Old BMS General Thread

[qdos]

Estimated range is good that works for me How about a means of inputting desired range and a Red Light Green Light kind of thing according to how much juice you're using or regenerating. It's easier to do in a petrol as there's feedback in the feel of the engine but not so sure with an EV, naff all motor noise and bugger all vibration I suppose I ought to look at an ammeter ? But if my vario on my hang glider can work out speed to fly at different sink rates and figure out the wind speed then all this should be possible really afterall the ground doesn't move much. Well not for some of us anyway LOL

[retepsnikrep]

How about a means of inputting desired range and a Red Light Green Light kind of thing according to how much juice you're using or regenerating.

That sounds a bit naff to me You are welcome to do it though, that's the beauty of open source

[qdos]

LOL Pushing a heavy EV home the last 3 miles is pretty naff to me

[retepsnikrep]

Touche Mon Amie!!

I don't have the pushing the dead EV home issue.

[qdos]

LOL You will have it shortly I hope. I'm waiting on a member of the committee to get their act together on an different article first though

[retepsnikrep]

Just some late night thoughts.

Fortuitously with a 28 to 1 chance I seem to have selected the right input pin on the Master Picaxe 28X1 chip for use with the master data bus which allows a background serial receive function Being woolly brained I don't remember selecting this pin deliberately due to this function, but perhaps subconsiously I did thinking it might be useful in the future

Anyway once set in motion this receives data from the slaves without user or program intervention (program execution continues) and stores it in the scratchpad 128 byte ram. I think I should be able to use this to increase the overall Master program speed, as at the moment i'm using a simple for/next loop and standard serial receive command. This multitasking may prove very useful later.

As a test today I am also increasing the Slave and Master Serial speed to 2400 baud from the present 1200 baud. If the data proves reliable I may go the whole hog and push the boat out to 4800, Woo I hear you cry, well we shall see 2400 baud is in the middle of the available speeds so seems a sensible setting, and it will ensure BMS in curent form meets my target of scanning 50 cells a second. It's just outside that at the moment

I might need to change R8 on the Master board to 4k7 (from 10k) more in line with TTL levels to ensure adequate pull down of the Master Bus I had an earlier issue with that on the slave comms which amused one or two

Any other software functions people would like to see on the Master?

Thanks to those who have already replied to this. No promise I will include any of them but you've probably got good ideas. I've got a few weeks left as well before install day

I'm looking at a better menu button system using one of the spare Master ADC inputs and a resistor matrix to allow half a dozen discrete buttons on one input. Two just isn't enough

[retepsnikrep]

Master and Slaves running at 2400 baud data comms for 12hrs,
no errors so I'll leave it at that for the moment as it has improved response to within my target specs.

www.solarvan.co.uk/bms/SLAVEDIGITAL0710 ... 00BAUD.txt

www.solarvan.co.uk/bms/MASTER071008_V37_2400BAUD.txt

Must try the background serial mode as well.

[retepsnikrep]

Roger Crier another UK Insight owner has asked about improved mph accuracy perhaps to 1/10 of a mile per hour. Sadly having looked at this now I think it's beyond me, the capabilities of the Insight VSS, and the Picaxe 16bit integer only maths.

It's not ultimate mph accuracy Roger is after but subtle changes in speed not reflected by a normal speedometer, these changes do affect the Insight when striving for ultimate mpg on a long journey. If speed starts to bleed off unoticed then more dramatic changes are reqd in fuel/throttle usage to counteract it than if it had been corrected earlier. Nerdy, yes, but worth a look

The Insight VSS provides 76 pulses per second at 60mph. It is not able to resolve part revolutions unless the time taken between pulses, or the length of pulses could be measured. Hey hold on! Why don't I try that.

At the moment the Master software simply counts the pulses in 250ms and multiplies it by 4 to get pulses per second.

This takes 250ms for a start if I measured the time between/length of a/pulses it would take only a fraction of that and perhaps give the accuracy we seek.

The Picaxe "pulsin" command can time pulses at 8mhz to within 5us and with a 0.32768s timeout period. I'll try experimenting with that using the sneaky pulse generating software I did a couple of days ago.

So I take that back at the begining I might be able to get more accuracy

Hey Roger do some digging/testing on the Insight VSS matey.

[retepsnikrep]

Going to added the extended menu button stuff today using one of the spare Master ADC inputs this will also free up Input 0 & Input 1 on the 28X1 chip for other uses.

www.solarvan.co.uk/bms/ButtonEncoder.jpg


Edit

OK, didnt get above done but did improve Master display refresh time by utilising the SV2000 video chip's ability to save the current screen to flash eprom and restore it later Now the generic display appears instantly and is then updated in the background with the current values, sneaky

Did look at the more accurate speedo issue and the new timing of the VSS pulses works well, for my 60mph 76 pulse per second test input the pulses were timed at 1318 units (Each unit is 5us so that equals 6.59ms) for the pulse length at 60mph.

The "pulsin" in command gives me the value 1318 for 60mph so could someone have a go at the calculation now to get the speed in mph from that, remember at 30mph the value would be 2636 so you can't just divide it Also you can only use 16bit integer maths! Matt where are you

[Roger Crier]

Taking a leaf out of the Autospeed $20 fuel ratio article, rather than making the speedo more accurate would it not make sense to use a led set to show if you are accelerating, staying at a steady state, or decelerating. Perhaps this would be done by comparing the present reading with the previous one and checking by how much it has changed so that the rate of change can be calculated. this could be added below the Autospeed readout etc and then you would just need to get every led readout showing green!