It's the BMS, Stupid!

[llile]

JohnRobHolmes can tell you I was stumped yesterday, but diligent troubleshooting has revealed the problem: Stupidity. In fact, that should be the first thing to check for. Look in the mirror - does the reflection resemble Homer Simpson? Then there is your problem.

Have a setup with two motors, two controllers, one throttle, one battery. Battery is two 36V PINGs in series for 72V. Wired it all up, hopped on, and took off. It went about 25 feet and stopped. Checked for blown fuses, dead batteries, everything seems to be fine. Plugged it all in, hopped on, went about 25 Feet and stopped - no power when I twist the throttle. Power returned when I unplugged all the batteries and plugged them back in. Same thing happened again.

JohnRob's first theory was a bad throttle. Not a bad theory, since throttle failures are common.

But here is a simple troubleshooting sequence that tells the tale:

Disconnect e-brake override plug, in case that is the problem. Didn't make any difference. Keep it simple.

Leave CA plugged in to two controllers. CA chows 79V on the batteries. Bike up on a stand, so motors spin free. Ease on the throttle slowly, motors work fine for some time. No problems.
Then goose throttle hard to WOT. Everything stops. A glance at the CA shows 38V on the batteries! Where did my 79V go?

Unplug all batteries and check terminal voltages. one is 39V, one is 38 V. Fuses all OK. Everything seems fine. Plug them back in, CA shows 79V again.

Now plug in just the throttle, to both controllers. Unplug the CA. Rev it up, both motors spin fine. Goose it to WOT, everything works fine. Check battery voltage. It is 38V again! But it worked.

Unplug all batteries, check voltages and fuses, plug them back in. Batteries seem happy. Now CA measures 79V on the batteries.

It looks to me like one of the batteries' BMS is cutting it off. Likely there is one charged battery and one discharged. BMS cuts out on low voltage, disabling the battery. When battery voltage drops below 61V, where CA is set, then CA inhibits throttle via throttle override to the controller. If the CA isn't connected, then controllers just see the throttle signal, without interference from the CA throttle override. The motors will still run on a stand at 38V or 79V, they don't care. They have no load, except bearing friction, so they spin fine on half voltage.

Sure enough, I put both batteries on chargers, and one of them lights up like a christmas tree right away, while the other one sucks up juice like a drunk on New Year's Eve. I have a rule - always charge batteries after every ride, so I always assume my battery is charged when I hook it onto the bike. But this time, I musta fergot to charge. In a two-battery system, with antiparallel diodes, even if one battery is disconnected, current still flows (this may be hard to understand but trust me on it) albeit at half voltage.

So this leads to a question: How exactly does a PING BMS behave under these conditions: Excessive current, and low battery voltage? Ping's documentation is characteristically brief on this point, mentioning a "high current override 40A" but not explaining exactly the sequence of operation. Once the BMS cuts out, when does it decide to cut back in - when ALL loads are disconnected? Seems like that must be the case. Who knows the sequence of operation for sure?

 

[dnmun]

you should order a 24S BMS from ping and rewire the two packs into one 24S pack. you have to use a soldered permanent physical connection from the top of #12 to the bottom #13 so the two packs would not ever become even slightly disconnected. you cannot use the anderson connectors or other connectors, it has to be soldered or two ring terminals bolted together. using the two 12S BMSs in series increases the power loss through the mosfets.

then you can hook up the charger in series and charge up the pack that way or make another charger for 89V with a power supply and charger combined in series.

there is a way to rewire the two 12S BMS into a series connection to make it so they will handle the 24S but it requires more experience and electronics understanding than you have now so maybe later you will know enuff to try. for now, you would be better off buying the 24S BMS and save the two 12S BMS for the time when you use the packs later as 36V packs.

 

[llile]

Charged up both batteries. Same result. See post below.

 

[llile]

there is a way to rewire the two 12S BMS into a series connection to make it so they will handle the 24S but it requires more experience and electronics understanding than you have now so maybe later you will know enuff to try. for now, you would be better off buying the 24S BMS and save the two 12S BMS for the time when you use the packs later as 36V packs.

