Crystalyte Controllers - Repair and Modification information

[BiGH]

after modification and side plate replacement i'm having the same one phase out like the above poster. Its the yellow phase (cosest to the exit of the controller. Will open her back up again and test.

ok the controller is stuffed - not sure whats wrong with it, need help as to how to test it. I'm not getting any voltage out of the yellow phase to the motor - the motor is tested fine now.


regarding teh new controllers: the fets in them are actually very good - and should be reliable for 72v 35a i've started a thread about the insides of the controller.



H.

[GGoodrum]

What is the difference between the 35A and 40A controllers? Is this new one I have the 35A or is it really the 4840 replacement? I'm confused.

[billvon]

Its the yellow phase (cosest to the exit of the controller. Will open her back up again and test.

Interesting. I wonder if there's a common problem with the yellow phase FETs due to their proximity to the edge?

>ok the controller is stuffed - not sure whats wrong with it, need help as
>to how to test it. I'm not getting any voltage out of the yellow phase to
>the motor - the motor is tested fine now.

Are you getting a gate drive signal to the FETs? If you can spin the motor with the remaining two phases you should see a solid signal to the gate of the bottom FETs and a PWMed signal to the top FETs.

Are you getting lots of leakage? With the motor disconnected, current draw should be very low. If it's high one of the FETs might be fried.

[rhubarb]

Just wondering whats the highest voltage anyone's successfully run through 4110 FETs?

They're nominally rated to 100V... I'm running 85V (3 x 28V LiPO) reliably in my delightful KMX Puma.

Is there any chance I'll be able to get 115V, or am I dreaming?

Has anybody run more than 85V successfully?

I could use 150V FETs, but the 4110s have a really low on resistance, so you can use them in the small (20 A) crystallite controllers without overheating.

[fechter]

I think Lowell's had them up to 108v. 85v should be no sweat. They seem to have some headroom in the voltage department.

The 150v FETs have a much higher on resistance, so they will get much hotter. If you wanted to run higher than 100v, but reduce the current limit, they might be OK. I wouldn't recommend it.

[CGameProgrammer]

I opened up my 7235 with the 150V FETs so air can pass right through as I'm riding, but they still get too hot to maintain 96V and can just about manage 72V in 80-degree weather without shutting off.

[billvon]

After blowing out my first controller pretty quickly, I added a clamp circuit to reduce FET Vds stress, in hopes of preventing another failure. So far seems to work well. However, I am now getting massive power limiting, I think due to temperature problems. (Operating at 72 volts.) I'll have to stick a temperature sensor on the case to see if that's the problem.

[The7]

An very good idea!
Seems that you are using 3-phase bridge rectifer circuit with an 160V 220uF Cap as the clamping circuit.


What is the diode's rating?
Did you know the normal dc voltage appearing on the Cap?

Where do you connect the red and black wires from the Cap to?

[The7]

However, I am now getting massive power limiting, I think due to temperature problems. (Operating at 72 volts.) .

Could you explain a bit more detail?
Also where are the red and black connected to?

[billvon]

However, I am now getting massive power limiting, I think due to temperature problems. (Operating at 72 volts.) .

Could you explain a bit more detail?
Also where are the red and black connected to?

Diode is STTH4R02RL, 200 volt 4 amp ultrafast diode. The speed is what I was looking for; the body diode of the FET can handle the bulk current. I just wanted to trim down the peaks caused by the flyback voltages.

Red and black go to the + and ground power buses on the controller board, as near the FETs as possible. I probably don't need that big cap, a ceramic would probably work just as well.

Power limiting - I will get about 5 minutes of operation at 500-800 watts, then it will fade rapidly to under 100 watts. I'll let it coast for a few moments and will get power back. It's annoying. I am guessing it's a thermal issue.

[The7]

[quote="
Red and black go to the + and ground power buses on the controller board, as near the FETs as possible. I probably don't need that big cap, a ceramic would probably work just as well.

Does the black go to the battey ground before the SHUNT or after the SHUNT?

[fechter]

Do the diodes or capacitor get hot when it goes weak?

[billvon]

Does the black go to the battey ground before the SHUNT or after the SHUNT?

Between the shunt and the FETs. Basically the + and - go to the power rails connected to the FETs (since that's what I'm trying to protect)

[steveo]

could i do this fet replacement mod to the 40 amp version of this controller?
http://endless-sphere.com/forums/viewtopic.php?t=764

i'd like to be running 100v+

thanks
Steve

[fechter]

could i do this fet replacement mod to the 40 amp version of this controller?
http://endless-sphere.com/forums/viewtopic.php?t=764

i'd like to be running 100v+

thanks
Steve

Sure. There's not much difference between the 40 amp and the 35 amp version. The new ones with six 4310's are unknown. With 4110's, you can safely run 100v, but some have pushed a few volts higher without smoke.

