Doctorbass
100 GW
Microbatman said:
That's great idea to post pics like that Microbatman, but please next time try to rasmble all those in something like 4 pics per file (2 x 2 pics). this page is Looooong to load! :wink:
Doc
The care and feeding of a123-based packs...
- Thread starter GGoodrum
- Start date
That's great idea to post pics like that Microbatman, but please next time try to rasmble all those in something like 4 pics per file (2 x 2 pics). this page is Looooong to load! :wink:
Doc
jeffkay
100 W
- Joined
- Sep 14, 2007
- Messages
- 115
Abot lead-acid charger: I can use a CC model that ends up making 4s just right at about 3.75 each cell. Low cells can be brought up using a 6v old school charger for a short time while wathing the volts. Or I can use my Volt Phreaks individual charger. As far as cycles, I have had no problems as long as don't let charge go over 4.2v and discharge less than 2.2v. I think if you go lower once in awile it is okay but you should charge them right away...
Thanks, so 3.75 isn't sufficient to float them into self-balance. hmm
Edit: reading on RC forums, seems if you use well matched cells, 90% of the time charging to 3.8v is adequate to keep a pack balanced, and you can expect to be at ~90% capacity by ~500 cycles.
Edit: reading on RC forums, seems if you use well matched cells, 90% of the time charging to 3.8v is adequate to keep a pack balanced, and you can expect to be at ~90% capacity by ~500 cycles.
GGoodrum said:
So, every cell from the 10s2p is paralleled with the corresponding cell on the other 10s2p to make a 33v 10s4p? If so, can you leave them in parallel when they're being charged with the bms/charger?
I'm thinking about building a battery pack and I want to make sure I thoroughly understand before I attempt to make one.
Thanks for the info!
GGoodrum
1 MW
After screwing around for three months with the LiFeBatt packs, only to have the rug pulled out from under me, I decided to go back to making use of the big pile of a123 cells/packs that I still have, so I can get my bikes back on the road. 
What I decided I needed, however, is a new version of the solderless a123 kits I used to do for the RC crowd. For those kits, I had basically 12, or 16 cells in series, configured in stacks of four cells. For this new kit, I decided it would be much simpler to create 4s4p 12V/9.2Ah "stick packs", each with its own 4-channel LVC board. Here's how they turned out:
What I plan to do for my 72V setups is to use six of these packs, connected in series. The outputs of all of the optos are connected in parallel, and plugged into the controller's ebrake input. I'm also doing a new inline active cutoff board, using Randomly's circuit, described in the BMS thread, that will take the same opto outputs and used them to cutoff a pair of 4110 FETS on the main negative power lead. I'll make these available as soon as I get the boards back, which should be before the end of the week.
Here's what is in the kit:
In between the 4-cell blocks are G10 plates with brass rings on one side, and nickel-plated springs on the other. There are copper strips that connect the four cells in parallel. All you do is hot-glue, or tape four cells together, and put the plates on either side. A threaded rod goes down the middle that compresses and holds everything together. No messy soldering required at all.
Wires are provided to connect each block to the LVC board. The threaded rod serves as the main negative pack lead. There is another 3mm stud coming off the LVC PCB that is the positive terminal. Finally, a 5-pin RC-styled balancer/charger plug is also provided, as is an option for a matching plug/pigtail.
Anyway, I'll post some more pics, after I get my six packs all together. The kits are available now, though, on my site: Solderless 4s4p a123 Pack Construction Kit.
-- Gary
What I decided I needed, however, is a new version of the solderless a123 kits I used to do for the RC crowd. For those kits, I had basically 12, or 16 cells in series, configured in stacks of four cells. For this new kit, I decided it would be much simpler to create 4s4p 12V/9.2Ah "stick packs", each with its own 4-channel LVC board. Here's how they turned out:
What I plan to do for my 72V setups is to use six of these packs, connected in series. The outputs of all of the optos are connected in parallel, and plugged into the controller's ebrake input. I'm also doing a new inline active cutoff board, using Randomly's circuit, described in the BMS thread, that will take the same opto outputs and used them to cutoff a pair of 4110 FETS on the main negative power lead. I'll make these available as soon as I get the boards back, which should be before the end of the week.
