Renard's BMW

[Renard]

Oh I'm so excited. I can hardly believe it!
The BMS worked first time. Oh my god, I didn't expect that.







The only hitch was wondering whether the bar on the tantalum cap signified pos or neg.

Thanks so much Neville.

(And apologies to Weber and Coulomb for not going with their -- I'm sure admirable -- system. Neville's is more at the level of my (in)competence.)

That first board took about two hours, but I hope to get more systematic.

[Johny]

Geez it's nice when things just go right. Well done.

[Renard]

Ten days on and I've just finished 120 modules. I had to take a day off half way through as I was going a bit crazy. Now, I just feel wacked.
I had ordered 120 boards and 120 micros, and of the 120, I managed to screw up only two, so that leaves me six spares over and above the 112 that I need.
The hardest part was trying to get the right temperature spots in the oven, as too cold, no melt, and too hot, the resistors tended to leap around like jumping beans. I can see that one needs an oven with a very even temperature distribution.

Including the master unit from Neville and the solder paste, it's cost me a bit under $7.50 per board, so yes it is pretty low cost.

With that huge task out of the way, I can now assemble the battery pack on the bench, for testing the whole pack, the chargers, the heater and the DC/DC converter.

[Stiive]

STRAYA.
Looks good - congrats.

How many boards can u fit in the oven at a time?

[Renard]

I could have put 18 to 20 boards on one tray, but in fact I did them about six at a time to vary the monotony of pasting and placing.

[woody]

Extreme DIY!

[4Springs]

Ten days on and I've just finished 120 modules.

Fantastic!
Are you going for top or bottom balancing?
I'm likely to attempt the same thing in a couple of months time, although a few less modules (probably 48). I have done a little SMD many years ago, but never with an oven, and had thought I'd just do it by hand. But I think you and Nevilleh have convinced me to look into buying an oven. I'll go and check out some second-hand shops...

If you have a list of where exactly all your components came from, and how you got the micros programmed etc., then I'd certainly appreciate you posting it. It would help me along on my attempt. Perhaps in the "Low cost BMS" thread?

[Renard]

OK 4 springs, comments on the low cost BMS thread.

[jonescg]

But I think you and Nevilleh have convinced me to look into buying an oven. I'll go and check out some second-hand shops...

Check out universities / research organisations for disused gas chromatography ovens. Usually the temperature programming on them is still highly accurate and the ramp times are very fast.

[4Springs]

Check out universities / research organisations for disused gas chromatography ovens. Usually the temperature programming on them is still highly accurate and the ramp times are very fast.

Now there's an idea!
We have GCs at work, and are just about to retire one. It still works fine, and would be perfect for the job. I'm sure I could convince them to let me have it for a $5 dontation to the social club. This will give me an added incentive to do the paperwork to retire it!

[Renard]

I'm just surfacing after the last ten days. We have been somewhat knocked about by the furious three day cyclonic storm, the power outage for four days, and the trees down on the lines and elsewhere. Then to cap it all, when I was installing comms wiring on the cell-top modules, I accidentally touched a wire somewhere and blew not only the cell top micro but also the master micro.
Oh my god, but Neville to the rescue…

To soothe myself I fitted the JLD404 amp-hour meter and the BMS master into the centre console.
The JLD404 just fitted after I had taken out the nasty plastic fake walnut base -- a regrettable lapse in BMW interior decorating taste -- and cut out material at the back of the frame. The BMS master sits oh so close, a millimetre to spare, beside it. So the amp-hour 'fuel gauge' and amp reading will sit alongside a voltage reading (and cell temperature range and cell voltage range) on the BMS Master.

The final photo shows how the console will look. The blue knob is attached to the regen. slider pot.









I also connected up all my cells and put the chargers onto them. I'm happy to say that the TC charger 2V to 5V control voltage exercises a pleasingly linear control on the current output from zero at 2V to full (5.3A) at 5V just as they claim in the manual. This is important for my charger control circuit which outputs a micro DAC control voltage according to the BMS master states.

[jonescg]

Great to hear about the TC chargers doing what they claim!

