Bulging caps causing poor performance with EB312?

[hillzofvalp]

I recently repaired this 12fet Sensorless controller after it had been programmed incorrectly--followed by the two phase wires fusing together in the controller. I fixed this and tested it with new settings. Worked great.

Now the thing has cut off under load and gradually got worse. controller smells funky, but not burnt. Cap near phases is bulging the most, but one Of the battery caps is also bulging.

Spins up fine no load.. But dies once reaches high speed (I assume that's max rpm Sensorless circuitry can handle).

It was very gradual. I had to push off the bike and have an extremely minute amount of throttle for it to kick on. Sl
Ow failure

I checked resistance between banks and they are all good.

 

[Spicerack]

I can't help with the fault on the controler but certainly any swelling on electrolytic caps means they're toast. Replace them and see if it fixes the problem- but why have they swelled in the first place...

 

[bigmoose]

... but why have they swelled in the first place...

Was you battery voltage right near the cap's rated voltage? Also high ripple currents could have heated them and caused the swelling.

 

[tedcs]

I'm dealing with a EB312-C-1 that is similarly afflicted. (Crystalyte CT7240E)
My troubleshooting ability is limited, but hardware doesn't intimidate me.

Testing on the bench, low speed was fine, but top speed resulted in a couple
of alarming pops and a cloud coming from the controller. Blowed-up electrolytic
caps smell a little different than they did in the '60's, but one would not mistake
the odor for something else.

With the controller removed from the enclosure, two caps connected across
the power input on the circuit board were distended and the tops had split.
The caps were rated at 470 µF, 63 VDC! WTF?!!
Freshly charged batteries supply slightly over 80 volts.

The caps voltage rating is obviously too low. Since this is rough service for
these caps, I wouldn't use anything less that 160 volt caps. I yanked out
the bad caps and put in a couple of 330 µF, 250 VDC caps I pulled from a
PC power supply.

I tried out the controller again and it worked alright, but the replacement
caps got hot and probably would not last long. Clearly these caps are
subjected to ripple current well in excess of what they can handle.

Endless-sphere to the rescue.
What we need is some fast capacitors.
These are NOT Radio Shack™ items.
"Low ESR" capacitors are described in earlier posts by others who are
vastly more knowledgeable than I on this matter... and everything else.

Wrong value capacitors? Wrong type? Defective?
Some or all of the above.

 

[hillzofvalp]

mine are 100V.. 470mfd. I replaced the big cap near the gets with a bigger one. I replaced the battery buffer with a 50V ~4700mfd cap cause I'm running 36V lead... Problem seemed to be not as bad, but still cutting out a lot. turnigy 80-100 sounds like a clicking, dying hard drive. There reason they failed is that the controller was not programmed properly upon receipt. Controller overheated... phase wires actually melted and fused together in one place. I wonder what the melting temp is on the insulation? the caps vary from 85-105C destruction temps.

I'm suspicious it's a faulty FET going into failure mode, but before I yank all those out I'm going to replace all the smaller caps on the board. bummer.


All my traces and wires are beefed up now (8 AWG and 12 AWG) so I don't think this controller will ever get hot again. I also finally figured out how to program it with my modified EB212 serial cable.

 

[dnmun]

why would you replace all the caps in the circuitry?

it is only the input caps and the cap across the S/D buss that see the high voltage. the others are all in the 12V section of the controller.

 

[hillzofvalp]

because they all must've seen ambient heat. The phase wire insulation melted... one of them is kinda nicked from hot phase wire.

Why are they rated for 25V and 50V on 12V side? Why not use 16V caps

 

[dnmun]

ambient heat should not kill them. i don't know why they did not use 16V caps, except that maybe they assumed that 16V is not that far from where the 12V can end up sometimes. and there is gonna be spiking in the voltage when the zener turns off on the high side. jmho

 

[hillzofvalp]

I just went in and noticed that a lead on one of the high current caps had burnt up.. I had to replace it at the time with a straight through cap (with a lead on each end) because that's all the the shop had. I replaced the thin lead with some 12 Gauge and now it appears to work fine.

Original issue was using a 80-100 180kv motor. This successful test was done on a 9C 7 turn.

Why doesn't the CA work with this board?

