TIM600 Explosion

Technical discussion on converting internal combustion to electric
Stiive
Groupie
Posts: 295
Joined: Wed, 31 Dec 2008, 08:26
Location: Melbourne

TIM600 Explosion

Post by Stiive »

All,
My name is Stefan and i am currently Co-Team Principal of team Swinburne Electric. We are developing a formual spec electric race vehicle for the formula SAE electric race on December 9th.

Unfortunately during recent testing our TIM600 inverter has blown up what looks like a snubber. Was wondering if there are any technical gurus in Melbourne who can come down and help us either replace the snubber or diagnose what went wrong.
ALTERNATIVELY IF ANYONE HAS A MES-DEA TIM600 INVERTER THE TEAM COULD BUY ASAP THAT WOULD BE IDEAL! HAPPY TO PAY FULL BRAND NEW PRICE, JUST NEED IT ASAP!

The inverter blew up before the motor ever span, we were just following the instructions of setting up the motor settings and went to run the tests which verify that the phases are plugged in the right way.

We are situated at Swinburne, Hawthorn.

Kind regards,
Stefan
Rgds,
Stiive
User avatar
Johny
Senior Member
Posts: 3733
Joined: Mon, 23 Jun 2008, 16:26
Real Name: John Wright
Location: Melbourne
Contact:

TIM600 Explosion

Post by Johny »

Hi Stefan
I have a feeling that there is more damage than just a snubber.
Can you give more details, perhaps a photo.
I don't know the MES-DEA TIM600 at all and your time frame appears to be very tight - but I can at least try to advise.

What were the symptoms when you realised something was wrong?
(post photos, size them < 100K)
User avatar
Richo
Senior Member
Posts: 3737
Joined: Mon, 16 Jun 2008, 00:19
Real Name: Richard
Location: Perth, WA

TIM600 Explosion

Post by Richo »

That's a lot of BEMF to blow a snubber up!

Try jack the car off the ground and take the brake off.
An inverter should still work with a snubber removed.
(just not under heavy load perhaps Image )

Either that or you have the phases plugged in the wrong way Image

Sorry but I would be suprised if there was a spare TIM600 in Austrlaia.
Worth a shot tho I guess.

What is your back-up plan?
So the short answer is NO but the long answer is YES.
Help prevent road rage - get outta my way!
Stiive
Groupie
Posts: 295
Joined: Wed, 31 Dec 2008, 08:26
Location: Melbourne

TIM600 Explosion

Post by Stiive »

Well we have a new TIM600 from CERES now and are still having problems.

When trying to do a test to determine if the phases are in the correct way (which we have no idea which way they should be in or why it would screw with the inverter and not just make the motor spin backward) the motor randomly goes backward or forward and makes a loud squeeling noise and draws about 400A!. The error then reads that there is a wrong connection of motor phase U, V, and W and that we should swap 2 of the wires which we have done multiple times.

Crazy, didnt think the inrush should be so high but apparently it could be for a very cheap way of the controller to measure the inductance of the motor
Rgds,
Stiive
7circle
Senior Member
Posts: 540
Joined: Thu, 04 Feb 2010, 02:18
Real Name: Ken
Location: Melbourne

TIM600 Explosion

Post by 7circle »

The Motor is Permanent magnet type?

If so, the phase windings need to mauch the hall effect or resolver direction.

Is there position feed back of some sort?
Stiive
Groupie
Posts: 295
Joined: Wed, 31 Dec 2008, 08:26
Location: Melbourne

TIM600 Explosion

Post by Stiive »

nah 3phase induction.. the inverter is vector controlled so yeh uses encoder and flux feedback to control the motor
Rgds,
Stiive
7circle
Senior Member
Posts: 540
Joined: Thu, 04 Feb 2010, 02:18
Real Name: Ken
Location: Melbourne

TIM600 Explosion

Post by 7circle »

The 2003 TIM manual has X3 with encoder A and B signals for feed back.

The software link will have a test procedure for checking the encoder is counting +ve for forward rotation.

The M2-AC30-L (if your still using it) is Async but the controller will use encoder feed back for Vector Control especially at low RPM's

What is the counts per RPM of the motor and ho many poles is the M2-AC30-L ?
I assume 4 pole like the AC50.

