[chris01290]

Hello, if anyone could help I would be most appreciative:- I have a Sevcon Gen 4 size 8 and it will not recognise the signal from the motor resolver (on pins 19,20, 31 & 32). In my experimenting I have determined that the controller will read a sin/cos signal on pins 19& 20 (sin/Cos) to pin 21 (encoder 0v) when the 'encoder' is set to sin/cos- but I can't use this as the exciter signal disappears after keyswitch toggle (and 5V appears across pins 9 & 21 which are there to power the encoder, when that is configured/set in the DVT). What is confusing is that I get a resolver exciter signal from pins 7&8 (when configured to single or multipole resolver), my assumption being that if it was not configured (in the hardware) for a resolver it would not give the exciter signal. I have been though every CAN address I have access to on the DVT and can see nothing that would inhibit the resolver signal into the controller (and have tried almost every combination of settings in an effort to figure it out). I am wondering if on that particular model they changed the pin designations for the resolver input. The controller model is 634H83210 and is not on the Sevcon drg I have, nor can I find any reference to it anywhere after a multitude of internet searches, any info on that would be gratefully received too. I am now trying to source a sin/cos encoder, but even that is presenting challenges so it would be nice to get the controller to recognise the resolver (fitted to a Meiden 60Kw motor from a Mitsubishi), but am presently at a loss. As I said at the beginning any help gratefully received.

I am having exactly the same problem, i have put the oscilloscope on the sine a cos and can see the return signal but in the DVT debugger the min and max voltages for the cos and sine are 0v,
Did you resolve this issue?

 

[remy_martian]

What's the level of the resolver outputs vs exciter output level?

 

[chris01290]

sitting at about 1.2v peak on the sine output. Resolver input is just under 4v peak, i would say its a 0.3 resolver. (see attached wave forms)

 

[remy_martian]

Are you using all four resolver inputs for sin/cos?

Looking at the manual, Gen4 Size 8 Product Manual V3_3 (RELEASED) V3 3

It sounds like OP was using only one of the paired wired for each signal and referencing it to Encoder 0V which is wrong.

For a resolver application, sin+ and sin- HAVE to be used and cos+ and cos- need to be used. Encoder 0V is irrelevant.

This may be obvious to you, or not...I have no way of knowing how you connected the leads of the scope. You will get resolver outputs and you will appear to get a signal at the Sevcon's inputs, but it's blind to them if you don't connect the resolver pairs to the resolver pair inputs and ignore the Encoder 0V input.

 

[Martin1599]

Hi all, thanks for getting a conversation going on this, much appreciated. To clarify for Remy above, I did have all four sin cos +/- connected (on pins 19,20, 31 & 32 ) when I had the initial problem. Everything after that was experimentation to try to understand what was going on with the controller- Chris, your problem does indeed look identical to my issue, and I assume you too have all four sin/cos connections correct into the controller, what is the part number on your controller please because if you have the same as mine (634H83210) it may be that those controllers will not read resolvers. I have since given up with the resolver and installed an encoder, which the controller is reading with no problem- the only step left now is to spin the motor at a high enough rpm to allow the controller to read the encoder offset- this little gem was 'hidden' in the DVT help (under motor configuration) but was nowhere to be found in the applications reference manual or the DVT manual. Essentially the motor needs to be spun at a minimum rpm (750 ish for a 3 pole motor, 85 ish for a 28 pole motor) in order for the controller to read the encoder/resolver offset- it has crossed my mind that maybe when I had the resolver fitted I need to have turned the motor over at a much higher rpm for it to read the resolver (it doesn't make sense but it is the only thing I didn't try). Chris, if you have a means to do that (i.e. push the vehicle along if the motor is fitted) that may be worth a try. Please let me know if that works as I may refit the resolver and give it another go (the resolver will be more robust than the encoder I have fitted). Hope this helps. Cheers, Martin.

 

[chris01290]

yes my controller has the same part number and i am using the same controller pins for the cos and sine inputs as you have said.
From the research i have done these motors us 0.29 resolvers and in the manual it says:
"Gen4 size 8 Beta controllers are built specifically to operate with a resolver transformer ratio of either 0.5 or 0.29. For a particular Gen4 size 8 model, the resolver transformer ratio cannot be configured to a value different from the value stated for that model."
Which might lead me to think we have the wrong controller version for these resolvers. I have contacted the supplier of the controller to see if they can find out the above.
If this is the case we might be able to increase the output using a step up transformer to increase the output from 0.29 to 0.5.
The car is on axle stands and have turned the motor over a number of times and the error has not cleared.
I have also put the scope in a slow trace while turning the motor and got the following outputs on cos and sin.

thanks
Chris

 

[Martin1599]

The transformer ratio also crossed my mind (don't know whether our controllers are 0.29 or 0.5) but I figured that even if the TR was incorrect I should still be seeing some signal on the controller. Re spinning the motor- it needs to spin at a a minimum RPM of 760 (for a 3pole motor) to register the offset, maybe the same applies to reading the resolver (I'm guessing here though)- which is what I will try if I ever decide to replace the resolver. I'll update when I have news regarding the final commissioning of the encoder.