 

[dnmun]

you wanna use radial type electrolytic caps. you want the legs of the capacitor to be as short as possible. that is the problem with changing out the caps since the legs are specified at a specific width depending on the voltage and size of the caps. i think it goes 10mm, 13mm, 17mm you can look in the mouser catalog, they have the leg width in the specs.

 

[hillzofvalp]

for now I'll leave it with the big wire going to it

 

[dnmun]

from your circuits class, you should be able to calculate the inductance of the long leg of your axial capacitor. i doubt if it is more than a few millivolt though.

 

[hillzofvalp]

Controller ran strong for about 5 minutes.. then the capacitor literally burst (boy do they stink.. is that what borax smells like?). Need to get a low ESR high ripple current replacement, you think?

 

[NeilP]

I recently repaired this 12fet Sensorless controller after it had been programmed incorrectly-.

What exact model of EB312 is it? I have an EB312-C-1 and I have been unable to program it

I'm dealing with a EB312-C-1 that is similarly afflicted. (Crystalyte CT7240E)
My troubleshooting ability is limited, but hardware doesn't intimidate me.

.

Tedcs, same question to you , did you manage to program it?

I have two Lyen controller leads and tried both of them, tried various combinations of wiring to the EB312-c-1 header but never got lucky..using Keywin..XPD, Lyen and other hacked versions of the software. Best luck I had that indicated that the program was communicating with the chip was that XPD came back with an error that said it had received the incorrect response from the controller.

 

[Xrain]

If you have damaged capacitors, Immediately stop using the controller.

I would highly recommend replacing them, especially since electrolytic capacitors are not very expensive.


We have damaged capacitors on a controller and the controller still worked, it smelled quite funny. But after we ran it another time at a very low load, the controller stopped working, shot a 4 foot flame out of the controller. Managed to extinguish it before anything got burnt up too bad.

Granted we had a much more powerful setup than what you had (176 volts 45 Ahr)

But I just cant recommend playing with damaged capacitors, not enough reward for the risk you take.

 

[Planky]

I recently repaired this 12fet Sensorless controller after it had been programmed incorrectly-.

What exact model of EB312 is it? I have an EB312-C-1 and I have been unable to program it

I'm dealing with a EB312-C-1 that is similarly afflicted. (Crystalyte CT7240E)
My troubleshooting ability is limited, but hardware doesn't intimidate me.

.

Tedcs, same question to you , did you manage to program it?

I have two Lyen controller leads and tried both of them, tried various combinations of wiring to the EB312-c-1 header but never got lucky..using Keywin..XPD, Lyen and other hacked versions of the software. Best luck I had that indicated that the program was communicating with the chip was that XPD came back with an error that said it had received the incorrect response from the controller.

I have a EB312-c-1 and have programmed it, though mine is a Lyen one. I used the software and lead he provided.

 

[hillzofvalp]

Same model. I know how to modify the original cable to work with EB312, as I just posted a thread about this recently. Just get rid of the last red wire, then splice over the white wire to replace the last red wire. The white wire has to be on each end

Leave the controller cable as is. I've only tested with lyen's software, but I'd like it to work with XPD so I can have a higher battery current than 65A!

I love Lyen. Thanks for a great controller!

 

[tedcs]

I'm dealing with a EB312-C-1 that is similarly afflicted. (Crystalyte CT7240E)
My troubleshooting ability is limited, but hardware doesn't intimidate me.

Tedcs, same question to you , did you manage to program it?

I have two Lyen controller leads and tried both of them, tried various combinations of wiring to the EB312-c-1 header but never got lucky..using Keywin..XPD, Lyen and other hacked versions of the software. Best luck I had that indicated that the program was communicating with the chip was that XPD came back with an error that said it had received the incorrect response from the controller.

NeilP, I've given little thought to altering the firmware.
Presently, the issue is improper caps on the battery voltage buss.
63 volt caps with 80 volts across them is NOT what I call 'headroom.' WTF?!?
That the input caps split and spewed is no surprise.

I'm quite used to undertaking too many things at one time ...
messing with the programming would probably be too many too many things, however.
I'm not saying never, but don't want to alter the firmware unless I've no choice.
I feel quite handicapped without a protocol analyzer.