Keep the current limits low!
Stiive
Groupie
Posts: 295
Joined: Wed, 31 Dec 2008, 08:26
Location: Melbourne

TIM600 Explosion

Post by Stiive »

we are guessing 4 pole, we do not know.

The test disregards any current limiters we put in the parameters.. seems to do its own thing.. We have encoders A and B hooked up and can spin the motor with our hand and see the value change on the screen, it has 64 pulse per rev
Rgds,
Stiive
7circle
Senior Member
Posts: 540
Joined: Thu, 04 Feb 2010, 02:18
Real Name: Ken
Location: Melbourne

TIM600 Explosion

Post by 7circle »

May as well keep all my comments in this post.
Can you see the signal change in the controller software that allows paramater change?

The wiring may be effected by noise.

Usually is paird A-/A+and B-/B+ then 0V/+5V each is run as a twisted pair.

The common point is at the controller for common wires. Could be +5 with A+ and B+ if A and B inputs sink current.

Like you said though looked okay with DSO. But was that with big currents causing noise and Voltage noise from PWM?


-- Previous Private Message --
Sent by : Stiive
Sent : 2010 December 05 at 10:55pm


yeh we have swapped the phases around multiple times with no luck, will try the encoder again i guess but last time it made it worse... it had a 4pin plug and a 2 pin plug for the NTC thermistor... double and tripple checked the outputs and even had them up to the CRO, everything seems fine
The test current limits are often separate to the operating parameters or a percentage scaling of them.

The 64 pulse per rev could be for electrical RPM so 128 pulse per mechanical RPM.

Other wise it could be only 32 pulses per electrical RPM.
7circle
Senior Member
Posts: 540
Joined: Thu, 04 Feb 2010, 02:18
Real Name: Ken
Location: Melbourne

TIM600 Explosion

Post by 7circle »

Is the 64 pulses per Rev the default value?

Did you check that 360 degree turn mechanically actually gives 64 pulses?

Is the computer software still in Italian?

Did the encoder come with the motor?
Is it built in?

Can the controller work in sensorless mode with No-Load on the axle?

You don;t have to answer all my questions they are just prompts for you in case you have not thought of them.

When motors are not rotating its often noise or wrong feed back.

You have wired through the current sensors on two of the phases?
7circle
Senior Member
Posts: 540
Joined: Thu, 04 Feb 2010, 02:18
Real Name: Ken
Location: Melbourne

TIM600 Explosion

Post by 7circle »

Found the project facebook page and it show the 4 pin encoder connector onthe CAD drawing.

if you can't change the wiring you could try and keep it away from the power cables.

It also shows only 3 wires from the motor.

The motor can be tested with 50 hz three phase.

If isolation is an issue three transformers can be used to give a lower voltage.

Your battery pack is at deadly voltages so just consider other options.

Three 240V lights wired in star to monitor the the motor voltage when at zero to 10 hz will show what's happening too.

The transformers (toroidal are very good for VTs) can be used to connect a 4 channel scope to the motor three phases.

The bulbs can be removeable from bayonet housings so you can remove the load.
The bulbs can also be used in series with the motor to limit current.

The power labs in the EE building had some gear for 4th year experiments, could all be virtual experiments now.
User avatar
woody
Senior Member
Posts: 1716
Joined: Sat, 21 Jun 2008, 02:03
Real Name: Anthony Wood
Location: Mt Colah

TIM600 Explosion

Post by woody »

Check the recommended number of pulses per revolution for the controller - Danfoss recommends 1024ppr, MES-DEA may be expecting a lot more pulses than it's getting.
Can it run encoderless?
Planned EV: '63 Cortina using AC and LiFePO4 Battery Pack
7circle
Senior Member
Posts: 540
Joined: Thu, 04 Feb 2010, 02:18
Real Name: Ken
Location: Melbourne

TIM600 Explosion

Post by 7circle »

Yeah I was suprised at such low pulse count but the old 2003 manual had 64 pulses per RPM and on the motor mechanical drawing.
Stiive wrote: ...When trying to do a test to determine if the phases are in the correct way (which we have no idea which way they should be in or why it would screw with the inverter and not just make the motor spin backward) the motor randomly goes backward or forward and makes a loud squeeling noise and draws about 400A!. The error then reads that there is a wrong connection of motor phase U, V, and W and that we should swap 2 of the wires which we have done multiple times.


