperG, the PIS controller basically performs the same functions as the CANView/BMS+, but in a slightly different way as described above. I'd say the CANView/BMS+ combo mostly has the advantage of being more configurable and giving lots of visibility into the internal operation of the stock and PHEV system, while the PIS has the advantage of being part of a fairly plug and play system. Rumor has it though that the CAN View is being phased out, and while BMS+ does have other applications it was primarily designed to manage several additional stock batteries being added in parallel to form the PHEV battery. As far as I know both the Gen 2 and and Gen 3 approaches are similar as far as how the PHEV battery is interfaced into the stock high voltage system. I want to say that there were changes made to the battery ECU communications making it impossible to use the BMS+ on a Gen 3, but I could be remembering that wrong. The PIS software may also behave differently on the Gen 3 than on Frog's Gen 2 I suppose.

SOC spoofing was introduced toward the end of the open source cal-cars project, and you can see their original description here:
State Of Charge Manipulation - EAA-PHEV

Rob

 

Hi Flan,

The ECM module spoofs the ICE ECU which is also known as a ECM (engine control module). It intercepts and modifies messages from the HV ECU to the ECM or visa versa so the ICE comes on less often. So less ICE is used and more electric assist is used.

I guess that you can now buy Kiety's wheel for $250 from PIS.

 

Your description of how the PIS system spoofs the SOC is odd. My BMSplus replaces the real SOC message with a fake SoC that varies between 65% to 99% as long as the HV battery voltage stays above a end voltage. The SOC is spoofed according to the gas pedal position or throttle. 65% SOC when gas pedal position is at 0% and is increased in proportion to 99% SOC when the gas pedal position reaches 25%. 100% is when the gas pedal is fully pressed. When the end voltage is reached it stops spoofing the SOC and then just calculates the SOC just based on voltage. But still it never let's the original SOC get through from the battery ECU to the Hybrid ECU.

So according to you the PIS just keeps sending a higher SOC and the HV ECU for some unknown reason only accepts the higher but fake SOC and ignores the lower SOC. if this is true then the PIS is very different from the BMSplus.

 

Technically the BMS+ does not spoof, at least in the common usage of the "SOC spoofing" term. The BMS+ has two CAN ports, so it can completely isolate the stock battery ECU from the rest of the system. It can then either chose to pass the stock values to the hybrid ECU or block them and substitute its own values. Spoofing controllers have only one CAN port, and cannot block the messages from the stock battery ECU. The essence of the spoofing method is based on the observation that when it receives two conflicting values for the same parameter in a short period of time, the hybrid ECU will ignore the first value and use the latest value received. So as long as the higher SOC value is always rebroadcast immediately following the stock value, the result will be the same between a spoofing controller and the BMS+ approach. At the time I think Norm's approach was considered cleaner and desirable as it keeps the conflicting traffic off the bus, but in practice I don't think its been found to really matter.

I think you already knew most of that, its just a matter of clarifying what we mean by "spoof."

Rob

 

FWIW, toward the end of the cal-cars project Chris Ewert did a lot of the work to demonstrate SOC spoofing as a viable control method in ~2007. He did quite a bit of work to implement an SOC spoofing control board to work with the CAN-View and cal-cars type conversions, but it was never really finalized and released. He didn't discover SOC spoofing, and there were other closed/commercial projects believed to be using it at the time, but he did a lot of open work on the subject which was very helpful to the DIY community. There's a ton of traffic on the topic on the old cal-cars yahoo group if anyone is really interested in a trip in the way back machine, such as this thread:
Yahoo! Groups

This is about the time the cal-cars open source group was starting to dissolve, with many of the principle developers moving on to their own private conversion company projects.

Chris Ewert and his brother went on to form Ewert Energy Systems:
Ewert Energy Systems

Robb Penthroe of course went on to form Plug-in Supply. The PIS design (and many others) are still essentially based on the cal-cars method, but of course they've evolved it far beyond the original proof of concept.
Plug-In Supply | Plug-In Conversions for Prius and Escape

I believe it was after the "break up" and the open SOC spoofing controller being shelved that Norm developed the BMS+ around 2008 as another evolution of the cal-cars original method.
blank

A nice pic of the "happy family" in 2007 before everyone went their separate ways
2007 Maker Faire - EAA-PHEV

Rob

 

Thanks for the info Rob.

If this is true then I should be able to spoof the ICE ECU using a device with a single CAN port and immediately resend a CAN frame to inhibit the ICE from starting. There is no need to use a dual CAN port device.

The thing that confuses me is that Kiety states that he intercepts the frame ( which is what a dual CAN port device does and requires cutting into the Can bus). If you only need to resend the modified frame soon after the original then why intercept? Maybe it was a bad choice of words by Kiety.

 

Yeah, not sure we really know. Pretty sure they don't intercept the battery ECU messages, just spoof, but I also don't think anyone has actually seen the Kiety ECM so we don't know how it hooks up or what its traffic looks like. I seem to recall someone saying that its just a plugin card on the PIS controller, which would imply its still just a single port device which implies spoofing.

Rob

 

It could certainly just be a difference in what was meant by intercept. I think we do suffer somewhat in the diy community from differing definitions of things we make up like "spoof." I take it to mean retransmitting a CAN value without blocking it. Buts its been used for all kinds of things, including the hardware engine temp hack. Intercept would mean to me block and send a different command/value with two CAN ports as you say, but it could just mean read, modify and re-transmit or spoof.

Rob

 

any updates on this thread? what happened with the ECM boards, are people able to drive on EV only above 65 mph? Gun toting owner, did you get your pack?

 

I have a gen 2 with one of the first (2010) PIS 10kw kits which I am upgrading including ECM board. Im also using the Torque app which I love. I have the eLithion BMS and not the PIS BMS and managed to get the PIDs for the eLithion but searched everywhere for the PIS Specific PIDs. Was wondering if anyone worked up a set of PIDs for the PIS system before I went the ghetto route and logged the whole shebang myself.

 

I think you are the first one. Can you give us some more info on the PIS ECM board?

 

Why not ask PIS or kiettyyyy for the PID's? They should share them if they have them worked them out...