HK Lipo Bricks - what do I choose and why?

[ElectricGod]

It's easily possible for anyone to buy cans that had reliable mfg and QC. I don't even know if hobby cells mfg to EV standards exist.

Hobby pouches get minimum screening and are still largely human hand manufactured. Fully automated 18650 lines are industry standard.

What about the similar pouch cells that Zero uses? are they fully automated?
I always assumed they were... automation makes it cheaper, but if it can be automated, why wouldnt hobby cells be?

Yes, fully automated and extensive post mfg QC like virtually all EV cells. Many hobby grade pouches get a voltage check after formation, and little QC beyond this.

I think the manufacturing of the individual LIPO pouches is probably automated. It's the assembly of them into packs that's manual I would think. I can't imagine why manufacturing 18650 would be anything but machine done. Uniform shape, uniform parts...that sounds like mass production to me.

The battery holders got here today. I took the ends and roughed them up a little with my sander and then super glued them together into 20S lengths. It worked pretty well.

Here's the cross connects on one of the 20S sections. The negative ends are in the lower half of the connections. I bought 15 feet of 18 awg wire thinking it would be plenty, but that's going to do like 3 or 4 of them so I have ordered 60 more feet of wire. I'm still waiting on the balance connectors to get here so I can't complete one until then.

 

[liveforphysics]

I think the manufacturing of the individual LIPO pouches is probably automated. It's the assembly of them into packs that's manual I would think.

Having toured a number of them around the world, I've still not seen an automated hobby pouch cell mfg yet. I've heard at least one exists, but it's not making the low $/Wh cheap folks buy.

 

[Hillhater]

Lipo's excel in these areas.
Less weight for the same charge capacity than LION...or that used to be the case.
Much higher discharge rates for the capacity of the cell, but that's an over simplification.

LION's excel in these areas.
Higher charge capacity per volume
higher charge cycles regardless of the cell.
Better resilience to abuse, damage and environmental effects.

Typically people use 18650 cells in parallel to get the C rates and capacities found in physically large LIPO bricks. 4 18650's in parallel is the same as a single 10000mah LIPO cell. The multiple cells in parallel also gives better over all discharge rates to the LION pack. Without parallel 18650 cells, you will never get the C rates of better LIPO packs, but then the 18650 is a small cell compared to the large, high discharge LIPO so on a single 18650 vs a single LIPO...keep that in mind. Large LION cells exist with high discharge rates. Keep in mind that a LIPO battery with the same volume as an 18650 would have much less capacity than 2600mah. It would be more like 900 to 1200mah and the C rate would be 2 to 5. LIPOS are great for R/C stuff because they are light weight and have lots of storage capacity. 10,000mah at 30C in a LIPO and the same amount of capacity in 2600mah 18650's you would possibly have about 10-15% more weight and be a much larger battery pack. But that's not all there is to the story. The LION cells will be charging and going strong well after the LIPOs have been thrown out. Also, the 18650 cells are much less prone to damage. On a scale that EV's tend to use the 18650's will also provide more wattage than the same volume of LIPOs. The disadvantage is making a bunch of LIPOs into an EV pack is relatively easy while making an EV pack out of 18650's is harder.

Hmm ?..a few misconceived ideas in that lot !..
Lipos do not have less weight than 18650's for any given capacity. In fact the opposite is true.
18650's with C rates of 20-30C have existed for several years...(ref A123 M1 cells, as well as the LG cells in the above table )
Discharge rates are not an issue with 18650's if the correct cells are chosen, unless extreme applications are being considered.
The biggest issue with 18650's currently, is the best ones are still expensive !

 

[Hillhater]

Here's the cross connects on one of the 20S sections. The negative ends are in the lower half of the connections. I bought 15 feet of 18 awg wire thinking it would be plenty, but that's going to do like 3 or 4 of them so I have ordered 60 more feet of wire.

Why didn't you alternate the cells polarity in the holders and save a few feet of wire. .?

 

[MrDude_1]

Here's the cross connects on one of the 20S sections. The negative ends are in the lower half of the connections. I bought 15 feet of 18 awg wire thinking it would be plenty, but that's going to do like 3 or 4 of them so I have ordered 60 more feet of wire.

