Finally bought one

Hi,

Using the Leaf modules it would be simple to make one series string with (for example) 224S 4P (120AH). Using the Volt sub-packs the closest you could easily come to that would be to use two series strings of 24s 3P each. How much more complicated will it be to build a pack with multiple series strings (the OP is planning 2 series strings of 48S 3P).
Click to expand...
This Batt-Bridge Battery Out-Of-Balance Alarm seems like an easy way to to check to monitor two equal series strings for a bad cell:
http://www.sunrise-ev.com/LeesEVs.htm#battbridge
An EV's pack consists of many cells or batteries. In theory, they are all identical. In practice, they aren't. There will always be a "weak link" somewhere in the pack. That's the cell that limits your range, and limits how much you can charge before damage begins.

But it is difficult to know if you have a weak cell. Total pack voltage won't tell you until too late. The amount of circuitry needed to monitor every single cell can get very complicated and expensive!

The Batt-Bridge is a quick-n-dirty "idiot light" to give you a good/bad warning when any cell in the pack goes undervoltage (dead) or overvoltage (overcharged). It works with all types of batteries; lead-acid, lithium, nicad, or nimh. If it lights up red, back off on the current until it goes out. If it stays on even at zero current, stop driving or charging until you find the problem and fix it!


How it works: The Batt-Bridge divides the pack in half, and compares the voltage of each half. When the two halves are equal (within 1 volt or less), a green LED is lit. If a cell goes dead or begins overcharging somewhere in the pack, its voltage typically changes by more than a volt. This imbalance lights one of two bright red LEDs to tell you which half of the pack is low.

D1 is a standard brightness green LED. D2 and D3 should be ultrabright red LEDs for best visibility. R1 and R2 should be identical resistors, chosen to provide about 10ma at your pack voltage. The current sets the sensitivity and brightness of the LEDs.

Construction: Mount the LEDs in a pilot light holder. Install it on the dashboard where the driver can easily see it, and where it won't be washed out by direct sunlight. Mount the resistors at the battery terminal ends of the wires, so they will limit the current in case of shorts.

Batt-Bridge parts kit -- $10. Includes LEDs, resistors, lamp holder, and detailed instructions. Specify pack voltage. $5 for US shipping and handling. For shipping outside the USA, email for details.
Batt-Bridge, Assembled -- $20. Assembled Batt-Bridge in panel mounted lampholder, with resistors and detailed instructions. Specify pack voltage. $5 for US shipping and handling. For shipping outside the USA, email for details.
Click to expand...
 
Hi,

yopappamon said:
I'm hoping between my self (industrial controls engineer) and my brother (retired Delphi Automotive electronics engineer) we can reverse engineer the CAN interface to the battery manager.
Click to expand...

yopappamon said:
I'm thinking no on the cooling since I'm not going to use huge amps compared to the Volt. I'll monitor the temperature and see it it's necessary.
Click to expand...
Some questions:
1. When charging/discharging are the cell level HVC and LVC controlled by the BMS hardware or controlled by the CANBUS software? It seems likely to me that this is controlled by the BMS, whcich means if you can live with the reduced capacity you are home free.

2. How is the temperature control implemented and what happens if it gets too low or high? If the BMS hardware only monitors the temperature (instead of shutting down the pack discharge if the temperature gets a few degrees too cold or hot), and you can read them it might be ideal?
 
Got the controller programmed for the higher voltage and was in the road at 2pm! Drove about 24 miles on a partial charge, Woho! Now need to reprogram the charger and start thunking about a BMS.

20140726_150131.jpg
 
Great news! Congratulations :mrgreen:
 
The module dimensions are 11" high and 9.5" wide.

The 30 cell module is 23" long without the cooling tubes.
The 24 cell module is 18.25" long without cooling tubes.
The combined 24 and 18 cell module is 33" long.
The combined module has the cooling tubes in the middle so it's wider with them, not longer like the others
 
Congratulations!

I'd love to see pics of your installation !

-JD
 
Hi,
http://www.diyelectriccar.com/forum...chevy-volt-drivetrain-p437818.html#post437818

Re: Attempting to hack a Chevy Volt Drivetrain
Originally Posted by kerrymann
Sounds good to me. Want to work on one for a volt? I am still trying to get a hold of a wiring diagram and hopefully it is all canbus but we will see. I have been busy in the barn dealing with permit issues upgrading my barn with 100amp service but that is another story.

