Group Buy? Dr.Bass Steveo found 7s3p Module! Awesome Price!!

[Ypedal]

Yes..

Peak charge voltage of 3.6v ... they drop to 3.3 resting voltage.. 1C ~3 C discharge mostly above 3.0v .. drops like a rock below 3.0 to 2.0v fully empty and drained flat.

[EVPowers]

So here is the type of usage I'm considering for buying these packs
http://www.roperld.com/science/SunRoperTrips.htm

You can see in the third graphic down, that there can be some high-current draws. (290a, 250a, 225a, etc)
I'm not sure three 7s3p packs will be able to handle the load.

The voltage would probably need to be upped a small bit by creating another small pack of cells as well.

[ohzee]

While this pack size is to much for anything I would need I enjoy following this thread. I woke up about 3am last night and the
dream I was having was about this thread and these packs.. lol Just thought I would share - it was the great pictures Oat posted
I think that somehow had me dreaming about these packs - what they looked like what you could do with them.

Just thought I would come clean thought it was amusing.

[oatnet]

Hey thanks everyone for the kind words about the pictures, and thanks to all those who shared pics of the 54xx and CroMotor for me to drool over while I was waiting for them.

You can see in the third graphic down, that there can be some high-current draws. (290a, 250a, 225a, etc)
I'm not sure three 7s3p packs will be able to handle the load.

The voltage would probably need to be upped a small bit by creating another small pack of cells as well.

A single cell theoretically supports 600a (20ahx30c), so a 3-parallel cell will support 1,800 amps. A mere 290a will not be an issue. If you are trying to fit the battery to an existing setup, the controller will determine what voltage you require, and how many cells you need in series.

3.3v is the rated nominal voltage - aka the cell's resting voltage when half of its rated AH have been consumed. Hot off the charger, resting voltage is 3.65v. Ypedal and I undercharge our LiFe - to 3.60v and 3.55v/cell respectively - to minimize damage the cell incurs at peak charge.


On another note - am I the only one whose BMS's smell really bad? I think the circuit board itself decomposing, do I need to worry about the chemicals in it? Looking at the picture above that compares a good vs decomposing, you can see how the PCB seems to be balloning up and around the traces...

And of course, that makes me wonder whether they were intended to decay, or if it was a bad batch of PCBs.

-JD

[SlyCayer]

that makes me wonder whether they were intended to decay, or if it was a bad batch of PCBs.

I have never heard of decaying PCBs, correct me if I'm wrong, but ain't PCB mainly glass based?

The smell might be in a "Gel" coating? I remember opening up a HID Conversation Kit Ballast, and it had a gelly/silicone filled box and it smelled like shit but was there to protect the PCB from any liquids or elements...

Maybe...

[davec]

i do have a request for anyone who has the time so we can get a good idea about what we're getting here

1. IR test on one of the cells in the module
2. 1C test on a cell in the module to see how many AH they deliver

would be awesome

[deVries]

On another note - am I the only one whose BMS's smell really bad? I think the circuit board itself decomposing, do I need to worry about the chemicals in it? Looking at the picture above that compares a good vs decomposing, you can see how the PCB seems to be balloning up and around the traces...

Circuit boards are layers of resin & fiberglass...

on the right one of the 2009 BMS's that STINK, are gooey, you can see the color and sorta matte finish on these.

Since you mention "gooey" too, I think this somehow got exposed to or misted in some chemical that attacked the CB. Can you poke or scratch into the CB resin to see if it is softer & not hard on or near the surface?

I know someone mentioned using alcohol to clean CBs that had gotten dirty & that it works well.

Are you intending to use the BMS? Has someone learned how to do this already?

[oatnet]

you guys did a really good job
thanks for the great pics and for reviewing these modules for us
i do have a request for anyone who has the time so we can get a good idea about what we're getting here

1. IR test on one of the cells in the module
2. 1C test on a cell in the module to see how many AH they deliver

would be awesome

...crickets...

[oatnet]

Ghetto-balancing the 7s pack that had the 2.75v cells... So glad I made up a bunch of plugs for the single-cell chargers that have alligator clips on the other end, they come in handy time and time again. I'm hoping that a nice, slow 2a charge on these 60ah cells will bring them back from low v, hopefully they were just at a low SOC when they got shipped, At 2a it will take 30 hours just to generate 60ah, probably another 15 hours just to make up for effeciency losses.

We need a smellavision so I can share my BMS aroma. It could just be a coating on the boards - but it really smells foul to me, hard to believe they would put it in a product.