Why didn't you alternate the cells polarity in the holders and save a few feet of wire. .?

If we're going to suggest different methods... why not run strings of them in parallel? less balance wires too. like this:

+++++-----+++++-----
-----+++++-----+++++

 

[ElectricGod]

Here's the cross connects on one of the 20S sections. The negative ends are in the lower half of the connections. I bought 15 feet of 18 awg wire thinking it would be plenty, but that's going to do like 3 or 4 of them so I have ordered 60 more feet of wire.

Why didn't you alternate the cells polarity in the holders and save a few feet of wire. .?

I thought about alternating cells and yes it would make the wires shorter, but I wanted all the cells oriented the same direction so there would never be any confusion which way the cells should be oriented. Being these are battery holders glued together in a 20S string, the cell markings for each slot are all labeled with + on the same end. I could have labeled each slot for correct orientation, but decided to go this route instead. It is a good point. less wire=less resistance. Yeah...Its a good idea...I should do it. I've only wired up a single strip of holders so far so not a huge thing to fix. Thanks for the input!

 

[ElectricGod]

Here's the cross connects on one of the 20S sections. The negative ends are in the lower half of the connections. I bought 15 feet of 18 awg wire thinking it would be plenty, but that's going to do like 3 or 4 of them so I have ordered 60 more feet of wire.

Why didn't you alternate the cells polarity in the holders and save a few feet of wire. .?

If we're going to suggest different methods... why not run strings of them in parallel? less balance wires too. like this:

+++++-----+++++-----
-----+++++-----+++++

If I understand this correctly...
1. It looks like you are showing alternate every 5 cells with 5 cells in parallel. The problem with that is then I really only have 4S. The battery holders are 4 cells each and I have length for 5 battery holders. I considered running them in 4 parallel strings of 20 cells. The battery holders are squares so orienting them so all the cells are vertical or horizontal changed nothing in and of itself. However, with all the cells oriented vertically I am able to get a 20S pack onto my back deck in a single length. If I orient the holders so the cells are horizontal, I have several issues. I have to isolate the wires and solder pins on each side from each other since I now have a 4P pack that folds over on itself. I also have to connect 2 of them together in series to get the 20S requirement. 5 of the battery holders fits in the length I have available. So horizontal orientation gets me 5S4P. Going back up the other side, gets me 10S4P. I would have to build two packs of 10S4P and run them in series to get 20S. I have enough length to run all the cells vertically in the holders and get 20S. The two 20S1P back to back packs can be 100% soldered together in parallel to make a 20S2P pack. I don't have to worry about isolating anything electrically and all the exposed connections are hidden between the back to back packs where they can't cause issues or over time short against something they shouldn't touch. This was the most reliable option that would fit in the space I had. I have enough width to do 7 20S2P packs, but not enough width to do an 8th pack. 8 widths of battery holders back to back would be needed to do four 20S4P packs. I only have room for 3 20S4P packs which gives me 12 parallel cells in total. Going with 20S2P by 7 packs gives me 14 parallel cells which is definitely better than 12 parallel cells.

2. Balance connectors...One of the things I want...considering that I am using 100% used cells from laptop batteries is a way to monitor as many of the cells as possible. Many cells in parallel "hides" the weak ones. Realistically 2P can still "hide" a weak cell via a strong cell, but the probability is the two cells will discharge more quickly than average due to a strong cell and a weak cell in parallel. This will be easily detected and then a little investigation will reveal the weak one and the strong one. So in this case, more balance connectors isn't a bad thing.

3. I have taken a single battery holder and wired each slot with a 30 watt 800 ohm resistor. This will be a simple testing rig for load testing suspicious cells. Drop a battery (or 4) in a slot and let them discharge for a few minutes at 3000mah. The weak cells will drop voltage quickly while the strong cells will last much longer. Hmmm...I need to get some cheap voltage meters to add to the rig so I wont even need my DMM.

Last night I took 20 cells and laid them side by side and a 20S pack. It's hard to see, but the battery holder is an inch longer than the 20 cells. Not too bad.

 

[ElectricGod]

Here's the cross connects on one of the 20S sections. The negative ends are in the lower half of the connections. I bought 15 feet of 18 awg wire thinking it would be plenty, but that's going to do like 3 or 4 of them so I have ordered 60 more feet of wire.