You are right, I really don't know much about electronics at all. Just a Novice Noob. I have a degree in electrical engineering and hold five patents on electronic devices. My first job involved fixing TVs and CB radios. I've built robots, computers, solar trackers, oven controllers, cryogenic controls, production test equipment, fake crickets, and god knows what else from a pile of parts and a custom PC board. I worked for ten years in an electronic prototype shop, after I left the TV shop. The first electronic thing I ever built, after the obligatory crystal set, had tubes in it, and the last one was programmed in C. I have a fully capable electronics shop sitting right next to my computer. I have PICs, Arduinos, oscilliscopes, power supplies, and a soldering iron sitting around handy in case I need them. Most fun I ever had working was when I worked on a piece of the Keck telescope, building cryogenic controls for some of the machinery that built the lenses. I don't suppose, however, I am qualified to understand anything about electronics, given such little experience, and am looking forward to learning a few basics.

I will take your advice look into the idea of the 24 cell BMS, and see what that entails. As for now, I'm using a diode network recommended by Ping for series'ing these two batteries. I'm pretty set on staying with this arrangement, as it allows me to switch the same set of batteries onto my 36V bike.

On to the real problem. More testing (simple testing, as that is the best kind) has seemed to reveal my BMS cutout hypothesis has some merit. Here are some more factoids to digest:

1. Setting the CA low voltage limit to 32 volts eliminates the problem of throttle cutting out and becoming unresponsive. It was the CA defeating the throttle all along, on low voltage cutout, through the throttle override signal, as I thought. Batteries still kick out though.

2. When I first hit the throttle there is a surge of current. I'm using a digital meter, without much ability to capture fast transients, so it is not easy to tell, but it looks like it is maybe tens of amps. No matter how gently i try the throttle, a surge occurs, and If I goose it, the surge is quite a bit larger, and then a battery kicks out. This fits with my theory that the BMS 40A current limit is being reached briefly, causing the BMS to go into shutdown. Once that happens, the pack is again delivering 36V instead of 72V. Might have to dig an old analog meter out of the closet to get a better fix on the current surge.

3. Made up a quickie set of anderson plugs that allow me to run at 36v. I'm seeing the same problem when I rewire everything to get 36V instead of 72V. Battery pack still cuts out if I goose the throttle.

4. Calibrated the RShunt setting in the CA. It wasn't far off, according to the ammeter. Now I know that the CA is really seeing the current it thinks it is seeing, so less GIGO is going into the throttle override calculation.

5. It is looking like the CA's amp limit override doesn't kick in very fast, allowing this brief surge of current to get through. Although I played around with very low amp limit settings (1 amp, at one point, when I was calibrating the RShunt value) I can still kick the pack into shutdown by twisting the throttle fast. Likewise playing with the CA amp gain settings, I could make it more difficult, but still likely, that one or the other battery would decide it is time to quit delivering juice. CA throttle override isn't fast enough on the draw [pun intended] to keep the surge under control.

So here is a distinct possibility: There is just too much motor for the battery. Dual HS3540s and series 15AH 36V Pings, 40A motor controllers. This could be an expensive problem to fix. The simplest, and most expensive, fix would be to buy a lot more battery. Yow. Might there not be an elegant set of controls/settings/hacks that can keep these racehorses in check? I spent a good part of today looking for one, and enjoying it thoroughly but unproductively.

The other possibility: using the CA to process the throttle signal, instead of sending the throttle signal directly to the motor controllers. Might that result in less of a surge? Hmmm.. I'd sure like to be using the CA's amp limiting function to keep the surge from happening in the first place.

 

[llile]

So here is the question in this long rambling thread:

So this leads to a question: How exactly does a PING BMS behave under these conditions: Excessive current, and low battery voltage? Ping's documentation is characteristically brief on this point, mentioning a "high current override 40A" but not explaining exactly the sequence of operation. Once the BMS cuts out, when does it decide to cut back in - when ALL loads are disconnected? Seems like that must be the case. Who knows the sequence of operation for sure?

