Charging Headways with 2 MK3x8SMT BMS units.

[SerendipitySeeker]

I assembled a pack of 16 40152 S cells I purchased from Manzanita Micro to repower a FalconEV scooter. To these, I've attached a pair of MK3x8 Digital Lithium Regulators also from Manzanita and a pair of CellLog8s to keep an eye on the voltage. To these, I connected a 48volt 4amp charger from EVAssemble which is supposedly rated for 58 volts max. After charging the pack overnight and putting about 400watt hours into it, the cells stabilized out at 3.329 to 3.394 volts across the board for an overall pack voltage of about 54 volts. I'm quite happy with the overall even balance of the cells, but the per-cell voltage and pack voltage seem a little lower than they should be. None of the regulators on the MK3x8 were going green or heating up so I expect that the ebike charger wasn't able to deliver the voltage necessary to bring them all up to full charge. The charger is supposedly rated for 58.4 volts (for a 48volt nominal pack) but never got above 55 volts at any time.

I'm at a bit of a loss at what charger I need to buy to bring this pack up to full charge. Does anybody have any recommendations? I'm not savvy enough to rip into the charger and start replacing components. I'd prefer to buy a reasonably priced charger but I don't know what voltage or specs to look for. Should I be looking for a 65 volt or 70 volt charger? Any experience would be greatly appreciated not only for myself but for the next noob following in my footsteps.

 

[nebriancent]

to me it sounds like your chager is set up for 15s lifpo4 or another chem altogether
i would take a reading of the open circuit voltage the charger outputs and go from there

please post resalts at this point so we can figure out where to go from there

it could also be that said battery system needs a few cycles to get into the flow of things

i myself noticed my headways resting a little low after the first charge or 2 but after i ballanced them out for awile they came right up

 

[SerendipitySeeker]

Now that I am able to talk to the BMS via the Rudman Bus from my computer, I asked both BMS units to regulate to 3.5 volts instead of 3.8, their stock setpoint. I did this so I could test the BMS units without having to change the charger yet (3.5v x16s is 56v which the charger may be able to do). Set to 3.5v, most of the BMS regs are now coming into play (shown by blinking green), but one board is showing hot so I shut down the test. I need to fabricate some heat sinks and put a fan on them before I attempt to fully charge all cells and let the BMS do it's thing. I'm not about to fry these costly bad boys. I'll post again after I make it to the next stage.

 

[nebriancent]

3.8v is way to hot for headways (IMO)
i ballance to 3.5 myself and have found that high enugh for me

at 3.5v i find they have 98-110% capacity

 

[SerendipitySeeker]

Test 2
I set the regulators to stop at 3.5 volts and put a huge fan on them. While they stayed cool enough, at one point the regs were unable to contain the voltage. Everything seemed fine for a while, but then four of the cells shot up with one going right over 4 volts and the regs weren't able to shunt enough power off to bring the other cells into alignment. I had to pull the power to keep from frying the cells. The 3.5 volt test with the MK3x8SMT and the 4 amp, 58.4v power supply from evassemble would have gone up in smoke.

Test 3
Set the MK3x8SMT maxvolts to 3.650v and the setover to 3.842v. Connect the power supply, turn on the fan and switch the power supply on while monitoring all cells. One at a time, the regs came into play blinking occassionally. Each prevented the cells from going hitting 3.7v. After about an hour, all regs were blinking like Christmas tree lights without any going full bore or going over 66 degrees f. All 16 cells stabilized out nicely. High was 3.651v and low was 3.629v. Perhaps it's due to the shunting, but the power supply never got over 58.08v so it never got to the critical 58.4v point where it may have shut off on it's own. I want to find a balance point where the cells will all get their full charge, but also where the charger will shut off on it's own. I have not found that point yet, but at least I can charge the pack fully now.

Test 4 will be to find the setmaxvolts point where the charger shuts itself off. It may not be possible, but it's worth some experimenting. I'd like to avoid having to cobble together relays and a Regulator Relay Output Board. While that would be fine in a car, for a little scooter or ebike, what I'm doing already feels like overkill.

 

[SerendipitySeeker]

Test 4,
Changed the voltage for the regs from 3.650 to 3.600 volts (01seth.3.600 and 09seth.3.600 and the overvoltage down to 3.700 volts (01seto.3.700 and 09seto.3.700). That done, I powered it up and waited to see what happened. Did I mention I had to change the address of one of the regs from 01 to 09 so I could address 1-16 correctly?

A few regs on each BMS came into play. About 15 minutes in, all of the regs were blinking and the voltages for each cell stabilized out between 3.600 and 3.690 volts. And, a little later, the power supply shut itself into standby on it's own, it's charge light turned green and the fan shut off. It was done.

So, it is possible to use two Manzanita Micro MK3x8SMT BMS regulators with a 16s 48v 15ah Headway pack and a 48v (58.4v max), 4amp LiFePO4 battery charger from evassemble AND have them play nice together. It just takes dialing the regs to 3.6vmax and 3.7vover and so far it's working. I'll report back once this setup has been built into the scooter and tested in practice. But, at least on the bench, it's working.