[steveo]

which fets would be best? at 100v 4110's as per what your saying? i haven't opened my controller yet .. but how many would i be needing?

if i wanted to lets say go all out and put 150v.. which fets would i be needing .. has there been may heat problem pushing the controller that far....

p.s. sorry for all the questions ..

here is my bike build
http://youtube.com/watch?v=l0j9PlsqZ20

thanks
Steve

[BiGH]

for 100+ volts i would suggest what Antonio (on these forums) has used - FDP2532 rated at 150V 79A.

edit: you will also have to change the capacitors as they most likely are only rated for 100v also.

Also please note that we're starting to get into very dangerous voltages. at 48v 30a you can feel the voltage quite noticibly if you touch the phase wires.

at 150v etc would extremely dangerous - i'd suggest upgrading all connectors too to something with no open exposure of metal (trailer plugs are a no-no).


Take Care,
Haydon

[xyster]

which fets would be best? at 100v 4110's as per what your saying? i haven't opened my controller yet .. but how many would i be needing?

if i wanted to lets say go all out and put 150v.. which fets would i be needing .. has there been may heat problem pushing the controller that far....

Lowell here has reported his bicycle is difficult to keep under control at 8000 watts -- the front wheel wheelies too easily. Sounds like the limit of useable power for a typical bicycle is around 5-7 kw, which can be achieved at a safer 90-100 volts with the better 100v IRFB 4110s. There have been overheating problems with the 150v. The original current limit shouldn't be exceeded with those FETs, whereas the current limit with the 4110s can be increased to 80-100 amps. So you'll get much more, and more efficient power with the 4110s at a higher current limit.

[fechter]

The best 150v rated FETs have about 4 times more on resistance than the 100v IRFB4110. At 40 amps, the heat dissipation will be a problem. It could work with additional heat sinking and possibly a fan, but I wouldn't recommend it.

The 'typical' controller has 12 FETs, and the newer model, as well as the 20 amp version have 6.

If you read the first few pages of this thread, you can see what it looks like.

[steveo]

I think i will end up staying on the safe side @ 100v using the 100v IRFB4110 fets; Will i have to change capacitors with those fets or can i use the originals?


p.s. thanks for all the quick responses guys; this forum is awesome for diy info

[steveo]

After blowing out my first controller pretty quickly, I added a clamp circuit to reduce FET Vds stress, in hopes of preventing another failure. So far seems to work well. However, I am now getting massive power limiting, I think due to temperature problems. (Operating at 72 volts.) I'll have to stick a temperature sensor on the case to see if that's the problem.

what is that potentiometer i believe its hooked up to the cycle analyst port if i'm correct; i'm taking a long shot but it that the restrict the power to the controller? what potentiometer are you using?

[fechter]

I think i will end up staying on the safe side @ 100v using the 100v IRFB4110 fets; Will i have to change capacitors with those fets or can i use the originals?


p.s. thanks for all the quick responses guys; this forum is awesome for diy info

It depends on which ones are in the controller to start with. This seems to vary quite a bit. Most of the ones I've worked on had 160v caps, so they did not need to be changed. It wouldn't hurt to upgrade the caps to higher quality ones, but space inside is limited.

[billvon]

what is that potentiometer i believe its hooked up to the cycle analyst port

That's the throttle port. It just sets a (low) speed so the motor runs continuously.

[steveo]

I think i will end up staying on the safe side @ 100v using the 100v IRFB4110 fets; Will i have to change capacitors with those fets or can i use the originals?


p.s. thanks for all the quick responses guys; this forum is awesome for diy info

It depends on which ones are in the controller to start with. This seems to vary quite a bit. Most of the ones I've worked on had 160v caps, so they did not need to be changed. It wouldn't hurt to upgrade the caps to higher quality ones, but space inside is limited.

i may open up the controller on the weekend; i'll take a trip to the local electronics shop to check if they have the fets before i get started ..

i think i should upgrade my motor before even considering this modification; i don't think my motor is very efficent; i'm running a 250watt motor; my 7.5ah cells start to crap out after 10km's of travels; i did 19kms in total on one charge.. does that sound very inefficent too you?..

thanks
Steve

[steveo]

what is that potentiometer i believe its hooked up to the cycle analyst port

That's the throttle port. It just sets a (low) speed so the motor runs continuously.

thanks for clarifying.. i though it was used to restrict power to the hub motor .. hmm i wonder if there is such a mod without having to purchase the cycle analyst ..