sub-C cells are not discharged at 0.2C in the tool. current draw will be more like 2.6A in a bosch, so capacity is much less than the nominal capacity. try to fit in Eneloop AA. when bosch made those tools eneloops weren't a thing, not known to be superior.
i could never find half decent 4\5 subc, i tried several different, none lasted long,
today the only half decent subc are made by tenergy, but even with that brand it is hit or miss, they either have several factories, or there are fakes. some of their cells worked great, but others died pretty quick, or maybe some distributors kill them by storing them in wrong conditions.
i had several packs that i build die due to being overheated externally, wonder what happens when you keep a container full, at a port in a hot summer sun for long period of time.
the best results i had was with subc nicd from tenergy, yellow wrapper, 2,4ah iirc. i replaced them in 14,4v milwaukee drill, they worked just as good as oem pack that is made in japan, no loss of torque under any condition. (i eventually killed them, by hooking them to 150w 15v halogen bulb, and forgetting about it for a day, discharged them flat, to 0v, they never recovered.) but it was a while ago, i can’t be sure if it is the case today.
tempo, I am aware, but the point of a 0.2C discharge is to have something to compare, a sort of benchmark. Like flashlight throw which, as defined by the ANSI-NEMA FL-1 standard, is defined for an illuminance level of 0.25 lux (around a full moon), which is clearly dim imho.
Thanks alpg88. For the moment, we'll see. I may send the seller a message or two concerning cell pack information.
I have some experience with sub c nimh. I was into RC and they came into popularity in the early 90’s before being surpassed by li-ion technology. The Tenergy 3000mah and 5000mah (genuine ones…lots of fakes out there that are junk) are pretty robust and are pretty LSD. However, they don’t like being deep cycled or overcharged. Keep charge rates low, like 200-300mah to get the most capacity. They don’t need a balance circuit either because they self balance. I rebuilt an inflator with some 3000mah ones (6S) and it pulls 7+ amps. I can’t run it for more than 5 minutes at a time if pushing it hard though for cooldown.
also nicd\mh need to be broken in, about 10 cycles of charge \discharge should bring it to optimal performance, also those type of cells do not like idling, deep self discharge shortens their life considerably
True. That’s why most everyone I know running the nimh packs had “advanced” chargers that did cell conditioning after every 5 charge/discharge cycles.
I am considering using four of these to replace the 12 sub-c cells in my old Roomba batteries. The other option I am considering is eight of these cells . I would pair up a couple of those 18650s in parallel then put each of those sets in series to get to 14.8V. On another forum I found where someone did that with good results; they were even able to still charge the Roomba with the stock setup. They appear to have used this PCB for protection.
@Barkuti
Thanks but they don’t have the form factor for my Roomba. Even if they did the battery packs look rather pricey. Amazon has Tenergyreplacement packs (with the shell/case) for my Roombas for about $18, but from reading the reviews I would only expect them to last a year at the most. Honestly I would be worried about the longivity of those AliExpress packs too!
bcm00re, if price is around the same go for the 4S2P HE2 setup, will perform a little bit better than 4S1P M50s. However, since the Roomba isn't really power hungry you could get some more runtime with decent energy cells like LG MH1s in 4S2P (in case cells need to be LG).
Wellp, after inquiring the seller he/she told me the above cell packs weren't 4/5 sub-C, I guess this means they use sub-C cells but @#$%, while I still haven't seen the GSR 14.4 VE-2 battery packs to my eye they look to have the same thickness than that of PSR 1200 drills, and these use 4/5 sub-Cs for sure.
If sub-C, the Turnigy sub-C 5000mAh (cough!) 4400mAh is the cell to get. Sold at Hobbyking, but whether standalone or in packs @#$% with the stock lately.
@Barkuti
Sorry to digress again, but those HE2 cells sold out right after I made my post. 
I am now looking at this LG F1L cell. They have a max discharge rate of only 4.9A, and doing 4S2P means the battery pack would have a max discharge rate of 9.8A — but I don’t know if that is sufficient for a 400 series Roomba. I know some aftermarket Roomba li-ion batteries (that are no longer available) used cells that had a max discharge rate of 12.0A.
