Cheap LiFePO4 BMS?

[nicknorman]

1 hour ago, WotEver said:

So about 3A for <100ms?   
Nicked from allaboutcircuits.com is this:

 

Experiment with the value of C1 (30uF would probably be about right) and the relay shown does the switching to the bistable. I think that’d work ok, waddya reckon?

Im sure it would work, but all a bit 1970s! I think I'd rather use a PIC.

[MoominPapa]

1 hour ago, WotEver said:

So about 3A for <100ms?   
Nicked from allaboutcircuits.com is this:

 

Experiment with the value of C1 (30uF would probably be about right) and the relay shown does the switching to the bistable. I think that’d work ok, waddya reckon?

Yes, in theory, but there are actually two coils, one to open and one to close, so you need two RC-schmitt trigger differentators, one to make a pulse on the low-to-high transition for one coil and the other on the high-to-low transition for the other one, so it's actually simpler than that. I used R1=680Kohm and C1=0.1uF and it works fine.

 

MP.

Edited September 14, 2020 by MoominPapa
Not so smart I can avoid multiple posts.

[WotEver]

12 minutes ago, MoominPapa said:

there are actually two coils, one to open and one to close

Oh, I didn’t realise. Yes, as you say that makes it simpler still. 

22 minutes ago, nicknorman said:

Im sure it would work, but all a bit 1970s! I think I'd rather use a PIC.

Sledgehammer, nut

 

One CMOS chip, one resistor, one capacitor... hardly complex. 

[pgrange]

4 hours ago, MoominPapa said:

12AH per day is 10% of our typical on board power use, I'd be looking at ways to save that if I could.

 

MP.

Mmmm when you put it like that it is a lot! Are we talking about latching relays? 

Peter 

Edited September 14, 2020 by pgrange

[pgrange]

1 hour ago, WotEver said:

Oh, I didn’t realise. Yes, as you say that makes it simpler still. 

Sledgehammer, nut

 

One CMOS chip, one resistor, one capacitor... hardly complex. 

This is interesting. Forgive me I'm completey ignorant when it comes to electronics. Would the 12v my BMS provides to close relays if put through this circuit would it latch a relay closed? Similarly when the BMS cuts the 12v the relay would latch open. Does that make any sense? ?

[nicknorman]

3 hours ago, WotEver said:

 

One CMOS chip, one resistor, one capacitor... hardly complex. 

One 8 pin PIC, no resistor, no capacitor, less complex.

[WotEver]

7 hours ago, nicknorman said:

One 8 pin PIC, no resistor, no capacitor, less complex.

I need to investigate this modern magic of which you speak...

9 hours ago, pgrange said:

This is interesting. Forgive me I'm completey ignorant when it comes to electronics. Would the 12v my BMS provides to close relays if put through this circuit would it latch a relay closed? Similarly when the BMS cuts the 12v the relay would latch open. Does that make any sense? ?

Not quite that circuit no, because I hadn’t appreciated that the bistable relay has two separate coils. But something similar or, as Nick suggests, a PIC. 

[MoominPapa]

7 hours ago, nicknorman said:

One 8 pin PIC, no resistor, no capacitor, less complex.

It's amazing to me that modern microcontrollers now need no external components, except an in-circuit programming header, of course. And, stretching the definition, a programmer, and a computer to compile the program, and a program.

 

MP.

[nicknorman]

57 minutes ago, MoominPapa said:

It's amazing to me that modern microcontrollers now need no external components, except an in-circuit programming header, of course. And, stretching the definition, a programmer, and a computer to compile the program, and a program.

 

MP.

You don’t really need an in circuit programming header if you use a non-surface mount chip, you can programme off-board and subsequently solder or plug in the micro to the application board.

 

You do need the compiler (free) and the programmer (not free, but a 1-off purchase around £80) and a pc.

 

You do of course need a 3.3 - 5v supply and a FET or suchlike to boost the output (max 25mA or so) to drive the Tyco.
 

About 20 years ago I had a Subaru with central locking but not remote locking. I got a Maplin keyfob remote system which pulsed a relay when you pressed the keyfob. I made (on veroboard in those days) a circuit with 8 pin PIC and 3 relays directly driven by the PIC, one for lock, one for unlock, and one for the indicators. The code alternated between pulsing the lock and unlock relays, and flashed the indicators either once or twice according to lock or unlock. It worked flawlessly for about 8 years until I scrapped the car. Much, much simpler and cheaper than trying to do it with discrete components/logic gates/555 timers etc

 

 

Edited September 15, 2020 by nicknorman

[Richard10002]

1 hour ago, nicknorman said:

You don’t really need an in circuit programming header if you use a non-surface mount chip, you can programme off-board and subsequently solder or plug in the micro to the application board.

