GT i-drive 2.0 Build

[Hyena]

Have you actually gotten it badly wet ? I suspect your "water" damage is actually capacitor juice that has vented.
I had one let go in a meanwell and holy shit, you should have seen the amount of smoke and juice that thing sprayed out.

I haven't had a controller just blow a cap before but I doubt having a cap go open or short circuit would cause the controller to emulate blown fet symptoms and cause the motor to cog. I don't know if it's the chicken or the egg but one likely possibility is the cap has vented and the juice pooled in that corner of the controller, shorting out the legs of that fet.

edit: ah too late, you snuck in another post and some more testing

 

[GCinDC]

no, not horribly wet. one day moderate rain. top covered. bottom edge (heatsink side) exposed... had no idea about the juice in a cap.

definitely looked like it'd pooled in the bottom corner. perhaps the cap blew and then the juice got burnt and cruddy when the fet(s) went.

 

[grindz145]

chicken or the egg, did the fets kill the caps or the other way around. Electrolyte is nasty stuff though.

 

[dbaker]

Wow Another blown controller Hard to imagine considering you have done no mods to the controller and kept the stock current setting What effect does the speed switch have on the current limit setting?

How did the 8 pole bullet connector work? Gotta see a pic

 

[Hyena]

What effect does the speed switch have on the current limit setting?

I've wondered that myself, given the discussion we had a while back about how much current multiplication is going on and how soaringly high phase currents can get at part throttle. My next build will be using 24S lipo for high speed and torque when necessary, but much of the time I'll be running in a 70% speed mode which I'm assuming will be hard on the controller.

"Sorry officer, I have to ride flat out otherwise it's too hard on my mosfets. 4110s don't grow on trees you know " :lol:

 

[amberwolf]

I take that back about the fets. The last bugger on the left looks like he's been smoking in the bathroom. But maybe that was the cap splat.
View attachment 1

FWIW, it looks like the one next to the last on left is split wide open, right at the plastic/metal join, and puked it's magic smoke out all over the one next to it and the wires. Might be teh only dead one, prbably not though.

The cap goo generallly does look pretty icky like tarry snot, after it's boild inside them a bit getting hot enough to burn the shrinkwrap off the cap before blowing out the top.

At a guess: Cap dies either quickly from being installed backwards in the first place (might just overheat and vent, just like this, rather than exploding), or slowly from spiking overvoltage or from being insufficient capacitance to prevent huge current flows in and out of it all the time, overheating it until it finally really dies.

Once cap dies, some big power spikes aren't suppressed anymore, and the FET(s) sees an overvoltage spike that's too high for it to take, and blows. If it fails shorted it could then provide a shootthru path for the other leg of that phase to short right across power blowing one or more of that leg's FETs. Probably the whole bank should be replaced at once, so you don't have to mess around with it more than once. If 18FET, then all six on that side. If you want you can replace only the ones that fail the basic tests, but if they are weakened arleady they may die later and psosibly take other already-replaced FETs with them.

 

[GCinDC]

excellent reading, amber. explains a lot. and i think you're right about the fet. one over there measured zero ohms across the legs, or near the yellow, can't remember.

still don't understand why i should remove 6, including ones that aren't broken. is the idea that they're worn now? i guess you said weakened.. hmm. i do have 13 spare fets on a tree out back, but it'd be a shame to use them if it's not gonna work.

anyway, i'm off the bed. the old 846 is back in place. phase wires connected w/ bullets.

new soldering iron arrived today! what a lifesaver that was.

no idea about current limiting btw. seems like it was either water in the controller or the shorted speed switch.

how do you waterproof your controllers, dbaker?

this is temporary, but so far, so good.

wish me luck for tomorrow's commute!

 

[TMaster]

still don't understand why i should remove 6, including ones that aren't broken. is the idea that they're worn now? i guess you said weakened.. hmm. i do have 13 spare fets on a tree out back, but it'd be a shame to use them if it's not gonna work.

The very left side of the controller was blown, I am just being safe in saying take the left 6 fets out and checking them. You may have more then 1 blown. 2 of my fets looked perfect and they were blown. If you really really don't wanna take chances you test them all. That basic test for testing the fets while they are inside the controller did not work for me. I had to take them out.

This is the best test and it worked perfectly for me: http://endless-sphere.com/forums/viewtopic.php?f=16&t=19645 , I replaced all the bad ones and the controller worked the first time after. You will need a multimeter with diode test setting, make sure you match the fets from the same batch in pairs. Like the farthest left 2 fets from the same batch number and the next 2 from a different one, or all 4 fets the same.