 

[DAND214]

Are the controllers, programable? If so cut it down to 30 or less to see where it shuts down.

Have you tried just 36v to find which pack is shutting down.

Just some ideas.

Dan

 

[dnmun]

ok, i did not realize you knew something about circuits. it is really really simple if you understand the way the HVC and LVC opto transistors work.

in that case you can use each charger for each 36V section, and then just have the current flowing through only the output mosfets on the bottom of the 12S BMSs.

if you look at the V2.5 BMS, there is small white triangle next to the sense wire plug with two through holes inside. the circuit current comes up to the outer through hole, and crosses over to the other through hole and then returns to the top of opto transistor string. and there is 14.75V or so on the collector of the top opto transistor. when all those optos are turned on then the voltage at the bottom of the chain that is on inside is used to control the LVC of that section and voltage should appear on the emitter of the bottom opto. you will need to attach a small, as in tiny, wire to that emitter on the LVC chain of optos and use a 10-12k resistor as current limiting for the wire. so you can cut the emitter leg on that bottom opto, and stand it up so you can solder a 12k resistor to the leg and then solder the wire to the end of the resistor and that wire goes to the bottom BMS in the two pack.

in the bottom BMS for #1-#12, this will be the BMS that is used for the current control, so only the one set of mosfets has to carry the current. so in order to carry the LVC control for the upper BMS or channels #13-24 on that wire from the emitter, you will use an opto transistor just like the ones that ping uses in his opto chains. this opto will complete the circuit for the circuit current in the lower BMS so the lower BMS will be turned on when the opto is turned on by the LVC signal from the upper BMS. when the LVC signal of the upper BMS disappears because of hitting LVC in the upper section #13-24, then the lower BMS will be shut off too.

in order to do that, you will use the two through holes in the small white rectangle on the lower BMS to hold the transistor or output side of the opto transistor. first you have to go to the back side of the pcb and cut open the trace between the two through holes that carry the circuit current normally. so when that trace is cut, the circuit current for the lower BMS will have to go through the opto transistor. to turn on the opto transistor, you use the current from that wire coming off the 12k resistor on the emitter of the end of the upper opto chain. which will normally be about 12-14V between B- of the upper pack and the emitter of that bottom opto. remember to install the collector of the opto transistor in outer, higher voltage one of the through holes, and the emitter will be in the hole closer to the sense wire plug. it will actually be turned upside down with the legs folded back to the top of the opto too so the cathode of the diode will be closer to the end of the sense wire plug.

the current limiting resistor allows you to carry about 1 mA, because the bias across the opto diode is about 990mV-1V. so the wire from the upper BMS comes down to the anode of the 817B in that through hole on the lower BMS. the anode is next to the dot on the opto. and the cathode is connected directly to the top of #12 in the lower BMS. you can solder that cathode over to the top post on the sense wire plug where #12 is connected. maybe use a solid wire to the back of the pin in the sense wire to attach the cathode leg, and that will stiffen the opto which will be standing upside down with the diode section up above the pcb so when you solder the wire to the anode of the led it will not flop around and break off from vibration. maybe hot melt glue to stiffen it too.

so the circuit current in the lower BMS is on when that opto is on.

so that is the wiring of the BMSs, and you will just make a direct connection between B- of the upper pack and the B+ of the lower pack so the cells will make one continuous 24S string without being interrupted by the mosfets in the upper BMS.

this is actually safer for the situation where you have two separate packs than using a 24S BMS with separate 12S packs. if the connection between the two sections is loose and the connection is open while current is flowing then the current normally flowing through the battery tries to go through the shunt transistor and burns up the shunt transistor on channel #13.

in this situation if the wire connecting B+ of the lower pack is disconnected from B- of the upper pack then the opto will turn off, which turns off the mosfets and so no current would be flowing, except there will be reactance so maybe a diode would help. haven't worked that out in my head.