 

You do need the compiler (free) and the programmer (not free, but a 1-off purchase around £80) and a pc.

 

You do of course need a 3.3 - 5v supply and a FET or suchlike to boost the output (max 25mA or so) to drive the Tyco.
 

About 20 years ago I had a Subaru with central locking but not remote locking. I got a Maplin keyfob remote system which pulsed a relay when you pressed the keyfob. I made (on veroboard in those days) a circuit with 8 pin PIC and 3 relays directly driven by the PIC, one for lock, one for unlock, and one for the indicators. The code alternated between pulsing the lock and unlock relays, and flashed the indicators either once or twice according to lock or unlock. It worked flawlessly for about 8 years until I scrapped the car. Much, much simpler and cheaper than trying to do it with discrete components/logic gates/555 timers etc

 

 

I think I can speak for many here that almost all of the above could have been written in Klingon and it would have meant exactly the same.

 

I, for one, would know where to start with electronics stuff. Would it be relatively easy to get to the point of understanding, and being able to do the stuff in your first 3 sentences, or is it something that would take a whole series of lessons and experience?

[nicknorman]

7 minutes ago, Richard10002 said:

I think I can speak for many here that almost all of the above could have been written in Klingon and it would have meant exactly the same.

 

I, for one, would know where to start with electronics stuff. Would it be relatively easy to get to the point of understanding, and being able to do the stuff in your first 3 sentences, or is it something that would take a whole series of lessons and experience?

No, not for the electronically challenged unfortunately. The point was not really that such things are feasible for non-electronic people, just that for me it is easier to do that than to use the circuit posted by Wotever.
 

That said, lots of people get into Arduino programming. The entry point is relatively simple but it takes a lot of investment of time to become proficient especially if you have zero prior programming (coding) experience. It is fascinating though (in my opinion!)

[MoominPapa]

2 hours ago, nicknorman said:

You do of course need a 3.3 - 5v supply

In this application, that's a relevant factor.  To avoid catch-22 at start up, the circuit has to be directly supplied from  the batteries, (so 12-14.x volts) and the standby current matters. With a PIC you need a regulator (so parts-count  win goes back to the CMOS solution) and you have the standing current of the regulator. My CMOS latching relay driver, which is rather more complex that the circuit above, pulls 10s of microamps when not actually energising the relay coils. The power supply is a simple RC filter from the battery supply.

 

 

[Dr Bob]

47 minutes ago, Richard10002 said:

I think I can speak for many here that almost all of the above could have been written in Klingon and it would have meant exactly the same.

 

I, for one, would know where to start with electronics stuff. Would it be relatively easy to get to the point of understanding, and being able to do the stuff in your first 3 sentences, or is it something that would take a whole series of lessons and experience?

I'm with you Richard.

I'm ok with the circuits and putting them together but I just cant get my head around the programming. I used to do a lot of visual basic programming 20 years ago but stumbled when I got my first Rasp Pi and never really understood it. Arduino is the same. I just cant get my head around it. It was tempting to follow MP's route with his Arduino but after quite a few days looking at his programming, I retreated in defeat and went for the 'off the shelf' approach which needs none of this high techy stuff.

If I was 20 years younger then following the programming route would have been really interesting and satisfying as I was really hooked on visual basic programming but age takes its toll.

My limit now was to get my Rasp Pi recording all the power data on the boat which is a great resource with only very limited understanding of the Pi coding. I reckon my graphs are better than MP's

[nicknorman]

9 minutes ago, Dr Bob said:

I'm with you Richard.

I'm ok with the circuits and putting them together but I just cant get my head around the programming. I used to do a lot of visual basic programming 20 years ago but stumbled when I got my first Rasp Pi and never really understood it. Arduino is the same. I just cant get my head around it. It was tempting to follow MP's route with his Arduino but after quite a few days looking at his programming, I retreated in defeat and went for the 'off the shelf' approach which needs none of this high techy stuff.

If I was 20 years younger then following the programming route would have been really interesting and satisfying as I was really hooked on visual basic programming but age takes its toll.

My limit now was to get my Rasp Pi recording all the power data on the boat which is a great resource with only very limited understanding of the Pi coding. I reckon my graphs are better than MP's

I used to do a fair bit of VBA programming. At age 63 I learnt to programme in C. You are never to old to learn new tricks, it just takes longer!

