how to program a sevcon gen4

General Discussion about large electric scooters and motorcycles and other things with no pedals.
c70r   10 W

10 W
Posts: 94
Joined: Jan 22 2019 12:26pm

Re: how to program a sevcon gen4

Post by c70r » Jan 13 2021 10:19am

Anybody have a 75-7 SPM DCF file they can share?

Seems as secret as the colonels spice recipe lol.

rider119   1 mW

1 mW
Posts: 16
Joined: Mar 17 2020 11:29am

Re: how to program a sevcon gen4

Post by rider119 » Jan 13 2021 3:14pm

Anyone have a file for a QS 120 or 138 motor they would like to share that would at least get me spinning? Would be much appreciated! Been busy with a remodel on the house and haven't had time to mess with this much lately.

frodus   10 kW

10 kW
Posts: 820
Joined: Mar 03 2008 2:24pm
Location: Portland, OR

Re: how to program a sevcon gen4

Post by frodus » Jan 14 2021 1:30am

c70r wrote:
Jan 13 2021 10:19am
Anybody have a 75-7 SPM DCF file they can share?

Seems as secret as the colonels spice recipe lol.
Size 4 or size 6?
Travis

c70r   10 W

10 W
Posts: 94
Joined: Jan 22 2019 12:26pm

Re: how to program a sevcon gen4

Post by c70r » Jan 14 2021 11:48am

frodus wrote:
Jan 14 2021 1:30am
c70r wrote:
Jan 13 2021 10:19am
Anybody have a 75-7 SPM DCF file they can share?

Seems as secret as the colonels spice recipe lol.
Size 4 or size 6?
Size 6 if you have that please.

Thank you!

evc_motorcycle   1 mW

1 mW
Posts: 15
Joined: Mar 15 2021 8:53pm

Re: how to program a sevcon gen4

Post by evc_motorcycle » Apr 01 2021 6:28pm

Hello everyone, I am a member of my university's electric vehicle club and we are using a Sevcon Gen4 with an ME1302 in an electric motorcycle project. Over the past several weeks I have been familiarizing myself with the Sevcon DVT environment and editing parameters for the driver. There have been many faults (as expected after reading many forum posts), most of which have been cleared. User Pajda on the Forum was generous enough to send a DCF with the ME1302 parameters that I have been working on editing to match the rest of our setup. I was making progress at the beginning, but I have unfortunately hit a roadblock and have been stuck for the past few weeks.

The ME1302 uses a sin/cos encoder and I am having trouble getting the controller to read the Ud and Uq values mentioned in a post on graphing the voltage vectors to commission a new encoder. When I graphed the trace as described in the tutorial I received complete random points instead of a linear formation I was expecting. Subsequent traces are now filled with nothing but zeros and the controller does not seem to read anything in the Ud or Uq in the DVT. Attempting a graph results in a "divide by zero" error.

I cannot clear the "io init" fault on the gen4 to save my life :lol: . It has been persistent since the start of the process. The fault codes spreadsheet says to "check RDPOs" or something along those lines. I have 4 RDPOs for the throttle, FS1, FWD, and RVS. I have no idea what is causing the error. The Raw digital and analog inputs tab of DVT reads changes when I flip the switches/move the throttle.

The only way I can get the motor to spin (well, try to spin) is with the "testbox" feature in the DVT. When I do this, the motor simply twitches and vibrates on the table. I have tried rearranging the order of the M1, M2, M3 wires and the sin/cos wires from the encoder with no luck.

I understand that using controllers like this is a difficult process of trial and error. I am 100% willing to put in the time to get this working, I just feel that some direction would be helpful as the motor has been in the same state for about two weeks now. If anyone has any advice on any of these three problems I would greatly appreciate it!

