Battery Vehicle Society

1) The tiny board about 2cm square is for the switch assembly. The small board about 5cm x 5cm is for the remote receiver. Again there is no parts list but a schematic has been published in these threads. The txd device is mentioned in same threads/posts. Do some searches

2) You can use the Master with a minimum of 2 slaves populated I believe. I never tried it with less than about 50
If you don't have any slaves connected the constant alarms will soon drive you mad.

3) Greg published a schematic for the slaves It's on one of these threads. The main thing to remember is that all the slave schematics are basically the same. The only thing that changed was i used a harder pull down on the slave bus input like 1K i believe, and a higher value R like 100k for the vref current source, this reduces slave current consumption.

Ok, making progress...

I've made up the little switch board, but without the RC network as despite multiple searches with various terms I can't find what value resistor to fit - I did find a reference to a 10nf capacitor though. This makes the button use a little "random" but I can at least get through the menus. I've populated and connected 2 slaves powered by a couple of AAA cells each and set the number of cells to 2, and it seems to be communicating quite happily. However... it's reporting a voltage of 1.82V on each, even though the meter says 3.35V or so

Any tips on how to diagnose the slave circuits - have I got the right slave software version (PicMultiSlaveV29_300510.hex)?

TIA

The switch resistors are 6 x 10k and 1 x 1M or you can use 6 x 1k and 1 x 100k for a bit more interference resistance. There are photos of the completed switch in the threads. Resistors should be 1% tolerance. The 1M/100k R pulls the switch up to +5v when no switch is pressed. I don't see hiow you can have the menus working without a completed switch attached. Perhaps the adc is picking up spurious noise and randomly activating them. Be careful with that as if you go into the update slaves menu and send random codes to the slaves you could change load/in out voltages etc without realising.

Once the switch is working correctly try sending commands 7 & 8 to see if you get the correct responses. I suggest also send command 6 & 13 once it is working. Then adjust as reqd.

'******* Slave Commands **********

' Command 01 = Send Cell Voltage on Master Bus
' Command 02 = Turn Off Loads
' Command 03 = Turn On Loads as Reqd

' Command 04 = Increase Load CutIn Voltage by 50mv
' Command 05 = Decrease Load CutIn Voltage by 50mv
' Command 06 = Set Slave Load CutIn Voltage Default

' Command 07 = Turn On Slave Load for 0.5 seconds (Flash Led)
' Command 08 = Report Slave Software Version Number

' Command 09 = Increase FailLow Cut Off Voltage by 50mv
' Command 10 = Decrease FailLow Cut Off Voltage by 50mv
' Command 11 = Increase FailHigh Cut in Voltage by 50mv
' Command 12 = Decrease FailHigh Cut in Voltage by 50mv
' Command 13 = Set FailHigh/Low Voltages to Default

Get the switch working correctly then report back.
I don't use the capacitor on the switch board personally.

By looking back through the software thread you should be able to find the latest version i believe it is "SLAVE_V09_070211_MPASM". You def need to be using one of the latest versions.

Have you made any changes to the master software?
If so post the code with each line you have changed followed by a comment ;hohisilver
Then I can search for that if reqd to find lines you have changed.

it's reporting a voltage of 1.82V on each, even though the meter says 3.35V or so

Just a crazy thought, but if the voltage reference ic were inserted back to front (not easy with to92 I guess) could that give you the dodgy readings you have?
(reverse voltage is 1.2, forward voltage at 15ma and 25celsius is around 0.8v)

Derek

As you were on the slave voltage problem - I reprogrammed all the chips with the software from BMSFiles090211Updated.zip, and I now get 3.32V and 3.36V

Switch works better with the 1M in place, but...

I now get "Watchdog Pulse" errors

I connected a 1K between watchdog 7 and master 9 as shown in the post near the one I got the zip file from - is there anything else I should have done to enable the watch 1S pulse?

I haven't done any software changes as I don't have PBP, this is all just to prove the hardware with existing software.

Has the watchdog chip got the 1S pulse software loaded? It's in that same zip file i think.

Note the 1S pulse goes into the master at pin 9 RA7 I wasted a whole day using the wrong pin.

Double check your connections. Try a lower value resistor say 330R

Right, I've dusted off the 'scope (not for the last time I'm sure!), and verified that there is no 1S pulse coming out of the watchdog. Looking at the source for the watchdog software I can't see any reference to a 1S output, so I'm concluding that the watchdog software in BMSFiles090211Updated.zip is not the latest, but I can't find an update to this

Correction. I've found it in the software now - that'll teach me to post before doing my homework

Guess the next thing is to disconnect the resistor and see if it starts toggling...

Does the watchdog chip alarm go off and reset the master if you leave the master in one of the menu options for a minute or so.

What is the watchdog code version you are using?

Try a different watchdog pic.

Solved

The watchdog PIC wasn't programmed correctly. It was warning me that it would be unreliable if MCLR was disabled while using the internal clock, which I foolishly ignored. Once I'd overridden the MCLR setting in the configuration it programmed correctly