Even Newer 4 to 24-cell Battery Management System (BMS)

[Malcolm]

24 series is about the safe limit for the onboard voltage regulator circuit

Aha. I hadn't realised this. I read through the thread and got the impression that the upper voltage limit was higher. I'm planning a 108V 60Ah pack for a small car. The pack will be split physically into three 36V sub-packs, each with its own isolated 36V/30A charger. I was hoping to use one board populated for 12 cells with each sub-pack. Would this work without making modifications to the boards (aside from adding a heat sink to the FET). I guess what I'm asking is would each BMS only see 36V or would it see the entire pack voltage? I'd be grateful for any advice.

 

[GGoodrum]

Aha. I hadn't realised this. I read through the thread and got the impression that the upper voltage limit was higher. I'm planning a 108V 60Ah pack for a small car. The pack will be split physically into three 36V sub-packs, each with its own isolated 36V/30A charger. I was hoping to use one board populated for 12 cells with each sub-pack. Would this work without making modifications to the boards (aside from adding a heat sink to the FET). I guess what I'm asking is would each BMS only see 36V or would it see the entire pack voltage? I'd be grateful for any advice.

I don't think there is a problem with a higher upper limit, but what Richard is saying is that if you are going higher than 24 channels/72V, you just need to have a small external 12V supply that can handle the full pack voltage.

With three separate 36V chargers, you can simply use three boards, each with 12-channels populated. Each board will work with each charger. You don't need a heatsink for the FET at all. I've yet to see one get too much above room temperature, even with a really dead pack. Depending on how much current your charger puts out, you might need to parallel a couple of FETs, but that's only if you are talking about something ridiculous, like 50A+, I think. Actually, I'm not sure about what you'd need because once the shunt/throttling logic kicks in, the current is going to be down close to an amp, or so.

If you want to increase the shunt bypass current above the 1/2A that the current setup will handle, you will probably need to to use some clip-on TO-92 heatsinks for the shunt transistors. These are already the long version of the TO-92 package, but they do get a bit toasty (190-200F) with the 6.8 ohm/5W shunt resistors we are using now. They can certainly handle more current, but not without a heatsink, I think.

I don't have any direct experience with how closely balanced the big TS cells will stay, but even if there is a 5% imbalance, that's like 10Ah, for a 200Ah pack. At 1/2A, it would take about 20 hours to fully balance charge the pack. A 5% difference, however, is a lot. With a123s, LiFeBatt/PSI/BMIs, and probably most of the Headway and Ping packs, I think healthy cells will stay within 1-2% of each other, once balanced completely at the start.

-- Gary

 

[fechter]

Right, if you split the pack into 36v chunks and use a separate charger for each, you will not have a problem. For 108v, it would be safer to use the separate 12v supply, but I think a single master control circuit will be well within the ratings for the FET.

 

[Malcolm]

Many thanks for explaining that, and thanks again to you both for all the effort you've put into this.

 

[ldoolit]

Hi -

I just spent a couple of evenings reading the big BMS threads from this past year. There's a lot of good material in there, even if my eyes glazed over before I finished.

Sorry my first post has questions instead of answers.

Are more of the famous Goodrum/Fechter kits in the pipeline?
http://www.tppacks.com/proddetail.asp?prod=EBKE-BMS-Kit
If so, when are they likely to ship? Although I have some design sketches for a fancier BMS system, I also need to get my motorcycle back on the road, and charging safely, with its new string of 16 TS-LFP40 cells. Hand-balancing is a pain.

Many switchmode power supplies of the kind suggested for charging a string of cells use an optically coupled feedback circuit to stabilize the voltage. For a simple example, see the sample circuit for the MAX5021 and MAX5022 at
http://www.maxim-ic.com/quick_view2.cfm/qv_pk/2998
Has anybody tried hacking into such a supply with their own feedback signal from a BMS? That would avoid the need for yet another high-current series element (the IRFB4110 in the Goodrum/Fechter design) to turn the current down (or offentirely) at the end of charging.

This looks like a good community. Even if I can't spend much time with it, I hope to contribute something.

- Larry

 

[fechter]

Yes, more boards are in the pipeline. See above post from Gary (back a page).

