Stuttering

[drsolly]

I have a bike that was using the Infineon 12FET 4110 that I had from em3ev.com; it was working well, but I wanted to see if I could use a smaller controller. And someone here explained to me that 3077 FETs are better (less power consumed by the FETs) at low voltages.

So I installed the 6 Fet 3077 in the same bike. It has a 24v, 250 watt motor, and I'm running it from 8S Lipo. At first, it ran OK, but after a few minutes, it was running rather jerkily under load. It was giving current to the motor, then stopping, then giving current, then stopping; each cycle was taking about one second.

I used the bike motor tester that I got from em3ev (it's the same as the Lyen one); it reported that the halls are OK, and that the windings work (but I had to turn the wheel backwards to get this reading, and maybe someone can tell me why this was?). It said that the throttle was OK, and that the controller was OK (it showed three LEDs that cycled round the circle of the tester).

So then I put on the 12Fet3077, thinking that maybe the 6Fet3077 wasn't powerful enough.

I got the same "stuttering" effect.

So then I replaced the 12F4110 that I had there before, and now the bike runs smoothly; this means that the bike motor and halls must be OK.

Can anyone suggest what the problem might be?

Thanks.

 

[wesnewell]

Bad phase or hall connection?

 

[teklektik]

Where did you buy the 6/12FET 3077 controllers? EM3EV, Lyen, or elsewhere...

 

[drsolly]

I got them from em3ev

I'm going to change the phase connectors, in case that's the answer.

 

[DAND214]

Maybe it's hitting LVC.

Dan

 

[Gregory]

Maybe it's hitting LVC.

Dan

Yes what's the LVC of the controllers?
LVC on a 36V controller is normally around 31V so right around the mid discharge of your 8S LiPo.

 

[drsolly]

LVC is 25.6 volts. No, it can't be hitting that, because I have a beeping battery monitor that would alert me.

I've replaced the connectors with 4mm bulllets, I'll give it a test tomorrow.

 

[d8veh]

You have to turn the motor backwards to test the halls ecause there's a free-wheel in the hub. The motor doesn't turn when you rotate the wheel forward.

The stuttering is probably because you have the wrong phase/hall sequence. You can't always go by the colours.

Look at the diagram in this thread:
http://endless-sphere.com/forums/viewtopic.php?f=2&t=54000

 

[nechaus]

id say its a hall or phase problems, or a short with the phase..

Spin the wheel, if dd, is there more resistance than usual ?
Should not be hard to spin with you hand... if its hard to spin, unplug the phase and spin it again, see what happens

 

[drsolly]

OK, I changed the phase connectors to 4mm bullets, so I'm pretty confident that I have good connections.

Under no load (wheel in the air), the motor runs smoothly.

Under load, the motor (nominal 24v 250 w, running on 8s Lipo, which is about 30v) runs smoothly at first, taking me up to as much as 40 kph on a slight downhill and 32 kph on a slight uphill. But after I've done several hundred yards, I start getting the on-off-on-off-on effect again, with the motor switching on and off about once per second. The fact that it runs smoothly for several hundred hards, makes me think that I have the connections correct.

I felt the controller, it's very slightly warm, but certainly not hot. The motor is cool too.

And still the whole thing works fine with the 12 fet 4110 Infineon controller, just not with the 6 fet 3077 Infineon controller.

So I'm a bit baffled.

 

[dogman dan]

Either

A one or both other controllers need a different wire order to run well.

B one or both other controllers are not connecting as well, on either phase wires or halls. A good squint at all connector contacts would be the first thing to check. Maybe you have a loose halls pin, but it fits better and does not shove out when you plug in the controller that works good.

 

[gwhy!1]

you say that the controller is getting slightly warm after a couple of hundered yards .. if every thing is setup correctly then take it from me that a 6fet fitted with 3077's will defo not get remotely warm let alone slightly warm.. is this a sensorless controller ? .. for ref I have rum many a 6fet with 3077's @90A battery current and even then it dont get excessively hot and has more than enough power at these power levels. What is the battery current and phase current limit set to on the 6fet ?. what is the actual peak current you are getting from the 6fet when its stuttering... what is the noload current at WOT ? what is your min volltage reading when accelerating ?