Here's what is in the kit:
In between the 4-cell blocks are G10 plates with brass rings on one side, and nickel-plated springs on the other. There are copper strips that connect the four cells in parallel. All you do is hot-glue, or tape four cells together, and put the plates on either side. A threaded rod goes down the middle that compresses and holds everything together. No messy soldering required at all.
Wires are provided to connect each block to the LVC board. The threaded rod serves as the main negative pack lead. There is another 3mm stud coming off the LVC PCB that is the positive terminal. Finally, a 5-pin RC-styled balancer/charger plug is also provided, as is an option for a matching plug/pigtail.
Anyway, I'll post some more pics, after I get my six packs all together. The kits are available now, though, on my site: Solderless 4s4p a123 Pack Construction Kit.
-- Gary
disndat
100 W
Hey Gary this looks great.Could you post the weight and dimensions of the packs.BTW are you selling any of your a123 cells?
GGoodrum
1 MW
disndat said:
The packs ends up being 2" square by 12" long, and weigh about 1.3kg/46 ounces. Sorry, but I'm going to be using all my cells. I'm re-configuring all my existing packs, and weeding out the abused cells that I've trashed, that have lost some of their capacity, mainly due to being over-discharged under 2V.
-- Gary
GGoodrum
1 MW
Here is my first new 12s4p 36V/9.2Ah a123 pack:
The dimensions are 2" x 6-3/4" x 12", and it weighs 8 lbs 13oz, or almost exactly 4.0 kg. I'm finishing up a second one of these, and will use them in series, for 72V/9.2Ah. These will be the main packs I use on my Dahon Mariner 20" folding bike, but I will also use it as an "add-on" capability for my Townie, when I need lots of range.
I'm very happy with how the packs go together. It really is pretty easy. Mechanically, they are solid as a rock and very compact.
After I finish the second 12s4p pack, I'm going to make up four more 12V/9.2Ah "sticks", and put them together in a 16s4p 48V/9.2Ah configuration, for use on my other folding bikes.
-- Gary
The dimensions are 2" x 6-3/4" x 12", and it weighs 8 lbs 13oz, or almost exactly 4.0 kg. I'm finishing up a second one of these, and will use them in series, for 72V/9.2Ah. These will be the main packs I use on my Dahon Mariner 20" folding bike, but I will also use it as an "add-on" capability for my Townie, when I need lots of range.
I'm very happy with how the packs go together. It really is pretty easy. Mechanically, they are solid as a rock and very compact.
After I finish the second 12s4p pack, I'm going to make up four more 12V/9.2Ah "sticks", and put them together in a 16s4p 48V/9.2Ah configuration, for use on my other folding bikes.
-- Gary
Doctorbass
100 GW
that look great and simple Gary! :wink:
How many amp would you expect to get with these interconnections between cells?
How many amp would you expect to get with these interconnections between cells?
GGoodrum
1 MW
Doctorbass said:
With my previous kit, which simply used round head screws to make contact with the cells, we used to pull 100A+ in RC helicopters, and that was in a single 2.3Ah "1p" configuration. Here, there are fours cells in parallel, each with its own nickel-plated spring contact, and there are three 3/4" strips of copper that inter-connect the cells in parallel. Altogether, I'm sure each of these "sticks" can deliver well over 100A continuous, although I'm only using 12-gauge output wires so I wouldn't actually want to try that much. My setups might hit an occasional 80A peak, but mostly the loads will be in the 30-50A range.
-- Gary
GGoodrum
1 MW
swade said:
For these 12s4p packs, I'll use a pair of Soneil 3610SRF 36V/4.5A chargers. Since I am only using healthy cells, that are reasonably matched in capacities, they stay pretty closely balanced already. To keep them that way, about every 5 cycles, or so, I'll either use RC-type auto balancers, or I'll use my setup of 12 Voltphreaks 2A individual CC/CV chargers. Each 4s4p "stick pack" has an RC-type 5-pin plug/pigtail that provide connections to each cell block.