Why didn't you alternate the cells polarity in the holders and save a few feet of wire. .?

I thought about alternating cells and yes it would make the wires shorter, but I wanted all the cells oriented the same direction so there would never be any confusion which way the cells should be oriented. Being these are battery holders glued together in a 20S string, the cell markings for each slot are all labeled with + on the same end. I could have labeled each slot for correct orientation, but decided to go this route instead. It is a good point. less wire=less resistance. Yeah...Its a good idea...I should do it. I've only wired up a single strip of holders so far so not a huge thing to fix. Thanks for the input!

I'm definitely going to rewire for alternating cells to keep the wire lengths short as possible. I will also have to make half of these mirror images of the other half so when they are soldered together all the cells are still in parallel...and then solder the correct pairs together.

 

[MrDude_1]

instead of making them alternate.... put together however many go in parallel, right next to each other... making them like one big cell.
THEN flip the ones next to it, so its a short connection, and another big cell.

Then the big bars that go across the ones in parallel can be tapped for balancing. and everything is neat and tidy with the shortest paths to everything.

 

[ElectricGod]

instead of making them alternate.... put together however many go in parallel, right next to each other... making them like one big cell.
THEN flip the ones next to it, so its a short connection, and another big cell.

Then the big bars that go across the ones in parallel can be tapped for balancing. and everything is neat and tidy with the shortest paths to everything.

I read this 5 times and I'm still not sure what you are suggesting. Can you draw a picture please?

If I understand what you are saying then that means I would have a single set of balance connections for the whole thing. I would also be limited to 8S rather than 20S. The way I am looking at this is put 2 battery holders back to back and solder all the plus pins together and all the minus pins together for a total of 8 cells in a 1S8P array. Put that in parallel with another 8 cell block or 1S16P. Put the 1S16P in series with another 1S16P block and so on. I can get 5 blocks of length or 35 total blocks or in this scenario it's 4 blocks per 16P or 8S. One of the issues with this is I can do 5 block lengths which means a 1S8P block has to run to the next 5 block length to complete the 1S16P. Maybe a re-organization where I maintain my 14P arrangement would be better. 14 cells (7 on each side) soldered together for a 1S14P block in sereies with another 1S14P block and so on. It would eliminate all the balance connectors and keep my wire lengths short. There is still one problem. I have 20S worth of length in battery holders so the third block of 14 cells would be split between. 8 cells would be on the end of the first 20S length and 6 cells would be at the beginning of the next 20S length. Then would come the next 2 1S14P blocks. That would leave 3 more battery positions to add to the next 20S length to make up the next 1S14P block. LOL! That definitely sounds more complicated and not tidy. Assuming this is what you meant of course. There would be lots of interconnects between 20S lengths.

If there's a way to get better than 14P at 20S with 20S lengths, I sure would like to find it! 20S2P by 7 is the best I have come up with and still be able to monitor/balance as many cells individually as possible with the best reliability.

 

[macribs]

I am with you Dawg. Re read several times still draw a blank. Curios to see if this is something I could use.

 

[ElectricGod]

I am with you Dawg. Re read several times still draw a blank. Curios to see if this is something I could use.

I'm pretty sure all of us can use this...somehow. What a great use for laptop batteries! Not all the cells in the battery pack are bad...just some of them. Start collecting batteries now. LOL

 

[ElectricGod]

I forgot something...the 20S2P blocks will be removable and separate from the power buss via an XT60 connector. It makes accessing the individual cells if needed easy to get at. Something I considered was a buss across the bottom, but that means permanently mounting all the packs to it via screws or solder. The gap between 20S2P packs is about 1/8" which makes accessing the individual cells for replacement impossible without unscrewing or unsoldering something. To make these things easy to replace the cells and easy to understand how to hook them together means keep the connections obvious and independent to each 20s block.

 

[Hillhater]

I'm pretty sure all of us can use this...somehow. What a great use for laptop batteries! Not all the cells in the battery pack are bad...just some of them. Start collecting batteries now. LOL

Have you read any of " DarkAngels". threads re building packs from laptop cells ?
If not, I suggest you search them out and read for a few hours .