Attached below is a pictorial representation of my wiring. I am happy to share any other necessary information about my setup if needed!
Attachments
IMG_20210401_191031491.jpg
Wiring
IMG_20210401_191031491.jpg (564.2 KiB) Viewed 409 times

User avatar
darseygodwin   100 mW

100 mW
Posts: 42
Joined: Jul 31 2016 5:51pm

Re: how to program a sevcon gen4

Post by darseygodwin » Apr 01 2021 6:58pm

Have you already gone through the sevcon app note? https://www.e-kart.fr/images/stories/te ... e-pmac.pdf If you can't use the link then google for "Sevcon app note pmac"

Basically, if you were trying to figure out the encoder angle to get started (including swapping out the phase order) then this is your best strategy from here forward. Basically, use a scope to log the physical angle of each phase, and compare that with the sin and cos signals. Together with that information alone, you can determine the rotation angle, which wire is sin and which is cos, which order your M1, M2 and M3 phases should be, and the relative angle of your encoder to the M1 phase.

Check back in with your progress! Good luck.

Darsey

evc_motorcycle   1 mW

1 mW
Posts: 15
Joined: Mar 15 2021 8:53pm

Re: how to program a sevcon gen4

Post by evc_motorcycle » Apr 01 2021 9:06pm

I have not tried that process yet. We've got a scope in the lab I can use so I'll be sure to try it out next time I'm in there!

evc_motorcycle   1 mW

1 mW
Posts: 15
Joined: Mar 15 2021 8:53pm

Re: how to program a sevcon gen4

Post by evc_motorcycle » Apr 03 2021 4:40pm

Hello everyone, I am currently in the lab working on the process of determining the phase order of an ME1302 PMAC motor. The manual seems to be missing an equation and I am having trouble finding it other places online. Does anyone know that the formula is? Or is there an easier way to determine the phase order of a PMAC motor?
Attachments
Screenshot 2021-04-03 174005.png
Screenshot 2021-04-03 174005.png (308.35 KiB) Viewed 377 times

evc_motorcycle   1 mW

1 mW
Posts: 15
Joined: Mar 15 2021 8:53pm

Re: how to program a sevcon gen4

Post by evc_motorcycle » Apr 03 2021 6:02pm

After a bit of time and research I was able to derive the equation needed for the offset, but my original problem still persists. I have verified using an oscilloscope that the sin/cos encoder is working. I have determined the phase order of the motor and made all of the (hopefully) correct connections. However, the DVT still shows nothing on the Ud and Uq axis when I rotate the motor with a hand drill. I am still greeted with an io_init error every time I power up the driver and the motor still just twitches when I attempt to spin it with the testbox in DVT. Is there a parameter I could have misconfigured that could be causing this?

User avatar
darseygodwin   100 mW

100 mW
Posts: 42
Joined: Jul 31 2016 5:51pm

Re: how to program a sevcon gen4

Post by darseygodwin » Apr 03 2021 10:02pm

Looks like you're well on your way even if you're not getting the right values yet. Three possible suggestions so far, sorry if they seem to rudimentary but we should check these boxes. 1. Are you spinning fast enough? A drill should do it. 2. Check the forward reverse directions. Simply flip your "forward" switch to be a reverse switch and take your trace again. 3. Does the sin cos encoder populate your max/min voltage values when spinning the motor? What are they?

Sent from my SM-N770F using Tapatalk


kiwifiat   100 W

100 W
Posts: 266
Joined: Mar 16 2013 5:49pm
Location: Vancouver, Canada

Re: how to program a sevcon gen4

Post by kiwifiat » Apr 04 2021 1:31am

evc_motorcycle wrote:
Apr 03 2021 4:40pm
Hello everyone, I am currently in the lab working on the process of determining the phase order of an ME1302 PMAC motor. The manual seems to be missing an equation and I am having trouble finding it other places online. Does anyone know that the formula is? Or is there an easier way to determine the phase order of a PMAC motor?
This is what you are missing:
Attachments
Sevcon encoder offset calc.PNG
Sevcon encoder offset calc.PNG (8.86 KiB) Viewed 360 times
“You never know if quotes on the internet are genuine.” - Abraham Lincoln

evc_motorcycle   1 mW

1 mW
Posts: 15
Joined: Mar 15 2021 8:53pm

Re: how to program a sevcon gen4

Post by evc_motorcycle » Apr 04 2021 12:21pm

Hi everyone, thanks for the responses!
1. we are using a hand drill to spin the motor, so I think out rotation is good.
2. when I take the trace I have the forward switch (we are using a three position switch to switch between forward, reverse, and
neither) When I take the trace I have the switch in the middle position, so neither the forward or the reverse pin is connected
to ground through the switch. Is this incorrect?
3. When I run the "commission new encoder" script in the DVT, there are max/min voltages that populate the the encoder tab in
the encoder section on the parameter tree. They are as follows:
-sin trough: 1.59853479853 V
-sin peak: 4.46153846154 V
-cos trough: 1.61758241758 V
-sin trough: 4.47765567766 V