Looks like Jack posted parts of my videos on EVTV but there wasn't much lead in. I was hoping that would help get people interested in cracking the wiring of the inverter but I think he has more projects then can handle as it is.
The Volt may or may not have multiple can buses. I'm told that the Leaf has three of them. The GEVCU hardware is intrinsically capable of connecting to two buses at once. So, what's left is to figure out how many buses the Volt uses and where they are. Almost certainly at least one exists and is on the OBDII connector. So, that's a start.
Click to expand...

I'm actually working on making an Arduino IDE capable canbus to USB dongle. It will essentially be a modified copy of the GEVCU hardware but without the hardware I/O and significantly cheaper. My initial plan is to have it under $100 and still be galvanically isolated and dual bus. So, that gives another option for monitoring. The bonus with that idea is that you don't need to repurpose a big ECU for reverse engineering. Also, the little dongle will have a DB9 connector and multiple plugs will be possible. This will allow for using it on a Leaf with the proper plug or a Volt with the proper plug, etc.

Yes, Jack has a lot of projects and it seems like he's backed off of the vehicle reverse engineering a bit but I haven't so we're still a go. Without his support the process might be a bit slower but it'll get figured out. And, if such things start producing fruit it'll make it more enticing for him as well.
Click to expand...
 
Thanks for that link, Mitch. I didn't realize that the Volt packs are 7 groups of 48v and 2 groups of 24v. I've never taken the tab covers off. This makes things a little more interesting when it comes to my group placements. I can rearrange groups to suit the space.

This video was in that thread and shows full disassembly of the pack down to the small groups.

[youtube]00tTckGUv7I[/youtube]
 
I could do this and eliminate the need for another wire front to back. Then use the left over 12 cell group for a nice long distance bike battery. That would put my voltage around 168v, closer to the original car voltage.



Or put the 2 - 12 cell groups up front and everything else in the back and run 192v
 
Hi,

Would you mind posting the height and width (bottom and top) of the sub-packs/cell groups?

yopappamon said:
That would put my voltage around 168v, closer to the original car voltage. Or put the 2 - 12 cell groups up front and everything else in the back and run 192v.
Click to expand...
Unless it causes a major problem I don't see any advantage to the 168V system. Too much range, or power or speed?
yopappamon said:
Any thoughts on a battery manager?

Would the Fechter Goodrum Zypher be a good choice? Putting all the cells in means 96 cells to monitor, could get pricey quick. Others have used the Elithion Lithiumate LIte.
Click to expand...
IMO the Fechter Goodrum BMS would be an excellent choice because it only controls HVC and LVC and leaves the current regulation to the controller, where it belongs. But you will probably need to build it yourself. You probably need to design/have your own boards built because I have only seen 24S boards.

PM from Fechter:
fechter said:
My BMS circuits probably won't handle that many cells without some kind of modification. I don't have time to build those either. I have had pretty good luck with the units from Bestech Power, but those would also need some kind of modification or interface to work with a big pack like that.
Click to expand...
If you do this and want to sell one or two assembled 24S BMS's please let me know.

If possible IMO if you can figure out how to use the OEM BMS, that might be the best solution (depending on how well you get it working). You need HVC for charging and LVC for discharging. When you divide the modules can you keep all the cells with their BMS chips? If not it's probably not useable?

But otherwise you could charge the pack after charging one parallel group a little high and monitor the cell voltages. Does the OEM BMS keep it in range (it should)?

Discharge is more likely to involve the CAN bus, but you could discharge the pack after discharging one parallel group a little low and if it cuts the power before the group gets low enough to get damaged. If it does not you could try the same thing, while monitoring the bus. It should be relatively easy to figure out.

According to their site you can only use the Elithion Lite with cells with bolt-on terminals (Thundershi*, etc.) and can not use it with packs with more than one serial string.
 
you could take the output and charging mosfet HVC and LVC signals off of each of the upper BMS for a series of 24S D131 in series up to the top of your available pouches.

then you could use that to control a stack of the D131 power boards (physically separated by cutting the connecting pins, from the balancing boards stacked in series up the battery) on the bottom BMS all controlled by the same HVC or LVC signal line so you could then have monitoring of all channels and then instead of leaving the power board section on that BMS above the bottom 24S D131, you could wire up all the power boards in parallel to handle the current. i have a total of 4 of the D131 between the two batteries in the ZENN car. a pair of 21S lipo D131 for the lipo pack, and a pair of 24S lifepo4 D131 for the lifepo4 section of the battery.

so between the two batteries, four BMS and a total of 480A output current capacity. 60 mosfets in parallel on the output.

you can use the brass standoffs between the balancing and the power board to hold the power boards together, and maybe something similar for the balancing boards.
 
