16 Cell Multi Slave Board

Threads relating to the BMS system begun by Peter Perkins

Moderators: GregsGarage, retepsnikrep

User avatar
retepsnikrep
Posts: 1387
Joined: Sat May 26, 2007 4:50 pm
Location: North Yorkshire England
Contact:

16 Cell Multi Slave Board

Postby retepsnikrep » Thu Jun 24, 2010 5:56 am

This board is for any style of cell, prismatic, pouch or cylindrical in the range 5-200ah.

This board is currently in use by myself in my second lithium upgraded Honda insight which uses 48 x 8ah Lifebatt cyclindrical cells.

It was originally designed to use the Picaxe 08M chip, but can now be used with Picaxe 08M chips or plain 12F683 pic chips according to your taste. It has on board programming pins for the picaxe chips system but does not support ICSP for pic chips.

Hardware

Cells share sense voltage sense leads and the BMS board can be remote from the battery pack if required. Or if your pack is split into seperate compartments, each one can have a slave board poplulated with as many slaves as are required for that compartment.

It can be used in the digital configuration only.

The digital configuration requires a Master Pcb but gives full individual cell voltage reporting and other features.

16 Cell Slave Digital schematic http://www.solarvan.co.uk/bms/16cell311009.jpeg

Note the pic shown 12F675 is incorrect, but was used as that was the only compatible part available in the schematic program.


16 Cell Slave Digital board assembled http://www.solarvan.co.uk/bms/16CellSlaveHR.jpg

Software

The current software for the Picaxe 08M digital pcb layout is http://www.solarvan.co.uk/bms/DIGITAL_S ... 00BAUD.txt

The current software text file for the plain pic 12F683 digital pcb layout is http://www.solarvan.co.uk/bms/pbpsla06.txt
The current software hex file for the plain pic 12F683 digital pcb layout is http://www.solarvan.co.uk/bms/PBPSLA06.HEX

Note the above software in blue does not contain or utilise the slave command functions.


The current software text file for the plain pic 12F683 digital pcb layout with commands is http://www.solarvan.co.uk/bms/PicMultiS ... 300510.txt
The current software hex file for the plain pic 12F683 digital pcb layout with commands is http://www.solarvan.co.uk/bms/PicMultiS ... 300510.hex

Note the above software in red is also used with the single cell slave boards and contains the full command functions. This is the recommended version.
Regards Peter

Two MK1 Honda Insight's. One running 20ah A123 Lithium pack. One 8ah BetterBattery Nimh pack.
One HCH1 Civic Hybrid running 60ah A123 Lithium pack.

User avatar
dillond666
Posts: 139
Joined: Mon May 26, 2008 7:52 pm
Location: Central Scotland

Re: 16 Cell Multi Slave Board

Postby dillond666 » Tue Jan 11, 2011 8:05 pm

Peter,

I notice you are using a 330 ohm resistor on the masterbus opto (led side).

Assuming a 2v forward voltage for the led you could be getting only 3mA current if your cell was at 3v.
Are you finding this to be sufficient? I have been wondering whether the transistor side would be properly saturated
and how that would affect the data transmission. Would it just affect the rise time of the squarewave or could it prevent the
wave having the amplitude to switch logic levels on the master board?

Sorry to hit you with all the questions but my wee brain is buzzing and I've been reading too many electronics books, too
much to take in aaargggh! :wink:

Derek

User avatar
retepsnikrep
Posts: 1387
Joined: Sat May 26, 2007 4:50 pm
Location: North Yorkshire England
Contact:

Re: 16 Cell Multi Slave Board

Postby retepsnikrep » Wed Jan 12, 2011 5:37 am

No idea Derek i never calculated anything, just picked values that looked reasonable, feel free to change any you want. You need to keep the operating current for the slaves down to a minimum to avoid draining the cells.
Regards Peter

Two MK1 Honda Insight's. One running 20ah A123 Lithium pack. One 8ah BetterBattery Nimh pack.
One HCH1 Civic Hybrid running 60ah A123 Lithium pack.

GregsGarage
Posts: 870
Joined: Tue Apr 01, 2008 5:27 pm
Location: Galashiels, Scottish Borders
Contact:

New slave schematic

Postby GregsGarage » Mon Mar 21, 2011 11:27 pm

I am working on redrawing the slave schematic, fixing some problems with the old one such as adding an ICSP header and labeling the components correctly. Here is a first draft showing just one slave circuit. Any obvious problems?

EDIT: updated schematic 24 Mar 2011
Attachments
g6730.png
g6730.png (22.24 KiB) Viewed 36488 times
Greg Fordyce

Daewoo Matiz
http://www.evalbum.com/4191

User avatar
retepsnikrep
Posts: 1387
Joined: Sat May 26, 2007 4:50 pm
Location: North Yorkshire England
Contact:

Re: 16 Cell Multi Slave Board

Postby retepsnikrep » Tue Mar 22, 2011 5:56 am

I think we will have to look carefully at the icsp parameters i don't think that layout will work. The icsp signals have to be isolated from the rest of the circuit or the rest of the circuit has to be so designed that any loads it imposes are within the limits of the ICSP programmer driver. The MCLR pin especially is driven upto 12v when programming so that needs careful thought and probably pin realocation for that part of the slave circuit.
Regards Peter

Two MK1 Honda Insight's. One running 20ah A123 Lithium pack. One 8ah BetterBattery Nimh pack.
One HCH1 Civic Hybrid running 60ah A123 Lithium pack.