It just occured to me that maybe they aren't actually BMS, maybe they are just monitoring sensors that feed run by an external BMS over canbus via that other jack that is on the board. If anyone who can read circuits can tell from those big (click on) pictures I posted, I'd love to know what you thing they do. those wires are pretty thin.... Still and all, I don't see another way of balancing those cells, and they are in fact in balance. I thought I saw something about a123's "active balancing system"

-JD

IMG_3332.jpg (170.69 KiB) Viewed 644 times

[deVries]

I got an update from Sutho saying that he will be getting his 22 modules of 7s3p next Tuesday, so by Wednesday next week we should have some good info about what a bulk order looks like.

i do have a request for anyone who has the time so we can get a good idea about what we're getting here

1. IR test on one of the cells in the module
2. 1C test on a cell in the module to see how many AH they deliver

Since the cells are welded together in a 3p configuration, I seriously doubt someone will "destroy" their module to extract a single cell from the 3p modules.

That's just too much risk, work, etc. unless one is doing a science paper or analysis of high numbers of test cells for that purpose. One cell from one module would not really tell us much about the 100's of modules and thousands of cells available, so it's statistically meaningless and useless information, IMO.

[davec]

a cba test on the whole pack itself at 1c

[JRP3]

As has been repeatedly stated the cells are welded in parallel groups of three, you cannot isolate an individual cell from the other two without physically cutting it out. Not going to happen, and not necessary. Give it up.

[oatnet]

a cba test on the whole pack itself at 1c

Give it up. After you have persistently trashed talked these cells and harassed those of us who are actually working with them, why do you think anyone is going to do testing for you? We all know if we do, you are going to latch on to some irrellevent factoid and imagineer it into a series of turds you can sit in your corner and throw at those of us who are actually contributing content.

-JD

[deVries]

a cba test on the whole pack itself at 1c

Just a suggestion to help newbies & those interested in reading the posts in sequence coherently on ES...

Please, when you edit your original post, leave your previous text & request or info in the post you edit too! (Don't erase the original text.) Guys, we all make mistakes or ask "the wrong question" - maybe it seems "dumb" after the fact. Etc. I've done this myself as a newbie. It's embarrassing some of the early posts I made too.

Bottom line is we need to learn from our mistakes or misinformed posts too, and that way a newbie can learn from our mistaken ideas or misinformed posts too.

So just type EDIT: (Put your edited text here, and leave your original text with the error or mistake too.) This way we can all follow any posts that follow your post too that also refer to the unedited text you posted originally too.

For instance, IF you had done that, then this follow-on post would be easily understood too...

As has been repeatedly stated the cells are welded in parallel groups of three, you cannot isolate an individual cell from the other two without physically cutting it out. Not going to happen, and not necessary.

Thanks!

For anyone that bought a module: Please do post anything you learn about the module that can help other people too. We need that data & info... PLEASE.

[zaxxon]

Been charging the module that came at 77.4 volts using a fixed 120 V power source and old power slide resistor. Charging at 2 amps which is max of the four 30 volt vicor modules I am using in series. Module voltage came up as one would expect. Put in about 20 AH and now all the cell voltage seems stuck at 3.32 V up from the initial 2.75 V . I don’t have a volt meter with more resolution.

I have some questions to those more experienced with these cells. Is the charge characteristic of these 20 AH cell so flat that at this low charge current that its normal not to see it change from 3.32 over a period of five hours ? I understand it will take about 30 hours to charge at this rate starting from 2.75 V.

Also would like to know even at this low 2 A charging current will the cell voltage eventually climb to 3.60 volts ? Or do you need to be charging at a 1C rate to see that? Just want to make sure I am not cooking them. I have put in around 36 AH.

[oatnet]

It is normal for LiFe cells to plateau during both charge and discharge, they have what is known as a "flat discharge curve". As they approach capacity, yes voltage will go up; wh is a better measure of energy because it takes into account voltage and its change throught the discharge cycle. 120v is way too high for a (assuming a) 28s module, which should have a peak voltage of 102.2. If you don't catch it and you charge your cells to 120v/4.28v a cell, this will degrade their cycle life. You need to buy at $15 (or more) Digital Multimeter to work on your 2k investment, because you picked a tough pack to learn on.

Charging is a lossy process so you can expect to put in 30%+ more than the pack needs to fill up its 60ah.