 

[TMaster]

Did you have to force the connectors in the housings backwards with the 10AWG wire like I mentioned in my video? I used a small flat head screwdriver and pretty much hammered right at the lip of the solder/joint where the wire was comming out. lol. I didn't want to force the wire itself for fear of damaging the braid of the wire.

 

[chroot]

GCinDC, Nah, I am from bay area in california living near Lyen and illa (San Francisco). Lyen suggested me get 24 fets controller and I went 18 fets controller instead due costs cheaper. Lyen said more fets in controller = more powerful.

Gotta go work 3AM and hope get your ebike back on ground. :thumpup:

ps. chroot, you're from russia? maybe your stockings are bigger than ours.... i don't think 18fets will fit in our xmas stockings...

 

[GCinDC]

Made it to work with the old 846 infineon! Hauled the kids up a steep hill. Going easy on the new lipos, and the controller too. I gotta take it easy till I get another backup. (so hard to take it easy!) The lipos are hard, very little voltage sag. Can't wait to give them a workout. I kicking myself that I didn't spend another $10 for the 45-90C nanos. Dumb ass. Not that I'd make full use of those C's; I just wouldn't have to worry about punishing them! Especially if I did, say, upgrade the controller to something with a few more AMPS. Ah, me...

No, I gotta go easy for a spell. I tried cycling the lipos with the iCharger. Takes FOREVER. And gets way too hot. Not cool. I know, I know, I should have a good 10A load... I built the lights in parallel thing, but only have low wattage lights.. . :lol: Draws 1A..

So yeah, time to fix the controller, get it fixed, or buy a new one... Tmaster, do you know of any good FET REPLACEMENT vids? Or anyone, for that matter? It would be a drag to replace them and have that not fix it. I'd beef it up too. Maybe a bigger replacement cap for starters... It sucks not feeling confidence in the controller. I went so long on that one!

Go easy, go easy. You should see how twitchy my damn throttle is now. It's the controller. Always was twitchy. Needs inline resistors on the magura, so I probably did a bad job. But it's a bit like a bucking bronco when going slow.

And The CA is retarded right now. Absolutely nuts. Amps all over the map. Absurd speeds. It's basically just an on/off LED, except the whole pack voltage works!

 

[GCinDC]

Did you have to force the connectors in the housings backwards with the 10AWG wire like I mentioned in my video? I used a small flat head screwdriver and pretty much hammered right at the lip of the solder/joint where the wire was comming out. lol. I didn't want to force the wire itself for fear of damaging the braid of the wire.

Not sure I know what you mean. I used flux to lube the Initial Passage into the Housing. The when it was still hot, I looped the wire around my fingers like dental floss, held the connector, tugged and it snapped in!

 

[GCinDC]

Have you actually gotten it badly wet ? I suspect your "water" damage is actually capacitor juice that has vented.

On second thought, I think you're right. And what's more, I think the 'dark stain' is not from WATER but from THERMAL PASTE. It did not wipe off nicely. It was heavy and thick. It's like the cheese on pizza that cooked too much. You can see the edges that didn't cook as much.

And it would make sense that, as the the board was slid into place, some paste would glob at the front and drip down into that spot. How conductive is it? Could it short the board? Hmmm....

 

[GCinDC]

modern loft in soho. scene of the crime. area closed off for forensics...


bulk charging at last? maiden voyage in progress. so far, in 20 min, the pack has gone from 3.83 to 3.87. laptop charger led always stays green. ping charger red and fan running. coulda sworn i saw a puff of smoke at the beginning... :|
View attachment 1

anyone notice how i've grown up an shrink tubed both charger wires and finished off the bullet end? my last one was 'temporarily' wrapped with black electric tape for a year...

 

[dbaker]

81.1/20=4.06 :| Don't you need a little more to get those cells up to 4.15v Is the Ping charger adjustable?