Hi dnmun,
dnmun said:
you could take the output and charging mosfet HVC and LVC signals off of each of the upper BMS for a series of 24S D131 in series up to the top of your available pouches.
Click to expand...
I have no idea of what you are talking about doing (for example what is a D131?). Would you please explain what you are describing?

Will this work with the op's pack (either two strings of 48S or two strings of 42S)?

My current decision on my ZENN is whether to buy LEAF modules (either one string of 24S-2P 60AH or one string of 24S-4P 120AH) or Volt modules (two strings of 24S-3P - 45AH per string). In your opinion will using two strings of Volt modules be harder to implement than one string of LEAF modules?

OP: How hard would it be to connect the cell modules of two modules in parallel (make a pack of one string 48S-6P, or 24S-6P)?
 
Hi Mitch and dnmun,

I wasn't clear what the D131 was either. Is that the fetcher Goodrum BMS? If I get what you are saying, I could just use the cell balancing circuits of that BMS and cut off the charging side. I think that would work for me.

Mitch: I posted the dimensions above, is that what you needed?

So the Volt cells are overall grouped as:

30 cells . Sub grouped within as 12, 12, and 6
24 cells. Sub grouped as 12, and 12
42 cells. Sub grouped as 24 and 18. Sub sub grouped as 12 and 12. And 12 and 6.

So with some new threaded rod and banding, you can use any of combo of 7x12 and 2x6. I'll post a pic later.
 
Group one. You can see the banding around the top, the jumpers between sub groups, and 4 rods on the bottom are all that holds them together.

20140802_140253.jpg


Group two:

20140802_140300.jpg


Group three.

20140802_140309.jpg


So it would be quite easy to regroup the cells to what you are looking for.
 
HCX-D131 from bestechpower. if you have 42 cells in series then you could use 21 channels on each of two 24S lipo D131 in series. or you could use 24S on one and 18S on the other one.

or you can use two in parallel for the lower one and two in parallel for the upper one to get up to 350mA of balancing current.

then if you consolidated the power boards of all 4 that would give you 480A max or about 320A continuous power.

the trick is to be able to carry the HVC and LVC signal from the upper BMS down to the lower BMS. since they are limited in how much voltage can be applied to the LVC and HVC signal line string of resistors since it could blow the controlling transistors on the bottom BMS that would control the output and charging mosfets. i think that could be solved.
 
Hi,

Don't know how much they cost, please look they will do everything we need:
http://www.orionbms.com/

Their support for multiple strings, in parallel sounds excellent.

Too expensive, $1600 for 96 Cells.
 
Hi,

yopappamon said:
I think I can do 96 cells with the Fechter Goodrum BMS for about $600. The quantity discounts on Mouser make it cheaper to buy 200 of a part than to buy 50.
Click to expand...
PM from Fechter:
fechter said:
My BMS circuits probably won't handle that many cells without some kind of modification.
Click to expand...
Do you know enough to modify the design as required?

How hard would it be to change the 6S and 12S 3P subpacks to 3s and 6S 6P configurations?

The 3P cell groups are arranged like this?:
+________-
-________+
+________-
-________+

It should be easy to flip every other group like this?:
+________-
+________-
-________+
-________+

Which might make it easier?

If you did that you could use 48 of the minibms modules (about $13 each, total $625) or their 4S boards (cost is about the same), with their head unit (about $48).
minibms.jpg

They even have a cable (which looks like it will work with 6P).
http://www.diyelectriccar.com/forums/showthread.php/minibms-product-update-87776p2.html
we can program miniBMS cell boards for LiNMC cells from Leaf or Volt, you just need to state such request in order notes, when placing an order at our Web store.

We typically program HVC=4.2V , Shunt=4.1V and LVC=3.1V for such cells. Some customers also asked for LVC=3.3V to leave deeper reserve, although it may cause premature LVC alarms under heavy load, so adjustment of LVC delay on the head board may be necessary.

We also have 4S version of cell boards, which can be placed near cells, instead of individual boards on top of cells. This option is good for cells with tab terminals like Volt's and even for Leaf's cells since they have closely spaced terminals with center tap in 2S block, so placement of 2 distributed boards on each block can be difficult.