How did you generate the 2-5 V signal? External supply or did you use the +12 V-Gnd-Enable arrangements as per the manual?

Edit - OK your BMS supplies the signal. But I assume you put this between Enable and Gnd on the 7 pin plug?

[coulomb]

I guess 5 A is pretty gentle for those 100 Ah (?) cells, but one commandment we learned is this:

"Thou shalt clamp whenever thou chargeth".

No permanent damage seems to come from swelling, but it's just so creepy and it doesn't feel good for the expensive cells.

Impressive pack, BTW.

[Renard]

How did you generate the 2-5 V signal? External supply or did you use the +12 V-Gnd-Enable arrangements as per the manual?

Edit - OK your BMS supplies the signal. But I assume you put this between Enable and Gnd on the 7 pin plug?

For temporary purposes I used a bench power supply and checked its voltages with my Fluke meter. And yes, the voltage is supplied to the Enable (black) and ground (green) lead.

But in situ, the voltage is supplied by the charger control circuit (see my post 26th Dec. for details.) And ultimately the 12V for that circuit comes from the auxiliary battery via a relay whose coil is activated by the 12V from the TC charger.
I used a transistor in the coil circuit, but that would be unnecessary if the relay drew less than 50mA.

[Renard]

I guess 5 A is pretty gentle for those 100 Ah (?) cells, but one commandment we learned is this:

"Thou shalt clamp whenever thou chargeth".

No permanent damage seems to come from swelling, but it's just so creepy and it doesn't feel good for the expensive cells.

Impressive pack, BTW.

Yes, it's quite a pile isn't it? 366kg with celltop furniture.

I checked out both of my chargers, but used only one for the 5A charge. There was no noticeable warming or swelling of the cells -- yes they are 100Ahr. (That rate would be equivalent to 2A for your cells.) It doesn't seem as though clamping is necessary at this charge rate.

[antiscab]

Curiously, I just tested my TC Chargers running from DC.

At 130vdc in, they run at 75% of rated power at 93% efficiency.

fairly useful chargers they are

[jonescg]

Matt, I gather they put out their rated voltage, but with lower output current? Good for "jump-starting" a fellow EV'rs vehicle

[coulomb]

But in situ, the voltage is supplied by the charger control circuit (see my post 26th Dec. for details.) And ultimately the 12V for that circuit comes from the auxiliary battery ...

Oh. But isn't the control circuitry connected with the negative end of the charger output? In other words, aren't you thereby connecting pack negative to chassis?

A quick check would be to use a multimeter to check for continuity between the green control lead and charger minus. I think that the relay disconnects only charger plus.

I believe that at least high voltage packs are supposed to be floating, likely called for in NCOP14.

But perhaps you have an isolating linear opto or similar not shown in the 26th of December circuit?

[antiscab]

Matt, I gather they put out their rated voltage, but with lower output current? Good for "jump-starting" a fellow EV'rs vehicle

Thats correct - although I got the efficiency wrong, one of my cycle analysts I forgot had the volt guage maxed out (100V unit and 130v battery)

actual efficiency is more like:
21A x 126V in and 15.8A x 141.6V out or 84.5% efficient

I was using two chargers, both 120V (153.3v) 10A chargers. 1500W TC Charger

running from the mains, the two of them put out 19.8A

Not sure how low a voltage before those chargers stop working or brown out.....

Matt

[Renard]

Oh. But isn't the control circuitry connected with the negative end of the charger output? In other words, aren't you thereby connecting pack negative to chassis?

A quick check would be to use a multimeter to check for continuity between the green control lead and charger minus. I think that the relay disconnects only charger plus.

I believe that at least high voltage packs are supposed to be floating, likely called for in NCOP14.

But perhaps you have an isolating linear opto or similar not shown in the 26th of December circuit?

You've caught me out there. I was hoping to sneak this one past because that annoying charger-control-neg./charger-output-neg. connection would only apply whilst charging and not at any other time.
Yes I should put in a DC/DC converter across that charger 12V output when I use it to control the relay which activates the Contactor and BMS/Charger control circuits.
Edit: Oh um. Then I have trouble referencing my input to the ENABLE line. More thought needed. That 26th Dec. circuit needs to use the charger-control ground and be powered by the charger control 12V, whilst the Contactor circuit and its relay should be the other side of the DC/DC converter.