======

Anyhow, regarding my own modules... My deeply discharged 7s pack completed the slow single-cell charge, consuming 2.5kw, as reported by my kill-a-watt. Its been off for a few hours and voltages are still at 3.67v/cell, so that is a really good sign, I'll let it sit for a few days and hope they hold steady.

I bulk-charged my 28s at 4.2a to 98.4v - 3.55v/cell - and it hoovered down almost 4kw hours. Most of the cells were at 3.58v, but 3 were lower. I marked them specially and put single-cell chargers on them, two came up to 3.67v really fast, the third one is still charging (finger crossed) so at best it was pretty far out of balance. In the worst case, its a lame cell that I kill on a deep discharge. I'd solder a pair of 8ga across the terminals (the buss bar supplies 2 perfect channels) of the bad cells, run 48s instead of 49s (shrug), and carry and extra 3lb of dead weight in a 196lb pack.

Anyhow, I'm also bulk-charging the remaining pair of 7s to 49.7v (3.55v/cell) to see if either of them have a low cell I should be keeping an eye on. Then I'm gonna single-cell charge them, let them sit a few days to look for self discharge, pop them into the camper-bus and take them for a test drive, and see what I can suck out of them.

They broke off a bolt in a terminal in one of my 7s. The 28s pack was warped in shipping but I fixed it. It was quickly obvious that pushing on the side of the pack would dent the cooling fins, so I gently rolled the pack to its side on a flat surface, lifted one end 6" and dropped it, same on the other end, and the pack was now straight. BTW I noticed the cooling fins on the bottom of the 28s are covered with thermal paste.

-JD

[miro13car]

zaxxon,
I charged at 2A and stil was able to bring cells to 3.6V.
It just takes so much time - more than 8 hours.
In theory 20 Ah means 2A for 10 hours.

[zaxxon]

Oatnet,

Thanks !! That’s what I hoped; a very flat curve compared to what I am use to. Now I can continue with yours and Miro13car’s input with less concern. I have experience charging our 300v 84 module Prius NiMH 13 AH pack we assembled at 4 amps. Since the 4 amps is a much higher percent charging current, and I suspect its not nearly as flat a charge curve as A123, the voltage increases steadily to the 8.4v module voltage level we charged to.

Your right I need to invest in a better meter and should lower the voltage incase I fall asleep, and or forget to check.

Your braver than me in the approach you used to square up your cells. I was thinking about using the mounting through hole with some rods and some ratchet straps to try to relieve some of the pressure before trying to ease them back into position.

[drgrieve]

Charging LiFeP04 basics.

There is a formula called CC/CV to charging these batteries. You can't pick and choose what parts of the formula is easiest - it doesn't work that way.

To charge to 100% you should charge at 3.65v at a constant current until cell or module voltage is 3.65 then decrease current holding voltage at 3.65 terminating charge when current is 0.05C (3 amps on 60h). After settling (24 - 36 hour period) the resting voltage will be 3.4V. This is 100% charged.

My main issues with some of the recent comments is as follows.

1. If you are charging at 2 amps you are below 0.05C already and you should charge to a lower voltage otherwise you are over charging.
2. Charging LiFeP04 batteries is NOT excessively lossy, you should get back 99% of the energy you put in. Counting AH is an effective way of measuring SOC if you know your starting reference.
3. A cell resting at over 3.4V is OVER charged and likely suffered a small capacity drop.

Note [except for capacity testing] I recommend charging so that the resting voltage is 3.33 or 3.34 for extended cycle life. Replace 3.65 with 3.5. Also even when testing capacity 2V is too low (not much room for error!), aim for 2.5v when capacity testing as ah between 2.5V and 2.0v is minimal.

[deVries]

Charging LiFeP04 basics.

There is a formula called CC/CV to charging these batteries. You can't pick and choose what parts of the formula is easiest - it doesn't work that way.

To charge to 100% you should charge at 3.65v at a constant current until cell or module voltage is 3.65 then decrease current holding voltage at 3.65 terminating charge when current is 0.05C (3 amps on 60h). After settling (24 - 36 hour period) the resting voltage will be 3.4V. This is 100% charged.

My main issues with some of the recent comments is as follows.

1. If you are charging at 2 amps you are below 0.05C already and you should charge to a lower voltage otherwise you are over charging.
2. Charging LiFeP04 batteries is NOT excessively lossy, you should get back 99% of the energy you put in. Counting AH is an effective way of measuring SOC if you know your starting reference.
3. A cell resting at over 3.4V is OVER charged and likely suffered a small capacity drop.