Out of about 10 traces taken, there have been two that actually record values. One appears to be random, so I won't include it here. However, one a seemed to make a circle shape on the plot and I've attached the picture. Both times that these traces were logged the Ud and Uq values were changing in the DVT window when spinning the motor. Nothing was different (to my knowledge) between these runs and all the others.
Attachments
IMG_20210403_191924284.jpg
IMG_20210403_191924284.jpg (2.82 MiB) Viewed 344 times

User avatar
darseygodwin   100 mW

100 mW
Posts: 42
Joined: Jul 31 2016 5:51pm

Re: how to program a sevcon gen4

Post by darseygodwin » Apr 04 2021 1:07pm

Interesting. At this point someone probably has a perfect answer but I don't know what's truly happening. I do feel like the trace was supposed to be taken while in either forward or reverse, but I could be wrong. Easy to try them though. Secondly, that ring seems like it's correctly taking readings but there's a mismatch between the encoder and the phases (making it record at new angles). Again, this part is beyond me so I'm just speculating but it's like the actual motors number of pole pairs is different that what's recorded in DVT. But even if that were the case I'd expect chunks of readings at different points in the circle, not a nice even ring like that.

I guess just check those couple things next (engage the switch in the direction you're spinning, and check the # pole pairs) The sin cos voltages are great, you're getting plenty of signal there. Now I'm just interested in knowing what causes that!

Sent from my SM-N770F using Tapatalk


Pajda   1 kW

1 kW
Posts: 424
Joined: Jan 24 2016 3:45am
Location: Czech Republic

Re: how to program a sevcon gen4

Post by Pajda » Apr 04 2021 2:53pm

Hi, your approach described in points 1-3 seems to be good. When capturing the data try also his procedure:

1) spin the motor to the full drill power (ca 1000 rpm is enough)
2) start recording data
3) slowly decrease the speed to the full stop
4) stop recording data

.. the idea is not to record constant speed.

evc_motorcycle   1 mW

1 mW
Posts: 15
Joined: Mar 15 2021 8:53pm

Re: how to program a sevcon gen4

Post by evc_motorcycle » Apr 04 2021 2:58pm

Thank you all for the feedback! I will try those different procedures next time I'm in the lab and see if I can get better results :D

4ndr344   1 µW

1 µW
Posts: 3
Joined: Apr 04 2021 6:42pm

Re: how to program a sevcon gen4

Post by 4ndr344 » Apr 05 2021 4:44am

Hello, i was planning to do an electric supermoto conversion using the Sevcon gen4 size 6 and an Montenergy ME1507 (or ME1616).

My problem arrived when I was going to program the Sevcon, I have the ixxat usb to can 1.6 converter, but I need the Sevcon dvt(sevcon dvt v1.03c). I really don't want to buy another ixxat converter and the new software License. It would also be helpful if anyone could send the DCF for the ME1507 or the ME1616 (I don't know if The DCF for these motors are The same)

Thanks André

Sent fra min SM-G975F via Tapatalk


Pajda   1 kW

1 kW
Posts: 424
Joined: Jan 24 2016 3:45am
Location: Czech Republic

Re: how to program a sevcon gen4

Post by Pajda » Apr 05 2021 8:32am

Hi André,

I think you mixed several problems together :)

1) IXXAT USB to CAN Interface is not connected with the Sevcon DVT software licences. You should get your DVT Customer licence for free from your local Sevcon representative. (I think that some eshops offer IXXAT Interface together with licence, but this is only their policy) The DVTC license is issued for only 2 years, but i think that for older versions of DVTC worked to change RTC date in BIOS to make it operational "forever", but you cannot reinstall DVT.