[nicknorman]

22 minutes ago, MoominPapa said:

In this application, that's a relevant factor.  To avoid catch-22 at start up, the circuit has to be directly supplied from  the batteries, (so 12-14.x volts) and the standby current matters. With a PIC you need a regulator (so parts-count  win goes back to the CMOS solution) and you have the standing current of the regulator. My CMOS latching relay driver, which is rather more complex that the circuit above, pulls 10s of microamps when not actually energising the relay coils. The power supply is a simple RC filter from the battery supply.

Ah yes I’d forgotten that the CD4000 series can be run from a 12v supply. But on the standby current, the regulator im using for the battery monitor board uses 3microamps with no load, and a small pic put to sleep uses around 1 uA, so less than 5uA altogether, which is orders of magnitude less than self discharge and general leakage of a large Li battery.
 

But actually I’m not sure it has to be powered the battery side of the isolation relay. The board’s function is to disconnect the relay, the connecting bit can surely be done by a manual push button - at least in my case, where the disconnect is an emergency situation only, not part of routine battery management.

[nicknorman]

Duplicate

 

Edited September 23, 2020 by nicknorman

[nicknorman]

Eeek some of my Li batteries arrived today from China (the rest coming tomorrow hopefully), so I'd better get on with it...

For my own clarity of thought I drew out the system.

 

 

 

 

Quite simple superficially but behind the home made battery monitor and alternator controller is a fair bit of software with lots of data whizzing around over CANBUS. In summary the battery monitor keeps and eye on (and broadcasts on CANBUS) the individual cell voltages and battery temperature. In emergency, it can disconnect the battery using the Tyco bistable relay, but hopefully that will never happen! It controls the Combi when it is in charging mode to ensure no excessive cell voltage and make it switch to float when the SoC reaches the desired value. It drives the FETs / resistors for top balancing.

The Alternator controller adjusts charging voltage to ensure no cell over-voltage, reduces alternator output if it gets too hot, and switches to a float voltage at the target SoC (adjustable by a 3 position switch: 50%, 80%, 100%). Also has a slow charge switch to avoid unnecessarily fast charging when cruising for the day.

The Mastershunt of course broadcasts such data as SoC, battery current etc over CANBUS

 

Ooh look, no LA batteries in sight! Standing by for incoming...

[MoominPapa]

Looks great. No need for LA batteries. (I assume the Combi can function in powerpack mode without a battery in circuit, should the relay open.) How are you driving the balancing MOSFETs? I used opto-isolators to do the level translation to the gates of each MOSFET.

 

MP.

 

[nicknorman]

43 minutes ago, MoominPapa said:

Looks great. No need for LA batteries. (I assume the Combi can function in powerpack mode without a battery in circuit, should the relay open.) How are you driving the balancing MOSFETs? I used opto-isolators to do the level translation to the gates of each MOSFET.

 

MP.

 

I’m not sure if the Combi would work without any battery in circuit, but hopefully that relay will never open in anger. If it does there will have to be some “manual intervention” (eg temporary connection to starter battery) to get everything back up.

 

The MOSFET drive is part of the AD7280 chip, it has outputs that are referenced to the -ve of each cell so easy peasy, just need a 10k resistor in series with each gate drive.

 

Edited September 23, 2020 by nicknorman

[Tom and Bex]

On 14/09/2020 at 22:13, pgrange said:

This is interesting. Forgive me I'm completey ignorant when it comes to electronics. Would the 12v my BMS provides to close relays if put through this circuit would it latch a relay closed? Similarly when the BMS cuts the 12v the relay would latch open. Does that make any sense? ?

There is always  commercial REC BSLRD (Bi-stable latching relay driver) which I'm using for my overcharge protection.

https://www.solar4rvs.com.au/rec-bi-stable-relay-driver-bslrd-for-bottom-end-sw

And here:

https://www.off-grid-systems.de/REC-Touch-Display-BMS-4S-16S-Q_2

 

Alternatively another low powered solution is a motorised battery switch as used by me (for my loads) and Dr Bob. No complicated electronic circuits to build and understand, but maintaining the advantages of low parasitic power drain.

https://www.bepmarine.com/en/701-md

 

If starting again though, I'd be looking for a BMS with direct support for bi-stable relays. 

 

Tom

[Dr Bob]

8 hours ago, Tom and Bex said:

There is always  commercial REC BSLRD (Bi-stable latching relay driver) which I'm using for my overcharge protection.

https://www.solar4rvs.com.au/rec-bi-stable-relay-driver-bslrd-for-bottom-end-sw

And here:

https://www.off-grid-systems.de/REC-Touch-Display-BMS-4S-16S-Q_2

 

Alternatively another low powered solution is a motorised battery switch as used by me (for my loads) and Dr Bob. No complicated electronic circuits to build and understand, but maintaining the advantages of low parasitic power drain.

https://www.bepmarine.com/en/701-md

 

If starting again though, I'd be looking for a BMS with direct support for bi-stable relays. 

 

Tom

Fully agree with all Tom says.

My BEP motorized switch works very well. Everything is available off the shelf.

I too would be going for a BMS with direct support if starting from scratch as the BEP switch is pricey.

[redwing]

Finally got my 3 Valences in. I bought them from James at the end of Jan and they sat unmonitored  for over 7 months on the boat whilst I was at home.  

I bought theRS485 cable as recommended by Richard. I messed up the Tyco connector (the A and B data pins on a Valence diag are the wrong way round) so bodged the pins together and was able to read their voltages etc via Valence software. 

Voltages read 13.340(battery12) , 13.352(bat9)  and 13.231(bat23) which was in line with my voltmeter readings. The spread was between 6-8, 2-4 and 1-4 mV. The SOC readings seemed way out @ 95.69 , 21.18 and 36.00.

I charged them individually with Iskra 175 alternator via refurb Sterling 60a B to B on custom setting of bulk 13.8 and float 13.5 and monitored by BMV712. 

I let them settle and plugged them in to software again. Voltages read 13.382(12), 13.627(9) and 13.480(23) with spreads of 6, 38 and 18. The balance was active on battery 9 and then voltage came down to 13.480 with a spread of 18mV. The internal balancing on the Valences works when not on charge but connected to software. 

Paralleled em up and charged again via Sterling . They settled to 13.4 ish and have been down to 13.2 in use (parallel with 440aH LAs) and I haven't charged them via Sterling since. Solar (set at 13.8bulk 13.4 float) has had them back up to 13.4 by midday for last few days but we've had some sun and I'm not a heavy power user. 

 

I have a BEP motorised switch but not connected to BMV yet as I can't find a wiring diag. Any chance Bob or Tom could point me in direction of an idiot-proof source

 

All ok so far I think. Any comments (polite ones) and tips welcome. Thanks to all for your knowledge and experience to get me this far

[nicknorman]

11 hours ago, Tom and Bex said:

There is always  commercial REC BSLRD (Bi-stable latching relay driver) which I'm using for my overcharge protection.

https://www.solar4rvs.com.au/rec-bi-stable-relay-driver-bslrd-for-bottom-end-sw

And here:

https://www.off-grid-systems.de/REC-Touch-Display-BMS-4S-16S-Q_2

 

Alternatively another low powered solution is a motorised battery switch as used by me (for my loads) and Dr Bob. No complicated electronic circuits to build and understand, but maintaining the advantages of low parasitic power drain.

https://www.bepmarine.com/en/701-md

 

If starting again though, I'd be looking for a BMS with direct support for bi-stable relays. 

 

Tom

The BEP seems to take 12mA when switched off. Maybe not a lot but surely an aim should be to have as close to zero power consumption as possible, if the battery has to be switched off due to low SoC. 12mA for a month is 9Ah.

[Dr Bob]

1 hour ago, nicknorman said:

The BEP seems to take 12mA when switched off. Maybe not a lot but surely an aim should be to have as close to zero power consumption as possible, if the battery has to be switched off due to low SoC. 12mA for a month is 9Ah.

Life's too short to worry about 12mA. My TV/sky uses 9Ah in a few hours. If you have Li's then you dont worry about power nearly as much.

[Dr Bob]

1 hour ago, redwing said:

I have a BEP motorised switch but not connected to BMV yet as I can't find a wiring diag. Any chance Bob or Tom could point me in direction of an idiot-proof source

Here is the wiring diag.

 

the black wire from the plug (the one with green/red/black) goes to the relay connector on the BMV and then back to the black connector on the switch ( I think - I hope!!!). ie you are breaking the black line from the plug to switch when the relay on the BMV kicks in.

 

Its easy to test. Wire it all up then change the relay voltage setting to a lower voltage than currently present and the relay should operate.

I seem to remember on 'glitch' and that was to ensure that the main battery +ve connectors were in place - ie if you isolate the output line from the BEP, it wont operate. Not sure if that was just me!