The charger is model KP4804JL, 90-120vac input.

 

[SerendipitySeeker]

OK, so I was a little cocky after my last post. I got distracted on my way out the door and left the charger on. It ran all day without ever shutting off. I got home and while the battery was well and completely charged and all the cells were still being regulated to the correct voltage. The MK3x8SMT did it job quite well and kept the pack in line even while regulating and holding a temp of 132 degrees with fan, all day. I'm pretty impressed even though I was a dumb ass for leaving it on the charger all day unsupervised.

Test 5. I had to find a way to reliably get the charger to self shut off when charging was complete. I figured that the charger simply had to get to a voltage where it would call it quits. So, I upped the high voltage limit to slightly beyond the recommended 3.650 for the Headway cells. I told the regs 01seth.3.680 which set it to 3.672v in actuality. Did the same for the other regulator 09seth.3.680. I set the over to 3.750 01seto.3.750 and 09seto.3.750 which resulted in both responding that their overvolt limits were now 3.750v. Good enough.

Now that all 16 of the regulators have been set to slightly higher limits, this should let the charger run up to just high enough voltage to shut off. In practice, it worked twice in a row reliably. The charger which says it goes to 58.4v shut off at 58.15v. My second charger which says it goes to 59.2v shut off at 58.48v. Both did so without making the regulators run very hard at all, which kept them cooler than with the lower voltage setting. For now, the system seems pretty stable. I'm sure I'll have to tweak further once the build gets finished and used for real commuting.

I'm documenting all these settings and commands because I couldn't find anybody else who had been down this path and documented it thoroughly. Perhaps this will save the next noob like myself some time and perhaps burnt bits.

 

[dnmun]

after a pack has balanced, there is only a very short time that it spends in balancing after that. so there should not be much heat produced by the shunt resistors once the pack has balanced. most people don't wanna spend the money for just a simple balancing BMS so i doubt if there will be many spending this kinda money to balance their pack.

i think the headway BMS is superior in any case, especially for the money involved.

 

[SerendipitySeeker]

I tend to agree that all I really need is a balancer, but my challenge has been finding one that works.

When I started learning for this project, I didn't want to spend much before I knew what I was doing. I started with four mismatched cells from a blown pack and started experimenting. I bought a MK3x4SMT Digital Lithium Regulator from Manzanita Micro for around $200. While I wished I could have made heads or tails out of it, I couldn't figure out how to talk to it through a PC or get the settings right. The online video explains how to connect wires in order. The printed documentation is pretty sophisticated. My missing knowledge fell neatly between those cracks. After beating my head against the technical documentation gap and online forums for far too long, I gave up on it.

Next, I bought a simple Capacitor-4S-30A-BMS from EVAssemble. It cost $27 plus about twice that in shipping and a month of waiting. No adjustments, it just worked perfectly. After hooking up an old laptop power supply to the pack and watching the voltages from the CellLog8s, I was amazed to see all four wildly mismatched cells come into alighnment and settle in within 10mv of each other and stay there. Now that I had spent some education money to get up to speed on what balancers would work for me, I was ready to spend the real money on my pack. Based on my experiences, you can guess what I bought:
16 x 40152 Headway cells from Manzanita Micro. Given the shipping costs and wait time from China, Manzanita Micro was far cheaper and faster than any Chinese alternative, and they brought them to the Seattle Electric Vehicle Association meeting so I didn't have to pay for shipping or wait anyhow. Big win.

Next, I bought a Capacitor-16S-30A-BMS from EVAssemble. While the BMS arrived in a few weeks, the charger took another month to arrive. Good thing this is a side hobby project and not transportation I need to depend on. After connecting everything according to the diagram, the BMS turned out to be a dud. A few of the cells shot up far far beyond what Headway would ever warranty. I caught it before anything exploded and pulled the plug. The BMS failed to do any part of it's job. No balancing, no overcharge cell protection, failed the undervoltage test. Given the perfect performance of the Capacitor-4S-30A-BMS from EVAssemble, I was hoping for equally great performance from the 16S-30A-BMS, but luck wasn't on my side this time. Given the shipping costs and lead times, it wasn't even worth attempting to begin a conversation about warranty issues. Buying directly from China without a US intermediary for warranty coverage is buyer beware. I lost on that gamble, but only $100 or so including shipping so it's pretty cheap as far as life lesson expenses go.

Given that this scoot project is going to see some heavy real-world use, I couldn't afford to fry my pack because I saved a few bucks on a BMS. While the Manzanita Micro had baffled me earlier, it was worth doubling down on them. So, I bought the two MK3x8 Digital Lithium Regulators, the Dongle Terminator and some harnesses for about $800. About 8 times the price of the EVAssemble equivalent, but clearly more sophisticated and with higher component quality and much more rugged construction with real screw mounting options. After almost frying my 16 Headway cells, I couldn't risk it again.

