Old BMS General Thread

[qdos]

That's a pretty fast vehicle Greg's building then

[retepsnikrep]

It was as easy to do 0-255 as it was to do 0-40 which may be a more accurate assesment of an EV's speed capability I might be able to test it to 110mph + given the right track/autobahn

[GregsGarage]

That's a pretty fast vehicle Greg's building then

I only require 2 digits for the speedo function, even displaying speed in kph.

Top speed is around 50 mph, 55 with a tail wind.

I am picturing the mph on the top 3 lines with more info below, say SOC, batt temp, lowest cell voltage, etc. But it is looking good, and could be used for other variables as well to suit other requirements.

[retepsnikrep]

After a Major rewrite of the Master display code I've now got the big numbers mph display sorted I think.

I've gone back to the original plan of having a display sub routine which collates all the BMS information, and then draws the screen in one go according to your specs. It operates in two modes at present.

1) Normal with all the info displayed standard font size. (I'm still tinkering with the layout of this)

2) Greg's Big 3x3 Numbers with mph displayed at centre top and a reduced selection of the other info on the lower lines.

The bottom line is reserved for error messages in both modes.

I'm looking at trying to scroll the bottom line in case of multiple errors occuring. Scrolling is not easy either.

Still need to tinker with BigNumber code to improve speed but it works.

I'm uploading a demo to youtube as we speak.

http://www.youtube.com/watch?v=o-nsbhmOyts

The much changed latest Master code is here

www.solarvan.co.uk/bms/MASTER271008_V59_2400BAUD.TXT

[GregsGarage]

Greg. Welcome back, can we have a sitrep with your setup when you get a chance. Have the cells died whilst you have been away etc!!!

Just a quick update with my setup. All seem to be well. I still want to adjust the slave software upper voltage cutoff by lowering all the values about 50 mv. At the moment the balancing seems to get to around 100mv per cell. lowering the cutoffs should get this closer with the charger I am using, which is designed for lead acid batteries. I just need to remove the under tray and open all the battery boxes to reprogram the boards, haven't had the time.

Was away on holiday and the car sat for 10 days. E-meter read around 15ah lost during that time which had to have been through the precharge resistor across the main contactor. I will change this so it is fed from my battery isolator switch.

Yesterday while going up a hill, about a mile from home, 45mph, foot to the floor, 3rd gear, 200 amps and the pack having nearly 70 of its rated 100ah drained I got a low voltage warning. This is set if any cells reads below 2.4 volts. Backed of a little on the throttle and the warning went off.
Of course I didn't really notice any significant drop in speed, so it just shows that with a little careful driving you can improve range without hurting performance much. Remainder of journey never showed the warning again. But at least I now that works as its been tested in anger and doesn't seem to suffer from the electrical noise generated by a motor and controller working at full throttle.

[retepsnikrep]

We can tick that off the list now Low voltage Tested OK!

Just a safety note on the Slave boards.

The Picaxe programming pins on the Slave boards are not isolated in any way.

So if you are programming a series string of cells, the pins will be at whatever voltage that point is in the battery in relation to the negative/positive terminals.

I don't recommend programming the slaves in situ, but I do it using the one hand technique and extreme care. I know Greg does as well.

No holding the pins and the negative/chassis etc please

I suggest use an old battery powered plastic cased laptop with a suitably insulated serial connector avoiding actual contact with the terminals/pins is best if you want to do this.

I'm sure most people reading this are fully aware of the danger of high voltage systems anyway.

[GregsGarage]

Also, if programming in-situ you can disconnect the pack from the vehicle and remove several link bars/cables so that you break the pack up into smaller lower voltage blocks.

Edit: Peter, just had a look at the latest video of the display and it is looking good. The only thing I might change would be to display the odometer on the bottom line as long as there are no faults, then replace it with a error message as required.

I now need to get back to working on my shopping list for the master board and order the bits as well as a suitable display.

[retepsnikrep]

Edit: Peter, just had a look at the latest video of the display and it is looking good. The only thing I might change would be to display the odometer on the bottom line as long as there are no faults, then replace it with a error message as required.

Hmm I'm still thinking about the display, and i'm probably going to break it down further into several screens worth you can cycle through. I don't much like my abbreviations for all the data "Vo" for `Pack Voltage` as an example

I would prefer a complete line for each item, which I had many versions ago. I think I'll use button three to go to a display selection sub menu where you can switch between the available display configurations.

1) Might be as is now compact single screen with abbreviations

2) Might be the same data expanded over two or more screens which rotate round at 2 second intervals.

3) Might be a single screen with your Big numbers and the selected abbreviated data underneath.

I will force the display negative (Black text on white screen) if an error occurs to help draw your immediate attention to it.

[GregsGarage]

Another update on the batteries. First frosty morning, really had to do some serious scraping. My Paktrakr read the temp as 32f, so a good test to see how the batteries behave cold. I have a hill to climb not far from home, so 3rd gear, right foot to the floor, pack already 25% discharged from last nights journey, 175 amps at around 42mph and the Packtrakr was showing around 10.6 volts per battery (it thinks its connected to 6x 12v batteries), so around 2.65 volts per cell and no low voltage warning.

A full charged pack will get me over the hill at around 47mph, I normally see around 45mph with my pack at 25%DOD. So the cold weather is having only a minor effect. The rest of my journey was a couple of mph slower due to the slightly lower voltage of the pack.

[retepsnikrep]

Code tinkering in the warm today Cells are at about 3C in my garage.

Error routine now turns screen negative to highlight issue and cycles between normal and error data whilst error exists.

Improved normal BMS display IMO and got rid of nominal message and some *please do not swear* abreviations

BMS has been running on bench for over a month now connected to battery packs, with me uploading different versions of Master code for testing most days. No unforseen errors and cells are all within 40mv of each other according to the bms, which is about the limit of it's resolution. Spot checks with dmm put them bit closer than that but BMS can't resolve it.

I'm letting the cells capacity drop down a bit before I fit them to the car as the Insight will try and force charge them straight away to work out the SOC for it's oem system. You can't really prevent this so I need a couple of ah headroom for the install in December.