The other two-wire JST plug is the combined opto outputs from the three 4-channel LVC boards. All my setups have a cable with a matching plug that connects to the controller ebrake line. I'm alos working on an small external cutoff board that will use the opto connections.
-- Gary
disndat
100 W
Hey Gary any plans for a 10s4p kit?
GGoodrum
1 MW
disndat said:
I did do a 10s4p pack before. It's buried in this thread somewhere. The problem is that a 10s setup really limits the charging options. By doing 12s or 16s, we can make use of many existing SLA 36V and 48V CC/CV chargers.
With the new kits, everything is in multiples of four. There's four cells in parallel and then four block in series, so 16 cells in the 12V/9.2Ah pack. You could cut the stack down, to 2s or 3s, or maybe add a block, to go to 5s, but that would mean adding a 5th channel to the LVC board. You'd still end up with the non-standard charging issue though.
When I was using 10s4p packs, I was using two in series. Since I had to go to a 72V controller to do this, it just made more sense to me to just go to the full 72V-equivalent setup.
-- Gary
Some of that is a bit over my head so please feel free to correct me below.
I understand using the two 36v/4.5a chargers to charge in series without balancing. That would be great for charging at the office etc with minimum hassle. Isn't the math a bit off on the voltage though?
4S: 3.3 nominal * 4 = 13.2 volts
4 *3.6 = 14.4 volts <- Isn't this what one would expect to charge each block to??
"I'll either use RC-type auto balancers" Using these balancers, the best we could hope for after a balance is that each 4s block would have all cells at an average of 3.6volts? Some potentially could be higher or lower than that? Would the same be true if you used your voltphreaks individual charger setup?
I guess ultimately, you could easily de-construct the packs and charge each cell individually every couple of months.
Whatever the case, pictures of the charging scenarious would be a nice addition to your website and/or this thread. I'm currently using A1234Life's charging setup. It was a lot of work but the results are great. The only downside is the initial labour and the effort to manage the connections -- I'm still trying to find the ultimate a123 e-bike solution...
I understand using the two 36v/4.5a chargers to charge in series without balancing. That would be great for charging at the office etc with minimum hassle. Isn't the math a bit off on the voltage though?
4S: 3.3 nominal * 4 = 13.2 volts
4 *3.6 = 14.4 volts <- Isn't this what one would expect to charge each block to??
"I'll either use RC-type auto balancers" Using these balancers, the best we could hope for after a balance is that each 4s block would have all cells at an average of 3.6volts? Some potentially could be higher or lower than that? Would the same be true if you used your voltphreaks individual charger setup?
I guess ultimately, you could easily de-construct the packs and charge each cell individually every couple of months.
Whatever the case, pictures of the charging scenarious would be a nice addition to your website and/or this thread. I'm currently using A1234Life's charging setup. It was a lot of work but the results are great. The only downside is the initial labour and the effort to manage the connections -- I'm still trying to find the ultimate a123 e-bike solution...
GGoodrum
1 MW
The 36V chargers actually put out about 43.5V, which is about 2.40V for each of the 18 SLA cells in a 36V pack. That also works out to about 3.63V per cell for a 12s a123. The 361SRF also has internal pots for the current and for the output cutoff voltage, which can be accessed by simply removing the cover off one end. I've "tweaked" mine up to about 44.5V, which is about 3.70V per cell, and the current up to 5A.
SLA cells have a lower relative "nominal" voltage than LiFePO4-based cells. Each SLA cell is rated 2.0V nominal, but need to be charged to 2.40-2.45V to get full. LiFePO4 cells have a nominal voltage of about 3.3V per cell and need to be charged to 3.65-3.70V per cell. If you just look at the charge voltages, 4 LiFePO4 cells = 6 SLA cells, but the nominal voltages are a bit different, 13.2V for four LiFePO4cells vs. 12V for six SLA cells. Since all this is very confusing, most just use the SLA-based nominal pack voltages (i.e. -- 36V, 48V, etc...).
Regarding the use of RC-style balancers, I just make special harnesses that use two balancers with one cell overlapping. Here's what one looks like:
I'll post some more pictures of my 12-cell Voltphreaks setup, plus some more diagrams, a bit later, aftr I get this second 12s4p 36V/9.2Ah pack done.
-- Gary
SLA cells have a lower relative "nominal" voltage than LiFePO4-based cells. Each SLA cell is rated 2.0V nominal, but need to be charged to 2.40-2.45V to get full. LiFePO4 cells have a nominal voltage of about 3.3V per cell and need to be charged to 3.65-3.70V per cell. If you just look at the charge voltages, 4 LiFePO4 cells = 6 SLA cells, but the nominal voltages are a bit different, 13.2V for four LiFePO4cells vs. 12V for six SLA cells. Since all this is very confusing, most just use the SLA-based nominal pack voltages (i.e. -- 36V, 48V, etc...).
Regarding the use of RC-style balancers, I just make special harnesses that use two balancers with one cell overlapping. Here's what one looks like:
I'll post some more pictures of my 12-cell Voltphreaks setup, plus some more diagrams, a bit later, aftr I get this second 12s4p 36V/9.2Ah pack done.
-- Gary
slayer
1 kW
Wow that is great work Gary ...to include the lvc for each 12 volts packs also is excellent.
i think i will rebuild my pack into this setup
i think i will rebuild my pack into this setup
GGoodrum
1 MW
I finally got both my 36V/9.2Ah packs complete. Below are some shots of these mounted on my Mariner 20" folding bike:
As shown, this is a very compact setup. It is also fairly light, at about 17.5 pounds/8.0kg. That works out to about 42 Wh/lb.
I've re-done the setup on this bike a bit. It used to have a 20s4p a123 setup. The controller has the 4110 mod, and has the current limit bumbed up to about 75A. With the jump up to 24s/72V, I went to dual WattsUp meters. I will do some test rides today. Before, with the 20s4p setup, I ws able to hit 6500W peaks, and it would smoke the front tire fairly easy, even while moving.
Now, I'm guessing I'll be shopping for some new tires later. 
I also tried these packs on my Townie yesterday. This bike has an even higher curent limit in the controller. I hit multiple 100A/7500W+ peaks. This beast accelerates like a motorcycle, now. Pretty scary, really. Anyway, the packs/cells were not even warm. Performance-wise, a123 cells are still king-of-the-hill.
-- Gary
As shown, this is a very compact setup. It is also fairly light, at about 17.5 pounds/8.0kg. That works out to about 42 Wh/lb.
I've re-done the setup on this bike a bit. It used to have a 20s4p a123 setup. The controller has the 4110 mod, and has the current limit bumbed up to about 75A. With the jump up to 24s/72V, I went to dual WattsUp meters. I will do some test rides today. Before, with the 20s4p setup, I ws able to hit 6500W peaks, and it would smoke the front tire fairly easy, even while moving.
Now, I'm guessing I'll be shopping for some new tires later. I also tried these packs on my Townie yesterday. This bike has an even higher curent limit in the controller. I hit multiple 100A/7500W+ peaks. This beast accelerates like a motorcycle, now. Pretty scary, really. Anyway, the packs/cells were not even warm. Performance-wise, a123 cells are still king-of-the-hill.
-- Gary
recumbent
100 kW
For comparitive sake, this is also 18 lb battery pack, 48V and 12 amp/hr, has enough power to climb any hill where i live but slows down some of course.
12 amp/hr is surprising plenty range, at least 2 hours, more if i peddaled.
12 amp/hr is surprising plenty range, at least 2 hours, more if i peddaled.
slayer
1 kW
I would love to see a viedo of the townie in the same setup you showed us before with the eagle tree on screen showing 7500 watts...is your townie a rear motor
GGoodrum
1 MW
recumbent said:
Thats a nice setup. Is this using Justin's new pack?
Let's see, 16 x 3.3 x 12 = 633Wh, which is pretty close to my 72V/9.2Ah setup, which is 24 x 3.3 x 9.2 = 728Wh. I'm going to be getting significantly less range, however, as I can't seem to do much in the way of throttle management.
slayer said:
I will definitely be doing some new videos, both of the Townie, and of the Mariner.
Yes, my Townie has a rear-mounted 5304, but if I had it to do over again, I'd do the front mount. The way I have it now, I have a 24s4p setup in the bag, on the rear rack, and I put these two 36V/9.2Ah packs in a basket, up front. It balances okay, but I don't like having all the weight up front so high up. I just like the front motor, with all the batteries in back. It handles a lot better, in my opinion, especially at fairly high speeds (above 40 mph...).
-- Gary
recumbent
100 kW
GGoodrum said:
Justin's 48V packs only have 15 cells in them. And yes this is one of them, which read 50V on the Cycle anayst fresh off the charger.
Hi Gary
Nice work on the packs, really like that idea a lot being able to strip out bad cells so easily and the re usability of the system makes it a no brainer for me, when I get around to it after all my packs dry up I will be ordering some of them off you! hopefully A123 cells may be easier to get hold by then. I saw your other videos (very impressive) have not seen the one with the on screen display yet, where is that one?
Looking forward to more videos from you and your bikes
Knoxie
Nice work on the packs, really like that idea a lot being able to strip out bad cells so easily and the re usability of the system makes it a no brainer for me, when I get around to it after all my packs dry up I will be ordering some of them off you! hopefully A123 cells may be easier to get hold by then. I saw your other videos (very impressive) have not seen the one with the on screen display yet, where is that one?
Looking forward to more videos from you and your bikes
Knoxie
GGoodrum
1 MW
knoxie said:
Yep, I'm very happy with this new configuration as well. I really like having the option to quickly re-configure packs, if I need to, and it makes servicing packs a ton easier.
Here is the video of my Mariner, back when it had a 20s4p a123 setup: http://www.tppacks.com/video/Mariner-20s4p-5303-11-02-07.wmv. It is the one with the on-screen overlay, from the EagleTreeSystems MicroLogger. The 20s setup was pushing it, for the max voltage the MicroLogger could handle, so I haven't been able to use one of these with my 24s/72V setups. Instead, what I'm doing now is to have both of my WattsUp meters, and the speedometer, visable in the bottom part of the frame. This actually works pretty well. Here's the last video I did of my Townie, using this method, but I only had one Wattsup hooked up, to part of the pack:
http://www.tppacks.com/video/Townie-24s6p-5304-12-15-07a.wmv. In this configuration, I had 20s6p in the bag on the rear rack, and a 4s6p sub-pack in the frame-mounted bag up front. The Wattsup was connected into the 4s6p pack. Since then, I've swapped controllers around and had one that I somehow manage to blowup. Bob repaired this one and upped the current limit a bit more. Now, the one on the Townie hits 100A+ peaks, and the one that was on the Townie when this video was made is now on my Mariner folding bike. It hits 65-70A peaks and with the new 24s4p setup shown above, it'll hit 5500W, which is still way too much.
I'm going to try and do a new Townie video, in the next couple of days. With this latest setup, I have two 12s4p sub-packs in the back bag, on the rack. Each of these are connected to separate WattsUp meters, and then connected in series at the controller. When I need extra range, which is most of the time, I can add the two new 12s4p packs from the Mariner, in the basket up front. These are then connected in parallel with the two 12s4p packs in the bag, just before the WattsUp meters. All the LVC outputs, from all four 12s4p packs are connected in parrallel with the ebrake line, going into the controller. Anyway, the balance is much better than with the previous setup, which had most of the weight in the rear.
-- Gary