In the earlier 2003 manual it says the alarms from the "First Test" is either wrong phase rotation (A14) or encoder and/or poles are wrong (A15).

This test in some other controllers can be done at low torque or at high torque. If its using high torque it is usually measuring the inertia of the system. The motor is only part of the system. So doing the test with the drive connected to the wheels and having them off the ground will come up with a higher inertia. And will also show up and play in the gear train.
Stiive wrote:Crazy, didnt think the inrush should be so high but apparently it could be for a very cheap way of the controller to measure the inductance of the motor
The term inrush usually applies to capacitors.
The 400A you measured was that on the DC side or on the Motor side phase wires?

As these motors are so light and with an encoder with only 64 pulses the measurement of Motor inductance would be very difficult so using larger peak. currents would be needed to give a longer time count of dI/dt Rate.

Image
7circle
Senior Member
Posts: 540
Joined: Thu, 04 Feb 2010, 02:18
Real Name: Ken
Location: Melbourne

TIM600 Explosion

Post by 7circle »

7circle wrote:This test in some other controllers can be done at low torque or at high torque. If its using high torque it is usually measuring the inertia of the system. The motor is only part of the system. So doing the test with the drive connected to the wheels and having them off the ground will come up with a higher inertia. And will also show up and play in the gear train.


After looking through the old manual the J (1st moment of inertia) doesn't appear to be measured. But as there are so many control parameter time constants it would become part of one of those.

Also:
"P t 114 : Test current for ΔVLS determination : 0÷100.0 : 50.0 % : I NOM MOT"
This may help in reducing the 400A value your seeing.
What do you have this at?

Lots of other variations for other test in Auto Tuning.

The Encoder setting allows for much larger counts per rev upto 2^15
User avatar
Johny
Senior Member
Posts: 3733
Joined: Mon, 23 Jun 2008, 16:26
Real Name: John Wright
Location: Melbourne
Contact:

TIM600 Explosion

Post by Johny »

What is the DC input voltage (battery pack volts) and the rated motor voltage?
User avatar
Richo
Senior Member
Posts: 3737
Joined: Mon, 16 Jun 2008, 00:19
Real Name: Richard
Location: Perth, WA

TIM600 Explosion

Post by Richo »

Stiive wrote: Well we have a new TIM600 from CERES now and are still having problems.


So since you only have two parts A controller and a motor and you have replaced the controller that leaves one part suspect.
The motor.

Check the internal resistance between each terminal for shorts.
Do a hi-pot to the chassis and terminals.

Probably not such a great idea connecting direct to 3-phase 50Hz mains if there is a problem Image
So the short answer is NO but the long answer is YES.
Help prevent road rage - get outta my way!
7circle
Senior Member
Posts: 540
Joined: Thu, 04 Feb 2010, 02:18
Real Name: Ken
Location: Melbourne

TIM600 Explosion

Post by 7circle »

Wondering if anyone has thoughts on cabling inductance to the controller that only has (~ 6 x 45uF/700V in parallel) then feeds to the motor.

Should DC cabling be kept as a close pair of wires.

And do motor phase wires need to be kept all the same length to keep the phases symmetrical?

How much DC bus capacitance would be typical of a EV controller for an induction motor?

Should the DC Bus have over voltage protection that dumps voltage when battery isolators are opened?

I'm thinking the extra capacitance needs to be added
For example, like some of the kelly controllers have a seperate DC Bus Cap box.

The circuit in the EV racer will have shorter leads but currently with the longer and variable legth leads the appears to be huge self resonances on the DC Bus.

Any ideas?
User avatar
Johny
Senior Member
Posts: 3733
Joined: Mon, 23 Jun 2008, 16:26
Real Name: John Wright
Location: Melbourne
Contact:

TIM600 Explosion

Post by Johny »

If this was an in-use failure then those considerations would be important but this is just initial commissioning. There is something more seriously wrong here. (IMO)
7circle
Senior Member
Posts: 540
Joined: Thu, 04 Feb 2010, 02:18
Real Name: Ken
Location: Melbourne

TIM600 Explosion

Post by 7circle »

I'm concerned that the oscillations are so high that the capacitor 700V limits can be exceeded.
But the other issue is the controller first test is checking for correct connection to the motor. So with 40% ripple voltage on the bus the current in the fields cannot be controlled to the correct AC sinusoid so the test fails.