[coulomb]

I was hoping to sneak this one past because that annoying charger-control-neg./charger-output-neg. connection would only apply whilst charging and not at any other time.

Unless both sides of the charger are disconnected when not charging, I don't think that's true. I don't know if you switch the negative output of the charger, or if the charger does it internally (I suspect the charger disconnects only the positive output).

So it might not be so easy to sneek it past. Besides, I'm sure you would want safety while charging, and a linear opto, while more expensive than standard optos, is still only about 6 dollars. Example:
http://au.element14.com/vishay/il300/op ... dp/1045385

[ Edit: oops, I see that this one only has an output voltage of 500 mV, so some gain would be required after it. That's a nuisance, so maybe choose a better one; it was the first one I found. ]

[jonescg]

Couldn't you just use the charger's own +12V to do the DC charge current manipulating? The BMS could simply activate relays / transistors to dial back the power... These would be isolated from the 7 pin plug, which happens to be hooked up to the pack negative?

[Renard]

Unless both sides of the charger are disconnected when not charging, I don't think that's true. I don't know if you switch the negative output of the charger, or if the charger does it internally (I suspect the charger disconnects only the positive output).

My charger(s) will be sitting on a garage bench almost always, so they will be physically disconnected from the car when not charging.
Still, I agree that electrical isolation is very desirable.

So I will re-arrange things so that the BMS charger control circuit is wholly powered by the charger's 0V and 12V, and that this also controls a relay (which operates the contactor circuit) via an isolating device. Then of course I have also to have isolators for the actual BMS master switching outputs -- those which tell the charger to throttle down to 0.5A etc.

[Renard]

So I will re-arrange things so that the BMS charger control circuit is wholly powered by the charger's 0V and 12V, and that this also controls a relay (which operates the contactor circuit) via an isolating device. Then of course I have also to have isolators for the actual BMS master switching outputs -- those which tell the charger to throttle down to 0.5A etc.

At last I have done this.
The TCCH charger's 12V powers the circuit which Enables the charger, and also powers the relay which activates the contactor and BMS circuits (whose power comes from the auxiliary battery.) This activation of the relay is via a 6N139 optocoupler.
The Charger's 12V output is kept well below the 50mA limit that TCCH specifies.
The two signals from the BMS master which determine the charger's output are isolated by being fed into a HCPL2731 dual optocoupler, thereby keeping the two circuits entirely separate. The Grounds too are also kept separate.

[Renard]

Something I haven't reported yet, but made some time ago, is the gear select mechanism. I realised that the front ashtray made an excellent spot for the controls. (The two little screws in the lower RH corner are to hold the ashtray cover fixed.)



The buttons connect to a little circuit which is mostly just a PIC micro. The software is set up so that neutral is the default, one push of green engages 'forward', two pushes of red engage 'reverse', and any button push whilst in a gear takes one back to neutral. The circuit connects to the Driver Controls Unit.

The gear state indicator LED in between the momentary-push buttons is tri-state, red for reverse, blue for neutral and green for forward. It is controlled by the Driver Controls Unit.

The other recent job is just a small re-working. I found that the electronic pressure gauge I had planned to use was a bit delicate for the purpose of indicating low vacuum, and so I bought a more robust vacuum switch through KDRyan.
This is marked 3"HG, which, when translated into civilised terminology, means it closes when the pressure is above 90kPa, and opens below 90kPa (or when there is less than 10kPa of 'vacuum'.)

The reservoir is made from 100mm Ø PVC pipe with end-caps. The Wasco vacuum switch for the purpose of indicating low vacuum, as required by NCOP14 §1.3 (ADR 31 \& 35), is on the right, and behind it is the activating switch for the pump. There is a little circuit designed to keep the pump running for about three seconds after the switch opens, consisting of a capacitor, diode and comparator.



The Wasco switch is on the right.