i know it seems confusing, but i think if you look at how the circuit current trace runs up the side of the pcb to that white rectangle it will become more sensible. maybe somebody will take pictures and photo shop the stuff i was talking about for you.

for charging you would connect the charger negative to the C- of each BMS, and the charger positive to the top of the cell in that section, either #12 or #24. each charger would be controlled by that BMS, and the circuit current for the LVC is not part of the HVC circuit for each BMS. so when the BMS sees HVC on one cell in one section of the pack, it will only cut off the charger for that section, and the other section can charge up while that section is off. or you could charge up the two sections one at a time with the same charger. you would put the charger on until the top section balanced, then move the charger to the lower section and charge it up, then both sections would be charged and balanced.

also since you know about the white rectangle now, you will realize that you can install a switch there to control the BMSs, either a dip switch mounted directly into the through holes or connect a long wire up to your dashboard and use a switch there to turn the circuit current on and off in the upper BMS, which will turn off the lower BMS too, so the mosfets will be off and the battery will be deactivated at the output terminals so it will not spark when shorted or connected to the controller. remember to cut the trace between the holes when you install the switch.

you will not need the P- terminal of the upper BMS since the upper output mosfets will not carry current. just the P- of the lower would connect to the controller.

i have some 817Bs too if you need one.

 

[dogman dan]

There's yer problem. 80 amps of potential amp spike on every start. Real world less, but easily could be enough to pop the bms to shut off.

Too much motor and controller for a battery that at maximum, is comfortable with 30 amps discharge. In reality, you will damage it some ( less cycles) with very much continuous discharge at higher than 25 amps, even if the bms lets you do it. One 30 amps controller has been done with sucess on that size pingbattery, but likely they don't actually discharge at 30 amps very much. I tend to recomend 25 amps as the max size controller for a 15 ah ping.

I've never had an overamp shutdown on my pings, but when you have a low voltage cutoff, unplugging and replugging the battery to the controller does something that resets it, if the voltage at rest is still high enough. I have run a 40 amp controller on a 15 ah ping without the bms cutting off, but I don't do it regularly because it will murder my battery. I just did it once as a test. It worked, generally pulling just 20 amps at speed, but the amp spikes near 40 amps on starts cut my range in half.

I can safely say, pingbatteries hate big amp spikes. I severely unbalanced one with repeated 30 amps spikes in a race, after just a 40% dod discharge.

Only quick fix I see, is remove one motor and controller. And I'd advise turning the remaining controller down to 30 max, if possible. Lyens sells a 72v 20 amps controller for less than $100. It will be pretty lame off the line, but at least you'd get 25 or more miles out of it, and likely still exceed 30 mph on the flat.

 

[llile]

Thanks, Dogman, that's one good approach. You seem to agree that I've got too much motor for the battery.


Here is another tack: i only really need about 25A at 72V to run up a hill with this rig. I had planned to use the Cycle Analyst to limit it to that. I am using 40A controllers on Justin's recommendation, in order to make sure they are robust. But that also means they will draw 40A apiecve if i ask them for it! The CA isn't fast enough to limit brief spikes of current, apparently.

Finally heard back from Ping Li. He verifies that the low voltage shutdown is auto-reset (as I have experienced), if any single cell goes below 2.1V the pack is cut off, but turns back on if the voltage climbs again. That's why you can still limp along if you hit the low-voltage-limit wall. But the overcurrent limit is not auto-reset (also as I suspected), it does not reset until all current has ceased, so you have to unplug it to reset after a current spike. That and I finally located a schematic for the Signalab BMS, which apparently is similar to Ping's. With a schematic I can finally noodle out how this thing ticks!

Ping has a high rate BMS that includes a 60A current limit. That might be enough, I'm asking him how much that would be. Certainly less than a bigger battery. Would it be a battery killer though?

So, if I decide on a bigger battery, what might folks recommend?

 

[llile]

in that case you can use each charger for each 36V section, and then just have the current flowing through only the output mosfets on the bottom of the 12S BMSs.

Long and techical explanation ensues ...