Yes, I had considered tapping into the PWM on the charger to eliminate the need for the big FET. There would be no limit to the bulk charging current this way.
If you can find the spot in the charger, you could still do this with our board by tapping the control circuit in the right spot. I was hoping I could use the temperature probe input on a Zivan charger like this, but testing by Patrick indicates this won't work. You'd need to tap into the optocoupler inside the charger. With some chargers, an interface circuit might be necessary.

It just made things a lot easier to use the external FET, which seem to be able to handle quite a bit of current without getting too hot.

 

[ldoolit]

Yes, more boards are in the pipeline.

Ah. Yes, I see that now. I told you my eyes glazed over.

I will definitely order one board. I, too, would like to see the BOM Real Soon Now so I can turn in my order. Virtually kitted at Mouser, right?

You'd need to tap into the optocoupler inside the charger

Right, or supply your own. I see that has also been discussed recently on this thread.
Oh for two. Sorry, guys. I'll come back when I have something real to say.
I'll go read the LTC6802 thread now.

 

[ejonesss]

how are you coming on the mouser b.o.m?

i am looking forward to getting the board soon.

The Mouser part numbers basically are the manufacturer's part number, plus a 3-digit code in front, in order to identify the manufacturer. In any case, putting in the complete part number will call up the details for that part, including a link, usually, to the manufacturer's data sheet, in PDF form.

We purposely tried to pick as common and as inexpensive parts as we could, and most will have many substitutions. The exceptions will be the TC54 voltage detector, and maybe the LM431 programmable zener, but the optos, the gate driver, the SCR, all the transistors, the LEDs and even the FETs all are quite common and have many substitution choices available. We also went to great lengths to minimize the different/unique capacitor and resistor values, so that a larger quantity of fewer discrete part numbers could be ordered.

I haven't had time to put the BOM list together yet, as I've been too busy with my "other" job, but I'll try to get to it this weekend.

I've been to several Dick Smith stores, both in Australia and in New Zealand, and I would compare them to a cross between a Radio Shack and a Circuit City. Radio Shack also used to have quite a selection of discrete parts, but not so much anymore.

The board order is in, but because of the larger quantity, it will take 5 full days for manufacturing, and another 2-3 days for shipping. They are due in on Friday,the 31st.

-- Gary

 

[GGoodrum]

how are you coming on the mouser b.o.m?

i am looking forward to getting the board soon.

I am going to try and finish that up today, along with the updated instructions.

 

[ejonesss]

message deleted because it upset a user a little

 

[methods]

message deleted because it upset a user a little

Thanks,
-methods

 

[GGoodrum]

Okay, the boards are here. I have taken the day off from my "other" job, so that I can finish the BOM, finish the instructions and update the website. I will post the BOM here as soon as I can, hopefully within the next hour, or so. Next will be the site update, so that people can actually order the boards. Last will be the instructions, as those can actually be emailed later, if need be.

-- Gary

 

[ahambone]

Enjoy your "day off" Gary!

I am patiently looking forward to the boards! Thank you again for all the work you have done. I've posted a new thread under the Battery Technology for the HWC LiFePO4 cells that I mentioned earlier. If all goes well, these are the cells I intend to use your BMS with.

Cheers,
--adam

 

[AndyH]

Okay, the boards are here. ... -- Gary

Fantastic!

I'm going to buy a batch of boards, Gary - some for my own use. I intend to kit some, and assemble/test others, both for resale. This should expand availability for the folks that don't choose to 'roll their own'. News at 11...

Version 1.5 works well with a slightly-tweaked LiFePO4 smart charger, BTW!

Andy

View attachment BMS_all_shunts_on.jpg

 

[michaelplogue]

intend to kit some, and assemble/test others, both for resale. This should expand availability for the folks that don't choose to 'roll their own'.

Actually, I was just going to ask if anyone would be willing to assemble one of these for me for a fee. I have very little experience in this sort of work, and would prefer to have a expert do it the proper way. Let me know when you are going to be ready to build, and I'll send you my specs.

 

[GGoodrum]

Okay, I've completed and tested the BOMs. I did separate text files for quantities 4, 8, 12, 16, 20 and 24 channels. I can't sem to be able to attach text files here, so I'll just post the info below. What you do is just copy the appropriate section, not including the title line (i.e. -- just the data lines...), and then paste it into the input area on the "Import BOM" screen on Mouser. Then just click the "Import" button and everything automatically gets added to your shopping cart. If you are going to order parts for multiple boards, the best way is to use the"Project" feature and then import the BOM into a new project. You would then order the whole project and you can enter in a total number of projects to order.

I'm working on updating the website now, so that orders can be placed. I will post something here as soon as it is enabled.


4-CH
647-UFW1H100MDD|6
581-SA105E104MAR|1
512-1N914T50A|10
78-TLUV5300|1
859-LTL1CHKFKNN|4
512-2N3906TA|2
511-STP160N75F3|1
512-KSA931YTA|4
299-1M-RC|1
299-100K-RC|1
299-10K-RC|2
299-4.7K-RC|1
299-820-RC|11
299-100-RC|9
299-10K-RC|1
270-75K-RC|4
270-180K-RC|4
286-6.8-RC|4
299-20K-RC|4
610-2N5060|1
512-LM78L12ACZXA|1
747-IXDF404PI|1
579-TC54VC2102EZB|4
512-FAN431LZXA|4
782-ILD2|4

8-CH
647-UFW1H100MDD|10
581-SA105E104MAR|1
512-1N914T50A|18
78-TLUV5300|1
859-LTL1CHKFKNN|8
512-2N3906TA|2
511-STP160N75F3|1
512-KSA931YTA|8
299-1M-RC|1
299-100K-RC|1
299-10K-RC|2
299-4.7K-RC|1
299-820-RC|19
299-100-RC|17
299-10K-RC|1
270-75K-RC|8
270-180K-RC|8
286-6.8-RC|8
299-20K-RC|8
610-2N5060|1
512-LM78L12ACZXA|1
747-IXDF404PI|1
579-TC54VC2102EZB|8
512-FAN431LZXA|8
782-ILD2|8

12-CH
647-UFW1H100MDD|14
581-SA105E104MAR|1
512-1N914T50A|26
78-TLUV5300|1
859-LTL1CHKFKNN|12
512-2N3906TA|2
511-STP160N75F3|1
512-KSA931YTA|12
299-1M-RC|1
299-100K-RC|1
299-10K-RC|2
299-4.7K-RC|1
299-820-RC|27
299-100-RC|25
299-10K-RC|1
270-75K-RC|12
270-180K-RC|12
286-6.8-RC|12
299-20K-RC|12
610-2N5060|1
512-LM78L12ACZXA|1
747-IXDF404PI|1
579-TC54VC2102EZB|12
512-FAN431LZXA|12
782-ILD2|12

16-CH
647-UFW1H100MDD|18
581-SA105E104MAR|1
512-1N914T50A|34
78-TLUV5300|1
859-LTL1CHKFKNN|16
512-2N3906TA|2
511-STP160N75F3|1
512-KSA931YTA|16
299-1M-RC|1
299-100K-RC|1
299-10K-RC|2
299-4.7K-RC|1
299-820-RC|35
299-100-RC|33
299-10K-RC|1
270-75K-RC|16
270-180K-RC|16
286-6.8-RC|16
299-20K-RC|16
610-2N5060|1
512-LM78L12ACZXA|1
747-IXDF404PI|1
579-TC54VC2102EZB|16
512-FAN431LZXA|16
782-ILD2|16

20-CH
647-UFW1H100MDD|22
581-SA105E104MAR|1
512-1N914T50A|42
78-TLUV5300|1
859-LTL1CHKFKNN|20
512-2N3906TA|2
511-STP160N75F3|1
512-KSA931YTA|20
299-1M-RC|1
299-100K-RC|1
299-10K-RC|2
299-4.7K-RC|1
299-820-RC|43
299-100-RC|41
299-10K-RC|1
270-75K-RC|20
270-180K-RC|20
286-6.8-RC|20
299-20K-RC|20
610-2N5060|1
512-LM78L12ACZXA|1
747-IXDF404PI|1
579-TC54VC2102EZB|20
512-FAN431LZXA|20
782-ILD2|20

24-CH
647-UFW1H100MDD|26
581-SA105E104MAR|1
512-1N914T50A|50
78-TLUV5300|1
859-LTL1CHKFKNN|24
512-2N3906TA|2
511-STP160N75F3|1
512-KSA931YTA|24
299-1M-RC|1
299-100K-RC|1
299-10K-RC|2
299-4.7K-RC|1
299-820-RC|51
299-100-RC|49
299-10K-RC|1
270-75K-RC|24
270-180K-RC|24
286-6.8-RC|24
299-20K-RC|24
610-2N5060|1
512-LM78L12ACZXA|1
747-IXDF404PI|1
579-TC54VC2102EZB|24
512-FAN431LZXA|24
782-ILD2|24

 

[ejonesss]

i see that part 647-UFW1H100MDD is going to backorder if i go through with the purchase now it is an audio capacitor


parts list clipped to free up page space after being pointed out by another user

sorry for the long quote

Okay, I've completed and tested the BOMs. I did separate text files for quantities 4, 8, 12, 16, 20 and 24 channels. I can't sem to be able to attach text files here, so I'll just post the info below. What you do is just copy the appropriate section, not including the title line (i.e. -- just the data lines...), and then paste it into the input area on the "Import BOM" screen on Mouser. Then just click the "Import" button and everything automatically gets added to your shopping cart. If you are going to order parts for multiple boards, the best way is to use the"Project" feature and then import the BOM into a new project. You would then order the whole project and you can enter in a total number of projects to order.

I'm working on updating the website now, so that orders can be placed. I will post something here as soon as it is enabled.

 

[michaelplogue]

Would we need different components depending on what cells we will be using? For example, I'll be using the BMI (Lifebatt cells). Would this list work for them?

 

[ejonesss]

i see that there are 11 of the

Mouser #: 647-UFW1H100MDD
Mfr. #: UFW1H100MDD
Desc.: Audio Electrolytic Capacitors 10uF 50V 20%
RoHS: RoHS Compliant


available for shipping and it will go into backorder.

being a 10 uf can we substitute other values in series or parallel or does it have to be that value?

like 2 5 uf in parallel or 2 20 uf in series?

correct me if i am wrong 2 capacitors in series divides the uf value and 2 in parallel doubles the uf value?

 

[GGoodrum]

Would we need different components depending on what cells we will be using? For example, I'll be using the BMI (Lifebatt cells). Would this list work for them?

Yes, these will work fine for any LiFePO4-based cells, including Pings, PSI/BMI/LiFeBatt, Headway, a123, etc.

 

[GGoodrum]

i see that there are 11 of the

Mouser #: 647-UFW1H100MDD
Mfr. #: UFW1H100MDD
Desc.: Audio Electrolytic Capacitors 10uF 50V 20%
RoHS: RoHS Compliant


available for shipping and it will go into backorder.

being a 10 uf can we substitute other values in series or parallel or does it have to be that value?

like 2 5 uf in parallel or 2 20 uf in series?

correct me if i am wrong 2 capacitors in series divides the uf value and 2 in parallel doubles the uf value?

I don't think the value is all that critical. Plus, I think I remember that there were lots of choices for these. I think I just picked the cheapest. You could also use a couple of 4.7 uF electrolytics in series.

 

[methods]

being a 10 uf can we substitute other values in series or parallel or does it have to be that value?
like 2 5 uf in parallel or 2 20 uf in series?

10uF is probably the most common capacitor available. Got to be thousands of substitutes. If you run caps in series/parallel your board will end up a mess.

-methods

 

[GGoodrum]

I will try and work on the website later tonight, after all the kiddies stop banging on our door. I'll enable the site in the morning. I can still ship these tomorrow.

 

[chas_stevenson]

I did some work and here http://ilbcnu.org/misc_docs/goodman_bms.html is the parts list with quantities and links to the parts to make it easier for you all to order the parts.

Gary,

If you wish this removed please send an email to: chas_stevenson@yahoo.com and I will remove it.

Grandpa Chas S.

 

[rkosiorek]

i used these capacitors instead. - mouser par # 647-UVR1H100MDD1TD

work fine on the board i am using with my LifeBat cells.