Thanks for the advice. I am back on nimh power for the time being and the car is back together as I need it for work.
I need to do a fair bit more work on the bench, and I'm fairly convinced a lot of the noise is RF so I need a screened aluminum case for the Master board or I need to move it outside the ipu compartment. I intend to revisit the whole layout and add some screened cable in appropriate places as well as some of the RC filters. I intend using an isolated power supply as well. So lot's to do but I don't want to spend all Christmas getting annoyed in the garage when I could be lounging around eating, drinking and watching TV.
What have I learnt so far?
1) Insight rear compartment is a HF noise nightmare.
2) Insight OEM BCM accepted cells without throwing any errors, assist/regen worked normally on my short test trip.
3) Digital slaves work OK even in the noisy environment.
4) Master worked for short periods but disprupted/reset by all the noise, when it was working for short periods It appeared OK.
5) No I haven't lost heart or binned the project, just got to sort my noise
I would have to research it to find the part no. Its been a while since I worked with them, but one that would work is in the 74XX series of chips...if you can use a 5V power buss. I dont know the exact number. I am leaving town for the hollidays this morning. As soon as I get back I will see if I can dig out an old data sheet.
__________________
Jim Isbell
2000, 5 speed, 250,000 miles
"If you are not living on the edge, well then,
you are just taking up too much space."
Peter,
If you trigger on the noise, and set the timebase so we can see the characteristics of a single noise pulse,I believe that the rise and fall times are orders of magnitude faster than your serial data. A simple single pole or double pole RC or RL filter should be able to attenuate the noise while not effecting the much slower serial data.
Also have you looked on the input side of your opto isolator on the slave cards to see if the noise is there as well as on the output?
If the micro is seeing the pulses, the excess voltage is going to pass the noise spikes to the slave board power supply via the protection diodes on each i/o pin, which can cause all i/o pins to see it.
Lowering the buss impedance may also help.
Wish I could go out and do some poking around with a scope on my car, but we got nearly a foot on new snow that I have to deal with and an unheated garage.
Good luck.
Thanks for all advice. i'm relaxing a bit as now. I'll tinker as time permits.
Another option is to fall back to analogue mode while I'm tinkering, as that would at least let me build up the experience with the lithium cells and get them in and running
Luckily my BMS digital slave boards can be programmed to output a simple on/off over V on the Master bus when a cell goes over Abs Max V, and although not quite as good as analogue slaves for the under V mode they can still pass an under V situation message amongst themselves and out of the last board again as simple on/off signal. Both these signals are opto isolated and could be used to drive a simple dash warning led/audible alarm. The over V signal can also control the charger via a simple relay, as soon as the first cell hits max charge will be cut. I can monitor battery voltage and current with a couple of simple analogue meters. I'm very used to using a simple voltage meter as a fuel gauge from my years of driving with Li-ion and pb in my EV's!
I know it's not ideal but it gets it up and running with fairly minimal effort and I might be able to manage that over the holidays. We will also find out about how far the Insight will go with the extra charge and how it interacts with the bcm in more detail. All useful stuff. The Master is the icing on the cake and can come on stream as things progress with screening etc.
As you know, I was thinking about how to monitor the IMA pack for some time now, and was thinking of a different way to scan the pack and bring the data to a common master board.
I was going to keep it analog, including the opto isolators like I did on the MIMA card.Each cell monitor board would have A good voltage refrence, an op amp or two,and one of the HCNR201 analog opto isolators.
The opto output of each cell is fed through a shielded wire to the master board.
An analog multiplexer on the master board could address each cell voltage, and send it into a single analog input of the micro. Since the cell signals are very low frequency,filtering out the HF noise will be easy. The rest of the analog I/O could be used for temperature, current and the other required functions.
The only part that I had not solved was the cell balancing.
Just food for thought.
They basically now just output a simple high or low signal on the Master & Slave buses when a cell goes over or under permitted voltages.
The Master bus (High V) is paralleled and is easy, but the Slave Bus (Low V) signal is cascaded through the chips.
Sorry to Greg on the BVS forum who built his slaves in Analogue configuration, you could perhaps have built them in digital config and just used this software for the analogue control. It only needs the addition of one final opto on the last Slaves output Oh well live and learn.
I'll use this to allow the install to proceed with a simple visual/audible alarm. The Master bus which has become the Over V bus can also be still used to control a simple relay to control the mains charger, I can do a manual balancing charge if I need to when I see how well the cells keep in balance whilst I'm using this analogue functionality. The slave boards still retain their on board 250ma balancing load capability and 10x voltage over sampling, I'll run them at the standard 4mhz. I'll just add an analogue voltmeter to allow me to monitor the pack voltage and I'll try the install again.
Obviously I'll work on the Master and the screening/filtering as reqd.
Peter,
If you trigger on the noise, and set the timebase so we can see the characteristics of a single noise pulse,I believe that the rise and fall times are orders of magnitude faster than your serial data.
Correct, you can see that in the scope pic I posted earlier.
Looking at the spikes in my scope pic earlier anyone good at maths want to give me a few values for the low pass RC filter components. I know the formulae but it's all double dutch equations to me.
The AutoGuide.com network consists of the largest network of enthusiast-owned enthusiast-operated automotive communities.
AutoGuide.com provides the latest car reviews, auto show coverage, new car prices, and automotive news. The AutoGuide network operates more than 100 automotive forums where our users consult peers for shopping information and advice, and share opinions as a community.
Re: Phev Insight