Thanks, Dnmun! That is a really good explanation, and is even more clear with a schematic. I don't have a diagram for a Ping BMS exactly, but do have a diagram for the Signalab BMS which is appartenly similar. A battery cut off switch, in any pack, would be a really handy feature. I may implement that in any case!

I used to have a boss that wasn't satisfied until I found 6 solutions to a problem. Then we could pick the best one. We'll keep your suggestion solution as one of the six. This has the distinct advantage of lots of soldering on tiny components. Fun! Probably, the correct set of 6 answers includes also spending a fortune on a new set of batteries. Yow!

 

[dnmun]

the v1 schematic is different from the v2.5 schematic. it does seem like you may be dragging it to overcurrent shut off with so much load.

i recently bot another two ping 48v15Ah packs and one has the 60A BMS on it. it has a bunch more mosfets and extra shunt wires to double the current since he uses the same 14 pin surface mount comparator.

so ping does allow 4C discharge with that BMS, yet it is the same pack as the pack with the smaller 40A BMS.

if you do wanna go with a bigger BMS, consider using the headway BMS. jimmyD at Headway-headquarters sells the 24S and it will handle all the current the ping can produce, has thermal shut down, and extremely accurate final voltage balancing. irfb4110s for the output mosfets too.

or you can solder up the shunt a little on the current v2.5 signalab and see if you can keep it turned on during the current spikes. add just a little solder to each end, about 10% shorter shunt wires between the solder mass at the ends and maybe you can get the current spikes inside the range of the deltaV.

so sorry you though i meant to diss you, thought you were a newbie.

i have only hacked the v1 signalabs into a series under the lower BMS, when i built up a 22S pack from older ping pouches. i was able to do that without using the opto but it required 4 wires because the LVC and HVC both needed connections to the controlling BMS. it had other hacks since i had to hack the top BMS from 16S down to 10S too. but i can see how it will work on the v2.5. would be neat for someone to try it since making serial packs is so common and people waste so much energy using the diodes which i consider unnecessary anyway. they really are not needed, but if you short the output of the two in series it can a problem because then the B+ of each battery is directly shorted to B- of the other battery. i used to think it was impossible for people to short the output of their battery into a dead short but i know now that they can, and do. so the 2X12S to 24S hack would avoid that too.

i looked through my piles of BMSs and found two of the 12S v2.5 singalabs and one of the 24S v2.5 BMS. the 24S has three of the four output mosfets blown open, i expect the fourth is shorted too. if you are interested i can hack the 12S BMSs for you to test, since i don't have two 36V packs to test it on, or you can use the 24S BMS and bypass the output mosfets altogether or replace them with some irfb3205s. or you can use a single low power mosfet on the output, bypass the mosfets with the output current, and use the small output mosfet to turn on and off an led on your dashboard. use some current limiting resistor and the 'battery status' led would be lit all the time the battery is running in limits and the led would go off when the BMS hit LVC or over overcurrent and shut off the ouput mosfets.

 

[llile]

i looked through my piles of BMSs and found two of the 12S v2.5 singalabs and one of the 24S v2.5 BMS. the 24S has three of the four output mosfets blown open, i expect the fourth is shorted too. if you are interested i can hack the 12S BMSs for you to test, since i don't have two 36V packs to test it on, or you can use the 24S BMS and bypass the output mosfets altogether or replace them with some irfb3205s. or you can use a single low power mosfet on the output, bypass the mosfets with the output current, and use the small output mosfet to turn on and off an led on your dashboard. use some current limiting resistor and the 'battery status' led would be lit all the time the battery is running in limits and the led would go off when the BMS hit LVC or over overcurrent and shut off the ouput mosfets.

