Multiple Subpack Balancer

[retepsnikrep]

All good ideas Mike but i'm going to persevere with the on the bench battery and individual 6 cell subpack approach.

I've started my Master software and done a bit more on the schematic.

With regards to the software I'm after some ideas on how that should function, not the code but the general principles/design.

The hardware allows access to one subpack at a time which can either be charged or discharged at up to a max of 10amps (Assuming I use 10amp relays)

The charge/discharge current is controlled by 2 x opto isolated PWM controlled mosfets at about 5khz. the mark/space ratio can be varied to control the current from 0-100%

The voltage will be constantly sensed and I'm hoping for a resolution of about 10mv over a 10 volt range.

The pack ptc sensor will also be monitored to detect a single cell exceeding relevant temps.

I'll count AH in and out of course.

So I think software will do a general self check and cycle each relay in turn to check pack connections and give some initial data and an overall pack voltage.

If everything is in order operator presses the Start button and comes back in a week!

The system then starts discharging the first stick at 1C until about 1v per cell is reached.

It then charges that stick at 1C until say 7Ah counted in, or voltage reaches X? or temp indicates an issue.

It repeats above say 4 times.

It then moves through pack repeating above and logging cycle data to the SD card.

After a full cycle of the entire pack it looks at results and then gives a couple/few extra cycles to any subpacks showing weakness/reduced capacity.

It then tops up all subpacks using a lower charge current say 500ma and waits 24hrs before performing the top up again and noteing any high self discharge rates or obvious deviation amongst the packs.

Any heat issues detected by ptc strips lead to the IMA fan being operated.

Any errors or problems lead to main relay power being cut.

Thoughts?

[Mike Dabrowski 2000]

All sounds good.
Be careful with your relay choice, as breaking a DC circuit under load can cause arcing, and contact burning. Just cut the current back with the PWM before switching and your relay choice will get easier.

[IamIan]

I hope it goes well... I am watching with interest.

[retepsnikrep]

Originally Posted by Mike Dabrowski 2000

All sounds good.
Be careful with your relay choice, as breaking a DC circuit under load can cause arcing, and contact burning. Just cut the current back with the PWM before switching and your relay choice will get easier.

Of course all relay switching will be at zero load, so I'm not looking for a relay with a huge dc voltage capabaility.

[retepsnikrep]

Still tinkering, schematic is mostly done now, might finish that tomorrow.

Mike, your opto isolated Mima voltage conditioning circuit, can that be supplied on the input side from a varying supply voltage? i.e. The subpack on test, or does it need a stable supply against which to measure the subpack voltage.

I could use a mini 12-12v dc-dc converter to supply it.

[Mike Dabrowski 2000]

Peter,
In the MIMA system, I run the HCNR201 from the TPS and MAP sensors regulated power supply, and on the current sensor from the +12-12 regulated sensor supplies.
I have not looked at the effect of varying supply, but in theory, the circuit should work fine with a supply that is unregulated, as it is basically a current servo.
I suggest you explore the data sheet and suggested circuits to make your own determination, or just breadboard the circuit and test it.
The data sheets are here:
High-Linearity Analog Optocouplers
Application PDF showing the many ways the part can be used:
http://www.avagotech.com/docs/5989-0804EN

[retepsnikrep]

I haven't forgotten about this as I have two packs under my bench that need refurbing and I am just remembering what a pain it is with only one smart charger!!!

I can see why Ron has half a dozen.

Anyway to make the cost and complexity a bit more manageable I'm dropping the capability down to seven subpacks at a time. So basically it does a row at a time.

I'll go for 1C charge/Discharge and 7 relays as that keeps the cost down a fair bit. The pcb will also be a lot simpler. A single picaxe 28X1 with a unl2003A relay driver and 7 x 10amp DPCO minature relays and the connectors. Because the volatge will now be kept down to around 50 volts I'm not bothering with any adc isolation and the unit will be powered by an isolated laptop 12v brick of some sort with a 7 amp capability.

I'm bining off the on board data logging but going to work on sending the data directly to a PC windows laptop which will sit next to the unit and recieve serial cell data direvctly into a spreadsheet/text file or something.

That's the plan anyway.

[jonnyvtec]

Good luck with that Pete, Im interested to se ehow it develops.

Im getting a recal once a fortnight on a monday morning, usually after a 3 day rest weekend.

After yesterdays recal in the morning and going home and back to work (120miles) the battery still hasnt got back up past 3/4

We had a technology fair here at work (JLR) and the chap from MIRA was showing me his cell buddies on the Jag XJ Limo green REHEV car. It had a 300V 20Ah LiFe Batt unit in the boot.

Basically on the balance charge the 3.6v cells all have a bridging PCB that opens the circuit when they see cell voltage rise above 3.75v where the current is then passed through some resistors to junk the energy as heat rather than charge that cell, the rest of the pack continues to see the current until you get the full pack up to 3.75v per cell. It was only 120mA though, there upgrading to a 500mA setup. I guess NiMH doesnt really lend itself to that, nor does the construction of the insight pack. Interesting non the less though.

[retepsnikrep]

OK

Here is my new schematic for the simplified 7 subpack balancer. It's not quite finished need to spec a few connectors etc.

So you can do a whole battery in three sessions instead of 20 with a single charger.

This version will output data straight to an excel speadsheet on a windows machine/laptop running alongside it. It will use the standard Picaxe usb/programming cable for the downloads/uploads/data logging.

I have one spare input/output left on the pic any ideas? Anything I've forgotten.

It has 5 simple buttons on the pcb for the basic functions, and the capability to accept a ps2 computer keyboard for data input/control. It also has the capability to drive a simple two line 16x2 lcd display for basic visual info.

Any comments on the values choosen? Especially for the darlington charge/dicharge transistors and associated drivers?

I will be specifying fairly cheap 8A-10A DPCO relays and the idea will be to use 1C charge discharge.

I'll probably be ordering 5 pcb's to start with anyone interested let me know ASAP.

The software is not written yet but I have started it. I'm away after next thursday so will get the pcb's ordered whilst I'm on hols should be here when I get back in mid June. The software will recognise the fact it needs to log the data for 20 cells so you get a complete picture of the cell data.

The unit will require a well regulated 12v 10A isolated supply.

The discharge load can be one of those 12v 200W car cup/kettle heaters/elements in a pail of water!

Comments asap please before I order it!!

http://www.solarvan.co.uk/insight/Su...20509_7Sub.jpg

Diagram is 600kb

[Hybrid-Battery-Repair]

Originally Posted by retepsnikrep

OK

Here is my new schematic for the simplified 7 subpack balancer. It's not quite finished need to spec a few connectors etc.

Peter, given that the pack is 20 sticks, can you come up with a circuit for a 5 charger (4 sticks = 30V each?) system? (just a charger, not a balancer) I want to start wiring in the connections on the relay board side and bring out a (safe) connector with 6 charging leads. I'd like to wire in a main relay disconnect and a fan connect as well.

Plugging in the connector would disable the main relays and turn on the fan.

Actually, this should have been in the grid charger thread.