It took a lot of hours to figure out how to send commands to the MK3x8 Digital Lithium Regulators correctly and get the monitoring working right with a laptop computer, but eventually I figured it out. Perhaps one day, I'll make a YouTube video so others can learn how to talk to them too.

So, you are correct in that I only need something like the Headway or EVAssemble balancer. That would be my preference due to the price, size and heat challenges with the Manzanita Micro BMS units. But my bigger need is for a BMS that is of high quality, can be bolted down properly so it doesn't come loose on a bouncy two wheeled jalopy, and is dependable enough for me to risk my investment in batteries and not burn down my house. Given my short history in these lessons, I can't say I'd make different choices than the Headway40162 Cells, the CellLog8s and the Manzanita Micro MK3x8 Digital Lithium Regulators. I reserve the right to keep learning as time goes on and change my mind.

 

[dnmun]

jimmyD is right there in seattle. he coulda sold you the headway cells too or even the newer headway cells, and he also has the 16S digital headway BMS. this is what paul uses on the cell-man A123 packs. digital controls, thin and high powered, they have 14 of the RUI chip 190A mosfets, 3.4mohm in parallel so the power lost in discharge @120A through the mosfets is only about 25mW.

go look at the 16S BMS he offers here:

http://stores.headway-headquarters.com/-strse-92/48v-50a-fdsh-100a-bms%2C-16/Detail.bok?category=BMS%2FPCM%2FPCB

 

[SerendipitySeeker]

Thanks, I've seen that BMS on the HeadwayHeadquarters site before, good to know that it has a good reputation. How would one go about properly bolting that to a flat surface? I don't see any hard mounting tabs or threaded holes. I'm sure the pancake design is fine if all you going to do it to double side tape it to something or shrink wrap it into a pack.

 

[JimmieD]

Thanks, I've seen that BMS on the HeadwayHeadquarters site before, good to know that it has a good reputation. How would one go about properly bolting that to a flat surface? I don't see any hard mounting tabs or threaded holes. I'm sure the pancake design is fine if all you going to do it to double side tape it to something or shrink wrap it into a pack.

Sorry to hear of the troubles that you were having but good to hear that all is working out and that on top of that you are taking it in stride (as best as possible) as lessons in learning about your battery system. There are a few options on mounting those PCM's/BMS's, The most common way that we use here is to use silicone adhesive such as GC Electonics Part No. 19-155 and mount it to a piece of polycarbonate which then can be mounted in a location. The other option that I have done for a battery pack for a competition go-kart was to remove the aluminum heat sink cover, use a thread tap and, a ribbon of thermal paste and affix an aluminum piece of 'L' metal with mounting holes drilled into it similar to what the MK3 has, re-attached the heat sink cover and the owner was able to affix it where and how he wanted.

Bolting it to a flat surface in lieu of 'L' metal, one could take a piece of flat aluminum .125" bar and attach it as with the 'L' metal and it may appear to be 'upside' down, but the backing of the PCM/BMS protect the underside of the PCB and components and even then, one could use silicone adhesive to attach a piece of polycarbonate or ABS for added protection.

You are correct though, there are no mounting tabs or threaded holes. Also, just for clarity, if one were to remove the aluminum heat sink top and affix mounting tabs as described above, it would void the warranty since dismantling and dis-assembly of the PCM/BMS would most likely be required. So a steady hand and good fabrication skills would be required to do this, er-hum, unauthorized mod. Simply using silicone adhesive on the 'underside' of the unit to affix to something would not void the warranty.

Double sided tape, shrink wrapped into a pack, yellow film tape, etc. not recommended! :wink:

Hope this helped with some ideas!

 

[SerendipitySeeker]

Thanks for the mount fabrication ideas Jim. It's great to have that information on the forum so many people can benefit. I'll try one of those methods on my next build. Given my fondness for sturdy, I'll probably opt for the unauthorized mods.

If you have an in with Headway the company, any chance you can ask them to make their BMS easily installable? I love the Headway modular battery concept because I can build a quality pack of any shap, voltage and amp hour capacity by snapping together the Lego pieces. I love it. They'd have the whole package if they could apply the same design talent to their BMS mounting as they did to the battery assembly design. It couldn't cost them much really to make their BMS with threaded holes and mounting tabs. The company can really benefit by differentiating themselves from every other BMS out of China by improving the physical design without changing the electronics. While they are at it, it would be super cool to have electrical mounting lugs so I could connect high quality wires of my choice (welding cable) with good sturdy lugs.

 

[dnmun]

no, you cannot use bolts or drill and tap. the aluminum plate is there as a heat sink for the mosfets and cannot be removed.

it has a cardboard fiber/plastic coated backing on the other side of the pcb to protect the traces on the underside from shorting.

headway doesn't manufacture the BMS, they source it from someone else from what i understand. paul uses it on his A123 packs too.