YAPiP - recreating pEEf's approach

Discussion in 'Prius PHEV Plug-In Modifications' started by jdh2550, May 23, 2012.

  1. NortTexSalv04Prius

    NortTexSalv04Prius Active Member

    Joined:
    Nov 14, 2009
    915
    111
    0
    Location:
    Texas
    Vehicle:
    2004 Prius
    Model:
    N/A
    jdh
    The realforce you have are they RFE 24vdc 80ah times two which is 48vdc ?? I see other posting/stating the RFE is 48vdc for one pack
     
  2. NortTexSalv04Prius

    NortTexSalv04Prius Active Member

    Joined:
    Nov 14, 2009
    915
    111
    0
    Location:
    Texas
    Vehicle:
    2004 Prius
    Model:
    N/A
    Oh plus I notice/ see that Enginer is currently only selling off website(denver based server)(michigan address) and dropped ebay listing
     
  3. jdh2550

    jdh2550 Co-Founder, Current Motor Company

    Joined:
    Nov 16, 2010
    317
    246
    0
    Location:
    Ann Arbor
    Vehicle:
    2005 Prius
    Model:
    N/A
    I had Enginer re-configure the packs as 40Ah / 48V instead of 80Ah/24V.
     
  4. lopezjm2001

    lopezjm2001 Senior Member

    Joined:
    Apr 14, 2009
    1,146
    403
    5
    Location:
    Sydney Australia
    Vehicle:
    Other Electric Vehicle
    Model:
    N/A
    The kits are listed on Amazon. I suspect Jack is avoiding using Paypal and paying a Paypal fee for every kit he sells.
     
  5. jdh2550

    jdh2550 Co-Founder, Current Motor Company

    Joined:
    Nov 16, 2010
    317
    246
    0
    Location:
    Ann Arbor
    Vehicle:
    2005 Prius
    Model:
    N/A
    A little bit of progress...

    I have the battery ECU hooked up on the bench*, powered up and spitting out CAN messages that are being read by the Atmel dev board. This just confirms that the bECU works - next up I'll feed in some voltages on B12 (the orange connector) and check that I can see corresponding data values in the CAN data. With that done it confirms that I can use the re-purposed bECU with my pack.

    * I've found a new job and so I've transferred my bench from my old place of work to the home office. Old job was electric motorcycles new job is mortgage software... I should have more time for YaPIP in a week or so.

    bench.jpg
     
  6. NortTexSalv04Prius

    NortTexSalv04Prius Active Member

    Joined:
    Nov 14, 2009
    915
    111
    0
    Location:
    Texas
    Vehicle:
    2004 Prius
    Model:
    N/A
    My 4kw version 6 (April2012) was BUYNOW ebay then billed thru google checkout never went to paypal...........
     
  7. lopezjm2001

    lopezjm2001 Senior Member

    Joined:
    Apr 14, 2009
    1,146
    403
    5
    Location:
    Sydney Australia
    Vehicle:
    Other Electric Vehicle
    Model:
    N/A
    Any photos of your RFE battery packs in the back of your Prius?
     
  8. jdh2550

    jdh2550 Co-Founder, Current Motor Company

    Joined:
    Nov 16, 2010
    317
    246
    0
    Location:
    Ann Arbor
    Vehicle:
    2005 Prius
    Model:
    N/A
    Hi

    I've been making progress and actually do have more time now. The new job is strictly regular hours. Yippee!

    Progress is nothing very "earth shattering":

    1) I've got my DuinoMite (DM) to a stage now where it will be ready to be a good "head unit" for display and logging CAN bus traffic. I've been working on getting a suitable set of input controls for use with the DM in the car (I dislike using a keyboard in the car - too unwieldy). That should be up and running this weekend. I'm thinking of adding a real time clock (RTC) to the DM so that I can properly timestamp the log files. This will involve some more MMBasic alterations.

    2) The proof-of-concept of re-purposing the battery ECU to monitor my new pack looks promising. I just need to finish off the final part which is to check I can take arbitrary signals in on the input side and re-interpret the messages sent out onto the CAN bus (which will now be the secondary bus). I should be able to do that in one evening. It's just a question of getting a round tuit.

    3) The RFE packs have been sitting idle for a while. I need to hook them up to the charger and also confirm that I can control the charger via the CAN interface. I think that the partial pack (I have 4 and 1/2 packs) needs an initial charge. The other thing is that I don't have a BMS solution yet (changing jobs means getting one from CM may prove more difficult!).

    4) Once 1, 2 & 3 are done then I can work on firmware (which will live on the Atmel dev board for now) that will implement "large pack handling" and allow me to install the pack properly and rip out the NiMH. At this point I'll have something akin to one of the "standard" conversions that don't have hi-speed. However, mine will be slightly different in that it will have replaced the battery ECU (well re-purposed the existing bECU and put it on the secondary CAN bus)

    5) Then to fully re-create pEEf's work is to transition the engine ECU from the primary bus to the secondary bus. This allows us to intercept hybrid ECU request to the engine ECU and to instead send our own messages to the engine ECU telling it to spin at higher speeds. And this is what allows hi-speed operation. This is the task that requires understanding the flags in the 3C8 and 348 messages.

    In terms of timing I'm in no particular rush - but hopefully I'll have 1,2&3 done in the next week or so.

    While doing all of the above I'm also paralleling a couple of other related tasks:

    (A) I'm looking for a suitable replacement for the Atmel dev board. A lower cost unit with two CAN channels. I may end up designing a shield for Arduino or TinyCircuits or Galego (the last two are Kickstarter projects I backed). Arduino already has a CAN shield - but just a single channel. This doesn't hold up anything as I can use the Atmel dev board.

    (B) I'm looking to replace the MFD and Audio (see my other thread) with an Android tablet talking via USB to an Arduino/Galego/TinyCircuits device (using the Android ADK library). This involves interfacing a new radio module and the tablet audio signal into the pre-existing AVC-LAN network. As well as allowing the tablet to control the A/C which is on the BEAN network. I also want to hook up a rotary encoder to the device to keep a traditional volume knob (well it will actually have multiple functions)

    And lastly, completely un-related to the Prius there are a couple of other hobby projects floating around! Never a dull moment...:)
     
    miscrms likes this.
  9. lopezjm2001

    lopezjm2001 Senior Member

    Joined:
    Apr 14, 2009
    1,146
    403
    5
    Location:
    Sydney Australia
    Vehicle:
    Other Electric Vehicle
    Model:
    N/A
    It all looks good. I was wondering if you have considered also implementing the original DTC fault codes into your new bECU.

    I also noticed that the HV ECU gets a NEO and GO signal(separate from the CANbus) from the ICE ECU and the HV ECU will probably throw a DTC if the NEO or GO signal does not match the RPM if you try to modify the Target RPM message going to the ICE ECU. Maybe there is an allowable percentage difference before throwing a DTC. So you may not be able to modify the Target RPM.
     
    [email protected] likes this.
  10. jdh2550

    jdh2550 Co-Founder, Current Motor Company

    Joined:
    Nov 16, 2010
    317
    246
    0
    Location:
    Ann Arbor
    Vehicle:
    2005 Prius
    Model:
    N/A
    I should be able to send a subset of the DTC's - some of the codes won't be applicable to the new system.

    I don't recall pEEf having any issue here - so I'm hopeful we'll be OK. We'll see.
     
    [email protected] likes this.
  11. lopezjm2001

    lopezjm2001 Senior Member

    Joined:
    Apr 14, 2009
    1,146
    403
    5
    Location:
    Sydney Australia
    Vehicle:
    Other Electric Vehicle
    Model:
    N/A
    Instead of using a two CAN channel device to modify the CAN messages you could try the PIS method of just spoofing the message. It only requires one CAN channel and is more fail safe. The spoofing is sending the modified message straight after the original and the HV ECU will ignore the first. So you would probably want a CAN device with a really fast processor.
    I am thinking of trying this myself with my Duinomite Mega by just preventing the HV ECU warm up request acknowledgment getting to the ICE ECU message - H348 flag byte. Just need to monitor the flag bits to find out which bit does the acknowledgment. Reading through pEEFs thread I could not see any specific mention of modifying the target ICE speed so maybe he just modifies the flag bytes only.

    So if spoofing works and it probably does when using the right hardware then the only advantage of pEEFs approach is replacing the OEM HV battery pack with one having a better weight/volume to power ratio which would further improve MPGs.
     
    [email protected] likes this.
  12. jdh2550

    jdh2550 Co-Founder, Current Motor Company

    Joined:
    Nov 16, 2010
    317
    246
    0
    Location:
    Ann Arbor
    Vehicle:
    2005 Prius
    Model:
    N/A
    Nope - I'm committed to pEEf's approach. I think it's better.
     
  13. jdh2550

    jdh2550 Co-Founder, Current Motor Company

    Joined:
    Nov 16, 2010
    317
    246
    0
    Location:
    Ann Arbor
    Vehicle:
    2005 Prius
    Model:
    N/A
    :oops: DOH!!!

    The solution has been staring me in the face. The perfect two channel solution will be a UBW32. The PIC32 already has an integrated CAN controller with two channels - the DM just exposes one. So, all I have to do is to replicate the transceiver circuitry from the DM times 2 and then hook those up to the UBw32* and then I've got my 2 channel device AND (& this is the important part!) I've already written low level CAN code for the PIC32....

    * Actually I may use one of the smaller versions of the DM (the mini or the standard instead of the mega). Whichever is most cost effective. I don't need the Mega because this will be the unit that's sole purpose is to control the two buses - I don't want anything non-essential running on this board. No VGA, or other stuff. I certainly won't be coding this part in BASIC - I'll be flashing custom code onto this instead.

    BTW, earlier in this thread I made a WAG that custom hardware would be $150(ish). Actually we will be able to have a mostly off the shelf solution for less than $50. This new guess is based on purchasing a PIC32 based development board (UBW or DM) and $10 (or less) of components to add the CAN transceivers.
     
  14. jdh2550

    jdh2550 Co-Founder, Current Motor Company

    Joined:
    Nov 16, 2010
    317
    246
    0
    Location:
    Ann Arbor
    Vehicle:
    2005 Prius
    Model:
    N/A
    Current Status of the BOM:

    Car: 2005 Prius Gen 2 - high mileage with failing pack - $5,000
    Pack: 72SP2 RFE Cells (~9kWh pack) - $4,500
    Charger: Elcon PFC-2500 w/ CANbus Adaptor - $770
    BMS: TBD (budgeting ~$750)
    Battery ECU: Made up of two parts (a) pre-existing Battery ECU to interface with HV pack & (b) new controller with 2 CAN channels. Part (a) is free (comes with the car!) & Part (b) hardware will cost approx $50 (the software is priceless! :rolleyes: )
    Springs: TBD (budgeting ~$300)


    Total cost for a 9kWh conversion: $6,370 (doesn't yet include cabling or mounting)

    A 4.5kWh conversion should cost about $2,250 less and a 13.5kWh conversion about $2,250 more (when calculating a different pacl size simply scale the pack cost - all other costs remain approximately the same)
     
    [email protected] likes this.
  15. bisco

    bisco cookie crumbler

    Joined:
    May 11, 2005
    94,763
    42,979
    0
    Location:
    boston
    Vehicle:
    2012 Prius Plug-in
    Model:
    Plug-in Base
    what is the max pack size you can get under the hatch floor?
     
  16. Flying White Dutchman

    Flying White Dutchman Senior Member

    Joined:
    Dec 29, 2007
    4,374
    305
    0
    Vehicle:
    Other Non-Hybrid
    Model:
    N/A
    I have seen around 15 kwh

    -Tapatalk
     
  17. jdh2550

    jdh2550 Co-Founder, Current Motor Company

    Joined:
    Nov 16, 2010
    317
    246
    0
    Location:
    Ann Arbor
    Vehicle:
    2005 Prius
    Model:
    N/A
    I don't know for sure - but this certainly sounds about right for LiFePO4. If you went with LiPo I bet you could get over 20kWh...
     
  18. Flying White Dutchman

    Flying White Dutchman Senior Member

    Joined:
    Dec 29, 2007
    4,374
    305
    0
    Vehicle:
    Other Non-Hybrid
    Model:
    N/A
    Dont think i want lipo in my car.
    Even more so a 20 kwh pack of danger....
    Lipo are a fire hazard

    -Tapatalk
     
  19. jdh2550

    jdh2550 Co-Founder, Current Motor Company

    Joined:
    Nov 16, 2010
    317
    246
    0
    Location:
    Ann Arbor
    Vehicle:
    2005 Prius
    Model:
    N/A
    There are automotive grade LiPo's from at least one reputable supplier (Kokam I believe) - there may be others. Any device storing 20kWh of energy be it gasoline or battery is a fire hazard. The trick with all of it is handling and monitoring it correctly
     
    [email protected] likes this.
  20. Flying White Dutchman

    Flying White Dutchman Senior Member

    Joined:
    Dec 29, 2007
    4,374
    305
    0
    Vehicle:
    Other Non-Hybrid
    Model:
    N/A
    Automotive grade?

    I read that any lipo when punctured will burst in flames where lifepo4 will not do this . Only smoke.

    -Tapatalk