Battery Vehicle Society

Sounds OK to try.

What may be better though is an extra layer of opto isolation. I used a little DC-DC and added another opto right next to the master bus input on master board. The dc-dc powered the slave master bus optos which drove the final opto right next to the master. That cured my problem. So in effect the master bus is now very short not snaking all round the pack.

The resilient simple opto circuit can be driven quite hard with say 12v to help add more immunity. Juts need a dc-dc, one current limit resistor (adjustable if you like to get stable operation with minimum power consumption) and one opto

The 12 volt master bus idea didn't work out, comms where actually worse, so I have put the master bus back to the normal layout although VR2 and VR3 haven't yet been refitted. I have found out what is going on though. It is my front 9 cell slave board. It looks like interference from the controller is causing one or more of the master bus optos to start conducting, taking out the master bus. D9 on the master lights continuously as I activate the controller. removing that slave board from the system and all is o.k. The rest of the slave boards are in the back of the car away from the motor and controller, they all work fine.

Got myself a decent o-scope today and had a look at the master bus with it. first picture is with the controller off and second picture is with the controller spinning the motor. You can see the switching pulses from the controller as 2 volt spikes at 16khz. This is with the master running on an isolated supply (separate battery ). The master is fine with this level of noise and dtill reports cell voltages properly. When powering it from the same 12 volt supply as the controller the noise gets worse and the data is lost in the noise. So my next steps are to earth the cable shielding to the chassis, move the front slave box inside the car and get an isolated 12 dc-dc and see how I get on.

Hi Gregg,

I'm hoping you are going to do these one at a time to see how much relative difference each option makes to the noise. I am still bench testing my system but getting close to buddying it up on top of the PakTrakr I have already installed but which suffers badly from noise - and also not being terribly suited to Li cells (in my opinion!).

Look forward to seeing more from you - those traces speak a thousand words.

Regards, Martin.

Try my simple master bus extra optocoupler trick with a pp3 battery for testing if you need too and a 330R resistor.

Martin, I used a paktrakr in my Fiat, connected to every 4th cell. It gave an indication of what was going on, but the biggest problem is that the paktrakr draws power from the pack in an uneven way. My first 4 cells where always a bit lower than the rest because of this. For this reason I wouldn't use one on a lithium pack, but for lead acid they are great. Shame they haven't come out with a better solution for lithium.

Peter, I like your idea of an isolated master bus, will give it a go. The noise doesn't seem to be affecting anything else with the bms, so maybe this is the way to go.

Is this the circuit for the master bus isolator?

Yes. When the master bus optos conduct they turn on the new led which is in the opto right next to the master input

Job sorted.

You can in fact use a pp3 3 x AA cells if you don't want a DC-DC as it is only being drained when system is transmitting data so current is very low and it is auto off LOL

Could be a candidate for the ACME "ElCheapo" fibre link
Probably not much difference in price from a dc-dc converter.

http://www.dillond1.pwp.blueyonder.co.uk/cheapo1.JPG
http://www.dillond1.pwp.blueyonder.co.uk/cheapo2.JPG

Cheapo link uses the black cable which is a TosLink cable with the ends cut off.
Vishay TEPT4400 ambient light receiver
Ebay ultrabright red LED with clear optics
Heatshrink

Derek

Lets see how he gets on with the opto, but yes derek your system will prob work as well.

Cheap dc-dc here http://cgi.ebay.co.uk/NME1212S-12v-DC-D ... 0585013572