(solved) EB218 18FET regen stutter/not working correctly

SOLVED

When I engage regen it stutters, as if it's only working on two phases or halls are going crazy.
I observe my CA speed reading also goes haywire. into the 30's to 50's if I'm only doing 20 and enable regen. Speed reading is fine during all other points of operation.

Everything else works well.

Same (solved) problem from past controller: I had an EB2xx 24FET that gave this same problem. One day it blew and I went to fix it. ... I replacement the usual mosfets and happend to observe a resistor around gate drivers that was not reading same value as same resistor on other two phases. I decided to replace that resistor. ...the controller's regen then behaved correctly but I can't quite say this resistor was the cause of regen stutter for that controller.
So going back to this new 18FET, I now decided to look at the resistors around gate driver area thinking there's a resistor badly out of tolerance but all seem fine.
My investigation ends here.


Why does this EB218 18FET stutter upon regen?
(running cromotor in 20" wheel).

Without knowing small details, it’s a guess.

The times I’ve had stuttering performance was where the insulation frayed against the framework and was shorting in damp weather, or I had a flakey ebrake switch; check the ebrake signal line for continuity. Check the ebrake switch to be sure it is performing solidly. Look for loose wires/cold solder joints inside the controller as well.

If that doesn't solve the problem we might need more information: Battery config # of S & Ah, Battery voltage when stuttering occurs, and typical velocity. Does this occur in both dry and wet weather? Have you applied any mods to the controller? How much power are you trying to push back into the battery? Does it occur on flat level ground or twisty roads or down hills?

Often the solution is very simple.
~KF

Kingfish said:

Without knowing small details, it’s a guess.

The times I’ve had stuttering performance was where the insulation frayed against the framework and was shorting in damp weather, or I had a flakey ebrake switch; check the ebrake signal line for continuity. Check the ebrake switch to be sure it is performing solidly. Look for loose wires/cold solder joints inside the controller as well.

If that doesn't solve the problem we might need more information: Battery config # of S & Ah, Battery voltage when stuttering occurs, and typical velocity. Does this occur in both dry and wet weather? Have you applied any mods to the controller? How much power are you trying to push back into the battery? Does it occur on flat level ground or twisty roads or down hills?

Often the solution is very simple.
~KF

Click to expand...

Dry weather, stock controller, 24S 20Ah HK LiPo

I'm not using ebrake switch for these tests. I have it automatically regen when throttle is released.
It happens in all riding conditions. Flat, hill, etc.. even when wheel is off the ground.

I will try two things that I can think of:
1. Get video to show it.
2. Test it with a different motor

Understood. OK, so tell me about the controller: Is it 100V or 150V rated?
Thanks, KF

Kingfish said:

Understood. OK, so tell me about the controller: Is it 100V or 150V rated?
Thanks, KF

Click to expand...

100V.

The other controller that I mentioned above was a 150V model of the 24FET eb2xx.

Haven't gotten chance to test out two things above but it's either linked with the eb2xx controllers or my motor/wiring.

I am thinking that you are literally up against the wall: Using a 100V controller with a 24S battery which at full charge (4.2V * 24) = 100.8 VDC. When you try to ebrake/regen – there simply is not enough headroom to make a braking effect.

I use a 100V-rated controller, though run with a 15S battery, AND have applied the R12 mod which gives me the whole 100 minus 63 space to play with. On 2WD, the braking effect is quite profound. But with you all topped off, you will have nothing, a stuttering effect as your controller (or perhaps the BMS) self-limits via HVC.

That’s the story with the 100V 18FET controller. I don’t know what’s going on with the 150V 24-fetter, unless the battery voltage was pushed in the same manner. Regardless, I can see how you would blow FETs as they are being overdriven beyond spec.

Traditionally, a good design application will allow for 25% more capacity than what is needed, so if running 24S then you’ll want the 150V controller which is good up to 30S.

My ½W, KF

Are you sure you're connecting to the correct wire? I only had regen stuttering once, and it was because I connected to the speedo wire which was a pulsed signal from a hall sensor.

There may be a HVC setting that turns off regen when the pack voltage gets too high. If this is set just above normal pack voltage, it might oscillate on/off trying to keep the pack below the setting. There may also be a regen current limit setting (sorry, I don't know the details of the settings on that particular controller).

Ok. Still haven't tested regen with another motor or gotten video of issue but will address some of the points made above.

Yea, 24S LiPo is at the wall of these FETs, as you say. But I charge my pack to 99.5V and drops down to 95V quickly. R12 mod applied, I have regen set to approximately 100V.
I have temp sensor on FETs. Temps are no problem.

Regen wire is connected to correct spot. I have the regen pad (forget what it is) connected to gnd to enable regen. Sounded like good advice, unfortunately not the case.
Ebrake wire is not connected to anything since I have slip charged enabled (auto regen on throttle 0%)

HVC is not an issue either. I know what regen feels like when it hits HVC and my case is different.

The closest I can describe it (before I get the video) is as if 2 of the 3 phases working towards regen and 3rd one missing. It makes a very audible stuttering sound which can be heard 10 feet away.

Okay. Getting close here!
I tested Cromotor with wheel off ground with regen. It stuttered.
Then I put my ht3525 in a vice and spun it up a little. Regen kicked it but I didn't hear a stutter.

I did measurement of hall red and black to bike frame/motor. Several ohms of resistance shown.
Edit: redid this. Continuity between halls and frame is no longer present but I do get around 4.5k resistance from hall ground or positive to phase wire of motor!

Ht3525 shows no reading for same test.
I think I may need to find and isolate the red and black hall wires from the motor?

Sounds like you found the problem, or at least the evidence of it.

I rebuilt my wheel and in the process also redid my cromotor wiring harness. Replaced a few hall sensors and added a temp sensor.
After it was all done I mounted the wheel, started my bike up as normal (with regen wires connected this time) and spun it up to test regen.
It works! Problem must have been phase wires leaking current to hall sensors (wiring harness was a damaged a couple times from when friends crashed my bike several months ago).

Nice tire!

Maybe design a crash guard for the wires. That seems to be a vunerable spot on most hub motors.