Lots of good ideas. Dealing with your excellent answers one at a time: Come to think of it I have a blown up Signalab BMS I got from Ping, I remember it has the signalab logo so it is probably a V1 BMS. Big hole in a mosfet, and you can still smell the magic smoke all over it. Shame to let that stuff out, it works better when it is inside. Never bothered to try to fix it since Ping replaced it no charge - replacing all the mosfets would be the first step. That could be a start on a 24S BMS. But conversion to 24S would also mean de-batterying my existing, working 36v bike, if I understand the process correctly you get one or the other (2)36V or (1)72V. Hmm.. I have no idea if the existing bike components are designed to function at 72V.

I am in two minds as to which way to go, weighing the expense and bother of getting a new batt altogether, designed from the ground up for this motor, against the interest and challenge of making the old battery work without disabling my existing, and functional, 36V E-bike. Ideally a new Lipo would be a better fit for this bike.

I started this thread because a search didn't come up with a similar problem on ES (Probably is, but I could not find it) and so i figured it would be helpful to someone else. Meanwhile, lots of pretty good options are being discussed, even if I don't go with some of them I think these are useful things to know are possible.

 

[llile]

OK, here is one good solution, likely to work. Use the existing batteries,
BMSBAttery.com 24S 50-100A BMS, $95 thanks for the tip Dnmun, never knew these guys existed.. This might not be a bad way to go, since PING's high capacity BMS is still limited to 60A and I could potentially pull 80. Gotta make sure I don't create a battery killer though. This Charger might not be a bad pick for the 15AH Ping battery, which wants less than a 5A charger. So this might be a $150 solution, and would also involve making new battery cases most likely, since the two batteries are in separate boxes now.

Although I can built/hack one, I'll probably stick with off-the-shelf stuff as it likely has better reliability.

 

[llile]

or you can solder up the shunt a little on the current v2.5 signalab and see if you can keep it turned on during the current spikes. add just a little solder to each end, about 10% shorter shunt wires between the solder mass at the ends and maybe you can get the current spikes inside the range of the deltaV.

Good plan!

I have a 6 digit ohmmeter capable of four wire measurements (Thank you, Ebay!) so I might be able to actually calibrate the shunt resistor accurately. This, so far, is the cheapest solution, although risky as I am not condifent the rest of the existing BMS is up to the task. i can check some FET specs and make sure. If anyone wants some splainin' about four wire ohm measurements I can go into it. Not a bad idea, as it doesn't involve a radical BMSectomy.

 

[llile]

i have only hacked the v1 signalabs into a series under the lower BMS, when i built up a 22S pack from older ping pouches. i was able to do that without using the opto but it required 4 wires because the LVC and HVC both needed connections to the controlling BMS. it had other hacks since i had to hack the top BMS from 16S down to 10S too. but i can see how it will work on the v2.5. would be neat for someone to try it since making serial packs is so common and people waste so much energy using the diodes which i consider unnecessary anyway. they really are not needed, but if you short the output of the two in series it can a problem because then the B+ of each battery is directly shorted to B- of the other battery. i used to think it was impossible for people to short the output of their battery into a dead short but i know now that they can, and do. so the 2X12S to 24S hack would avoid that too

It seems like the 12S to 24S BMS hack would still need some shunt surgery, because otherwise the 40A current limit would be unchanged. What this 12S to 24S buys is a more efficient system overall, with less series components. Maybe hacking the shunt to create a higher current limit would be a simple place to start. Although there will be a few surges, i plan to limit the average current out of the battery to well within its capabilities, using a Cycle Analyst setting. This is starting to make some sense now, and isn't a budget buster. Hopefully it isn't a battery killer. [pulls hair and gnashes teeth]

 

[johnrobholmes]

Glad you got it figured! The CA showed the tale.


I could reprogram the controllers for you, or take out a shunt and recalibrate the CA. Turn it down to 30a or so, should make the pings happy.

 

[llile]

Glad you got it figured! The CA showed the tale.


I could reprogram the controllers for you, or take out a shunt and recalibrate the CA. Turn it down to 30a or so, should make the pings happy.

Fantastic! We'll start with the controllers. I've already exhausted possibilities with the Cycle Analyst. I'll drop by your shop this week.

 

[dnmun]

you can kinda estimate how much the shunt resistance will change as the length of the exposed shunt wires decreases as the space is filled with solder. should be almost linear.