 

[chris01290]

Emailed the manufacturer and got confirmation the 634H83210 is a 0.5 resolver, so going to investigate using a line transformer to step the signal up. Just waiting for them to arrive. As soon as I have more info I will update the post

 

[remy_martian]

Good as an experiment, though you'd want to use a signal transformer with good linearity over the frequency range in practice.

 

[chris01290]

ok so a bit of a update, the transformer seems to solve the problem the error has cleared. i can now see the RPM's increase in the test byte in the helper. So now i can put the motor into drive i can feel there is more resistance when i turn the motor backwards then forwards, but when i press the throttle the motor does not spin and the controller then errors. viewing on the helper is looks like its increasing the throttle but the motor does not spin so its erroring with w throttle error.

I have gone through the commissioning in the help, has anyone got any idea's of what the motor parameters should be for the meiden Y61 outlander motor should be?

 

[chris01290]

ok so a bit of a update, the transformer seems to solve the problem the error has cleared. i can now see the RPM's increase in the test byte in the helper. So now i can put the motor into drive i can feel there is more resistance when i turn the motor backwards then forwards, but when i press the throttle the motor does not spin and the controller then errors. viewing on the helper is looks like its increasing the throttle but the motor does not spin so its erroring with w throttle error.

I have gone through the commissioning in the help, has anyone got any idea's of what the motor parameters should be for the meiden Y61 outlander motor should be?

Mg migit with outlander rear electric motor sevcon controller

youtube.com

Ok I works added a short video to you tube, looks like it was not running because the max battery current was set to zero

 

[chris01290]

Currently all appears to be running smoothly my Best solution was to put one transformer on each of the sine and cos signals into the controller, is was easier to match the output impedance that way. This is the final transformer i settled on:
42TM011RC - 1.5K:500R windings ratio 1.8:1
ebay Audio Matching Transformer, Centre Tapped on Both Sides, Various Types, Xicon | eBay

Note: this is the solution i am using as my sevcon controller requires a 0.5 resolver and the motor resolver on my motor was a 0.29 but I do not give any warranty with my advice You should check the voltage levels from the transformer before connecting to the controller.

 

[Martin1599]

Chris, thanks for the update, very much appreciated. I'd shelved my project for a while due to other commitments but will dust it off and continue soon, (possibly reverting back to the resolver, or sourcing a new 0.5TR resolver in its' place) using your experiences as guidance. I'll update this thread later in the year when get the motor running (I presently have an encoder in place so may configure that temporarily). All the best, thanks again.

 

[hummbug]

Maybe those following here can help me. I have the Sevcon Gen4 Size 8 as well. Wondering how this powers up: pins on the 35 pin connector to hook up in my case:
Pin 1: 12v key input
pin 3: contactor out 1
pin 11: can H
pin 17: pot 1 wiper in
pin 18: pot 2 wiper in
pin 22: can L

I am just trying to see if it is alive and transmits CAN, not to spin a motor yet.

Since the b+ is interrupted by a contactor to start with, how is it getting any voltage at all to start up?
Does the controller not run off 12/24v low voltage, but runs off the HV?

 

[Martin1599]

I'm a bit rusty as my project has been on the shelf for a while......... but it looks as though you're on the right track. The unit is powered up via pin 1 (+12v) and is not powered when there is no supply to that pin, and all configuration settings are retained when it powers down; and as far as I am aware it does not run off the HV in any way (at the end of every day I disconnected all power both HV & LV just to be super safe, and it always had kept the saved settings/configuration.). Hope this helps but please feel free to post more questions- If I can help in any way I will, but may take some time to reply as very busy atm! . Atb, M

 

[chris01290]

I have got my project to the point the motor spins up great with the wheels in the air. and i can see the motor taking upto 30 Amps, probably more if i was brave enough, but when i put it down to the ground to test it seems like the motor just does not want to go and give's a 'DSP PF fault' pretty much as soon as the current gets over 4 amps.
Any idea's any one?

 

[Martin1599]

Looking through the fault code list I've found this:-

Maybe there's a conflict in one the settings?

 

[chris01290]

Looking through the fault code list I've found this:- View attachment 132252
Maybe there's a conflict in one the settings?

ok that problem is now solved basically it was down the Kp and Ki gains been set too high and the controller over shooting and it cutting out hitting the current limit.

 

[chris01290]

So now the next question. This might seem a stupid question or thing to do but it was the way i tuned the Kp and Ki so the motor did not cut out.
I have the controller set to max 195nm torque which is the rated for outlander Y61 motor. the max motor current at 320A.
If i block the front wheels and accelerate, It seemed to me the motor did not seem to have that much go from 0.
so at full throttle the battery current is around 15A (metered from the battery current sensor)
Data from the controller:
the AC motor current is 320A
AC motor Voltage is 12v
Torque 195nm

So my question does this seem correct?

I suppose even though you get max torque from 0, it could be the effect of not having a fly wheel and a clutch like a normal car would.