2) IXXAT USB to CAN Interface up to 1.6 version should be compatible with all DVT software versions, the problem is only with the newest IXXAT USB to CAN V2 version, which is supported only by the newer DVT software release (but I am not 100% sure about that)

3) ME1507 and ME1616 are electricaly the same motors, they differ only in cooling system. Also ME1507 seems to have the same stator as older Zero 75-7 motor with SPM design rotor. It differs in rotor design, which now utilized the IPM design (as all Zero motors from at least MY2019) but it still have the same amount of 5 pole pairs. I think that .dcf file from Zero 75-7 should work for ME1507, at least for the beginning of motor tuninig.

4ndr344   1 µW

1 µW
Posts: 3
Joined: Apr 04 2021 6:42pm

Re: how to program a sevcon gen4

Post by 4ndr344 » Apr 05 2021 8:43am

Pajda wrote:Hi André,

I think you mixed several problems together :)

1) IXXAT USB to CAN Interface is not connected with the Sevcon DVT software licences. You should get your DVT Customer licence for free from your local Sevcon representative. (I think that some eshops offer IXXAT Interface together with licence, but this is only their policy) The DVTC license is issued for only 2 years, but i think that for older versions of DVTC worked to change RTC date in BIOS to make it operational "forever", but you cannot reinstall DVT.

2) IXXAT USB to CAN Interface up to 1.6 version should be compatible with all DVT software versions, the problem is only with the newest IXXAT USB to CAN V2 version, which is supported only by the newer DVT software release (but I am not 100% sure about that)

3) ME1507 and ME1616 are electricaly the same motors, they differ only in cooling system. Also ME1507 seems to have the same stator as older Zero 75-7 motor with SPM design rotor. It differs in rotor design, which now utilized the IPM design (as all Zero motors from at least MY2019) but it still have the same amount of 5 pole pairs. I think that .dcf file from Zero 75-7 should work for ME1507, at least for the beginning of motor tuninig.
Thanks for the info, i'm looking for an older version og the dvt software. They didn't have licencing in the versions 1.03 and below

Sent fra min SM-G975F via Tapatalk


john61ct   100 GW

100 GW
Posts: 7832
Joined: Dec 18 2018 2:06pm

Re: how to program a sevcon gen4

Post by john61ct » Apr 05 2021 9:04am

In my experience asking dealers, they only supply the software and license to paying customers, either buying the hardware or retaining them on an annual service contract.

My impression was the goal is to reduce the value of secondhand units in circulation.

I don't understand why there aren't freelance Sevcon techs offering their services remotely worldwide.

Pajda   1 kW

1 kW
Posts: 424
Joined: Jan 24 2016 3:45am
Location: Czech Republic

Re: how to program a sevcon gen4

Post by Pajda » Apr 05 2021 1:57pm

john61ct wrote:
Apr 05 2021 9:04am
In my experience asking dealers, they only supply the software and license to paying customers, either buying the hardware or retaining them on an annual service contract.

My impression was the goal is to reduce the value of secondhand units in circulation.

I don't understand why there aren't freelance Sevcon techs offering their services remotely worldwide.
From my view the small customer support from Sevcon decreases to almost zero with the rise of this company. But this is not the problem of Sevcon alone but many other companies. They are existing Sevcon tech ops around the world, but they are very expensive for end customers as the knowledge about electric tractiom motors, FOC and other necessary control algorithms is highly specialized. If you ask your local Sevcon representative for a quote to pair your motor with Sevcon controller you will most probably get one.

User avatar
everythingisawave   100 mW

100 mW
Posts: 49
Joined: Sep 13 2016 8:01pm

Re: how to program a sevcon gen4

Post by everythingisawave » Apr 05 2021 3:25pm

evc_motorcycle wrote:
Apr 04 2021 12:21pm
Hi everyone, thanks for the responses!
1. we are using a hand drill to spin the motor, so I think out rotation is good.
2. when I take the trace I have the forward switch (we are using a three position switch to switch between forward, reverse, and
neither) When I take the trace I have the switch in the middle position, so neither the forward or the reverse pin is connected
to ground through the switch. Is this incorrect?
3. When I run the "commission new encoder" script in the DVT, there are max/min voltages that populate the the encoder tab in
the encoder section on the parameter tree. They are as follows:
-sin trough: 1.59853479853 V
-sin peak: 4.46153846154 V
-cos trough: 1.61758241758 V
-sin trough: 4.47765567766 V

Out of about 10 traces taken, there have been two that actually record values. One appears to be random, so I won't include it here. However, one a seemed to make a circle shape on the plot and I've attached the picture. Both times that these traces were logged the Ud and Uq values were changing in the DVT window when spinning the motor. Nothing was different (to my knowledge) between these runs and all the others.