EDIT:
After some deeper inspection I think it would be overly complicated to fit eight 18650 cells inside a 400 series Roomba battery case (along with the protection circuit). So I have now turned my sights to this 21700 cell with the plan to use four of them. It is spec’d at 3750mAh but testing shows it is more like 3500mAh — but I think that is at least a good as the original Roomba battery. These have a max continuous discharge of 30A which means there is no concern about these cells providing enough current. 
i never found 4/5sc cells that were worth a damn unless i bought sanyo/panasonic $$$.
it was always easier to find the full size packs and rebuild with sc.
and even then it was sanyo/panasonic to get an end product that held up.
i watched several friends go through the crap aftermarket dewalt packs over and over.
while the ones i still have were rebuilt with old stock sanyo 1900’s a friend got in huge quantity cheap.
these were about 10 years old at the time.
those are still going while all those cheap aftermarket packs are in my rebuild pile for their cases.
That Eizfan cell is a rewrap of some chinese OEM, probably Lishen. It's an 0K choice but, why these cells? I mean, the Roomba doesn't really needs that much current. You would be better served by, coincidentally, a super affordable Lishen cell: the LS2170SA. The LS2170SA is sold by LiitoKala as its Lii-40A, check it reviewed at lygte-info.dk. Can be bought with tabs, a convenient thing for battery pack assembly. At the liitokala Official Store, for example. Advice: price is right now a bit higher than what it should, there's an AliExpress sale in close to 4 days and some sellers bloat their prices for it to seem they are making discounts; I bought 2 of these cells without tabs a few weeks ago and I just paid €5.32, they're at €7+ now LoL.
Also, if you can fit 4x 26650 cells you can enjoy 5000+ mAh capacity for very little additional money, with the Lii-50A (Power Long Battery INR26650-50A rewrap). 
I hear you snakebite. The above Turnigy sub-C cells are performing great, though.
@Barkuti,
I didn’t notice that Efan was a Chinese cell — that does give me some pause. I can get the Efan cells for really cheap on clearance from a store in the US. That said since they are really 3500mAh and not 3750mAh like the wrapper says, that means the bang for buck isn’t what I thought it was. Taking that into account the LG 50T cells are actually a little better (mAh/
.
I was also leaning towards the Efans because they’d be more in line with the capacity of the original battery. Part of me wonders if I went to the 5000mAh LG batteries, then I could be running the robots considerably longer than they were designed for. I fear that could lead to wear problems on gears, motors, etc.
Lastly, I am planning on using one of the 21700 battery holders I mentioned in this thread: 21700 battery tray recommendations
Thanks again for all your input!
@Barkuti
So one of my BMS circuits finally arrived. It is the 2S one for my Eureka stick vac. I assembled the new battery pack using two LG 50T 21700 cells to replace five sub-C NiCD (or maybe they are NiMH) cells. I connected everything properly but no power. I then “woke” the BMS circuit by applying 8V briefly then I read power on the input/output terminals, but when I try to turn on the vacuum it only very briefly turns the motor then it shuts down and there is no voltage. When I bypassed the BMS and the vacuum starts and runs just fine on just the two 21700 cells. The BMS is rated to do 8A with 15A cut-off, and I measured the vac pulling around 5A when running and a bit higher during start-up. So I don’t know what is going on. I then hooked the new battery pack (with the BMS) to an old 12V drill and it runs that fine. I even applied some pressure resistance to the drill while using my clamp meter and confirmed the pack could supply peaks of over 7A and 6A sustained (kind of briefly). Anyone have any idea why the new battery pack and BMS won’t run the vacuum?
To be clear, the original Eureka battery contained only five sub-C cells nothing else at all — and it only has two connections/terminals.
bcm00re, all I can say is that the Eureka stick vacuum cleaner must be pulling a very high motor startup surge current, and the BMS shuts off. Get a bigger BMS, or use another one of the same type in parallel with it (and preferably thermally coupled).