 

[GCinDC]

yep, after about 2 hours, i heard the ping fan shut off and all the cells were ~4.07V. pretty well balanced, but not identical...

this was a test, and am happy it works! it'd be nice to get 'em higher, but charging to only 4.07, they'll last longer, and there's some cushion if one gets high.

just did some research to verify this and actually, it looks like 4.05 is closer to 80% capacity:

Voltage in lipo cells can be broken down like this:

4.20v = 100%
4.03v = 76%
3.86v = 52%
3.83v = 42%
3.79v = 30%
3.70v = 11%
3.6?v = 0%

 

[Hyena]

I don't really agree with that as a blanket rule, I think it'd be highly dependent on the age of the cells and their C rating.
Some cells drop close to nominal voltage after only 30% discharge and others will deliver 100% capacity and still only read nominal voltage. I had one old crappy early gen lipo pack that'd drop to ~3.3v/cell as soon as you set out, but would still deliver 90% of its rated capacity before cells hit 3v

 

[dbaker]

yep, after about 2 hours, i heard the ping fan shut off and all the cells were ~4.07V. pretty well balanced, but not identical...

this was a test, and am happy it works! it'd be nice to get 'em higher, but charging to only 4.07, they'll last longer, and there's some cushion if one gets high.

just did some research to verify this and actually, it looks like 4.05 is closer to 80% capacity:

Voltage in lipo cells can be broken down like this:

4.20v = 100%
4.03v = 76%
3.86v = 52%
3.83v = 42%
3.79v = 30%
3.70v = 11%
3.6?v = 0%

And the Ping charger light turned green Great; you are good to go with the best value bulk charging system 2 hours and only $60

 

[GCinDC]

ping charger is red when charging, usually w/ fan on. then goes green when finished... rated for 2A, it was putting out at least that. got quite hot though, and gently moved the laptop charger to the side, rather than on top...

Now I wonder if I should order another 20V laptop charger and add a pair of the old batteries to go 25S. The old ones will sag heavily and keep me under 100V, but will they charge out of balance from the others?

Just whispering out loud...

 

[GCinDC]

couple days of commuting. close calls around 1:20.
[youtube]AhgVMwmkK4Q[/youtube]
used a fade to black transition. is that annoying?

 

[ambroseliao]

GC

You should create a "Close Calls Greatest Hits" video!

What was the guy on the bike telling you at around 2:35?

I didn't find the fade-to-black transition annoying.

Ambrose

 

[GCinDC]

yes, i've thought about the close calls highlights vid. slight logistical challenge, since my macbook hard drive is so small, i save a high res export, then wipe everything after uploading it...

i've got exports of all my vids w/ the gopro, but may not have the others! i want to make a 2010 highlights reel. Ebiking scenes through the seasons... All the mounts, etc. Too bad the stuff from last winter is jerky low res stuff.

What was the guy on the bike telling you at around 2:35?

GUY: Your bike very good, eh? Very Strong.
ME: You've seen it?
GUY: YEAH! (don't understand the rest; something about made it w/ electric?)

 

[GCinDC]

I don't really agree with that as a blanket rule, I think it'd be highly dependent on the age of the cells and their C rating.
Some cells drop close to nominal voltage after only 30% discharge and others will deliver 100% capacity and still only read nominal voltage. I had one old crappy early gen lipo pack that'd drop to ~3.3v/cell as soon as you set out, but would still deliver 90% of its rated capacity before cells hit 3v

Were you mixing and matching them much? Mine were paralleled as 5s2p and stayed balanced for a year or so. What I noticed is that at the beginning, they're hard! Very little sag. Little bulls, raring to go. At half life, 300 cycles, they'd 'matured' and dropped a bit more, didn't give quite as much... And at the end, the little warriors came out to fight, but tired quickly, and all sagged substantially. And yes, some much worse than others. But they did keep their stations, not as long as before, but honorably. :lol:

Bulk charging under the desk. Not sure which is more of a pain: dismounting pack and carrying to desk OR dragging meanwell everywhere and plugging/unplugging the 8X bullet plug! I am trying to bring the pack in at night anyway tho...

In any case, charging up only to 81V might not cut it. On the ride this morning, it felt like I was taking off at 75% capacity. And halfway through the ride, (after say 2Ah), the pack felt half gone, resting voltage was 76-77V?

Maybe I can find a 21-22V (even 2A) laptop charger... to replace the 20V...

Then again, I did forget that I was supposed to be breaking the pack in gently...

 

[TMaster]

Has your wife seen these videos?

 

[GCinDC]

Sure! We sit down with popcorn, shoot 'em over to the big screen and laugh!

Just kidding. No, she's seen some but doesn't seem interested. I do lots of family vids, etc.

She got pretty irritated with my ebike habit at the beginning but has been more supportive lately of my breakdowns, etc, especially as they interfere with our daily logistics w/ the kids. She did mention she wished she had a similarly interesting project, like sewing or something... :lol: