Version 2 Crystalyte Controller information

[fechter]

i have a shorted FET on my 12 FET v2 controller. do i just need to replace the 4 FET's on the phase that has shorted or the whole lot? is there any thing else that will have to be replaced like the gate driver etc?

Yes, check the gate resistors. Those usually blow. The one I have uses 51 ohm resistors.

If the resistor is blown (open), you can bridge a regular resistor across it. If it's blown, there's also a chance that a gate driver transistor is blown and possibly the MCU.

If you put your meter in the diode check mode, you can compare measurements between a good phase and the blown one. You can measure across the driver transistors and look for shorts.

 

[monster]

where are these components on the board, could you post a jpg? i couldn't see anything about them in the thread.

 

[The7]

There is a "reverse" switch in the V2.

The color wiring must be completely matched for Crystalyte motors in normal running.
In "Forward", C-motor runs forward.
In "Resevse" C-motor runs in reverse.

If the V2 is in "Reverse", is it possible to find one or more (from the 36 combinations) to run the C-motor in forward direction? Most likely "YES".

If YES, could we find one particular combination that will give identical performance as if the C-controller and C-motor are connected normally?

Such questions are asked because it seems that:
a) for driving one motor in forward dirction, there could be only one best combination when the C-controller is set in "forward";
b) for other motor in forward direction, there could only one best combination when the C-controller is set in "reverse".

 

[fechter]

If the controller is set for reverse, there is one combination of wires that will make it run forward and vice-versa. The performance will be identical in either case.

Monster:
The gate resistors are labeled E1, E2, E3, etc. The actual resistor is marked 510, which is 51 ohms. There is one resistor for each pair of FETs (stupid).

 

[monster]

cheers fechter

 

[monster]

so are the gate drive transistors these things? marked T1, 2, 3, 4 etc?

 

[monster]

all E1 - E6 measure 51.4 ohms. thats good right?

 

[calinb]

so are the gate drive transistors these things? marked T1, 2, 3, 4 etc?

Yes--what are those things! My T4 just went pop. At least I think it's T4 from memory. The silkscreen is so scorched, it looked like an "F1" to me, but now I see from the photos posted here, there's no F1 there. The region around the three horizontal throughhole vias, just to the left of A2, has also been very hot. The middle via is just to the left of the T3 blue arrow head in the above photo.

I have a 5304, stock version 2 w/ 6 FETs, Yesa "48 V" pack (16 cells ~= 53V nominal) and I heard it go "pop." The motor produced no power at all after the pop, but the power LEDs on throttle were on. After disconnecting the battery, I still felt a cogging vibration that I've never felt previously, while coasting back down the 7% grade that I was climbing (pedal assisted) at 20 mph when T4 popped. I had to disconnect the wires at the motor to keep it from vibrating.

Is this thing worth attempting to repair, or should I try to get a Kelly. I've only had the bike 6 months so I'd like to try to repair it, for sake of e-bike economy. Ebikes aren't very economical, compared to my Japanese motorcycle, if reliability is this poor!

Thanks for any advice or debug strategy.

-Cal

 

[The7]

If the controller is set for reverse, there is one combination of wires that will make it run forward and vice-versa. The performance will be identical in either case.

Fether.
Wonder if you could use the Shezhen controller to drive your MAC-BNC motor at around 200 Hz (with partial throttle) in forward and reverse directions at no-load with the best possible combination of connection.
Take a picture of one applied motor voltage and one Hall signal each at these conditions.
Would like to see the time advance of the applied voltage wrt the BEMF at these condition.
Thanks

 

[calinb]

It is F1 that's blown. If it's a fuse resistor, does anyone know what it protects or the value of the resistor? Notice the trace was very hot under the three through holes above D1.

 

[The7]

There are two versions of 6 FETs for V2:
6 FETs at EVEN no positions; and
6 FET at ODD no positions.

Both have hot "plated holes" between the No 3 and No 4 position.

I think these plated holes are used to carry the high current ( say 25A or higher) )from the sourse leg of low side FET to the -ve bus.

 

[calinb]

Both have hot "plated holes" between the No 3 and No 4 position.

I think these plated holes are used to carry the high current ( say 25A or higher) )from the sourse leg of low side FET to the -ve bus.

My three "singed" vias are also probably placed to carry high current from one side of the board to the other too. They got hot and the solder started to melt out of them with my F1 frying.

The7, would you mind reporting the numbers on the top of your F1. I'd like to try replacing it, if I can find the value. F1 is visible in my photo, above. It's at approximately the "7AM" point from C11. As you can see, it is fried.

If I can find a replacement--even a leaded, through hole replacement, I'll solder it in, refill the via holes with solder, power it up, and see what happens.

