Page 1 of 1

Controller delivers too less current?

Posted: Sat, 03 Mar 2012, 08:11
by Polmage
Hey,

It seems like I have a problem with my controller.
Its a Kelly KDH12601e (600A max.).

System:
Battery: Headway 40160SE 16 seriel, 3 parallel
Motor:   Agni 95R
Controller: Kelly KDH12601E

In the software I can adjust percentage of maximum battery and motor current. I chosen always the same.

So, first I set it to 10%, which should be 60A, but I could measure almost no torque at the AGNI95 motor.

Fany from kelly said to me, that controllers just work right after ca. 40% because "affecting the sampling of the datas".

So I started with 25% motor/battery current and measured the torque in 5% steps. Torque shouldn't be exactly right because its at different rpm. I used a breaking-torque curve similar to the effects of air resistance:

25%: Should do: 150A -> 19Nm   Torque measured: 7Nm
35%: Should do: 210A -> 27Nm   Torque measured: 8Nm
40%: Should do: 270A -> 35Nm   Torque measured: 11Nm
60%: Should do: 360A -> 48Nm   Torque measured: 13,5Nm
80%: Should do: 480A -> about 65Nm, not on diagram anymore because motor-limit is 400A    Torque measured: 16Nm

I would think, the Torque-Sensor works right, because batteries and also motor didn't get warm (let it run 10-20 sec). With 480A the motor power should have been that high, that the asynchron break wouldn't have bee able to break it.

Now I'm not sure whats maybe my failure or defect. Sadly I couldn't measure the current because the school just had a amperemeter with a frequency of 9kHz. Controller pulses with 16,9 kHz I think.

Anyway, I tried to measure it, hoped it wouldn't pulse anymore with fully opened throttle. But what I got was just bullsh*t, e.g. 95A @ 80% controller current, which should have ended in about 12Nm @ the motor. So it was always less then my expectations.

What hapened at the controller: a nut of the RS232 fell into the case, during delivery, but don't think it shunts out sth.. Theres no failure code (just sometimes at starting up it says overcurrent, manual also says this can happen (Code ... ..)). KeySwitch off/on again and it works normally.

Do you think some undervoltage of the batteries can effect this? Didn't use some monitoring for it. But after driving I used a multimeter at every cell and all were at 3,31V.
Or you think the Agni 95 is damaged?

Problem now is, that I had to demount the dynamometer (its a study-project). Not sure what I should now write into my report.

Greetings

martin

Controller delivers too less current?

Posted: Sat, 03 Mar 2012, 15:11
by coulomb
Polmage wrote: Sadly I couldn't measure the current because the school just had a ammeter with a frequency of 9kHz. Controller pulses with 16,9 kHz I think.

That should not matter much. The inductance of the motor should smooth out the current fairly well, with just a little 17 kHz triangle wave component on top of a lot of DC. The ammeter may have say a 10% error because it can't properly handle the triangular component, but it should still give you a good idea of the current.

Is the pack sagging badly in voltage perhaps? Maybe due to a poor cell-to-cell connection somewhere?

Controller delivers too less current?

Posted: Mon, 05 Mar 2012, 01:38
by Polmage
mmh, I used a cut off Voltage of 48V as controller setting, because there is is no undervoltage monitoring.

I always thought it sets the possible current to zero.
But here it sounds like, it would just decrease it.
But if it would decrease it to the lowest possible current, that voltage is still over 48V:
Why the measured current (if it is right) and torque is growing with higher max. current settings.

Controller delivers too less current?

Posted: Mon, 05 Mar 2012, 04:09
by weber
There's not enough information to determine what is going on, but I think it unlikely that there is anything wrong with either the motor or the controller since they are both new.

Ideally you would measure battery voltage and current, motor voltage and current and motor speed and torque, at the various settings.

I've never heard of a DC ammeter having an upper frequency limit in kHz. Makes no sense to me. Sounds like you're trying to measure AC current, which would be pointless.

Controller delivers too less current?

Posted: Mon, 05 Mar 2012, 08:05
by antiscab
3.31v at no load doesn't tell you a whole lot - thats anywhere between 20% SOC and 90% SOC

the 95A I presume is motor current?

out of curiosity, why are you changing both motor and battery current limits?

I would try lowering the low voltage limit - 48v between 16 cells is 3V ea
on headways, you would be limiting yourself to about 2-300A

what are you using as a load to limit rpm?
If the rpm is too high, the back-emf will prevent significant current to flow (motor voltage = battery voltage)

what voltage is the throttle putting out? (if the throttle doesn't put out above 4.5V, and the controller takes above 4.5V as full throttle, you won't get full throttle)


Matt

Controller delivers too less current?

Posted: Mon, 05 Mar 2012, 09:22
by Polmage
no, I'm definitely not measuring AC.
This is the Ammeter I have:
http://www.tequipment.net/pdf/LEM/2050_man.pdf
On site 5 there is a sentence: "all measurements DC and 10Hz to 1kHz"
Is this the frequency it can measure? I don't get it....
Therefore I can't find any information about measurable frequencies with clamp-on ammeter, I'm going to believe that it either doesn't matter or nobody needs it.

Controller delivers too less current?

Posted: Mon, 05 Mar 2012, 15:18
by weber
Polmage wrote: no, I'm definitely not measuring AC.
This is the Ammeter I have:
http://www.tequipment.net/pdf/LEM/2050_man.pdf
On site 5 there is a sentence: "all measurements DC and 10Hz to 1kHz"
Is this the frequency it can measure? I don't get it....

OK. This just means that when it is set to measure RMS current (which may be a combination of DC and AC) it can only measure AC components in the range 10 Hz to 1 kHz. When it is set to measure DC current only, that frequency range should be irrelevant. Even if it was relevant it would be good that it cuts off at a lower frequency than the PWM frequency.

As Coulomb said, there should only be a small amount of AC PWM ripple on top of the DC. And we don't want to measure it anyway. We only want to measure the DC.

So provided you had the meter on the correct setting, and you unclamped it from the wire and zeroed it before each measurement (or at least before the first measurement), then those current measurements are almost certainly real.

Controller delivers too less current?

Posted: Wed, 07 Mar 2012, 20:31
by woody
Tell us more about the torque measurement.

Can you do some short duration stall tests? e.g. attach a 2x4 to the shaft and have it pushing against a bathroom scale at a set distance from the centre of the shaft.

Controller delivers too less current?

Posted: Thu, 08 Mar 2012, 06:07
by Richo
16S3P is ~ 43mR
So
75A=48V
150A=44.8V
210A=42.2V
270A=39.7V
360A=35.8V
480A=30.8V

So as Matt points out the controller is probably cutting the current to keep the pack above 48V which is less than 75A on the pack side.
So the torque will be significantly less than expected and is dependant on the RPM used.

Controller delivers too less current?

Posted: Wed, 25 Jul 2012, 17:59
by cruisin
KELLY is junk and you shouldnt be suprise of its shorcomings. You bought it because it was cheap, and now you are disappointed. I wonder why. See what happens when you buy price?