 

[ElectricGod]

I'm pretty sure all of us can use this...somehow. What a great use for laptop batteries! Not all the cells in the battery pack are bad...just some of them. Start collecting batteries now. LOL

Have you read any of " DarkAngels". threads re building packs from laptop cells ?
If not, I suggest you search them out and read for a few hours .

I ran across him some time ago and its "drkangel". Part of the reason why I am doing battery holders is so I can swap out weak cells easily. Also, I have my LIPO packs to run off of too. It's not like Im going to use the LION packs exclusively...at least not right away. I think if I load up all the packs with cells and then run on them for a while, I will quickly discover the bad cells in the packs because they will run down fast. Even a cell with a 1C discharge that's good will run longer than a bad 5C cell. The rest of the cells will still be OK. All I have to do is keep doing that until all the cells are "tested". Then I can load up whatever cells I have left into however many packs I can and charge them up and use them like normal. I don't expect perfection ever, but I do expect to increase my overall capacity quite a lot. I'm also aware that some cells will have better C rates than others so by running many in parallel (14P), I can minimize the problems with lower C rate cells.

Maybe later...if I run out of cells from laptop batteries, I can replace them with new 18650's, but that's a good ways off. I have access to probably 200 more batteries that will come my way 10 to 20 at a time over the next few months. I'm just saying that's a load of 18650's.

 

[MrDude_1]

instead of making them alternate.... put together however many go in parallel, right next to each other... making them like one big cell.
THEN flip the ones next to it, so its a short connection, and another big cell.

Then the big bars that go across the ones in parallel can be tapped for balancing. and everything is neat and tidy with the shortest paths to everything.

I read this 5 times and I'm still not sure what you are suggesting. Can you draw a picture please?

If I understand what you are saying then that means I would have a single set of balance connections for the whole thing. I would also be limited to 8S rather than 20S. The way I am looking at this is put 2 battery holders back to back and solder all the plus pins together and all the minus pins together for a total of 8 cells in a 1S8P array. Put that in parallel with another 8 cell block or 1S16P. Put the 1S16P in series with another 1S16P block and so on. I can get 5 blocks of length or 35 total blocks or in this scenario it's 4 blocks per 16P or 8S. One of the issues with this is I can do 5 block lengths which means a 1S8P block has to run to the next 5 block length to complete the 1S16P. Maybe a re-organization where I maintain my 14P arrangement would be better. 14 cells (7 on each side) soldered together for a 1S14P block in sereies with another 1S14P block and so on. It would eliminate all the balance connectors and keep my wire lengths short. There is still one problem. I have 20S worth of length in battery holders so the third block of 14 cells would be split between. 8 cells would be on the end of the first 20S length and 6 cells would be at the beginning of the next 20S length. Then would come the next 2 1S14P blocks. That would leave 3 more battery positions to add to the next 20S length to make up the next 1S14P block. LOL! That definitely sounds more complicated and not tidy. Assuming this is what you meant of course. There would be lots of interconnects between 20S lengths.

If there's a way to get better than 14P at 20S with 20S lengths, I sure would like to find it! 20S2P by 7 is the best I have come up with and still be able to monitor/balance as many cells individually as possible with the best reliability.

heres is a pic to try to clear the idea up.
instead of making two 20s packs, and putting them in parallel, you make two 10s 2p packs and run them in series.
each one of these bars can have 1 wire for your 10s balancing. super simple.
each one can be soldered together easily. I would probably do it with bare solid core wire myself, so its just a big bar.
blue bars are interconnects.
red bars on each end are the positive and negative.
this method will have the least resistance, shortest distance, and smallest number of balancing taps.

 

[MrDude_1]

by the way, if I read that wrong, and its not a 20s2p goal sorry. It was hard to find what your goal was above.
but the concept remains the same.

put them all in parallel for however many cells you plan to use.
then flip them and repeat the same number. over and over until cell count is reached.

physically when you hit the end of a row(your plastic blocks in this case). you just loop it across to the next group of plastic blocks...

 

[ElectricGod]

by the way, if I read that wrong, and its not a 20s2p goal sorry. It was hard to find what your goal was above.
but the concept remains the same.

put them all in parallel for however many cells you plan to use.
then flip them and repeat the same number. over and over until cell count is reached.

physically when you hit the end of a row(your plastic blocks in this case). you just loop it across to the next group of plastic blocks...