Here is a picture of the BMS wire harness available for Volt modules to be used with BMS systems available from Mini BMS, for LiMn.
Click to expand...
attachment.php


A problem with the Fechter Goodrum BMS, is there has never been enough demand to justify the production of finished boards. That problem would be mitigated with a cell level BMS (you want 48 or 96, I want 24 or 48) similar to the Mini BMS. It would be nice ES effort to develop something similar and do a group purchase. There is some open source information available on the minibms:
http://www.diyelectriccar.com/forums/showthread.php/open-source-bms-55227.html
Open Source BMS
Some time ago there was a thread about "open source BMS" this led to what is now a proprietary BMS offered by Clean Power Auto. An issue with nearly all proprietary stuff is that secrecy masks deficiencies I'm not saying there are any deficiencies in the MiniBMS being marketed today. Is there interest here in continuing the open source development, right through a group buy? If so we can restart the activity. I have taken the last posted schematic as a conceptual start and fixed the many problems I found in it. I have also laid out a PCB. I chose a DC rated fuse rated at 300 volts and added a large TVS to guarantee no fires. I added hysteresis to the HVC and tightened tolerances on it and the onset of shunting. I changed the inductive shunt resistor to an array of SMD resistors to avoid long term issues with the shunt transistor. The design I've done is focused on CALB batteries. It assures that HVC of all cell modules is 3.6volts +/- 1%, that shunting begins at 3.46 volts +/-1%, that any cell that reaches HVC must fall about 70 millivolts to reset the HVC status. LVC is temperature sensitive and 2.5 volts +/-2% at zero degrees C and 2.7 volts +/-2% at 50 degrees C and once LVC is set the cell voltage must rise about 100 millivolts to reset it. The cell module is 1.225" x 1.875". The current loop conects to the cell module through " 250 Faston " connectors.
Click to expand...
 
Hi,

fechter said:
Zenid has been making small runs of one of my board designs and selling them.
see: http://endless-sphere.com/forums/viewtopic.php?f=14&t=26554&p=928377#p928377

These are really targeted to bike/motorcycle size packs. These are unpopulated bare boards made for through hole components.

To get any kind of real volume, the boards would need to be fully built. This would be far better if done with surface mount parts, which would require a full redesign of the board layout. It also makes a batch of boards more expensive and mistakes on the layout are more expensive.

I'm sort of inclined to use off-the-shelf boards from Bestech Power and make a simple interface board that allows them to work with EV sized setups. This is way less work, but not fully proven yet.
Click to expand...
 
:) Hi,

fechter said:
Zenid has been making small runs of one of my board designs and selling them.
see: http://endless-sphere.com/forums/viewtopic.php?f=14&t=26554&p=928377#p928377

These are really targeted to bike/motorcycle size packs. These are unpopulated bare boards made for through hole components.

To get any kind of real volume, the boards would need to be fully built. This would be far better if done with surface mount parts, which would require a full redesign of the board layout. It also makes a batch of boards more expensive and mistakes on the layout are more expensive.

I'm sort of inclined to use off-the-shelf boards from Bestech Power and make a simple interface board that allows them to work with EV sized setups. This is way less work, but not fully proven yet.
Click to expand...
 
Mitch, the 6s and 12s x3p groups are as small as you can go without cutting apart the welded tabs. You could jumper each 3p set between two groups of 12s to get 12a x6p.

My intention with the Zypher BMS is to not use the charger circuits, but only the cell HV and LV portion. Then run a warning buzzer for LV while running. And make my own charger cut off. My main concern is keeping the cells balanced.

I have a mini BMS for Lifepo4 and I don't want a cell module to mount. With the Zypher I can wire off the existing BMS connectors.

I also like that the Zypher has a balance current of 1 amp.
 
You must log in or register to reply here.

Similar threads

A
Sevcon DVT Regen setting
Replies
1
Views
157
arf8
A
A
Upgrading to a Baserunner with CA3, feasibility questions
Replies
6
Views
254
E-HP
E-HP
branson.cz
alternative battery options
Replies
9
Views
400
branson.cz
branson.cz
rfndzc
Honda CB350F rebuild, belt drive and other upgrades
2
Replies
27
Views
1,480
rfndzc
rfndzc
X
  • Question
Troubleshoot! Need help! Turnigy !
Replies
0
Views
158
Xprotojay1993
X
Back
Top