Encoder Voltage too low on Sevcon gen4size6

[thriftybuilder]

I have a Sevcon gen4size6 controlling an ME1616 motor. It is now working, after I found an issue with the Encoder voltage coming out of the Sevcon. I have DVT set to 5V sin/cos sensor, but the I measured the voltage between pins 26 and 15 at ~1V (spec sheet says these pins power the encoder). This obviously was not working (motor was making noise but not spinning). Voltage from pin 26 to B- (ground) was ~5.4V, so I simply connected encoder negative to B- instead of 15, and now everything is working.

So my questions are:
1. Has anyone else had encoder voltage issues between pins 26 and 15?
2. Why might pin 15 not be tied to ground? When I measured pin 15 to ground, it was ~4V, but it should be 0V.
3. Is there another relevant setting in DVT other than the obvious encoder voltage in object 0x4630?
4. Do you foresee any issues with my current wiring? Maybe noise, voltage spikes, or floating voltage.

DCF and wiring diagram attached. Disclaimer this DCF has various settings that could damage your equipment, so do not use it unless you know how to use DVT to update relevant settings, understand the risks, and you take full responsibility / liability.

 

[amberwolf]

Are you measuring pin 15 at the pin itself on the controller, or at the contact in the wiring harness?

If the contact is not touching the pin on the controller, or the wire is not crimped into the contact correctly, you could have a floating voltage on the wire/contact, even if the pin itself is correctly grounded.

Other than that, without knowing hte circuitry inside the controller itself for that pin, I don't know what could internally cause the problem (beyond poor solder, physically broken pin, or broken trace to the pin, etc)

 

[thriftybuilder]

Are you measuring pin 15 at the pin itself on the controller, or at the contact in the wiring harness?

I am measuring at the encoder connector, controller side. Since the controller needs power to turn on, it would be kind of tedious to only power up the controller and measure pins 26 / 15 directly, although you make a good point that it might be an issue with the harness. I did not think of this, thank you.

I will check the crimps and resistance through the harness.

My distributor suggested that the PWM circuit in the controller might be fried. They have seen encoder shorts cause damage in the Sevcons. I do not know the full history of this particular Sevcon unit, so this may be the case.

I might put a fuse and MOV on the encoder power to prevent encoder damage, since I now have atypical wiring and am slightly worried about voltage spikes damaging the encoder, or encoder shorts damaging the Sevcon.

Another question I have: during my debugging process, I swapped the the Sin/Cos based on bastens_com recommendation on a different thread. The system is working with these swapped. What does swapping sin/cos really do? Does it simply reverse the direction, or will swapping them back definitely not work? I'm tempted to swap it back and see how it works, but this is yet another variable that might waste a lot of time during my tuning.

 

[amberwolf]

All I can tell you about swapping Sin/Cos is that each is a sinewave signal from a pair of hall sensors that read a circularly-magnetized ring, offset 90 degrees from each other. So, AFAICT, swapping them just means the timing signal is 90 degrees or 270 degrees from where it would be normally.

How that affects actual operation, I couldn't say.

Regarding tediousness...many troubleshooting steps or sets of them are that...but often necessary to eliminate specific problems when other simpler tests didn't give a conclusive result, etc.

The thing about troubleshooting is knowing how something works and what specific things could go wrong with it well enough to know what *not* to bother testing, and knowing what simple tests to try first to eliminate all the easy stuff, before getting into the stuff you have to start taking things apart for, or other more complicated (and potentially damaging) tests have to be done.

Not knowing a specific system and setup means not being able to do those very accurately, and having to be much more general in the types of tests to suggest.