There's several things I want out of this build.

1. Monitor as many cells individually as possible since they are all used and may have issues
2. Build battery packs for as cheaply as possible that actually work reliably
3. Since I have access to hundreds of free laptop batteries, leverage them in an EV battery pack build
4. Make the cells easily removable since individual cell failure is probable
5. Make the individual battery packs as robust and as reliable as possible
6. Make the packs removable and individually isolatable from the system for maintenance
7. Make them so they can be used long term and for a wide variety of current loads

 

[ElectricGod]

I used much less wire doing things this way, but still 15 feet isn't enough to finish all 14 of these. I'm about 1 foot short. I ordered 60 feet...I think I might have some extra 18awg silicon wire! These are left and right sides of a 20S2P pack. I have 6S balance cables and 2S cables coming. I thought about using 5S cables for everything so that everything was in units of 5 cells, but since I can charge 6 cells at a time, I thought it might go faster for balancing to do 6 cell chunks rather than 5 cell chunks. In retrospect it is still 5 chunks of cells no matter what. I should have gone with 5S balancing for everything. I'm going to postpone finishing these for 5S balance cables. THe 6S and 2S cables will get used "somewhere" else. This wont be my last battery build. These two slabs will go back to back and then the pins will be soldered together. That's the shortest distance between parallel cells I can get. Thanks Hillhater for pointing that out.

 

[ElectricGod]

I was considering integrating balance charging into the battery packs. I found these inexpensive LION 5S balance boards on ebay. Anyone have a better option?

http://www.ebay.com/itm/5S-10A-Li-ion-Lithium-Battery-18650-Charger-Protection-Board-18-5V-21V-Cell-/331728997076?hash=item4d3c96cad4:g:4PAAAOSwwE5WaJn-

Looking in another thread, I ran across this, but don't know anything about it.

https://endless-sphere.com/forums/download/file.php?id=44290&mode=view

 

[nutnspecial]

Slightly ironic how this lipo thread is now mostly about li now lol.

I am reliably pulling 30amps from a single pack. I do think 70amps is a little sketchy since 30 amps gets a tad bit warm (95F) occasionally.. For this price, why not hook them in 2p for 32ah and a totally realistic 2.2C (This is what I am doing once they arrive).

Thanks for the confirmation lyonnightroad, I was afraid of that. I'm very happy with 11.6ah 30c and 70a, so I hoped to for 16ah and similar draw with better value, but 32ah is way too much for my taste.

 

[liveforphysics]

[
2. Build battery packs for as cheaply as possible that actually work reliably
3. Since I have access to hundreds of free laptop batteries, leverage them in an EV battery pack build

There is an old saying that traditionally applies to hotrods: "Cheap, fast, or reliable, pick any two."

For EV batteries, it's similar. "Cheap, fast, or reliable, pick any one".

 

[Ohbse]

There is an old saying that traditionally applies to hotrods: "Cheap, fast, or reliable, pick any two."

For EV batteries, it's similar. "Cheap, fast, or reliable, pick any one".

Beat you to the punch on that one a few pages back!

Price - of course! This can be broken down in a number of ways, however as Allex touched on the best way to evaluate is on a wh/$ basis. Larger capacity cells (in the same volume) may be more expensive, however you may require less of them.

Power - another way this can be thought of is internal resistance. Higher iR means less power delivered, more heat. The equivilent in Lipo is the advertised C rates. You can actually reverse engineer the HK 'rated' C rates to obtain an approximate cell level iR which you can extrapolate to forecast pack performance.

Capacity - this is how much energy the cell contains. More is better, but there is a compromise to be made - generally as capacity increases, internal resistance also increases. This is why 'power' cells generally aren't right up at the 3500mah level but 3,000mah or less. This comes down to cell chemistry and is a complex topic unto itself. With time has come cells that have both MORE capacity and LOWER iR. These are just plain better. A recent example is the Samsung 30q cell with 3,000mah - these actually have exactly the same or even slightly lower iR as the older Samsung 25r cells with 2,500mah. However of course this is reflected in the price :lol:

Remember that old adage, Fast, Cheap, Good - pick two? This pretty much applies to battery packs too with the above metrics. There are additional concerns around maintenance, reliability, lifespan etc but will try to keep this relatively simple.