GregsGarage
Posts: 870
Joined: Tue Apr 01, 2008 5:27 pm
Location: Galashiels, Scottish Borders
Contact:

Re: 16 Cell Multi Slave Board

Postby GregsGarage » Tue Mar 22, 2011 10:57 pm

Peter, you are probably right, but I thought it would be worth trying to build the circuit and see if it would program. I should have enough spare bits to cobble something together on a bread board. It would be nice to be able to leave all the pin assignments the same, otherwise we end up with 2 different versions of software, which isn't the end of the world. I am not sure if I will get time this week to do this, it might be next week.
Greg Fordyce

Daewoo Matiz
http://www.evalbum.com/4191

steiner
Posts: 89
Joined: Mon Sep 08, 2008 1:50 am
Location: Florida

Re: 16 Cell Multi Slave Board

Postby steiner » Thu Mar 24, 2011 3:44 am

Were you considering using the 12F1822 since it has an internal voltage regulator. We should then be able to eliminate one component. This chip also has the ability for ICSP to be accomplished with either 5v or 12 volts depending on a setting. I was thinking we could use jumpers for the traces to the bms and then simply remove them when programing.

Rick

GregsGarage
Posts: 870
Joined: Tue Apr 01, 2008 5:27 pm
Location: Galashiels, Scottish Borders
Contact:

Re: 16 Cell Multi Slave Board

Postby GregsGarage » Thu Mar 24, 2011 11:42 am

Rick, I am going to stay with the 12f683 and Peter's basic lay out for now. Mainly to keep software changes to a minimum, never seem to have enough time to do everything I want to and don't want to have to completely rewrite software or come up with a complete new layout. I did eliminate one component though, the present design uses two 330R current limiting resistors for the slave and master optos, I have just one current limiting resistor for both optos. Anyone see any reason why that won't work? As a side note I spotted 2 errors on the schematic, the current limit resistor isn't actually connected and I have a junction that shouldn't be there at the bottom of the schematic, will correct that at some point.

It shouldn't be necessary to use jumpers for programming and I would want to avoid this. Ideally I would keep the pin outs the same on the pic, but if we have to change it then we do. The software changes would be minimal, just comment out the current pin allocations and add the new ones for boards with the ICSP header.

Here is what I want to achieve with the new boards, mainly smaller size. I have recently found out about http://iteadstudio.com/store/ . They have a pcb prototyping service that is very cheap. Up to 5cm by 5cm is just over £10 for 10 boards delivered to the UK. 10cm x 10cm is a bit more. So that got me thinking, how many cells could be squeezed onto those sizes of boards. I need a 4 cell board, so my goal is 4 cells on a 5x5 board. Rapid do a range of extruded aluminum cases that the boards can slide into which means we can mount components on both sides. So the idea is, smt on the bottom side and through hole on the top for the load resistor, filter caps and d-sub connector, etc. If I can make it all fit then 16 cells on a 10x10 should also work. Since the boards would only cost £10 for ten, its worth a shot. I have already placed one order from ITead for a Allegro current sensor board, it's been shipped, so will post pictures when I get it. Now where did I put that current sensor? :shock: :shock:
Greg Fordyce

Daewoo Matiz
http://www.evalbum.com/4191

User avatar
retepsnikrep
Posts: 1387
Joined: Sat May 26, 2007 4:50 pm
Location: North Yorkshire England
Contact:

Re: 16 Cell Multi Slave Board

Postby retepsnikrep » Thu Mar 24, 2011 4:54 pm

The ICSP pins on your connector are wrong.

'1) MCLR
'2) +5V
'3) GND
'4) Data
'5) Clock

I think the single 330R will work good spot.

I'm sure we will have to reasign pins. But if we could use the newer chip we could perhaps do away with the voltage ref and the 22k as well.

The whole thing really needs a redesign so it uses current loops for the master bus which is the weak point especially if it trails all round a car and a number of boards in different places.

But the newer chip with icsp and surface mount with less components means you could fit quite a few on a single board. I'm quite keen on the idea of a 16 or 62 cell slave board and having it made up in china!
Regards Peter

Two MK1 Honda Insight's. One running 20ah A123 Lithium pack. One 8ah BetterBattery Nimh pack.
One HCH1 Civic Hybrid running 60ah A123 Lithium pack.

GregsGarage
Posts: 870
Joined: Tue Apr 01, 2008 5:27 pm
Location: Galashiels, Scottish Borders
Contact:

Re: 16 Cell Multi Slave Board

Postby GregsGarage » Thu Mar 24, 2011 9:03 pm

retepsnikrep wrote:The ICSP pins on your connector are wrong.

'1) MCLR
'2) +5V
'3) GND
'4) Data
'5) Clock


I'll fix that.

I'm sure we will have to reasign pins. But if we could use the newer chip we could perhaps do away with the voltage ref and the 22k as well.


I'm starting to come around to the idea of the newer chip, but I just checked stock at Farnell and they have 179 in stock, and the 12f683, they have tens of thousands in stock. :shock: For the moment I might stick with the old chip.

The whole thing really needs a redesign so it uses current loops for the master bus which is the weak point especially if it trails all round a car and a number of boards in different places.


Peter can you explain current loops a bit more? How about we drive the master bus with 12 volts instead of 5. Should make it a bit more tolerant to noise and the slaves can stay as they are. Looking at the master V2 schematic (Master20082009) connect pin 1 of J11 to 12 volts instead of 5 and adjust the values for R27, R8 and R34 so the pic only sees 5 volts.

But the newer chip with icsp and surface mount with less components means you could fit quite a few on a single board. I'm quite keen on the idea of a 16 or 62 cell slave board and having it made up in china!

Not sure about a single 62 slave board, but the 16 cell slaves sounds great.

EDIT: schematic updated on previous post.
Greg Fordyce

Daewoo Matiz
http://www.evalbum.com/4191


Return to “BMS thread”

Who is online

Users browsing this forum: No registered users and 20 guests