On another forum someone pointed out that a DC motor that pulls 4-5A continuous should pull much more than 6A at startup; so it appears my meter wasn’t catching the peak startup current correctly. I confirmed this by attempting to power the vacuum with my variable power supply (that I found after a little research is rated to supply 15A with peaks of 18A) and found it cannot start the vacuum either! I was able to use the new battery pack and my power supply together to get the vacuum started, then confirmed just the new battery pack runs it just fine. So I need a much, much larger BMS circuit — Barkuti suggested.
no meter can catch the peak correctly, it happens in milliseconds and can get up to 10x of running current, with some motors,
Yes bcm00re needs a BMS capable of handling more current. And hope the batteries are not the culprit, because M50T cells aren't high current, and if the vacuum cleaner pulls a very high startup current the voltage from the cells may drop too much when state of charge is low. If you look at HKJ's M50T review, he only tests at up to 15 A which already is (continuous) overboard for the M50T. At 20 or 25 A of inrush with cells at low state of charge, the voltage drops below 2.8 V. Henrik's tests are done at 25 °C, but bear in mind cells don't perform that good if ambient temperature is colder (higher resistance when cold and even more voltage drop). Shouldn't be a problem, but see my point.
The following is the BMS type I once used for a small electric screwdriver with 2S Samsung 25R cells, bought them from sunmall back then but I see it cheaper now at HeltecBMS:
I installed two BMS boards in parallel, that way I made sure I wouldn't encounter any unwanted “inrush” or overload cut-off (the overcurrent protection rating actually is low for 25R cells). And the stuff worked very well.
Also, in case you may want to know the only wires to the BMS which need to be thick are the one from the battery cathode to the BMS (from the upper cell's negative/cathode to the BMS “0V” pad in the picture), and the one which comes from the BMS negative output (the black colored line). Neither the wire from the lower cell's positive/anode to the BMS “8.4V” pad nor the red line need to be thick, this is because the BMS only needs a very tiny little bit of power from the battery. Thenceforth, only a small wire is needed from the battery anode (lower cell's positive pole) to the BMS's “8.4V” pad, and no red line. The thick positive wire to the motor is directly wired to the battery anode (lower cell's positive pole) in this case. But if you do like in the picture, then yes both these wires need to be thick.
As I mentioned earlier in this thread, I tested the LG 50T cells wired directly to the vacuum and they had no problem starting it (or keeping it running of course). I fully charged them and didn’t let it run very long so as to be sure I didn’t over discharge them since I was running them unprotected.
Thanks for the other info! I ordered one of these BMS boards last night — it looks very similar to the standard one from sunmall that you linked (which I hope will be fine since I won’t be using it for charging). I also ordered a pair of these too because one thing I don’t like about the BMS board that I have (and I suspect the others might might be the same way) is that you have to apply some power to “wake” it, but I plan to removed these cells and charge them on my XTAR VC4 so that means I’ll have to wake them each time I re-install them. I am thinking those 1S protection circuits might not have that requirement. Both items won’t be in until May or June so it’ll be awhile before I get to experiment.
Without BMS 0 problems can be. No BMS means no complaints if the voltage drops a little bit below whatever thresold upon startup.
You also didn't needed to be so over-caring with the cells, even if the vacuum cleaner drained them at an average of ≈10 A you would still have close to 27 minutes of runtime until depletion from full state of charge. You could even let it do it, at some point the device would notice a more or less sudden voltage drop and its electronics would likely complain and stop working even before it is really dangerous for the cells. I have recharged momentarily over-discharged cells after certain tests, this is harmless because you don't leave the cells to rest in that condition.
It is always wise to use proper battery management systems, in any case.
That eBay 2S BMS is the same one I linked from AliExpress sellers, non-balance version.
Well, BMS boards require to be “awakened” to operate after they're connected to cells, the controller chips are designed to work this way. I once also bought an 1S board pretty much like the one you link, with 6x 8205s, for my hair clipper mod. I ended up uninstalling and throwing it away, didn't knew :facepalm: I needed to activate it. What times those 
were my God! The hair clipper works fine without BMS, it has my love and light :-) though.