Hi, joining this conversation late. If my memory serves me correctly, when doing this procedure you need to have the inverter actually on and gating, but zero current command. Point is that the controller outputs voltage to counteract the back emf of the spinning motor such that there is zero current. This can only be achieved when the true value of voltage output is in line with the back emf--Vd=0 and Vq=back emf, so if the controller is outputting an angled value then this will show you the angle mismatch. If the motor is capable of some rotation you can spin it up and then let off the throttle, and measure the Vd Vq as it is coasting.

***As always, be careful as this is rotating machinery, and consider that a motor sometimes can get out of control and runaway to max speed etc.***

evc_motorcycle   1 mW

1 mW
Posts: 15
Joined: Mar 15 2021 8:53pm

Re: how to program a sevcon gen4

Post by evc_motorcycle » Apr 05 2021 4:12pm

Hi everythingisawave- what does it mean to have the inverter on and gating? Unfortunately, the motor cannot spin up to measure the values while coasting so we have been using a hand drill to accomplish this. Thanks for the concern as well- we have the motor secured to the table we are resting and someone is always on the keyswitch when attempting to spin up. Unfortunately we're still waiting for the moment that we actually get some rotation :lol:

User avatar
darseygodwin   100 mW

100 mW
Posts: 42
Joined: Jul 31 2016 5:51pm

Re: how to program a sevcon gen4

Post by darseygodwin » Apr 05 2021 5:18pm

Mind inserting a screenshot of your motor parameters window? Got 4 pole pairs in there and not 5, 6 or 8? I'm guessing now your "random" image included accel or decel speeds, and your "ring" trace was at a relatively constant speed, but both are regularly rotating in perceived angle for some reason.

Sent from my SM-N770F using Tapatalk


User avatar
everythingisawave   100 mW

100 mW
Posts: 49
Joined: Sep 13 2016 8:01pm

Re: how to program a sevcon gen4

Post by everythingisawave » Apr 06 2021 1:28pm

Would be useful to see the encoder settings page, and if there are any errors.

You might also need to do that calibration step when the inverter is in forward mode. However if it is not rotating at all I would advise to check your connections first. Also, regardless of alignment, you should read stable speeds>> when the motor is not moving, DVT should be able to show consistent 0rpm, then try rotating forward and backward, see if the speed direction is stable and if you get sensible values for rpm.

evc_motorcycle   1 mW

1 mW
Posts: 15
Joined: Mar 15 2021 8:53pm

Re: how to program a sevcon gen4

Post by evc_motorcycle » Apr 07 2021 9:17pm

Hello everyone, thank you so much for the suggestions and feedback! Sorry about the few day break- this week has been kind of crazy with midterms and other assignments. Updates since the last post are as follows:
- I reflashed the gen4 with new firmware SN0058.28 (turns out the one it was running when we got it was not for PMAC)
- a new DCF file has been loaded and configured with the data for our setup on the new firmware (battery voltage, contactor, I/O, etc.)

Unfortunately I have not been able to try out the new suggestions for getting the encoder to line up because I now have a precharge fault on the controller that I cannot get rid of. I'm running a nominal 51.1V pack with the cutoffs set at the appropriate values for Lithium cells. The parameters are the same as they were in my previous DCF but the capacitors are only charging up to about 45V max which isn't the 85% needed for the contactor to close. Does anyone have any insight on to what could be causing this issue? I feel that I can't continue with motor testing until I get this bug figured out :/
The positive battery terminal is connected to pin 1 on the gen4 through a keyswitch and a fuse (unchanged since the beginning) and the battery is charged to ~56V so there should be no issues there. I have gone through the entire "tree" section in DVT and changed every value relating to voltages and cutoffs I could find as well.

Post Reply