72V is more than most controllers will handle too, it is 88V DC and i think the pie has 75v mosfets, itty bitty bare unsinked mosfets.

i decided to take a 14S pack apart i had built before and is hard to balance. i can assemble enuff pouches with this one and th eones i got from docnjoj to see if can make two 36V12Ah ping packs and then i will hack up a pair of these v2.5 signalabs to show how it works.

i have been able to repair a lot of the signalabs. i have replaced all the parts except for the small 603 surface mount shunt resistors. i use the irfb 3205 for the output mosfets if yours have burned, try them. i got a bunch cheap on ebay once for 50 cents each. in my pile of signalab BMSs i found another 16S v2.5 with a lot of the shunt transistors damaged by someone shorting through them. i may try replacing them, about 4 are damaged, but the shunt resistors are burned out on channel 3 so i am gonna have to wire it so i can jumper over channel #3, and maybe jumper over #14 too since it is really weird acting and then i will have a 14S replacement. i like to try to repair and reuse them. most people like to throw them away.

if you get interested and start probing as one who does electronics repair will do, just for the fun of it as they say, then be careful when probing the legs of the opto isolator chains. and don't let metal tools fall on them either because the 14V used on the output side of the inner row of optos is adjacent the input or diode side of the adjacent opto and if you short the legs at 14 V to the opto diode leg on the adjacent opto thene you will short the HVC/LVC comparator on the underside that drive the opto transistor diode. and that diode leg will be at the battery potential of that spot in the series.

i think i just did that to my 14S v1 signalab. to repair it i replaced the tiny little 6 pin SOIC on the underside that detects the LVC or HVC and triggers the opto. smallest thing i ever soldered. ruined three getting them unsoldered from the spare parts BMS. then replaced the shunt transistor mosfet with another also and replaced the opto too but it is still not allowing the LVC signal to pass the repaired channel. damn, if it wasn't for breaking everything i would never learn how they work or what the parts are. the shunt transistor is a p channel mosfet, si2333.

 

[DanSFIV]

Hi, I have the same problem ..

http://endless-sphere.com/forums/viewtopic.php?f=14&t=46589

im using 45a controller

 

[dnmun]

no, you do not have a ping pack. you have a limn2o4 pack. it does not have the ping signalab BMS.

 

[dogman dan]

Yes, turning down the amps with a CA or reprogramming the controllers will help. But you still have two motors, so turning it down to 30 amps each would still allow up to 60 amps. OUCH!

Yeah, still a battery killer. The 60 amp bms would be fine if the battery was huge, but it's not.

I think to still use both motors, you'd still be flirting with early death of that battery even if you turned it down to 20 amps per controller.

You might look into something I was trying, paralelling some more stronger lifepo4 with your ping. My plan was to paralell some A123 cells with my ping, but the plan hit a wall when I bought bad cells. With more battery capacity, you could then discharge at 40 amps total a lot safer.

 

[johnrobholmes]

he is using two batteries in parallel I think, 30a per pack will be seen.

 

[dogman dan]

I read' two batteries in series, for 72v" in the first post.

I think he's running 72v 15 ah of ping cells at potentially 80 amps. He needs it to be 30 max. Two motors will back off the amps some on the starts, so a pair of 20 amps controllers might work if he feathers the throttle some taking off,or sets his CA to 30.

 

[johnrobholmes]

His other thread stated 4 pings 2s2p, Lile will need to clarify whats being run at the moment.


Turning it back to 20 amps will be just as easy, we can stab at the settings a few times to see.

 

[dogman dan]

Ah, as Emily Latila often said, "thats different".

I havent seen the other thread. I just saw what he said on the first post of this one, that made me think 72v 15 ah. 72v 30 ah should be easily able to handle 40 amps of draw total. Especially since real world continous draw will mostly be in the under 30 amps ballpark.

Gets hard to give correct advice if the info you are given is not right. Easy to make a typo in a post. Bet I do it daily.