I'm not sure how smart they are trying to be with the test. The explanations are minimal.
7circle
Senior Member
Posts: 540
Joined: Thu, 04 Feb 2010, 02:18
Real Name: Ken
Location: Melbourne

TIM600 Explosion

Post by 7circle »

The Ripple is at a few frequencies 6kHz and 2.5 kHz.

If the DC bus is not rock solid the Inverter has poor hope of generating the 3 phase current waveform.
User avatar
Richo
Senior Member
Posts: 3737
Joined: Mon, 16 Jun 2008, 00:19
Real Name: Richard
Location: Perth, WA

TIM600 Explosion

Post by Richo »

Try disconnect some of the batteries for reduced power and voltage during testing.
So the short answer is NO but the long answer is YES.
Help prevent road rage - get outta my way!
7circle
Senior Member
Posts: 540
Joined: Thu, 04 Feb 2010, 02:18
Real Name: Ken
Location: Melbourne

TIM600 Explosion

Post by 7circle »

Have a look at this Torque current in induction motor document.

It explains what happens when V/Hz is incorrect and about voltage boost for start torques.

YASKAWA - Induction Motor Speed Torque Characteristics
7circle
Senior Member
Posts: 540
Joined: Thu, 04 Feb 2010, 02:18
Real Name: Ken
Location: Melbourne

TIM600 Explosion

Post by 7circle »

The Motor Name plate is:

Motor type: M2AC30L
Service: S21H
P: 30kW
V: 210V
I: 105A
RPM: 5000RPM
Freq: 175Hz

The Motor was tested by the manufacturer www.electro-vehicles.eu
using a TIM400

The motor description they used was "M2 AC30 @ 96V dc s2-1h"

They supplied at table and chart of results
"speed":"torque":"power"
"rpm":"T (Nm)":"P (kW)"
0:119.3:0
500:118.4:6.2
1000:117.5:12.3
1500:116.1:18.2
2000:115.3:24.1
2500:112.5:29.5
3000:101.4:31.9
3500:87.2:32
4000:75.4:31.6
4500:65.4:30.8
5000:56.1:29.4
5500:46.2:26.6
6000:36.2:22.7
Image

The test was done with a TIM400 at 96Vdc but the project is using a 300V dc Battery pack and and the motor Name plate is for 210V ac.

This chart doesn't help detiremin the V/Hz ratio for high performance.
As it is only showing continuous torque and power ratings for 1hour operation without overheating.

If anyone knows of helpful links to info to detiremine this from motor spec that would be great.

Ken
7circle
Senior Member
Posts: 540
Joined: Thu, 04 Feb 2010, 02:18
Real Name: Ken
Location: Melbourne

TIM600 Explosion

Post by 7circle »

The name plate doesn't show number of poles
So if it was 4 pole the P=Tw shows for
w = 1000RPM : rad/sec = 33.33

The the
P = Tw = 117.5 x 33.33 = 24.61 kW

if its 2 Pole
P = Tw = 117.5 x 16.67 = 12.30 kW

So the motor looks like 2 pole if the Chart supplied is correct.
User avatar
Richo
Senior Member
Posts: 3737
Joined: Mon, 16 Jun 2008, 00:19
Real Name: Richard
Location: Perth, WA

TIM600 Explosion

Post by Richo »

It's worse than you think!
They tested it at 96V DC (not AC) Image

It look like the RPM/Power is linear to 2500RPM.
IE the V/F ratio cannot be maintained past 2500RPM.
Now 96V DC would be 67.9V AC
We'll say 65Vac for marginal losses
Now 5000RPM=175Hz (from plate)
So 2500RPM should be 87.5Hz.
So V/f ratio would be 65/87.5=0.7428

Are you sure the name plate says 210VAC?

I do agree that it sounds like a 2-pole.

As long as the inverter gives something like the V/f your DC bus won't matter too much.
So the short answer is NO but the long answer is YES.
Help prevent road rage - get outta my way!
Post Reply