Thanks a lot,

-Cal

 

[calinb]

F1 is a 1.5k. I found the circuit and layout doubled elsewhere on the board and I could read the value.

I guess there's no harm in just filling the vias back up with solder and replacing the resistor, at this point. If the wheel powers up, I'll look into populating the other 6 FETs too. Otherwise, I guess I'll be shopping for a Kelly controller.

Has anyone else had this F1 (or F2, it's evil twin ) failure?

-Cal

Update:
I just saw Knuckles' controller thread. I think I'll get one of his controllers, regardless. Just in time!

 

[monster]

i think i'm just going to replace my FET's and see what happens. my gate resistors are ok.

does any one have any irfb4110's spare? i don't want to pay RS £10 delivery if i can help it. im in oxford UK. PM me if you do.

 

[The7]

F1 is a 1.5k. I found the circuit and layout doubled elsewhere on the board and I could read the value.

Yes, as shown in the picture.

 

[fechter]

i think i'm just going to replace my FET's and see what happens. my gate resistors are ok.

does any one have any irfb4110's spare? i don't want to pay RS £10 delivery if i can help it. im in oxford UK. PM me if you do.

Yes, try that, but test with a small fuse at first.

The7:
The I ran the puma both forward and backward by changing the wiring. The timing was the same in either direction.

With the BMC motor and the analog Crystalyte controller, the reverse switch will not properly reverse the motor. The motor can be rewired and it will run properly in reverse.

 

[The7]

The I ran the puma both forward and backward by changing the wiring. The timing was the same in either direction.

Thanks Fechter for doing the test.

The puma itself has the same phase relation of 80 deg between the phase BEMF and the Hall signal in forward and reverse directions. So the same controller could drive it in both direction by changing the wiring.

With the BMC motor and the analog Crystalyte controller, the reverse switch will not properly reverse the motor. The motor can be rewired and it will run properly in reverse.

When BEMF and Hall signal for Crytalyte 4000 were determined from the windings, the skewing of the stator poles and the location of the Hall sensors. it was found that there is an difference of 60 deg for the BEMF in forward and backward directions. If the BEMF In forward is +30 deg wrt Hall signal, then the BEMF would be -30 deg (or vice vera).
I think that the "reverse" switch in the analog C-controller is designed for compensating such difference for C-motors..

So this reverse switch may not be suitable for other make motors for the purpose of reverse.

Without using the "reverse" switch, it is always possible to one best combination for forward and another best combination for reverse.

Rotating in the reverse direction will also give an equivalent 180 deg in addition to the timing space difference.


.

 

[calinb]

Thanks, The7.

i think i'm just going to replace my FET's and see what happens.

I found that the source and drain of one of my six 4130s are a dead short. I may try replacing just that one FET. If it works, I'm interested in placing all twelve of 'em--perhaps with 4110s. Does anything else need to be done to add the additional six FETs that, apparently, should've been there in the first place?

Thanks,

-Cal

 

[solarbbq2003]

There is a "reverse" switch in the V2.

which pin is used for reverse function? connect to ground like abs and 60/120 pins?

 

[The7]

Are these Version 2 Controllers any more waterproof then old version controllers?
Got this one now [not installed yet]
It don't seem waterproof at all.

Is it a V2?

If so, V2 would have an "reverse" pin on the board because it has the "reverse" switch.
Could anyone locate this pin on the board?

 

[rkosiorek]

way back on page 4 of this thread is a list of how to hook up all of the connections to the board including the reverse switch. (and everything else that needs to be hooked up)

but for those who need a picture. i have borrowed someone elses photo from this thread and circled the "DSP" pad. shorted to ground this is the reverse connection.



rick

 

[The7]

Rkosiorek, thanks.

 

[calinb]

Yes, check the gate resistors. Those usually blow. The one I have uses 51 ohm resistors.

Two of my six 4310s are blown (V3 and V5). I de-soldered them and the leads are all one big short. I replaced the blown 1.5k F1.

Guess I'll probe around and try to find the gate transistors and resistors too.

 

[monster]

i narrowed it down to two FET's that have shorted (shorted across all legs)

unfortunatly i also found a discrepancy on T5. compared to all the other T's it has a much lower number across top and right legs when the meter was set to diode. don't know what this means can anyone shed some light?

 

[fechter]

i narrowed it down to two FET's that have shorted (shorted across all legs)

unfortunatly i also found a discrepancy on T5. compared to all the other T's it has a much lower number across top and right legs when the meter was set to diode. don't know what this means can anyone shed some light?

Err... that usually means it's shorted. If you don't fix it, you might blow up your new FETs.