Note [except for capacity testing] I recommend charging so that the resting voltage is 3.33 or 3.34 for extended cycle life. Replace 3.65 with 3.5. Also even when testing capacity 2V is too low (not much room for error!), aim for 2.5v when capacity testing as ah between 2.5V and 2.0v is minimal.

Hi,

Since you just joined ES to literally make this 1st post, could you give us some more reference points about this information? You seem to be speaking from a position of experience and/or a good knowledge base about these cells & charging correctly.

How you done any long-term testing with these types of A123 cells?

Thanks.

[JRP3]

2. Charging LiFeP04 batteries is NOT excessively lossy, you should get back 99% of the energy you put in.

Not unless you have a charger that is 100% efficient, which you don't. 90-95% efficiency is probably normal including the charger. Losses will probably be higher if you are using resistors to do top balancing.

[zaxxon]

Charging LiFeP04 basics.

There is a formula called CC/CV to charging these batteries. You can't pick and choose what parts of the formula is easiest - it doesn't work that way.

To charge to 100% you should charge at 3.65v at a constant current until cell or module voltage is 3.65 then decrease current holding voltage at 3.65 terminating charge when current is 0.05C (3 amps on 60h). After settling (24 - 36 hour period) the resting voltage will be 3.4V. This is 100% charged.

My main issues with some of the recent comments is as follows.

1. If you are charging at 2 amps you are below 0.05C already and you should charge to a lower voltage otherwise you are over charging.
2. Charging LiFeP04 batteries is NOT excessively lossy, you should get back 99% of the energy you put in. Counting AH is an effective way of measuring SOC if you know your starting reference.
3. A cell resting at over 3.4V is OVER charged and likely suffered a small capacity drop.

Note [except for capacity testing] I recommend charging so that the resting voltage is 3.33 or 3.34 for extended cycle life. Replace 3.65 with 3.5. Also even when testing capacity 2V is too low (not much room for error!), aim for 2.5v when capacity testing as ah between 2.5V and 2.0v is minimal.

Did you mean to say or was there a typo: "To charge to 100% you should charge at 3.65v at a constant current until cell or module voltage is 3.65 then decrease current holding voltage at 3.65 ..." I am not sure your required or would want to try to drive current to a level to force the 3.65 v during the constant current portion of the charge.

Shoud this have said: To charge to 100% you should charge at constant current value greater than 0.05C and less than or equal to the max recommended by the manufacture until a maximum voltage of 3.65 is reached, and then decrease current to hold the voltage at 3.65v and terminating charge when current is 0.05C (3 amps on 60ah).

[miro13car]

I do charge to lower voltage at 2A, never reach 3.65V , usually 3.6V.
My resting volts are never higher 3.32V
Better for longevity of cell

[emiyata]

I think he is talking about battery efficiency to accept a charge. I don't think he is talking about charger efficiency.

Steve

2. Charging LiFeP04 batteries is NOT excessively lossy, you should get back 99% of the energy you put in.

Not unless you have a charger that is 100% efficient, which you don't. 90-95% efficiency is probably normal including the charger. Losses will probably be higher if you are using resistors to do top balancing.

[oatnet]

1. If you are charging at 2 amps you are below 0.05C already and you should charge to a lower voltage otherwise you are over charging.

I suppose you missed the part where I bulk charged to 3.55v/cell, and then used voltphreaks single-cell LiFePO4 chargers handle the taper charge.

2. Charging LiFeP04 batteries is NOT excessively lossy, you should get back 99% of the energy you put in. Counting AH is an effective way of measuring SOC if you know your starting reference.

I wish I had chargers that were as efficient as yours.
Amps are never a definitive measure of current, because they don't consider Volts, thats why we use watts. V x A = W

3. A cell resting at over 3.4V is OVER charged and likely suffered a small capacity drop. Note [except for capacity testing] I recommend charging so that the resting voltage is 3.33 or 3.34 for extended cycle life.

You make sweeping statements about LiFePO4 as a chemistry, that would better apply to a specific powder that you may have experience with or read about. Take the old t-sky LiFePo4 for example, which ran at a higher voltages, and would be well on its way to depleted at 3.4v. I think it was goodrum I learned the old a123 M1 test from, charge it up to 3.65v, and if it stays above 3.60v for a week it is a good cell. I expect my headways to drop to 3.37v in a few days. I have junk that sits at 3.26v. So these characteristics vary widely.

Anyhow, was the experience you are quoting from working with the a123 20ah cells in general, or these specific modules?

-JD