[etischer]

I spent 3 days wiring up my BMS! Here are some pics

I have 36 distributed BMS boards, and 63 cells connected to the centralized BMS. I'm really happy with the centralized BMS, Dimitri did a nice job keeping it compact. Such a great feature being able to stack them, and have the serial link pre-wired. I did a little write up on my blog http://etischer.com/awdev/minibms.html

 

[TheSGC]

Ya know, your trunk wiring looks exactly like the network closets at work. It's so neat it's scary. You can give most of these auto manufacture's a lesson or two in properly securing wires.

 

[Pirape]

Man nice work! I just cant believe how neat it is

Can you describe how everything is connected? I am making the plans for my EV and it seems so complicated

Thanks and again nice work!!!

 

[mmark666]

Eric,

nice job! Why did you mix the individual boards with the centralized version? And aren't you afraid of overheating your stack of centralized BMS boards? If I counted correctly, there are 64 5W resistors tightly stacked together. So theoretically you could dissipate around 300W there...

But again, it looks very neat!

Markus

 

[etischer]

The resistors are only dumping 2w each, even though they are rated 5w. With the plastic mounts I machined to hold the boards together, I now have an air tight seal on the sides. It would be easy to hook up a fan to blow air across the group of the resistors, if it is even needed.

I plan to keep my charging voltage where it has been, and only raise it high enough to shunt once a month or so when I want to balance. I got the BMS for the peace of mind, it's nice to know my charger will shut off if any cell decides to shoot for the moon.


Oh yea, I chose distributed for the front pack since that was the easiest to mount. The rear cells are mounted up under the car where the gas tank was, in the spare tire well, and behind the rear seat. Cell mounted BMS boards wasn't feasible, especially in the rain! So central BMS was the best option.

Eric,

nice job! Why did you mix the individual boards with the centralized version? And aren't you afraid of overheating your stack of centralized BMS boards? If I counted correctly, there are 64 5W resistors tightly stacked together. So theoretically you could dissipate around 300W there...

But again, it looks very neat!

Markus

 

[etischer]

You are correct, a fan is absolutely needed. I hit 200F shunting only 3.44V in a matter of 10 minutes. I quickly added a fan, temperature dropped to 100F. Ambient temp is 70F. I measured 0.6v across the 4.7 ohm resistor, so a mere 600mw.

There is no cooling at all so the temperature just keeps rising. I don't think these resistors were designed to be packed so close, they don't get the natural convection they normally would. The resistors on the distributed board had a voltage drop of 1.5v and they were only 90 deg f.

Eric,

nice job! Why did you mix the individual boards with the centralized version? And aren't you afraid of overheating your stack of centralized BMS boards? If I counted correctly, there are 64 5W resistors tightly stacked together. So theoretically you could dissipate around 300W there...

But again, it looks very neat!

Markus

 

[Dave Koller]

Beautiful job Eric!! ............

 

[Crash]

Nicely done! Looks very clean. And like mentioned before, it looks like a networking closet.

Good job.

 

[tomofreno]

Nice work Eric. I am curious as to why you chose to use an ssr to shut off AC power to the charger rather than using the minibms relay to short pin1 to pin2 on the regbus to shut off the charging current? Consider the latter (edit: whoops, should have been "former") more foolproof?

Are you still using your timer to turn on the charger at night? If so, I would like to know how it works with the minibms. I took out the optocouplers on 4 cell level boards and the comparator chip on the main board. Never have determined the cause. Dimitri said they are sensitive to voltage spikes. After replacing them it is working fine. I hadn't check operation of the minibms for several days so don't know exactly when it occurred. I noticed it right after I used the timer to turn on the charger, so immediately suspected the timer of causing a voltage spike. Could just be a red herring, likely will never know, but I haven't been able to get myself to try it again. - don't want to have to desolder/resolder that comparator chip again. If yours works fine, I may try it again. Our power goes out several times a year here, and my ups for my pc beeps occasionally, so it could have been just a voltage spike while charging I guess. If so, and it happens while charging at night, I wouldn't catch it from overcharging, though I could limit the overcharge by setting the timer for slightly longer than the expected charging time.

 

[etischer]

On Elithions website, they had a comment that the manzanita occasionally would not turn off using the rj45 connector. I think the SSR provides an added layer of redundancy, even if the charger is going crazy, the SSR is a stand alone system that can kill power.


I am using a timer to charge starting at midnight. It has dry contacts to turn the power on. The charger is set to just be finishing CV when I wake up. This allows me to check cell voltages each morning if I choose. I don't know how the charger would cause voltage spikes. My charger is quite noisy to begin with, I'll have to upload a screen shot of my scope to see if you guys see the same thing.

Nice work Eric. I am curious as to why you chose to use an ssr to shut off AC power to the charger rather than using the minibms relay to short pin1 to pin2 on the regbus to shut off the charging current? Consider the latter more foolproof?

Are you still using your timer to turn on the charger at night? If so, I would like to know how it works with the minibms. I took out the optocouplers on 4 cell level boards and the comparator chip on the main board. Never have determined the cause. Dimitri said they are sensitive to voltage spikes. After replacing them it is working fine. I hadn't check operation of the minibms for several days so don't know exactly when it occurred. I noticed it right after I used the timer to turn on the charger, so immediately suspected the timer of causing a voltage spike. Could just be a red herring, likely will never know, but I haven't been able to get myself to try it again. - don't want to have to desolder/resolder that comparator chip again. If yours works fine, I may try it again. Our power goes out several times a year here, and my ups for my pc beeps occasionally, so it could have been just a voltage spike while charging I guess. If so, and it happens while charging at night, I wouldn't catch it from overcharging, though I could limit the overcharge by setting the timer for slightly longer than the expected charging time.

 

[JRP3]

So you're turning on the charger with the current turned up? To what level? I thought the recommendation was to turn on the charger with the current at 0 and then turn it up.

 

[etischer]

I usually charge at 8A, 230v over night.
At work I usually charge 16A, 115v.

So you're turning on the charger with the current turned up? To what level? I thought the recommendation was to turn on the charger with the current at 0 and then turn it up.

 

[JRP3]

I see. Probably not an issue at such a low current.

 

[etischer]

What is the issue at high current?

I see. Probably not an issue at such a low current.

 

[JRP3]

Supposedly setting the Manzanita's to a high current output and then switching them on can cause damage, though I have not found a definitive confirmation of this. I should probably just email Rich and get the facts.