Edit: but it do sound like a incorrect wire combo... and your lvc is kicking in as a result.
but as you bought the controllers from the same supplier then all the wiring should be identical :? ... but I would defo check

 

[Tench]

You are forgetting that the controller has its own low voltage cut off, it was pointed out above, this could be cutting off the power well before your battery low voltage alarms go off and sounds to me the most likely cause of your problem.
Fully charge your battery and see at what voltage it starts to cut off, then charge the battery again, I bet it will run fine until you hit that voltage again.

 

[d8veh]

Is it a sensorless controller, and is it a sensored motor? Some motors don''t like to run sensorless, in particular the Q100.

 

[drsolly]

The controller is sensored, and so is the bike.

I have LVC set to 25.6 volts, the battery is fully charged 8S Lipo, so it is 33 volts.

When I say that the controller is "very slightly warm", that could have been because it was in full sunlight, it was that slight an amount. I'll do some measurements on current and voltage later, as suggested by gwhy

Next, I'll try experimenting with the order of connection of the phase wires. That'll probably be Monday.

 

[999zip999]

Just use light throttle or not wot. Fine out your no load amp draw before twristing full throttle.

 

[drsolly]

I get the same stuttering at low throttle.

I just got an email from em3ev; they point out that the controller is rated as 36-60v, and I'm running it at 33v. But I did reprogramme the LVC to 25.6v. And the bike works fine with their 12F4110 controller (12 4110 fets) which is also rated at 36-72v

Tomorrow, I'll try running it at 12s, that's about 50v, if I still get the stuttering, that eliminates that possibility.
If it doesn't stutter, then that eliminates the possibility that I have the phase wires wrong. Either way, it'll narrow it down.

When I get this bike working, it's going to be rather nice, because it's a lightweight folder, which is ideal for geocaching.

 

[Gregory]

Sounds plausible.

If that turns out to be the cause, you can modify the power resistor (which sits just before the voltage regulator that supplies 5V ) to drop the compatible battery zone from 36/72V to more like 24/60V.

 

[drsolly]

Some good news and some bad news.

I put 12s on the bike, and it worked fine like that. So it was the LVC acting.

I had thought that, when I programmed the LVC to cut out at 25.6v, then it would do that. I think it was actually cutting out at maybe 30v. I still don't understand why the 12F4110 doesn't give that problem.

Now the bad news.

While testing the bike, something terrible happened. I was running it up a steep local hill for the test, and halfway up, the motor started making a grinding noise. I switched off power, and cycled home under leg power.

When I had a look, the motor won't turn. Because I was running it at 50v, and it was pulling maybe 30-40 amps, I'm wondering if I blown something.

It's a 250W, small hub, so I'm guessing it's geared, I haven't ever opened it up. I can wheel the bike forward, but the back wheel (with the motor) totally refuses to go backwards.

What have I done, and is this going to be fixable without a total motor rewind?

 

[Tench]

You are going to have to open it up to find out, sounds like time for a motor upgrade!

 

[gwhy!1]

sounds like the gears are mashed.. but the motor is prob still fine.

 

[drsolly]

Internal gears is what I was thinking.

But I can't even get the cassette off. I've been hammering away at one of the indentations on the thing that keeps it together, but it won't budge. And I can't use the usual spline tool, there's no way to put it on.

 

[DAND214]

A 6fet EB306 needs the low voltage resistor to run at that low voltage. The 12 doesn't, it self adjusts.
Take the cap of the case, the one without the wires. Look at what resistor is there. 330 ohm abopve 36v to 48v and then you double it again for higher voltage.
This is what I have run into on 6fet EB306 Infinion.

Dan

 

[gwhy!1]

Internal gears is what I was thinking.

But I can't even get the cassette off. I've been hammering away at one of the indentations on the thing that keeps it together, but it won't budge. And I can't use the usual spline tool, there's no way to put it on.

I dont think you need to take the cassette off to look inside the motor, just remove the screws from around the outside of the motor and the covers should be able to be pulled off.