EG - you're gonna have a bad time with what you're putting together. Do some small scale testing, extrapolate that to pack scale and work out what's going to happen.

Price - Free! hard to beat, assuming you value your time at nothing. You will invest a lot of time.

Power - rubbish. Used laptop cells vary wildly, however at best you're looking at a cell rated for full capacity at 0.2c. I have harvested ~1300 cells from laptop packs - I know. They are not very useful in anything other than huge numbers on long range, low power builds. Or in my case I used them for ~7kwh of stationary storage for some solar panels.

Capacity - again, lackluster. Depending on the vintage of your cells, you're probably going to get 2.2 - 2.6ah at best. Compare that to new cells that will do 2.9ah at 20a rather than 1a that Luke mentioned.

But mostly why you're going to have a bad time is exactly what you yourself raised earlier - complexity. Now you have many, many more failure points than a comparable spot welded pack. Vibration will ensure this doesn't work very well over any appreciable length of time. Please don't take this as me ripping on you, just calling out reality - there's a reason why there aren't a ton of packs built in this way.

 

[Samd]

There is another way here.
30q cells have better ampacity than Sanyo 18650ga.
But there's nothing to stop you running a higher voltage and lower amps for the same power level if you can find a controller that suits your needs.

All motor winding variants exhibit the same torque for the same power (combination of volts and amps).

 

[ElectricGod]

ohbse...
I agree with most of your points. The laptop batteries are going to vary a lot...admittedly they ARE used cells for laptops. I currently have 150 working cells so I'm about halfway to filling all the packs...it's a start. Right now I can't afford to drop $1000 to buy a stack of new 18650s to get the voltage, amps and capacity I want. These battery holders can be reused once I get 30Q's or something else. Who doesn't need a stack of 72 volt battery packs? To me they aren't a waste of time to build. Regarding vibration issues...the individual cells are really secure in the holders. You have to pry them out with a screw driver! When all 7 packs are installed in the box, the cells can't vibrate out of their holders because the packs are too close together for that. I'll also add about 1/8" of foam between the packs that will suck up any slop between them and hold the cells in the holders even more securely. My hope is that they will do reasonably well considering the cost and what I am dealing with for batteries. I agree that spot welding the cells together would be MUCH better, but that's also assuming I had new cells that had a decent C rating, IR and capacity and obviously these wont be that good. LOL! Don't worry, my expectations are low. I'm not looking for miraculous performance. The LION battery holders are intended to supplement my existing LIPO packs so the LION packs wont be running everything by themselves. Later when I have more $$ to drop into lots of batteries, I'll build up a real pack and do it right later. The disadvantages of battery holders are more like advantages since the crap cells can be easily replaced for free without any rebuilding or disassembly of anything. All said and done I have about $100 or so into these things so I'm not overly concerned about the money I've put into them or the time. This is an interim solution to supplement my LIPOs without buying more of anything. I can get tons more laptop batteries so not a huge concern there. Since I have so many laptop batteries to work from, I can always build more packs to increase capacity. The individual C rates don't really scare me much. Let's pretend that they are all .5C, 2000mah cells that's still 14 amps and 28000mah of capacity. If I get more than that well fantastic and if it's less than that...well that's OK too. I currently get 15 miles off the LIPOs. Rebuilding my 48 volt LIPO packs into 72 volt packs is going to cost me some capacity so I'm supplementing them with the LION packs to get the capacity back. Anyway...I'll take what I can get from this $100 build. If it ends up being great, well then I can brag about how well it worked out. If it completely sucks, I'm out some time and money and I will still agree with you that it's not the ideal solution or the best way to build a battery pack. Either way, It's been fun so far and been worth the effort to me. I could have dropped $100 on lots of things and gotten less enjoyment out of it. More to come as this continues. Obviously I am going for cheap for now.

I'm using a Kelly KBS72221 controller. It's good for 72 volts and 130 amps continuous. It will do what I need for now. I also have a Kelly KLS7230 controller which is rated similarly. I'm pretty much set for now.