motor is accelerating slightly over the course of a few hours on its own

[gadfly56]

See post #18. The real moral of your story: Engage an expert to solve a technical problem. He is more likely to pinpoint the problem immediately.

It is a capital mistake to theorize before one has all the facts. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. - Sherlock Holmes

 

[Besoeker]

See post #18. The real moral of your story: Engage an expert to solve a technical problem. He is more likely to pinpoint the problem immediately.

Don't you think that the OP was trying to tap into the expertise here?

 

[Jraef]

A wee tale for you....

We were commissioning some VSDs in a cement works. One was on an inclined conveyer, quite a common piece of kit in such plants. The conveyer was visibly varying in speed. Of course the drive and motor were immediatly blamed. Nothing unusual in that. So we checked the VFD frequency and motor speed. Both were as steady as a limestone rock............... Those were the bits we supplied and were responsible for. Job's a good 'un as they would say there - English Midlands

But there was no denying that the conveyer speed was varying. It turned out that there was a fluid coupling between the gearbox output and the conveyor drive shaft. Nobody seemed to know about and hidden inside machine guarding. Not a drive or motor problem.

The moral?
Don't jump to conclusions - sometimes they'll jump back at you.

Along that same line, albeit the opposite direction.

I once did a project retrofitting a VFD and motor to replace a diesel engine driving a WWII era portable rock crushing machine (used for making runways in the Pacific). Everything worked great at first, but the longer the machine operated, the slower it ran until eventually it would just stop. Because I provided the VFD and integration, I was called to solve it. The thing was, the drive and motor were NOT slowing down, they were perfectly steady, yet I witnessed the machine steadily getting slower and slower. Everyone kept blaming my drive or my control system, but I was looking right at it, dead steady. I even put a hand held tach on the shaft to prove the motor was at the correct speed. Turned out of course that it was a belt drive system and the belts were slipping, which got progressively worse as they heated up. They were actually smoking, but the machine made so much dust that nobody could tell. I smelled the burning rubber, nobody else did but once I pointed them to it, it was very obvious. Turned out I had grossly over estimated the HP conversion from engine to electric even though my math was correct, because an electric motor develops torque instantly, an engine has to build it slowly. The belts that had worked fine for 50+ years couldn't handle that.

The point is, it's NOT automatically a control loop problem. By the way, I've experienced the fluid coupling issue too. I also had that on a magnetic particle clutch where voltage drop from the main motor drive was allowing the clutch to slip for years, then when I put in the VFD, I also changed the transformer and fixed the voltage drop, so the particle clutch fully engaged and the machine ran faster.

 

[Besoeker]

Along that same line, albeit the opposite direction.

I once did a project retrofitting a VFD and motor to replace a diesel engine driving a WWII era portable rock crushing machine (used for making runways in the Pacific). Everything worked great at first, but the longer the machine operated, the slower it ran until eventually it would just stop. Because I provided the VFD and integration, I was called to solve it. The thing was, the drive and motor were NOT slowing down, they were perfectly steady, yet I witnessed the machine steadily getting slower and slower. Everyone kept blaming my drive or my control system, but I was looking right at it, dead steady. I even put a hand held tach on the shaft to prove the motor was at the correct speed. Turned out of course that it was a belt drive system and the belts were slipping, which got progressively worse as they heated up. They were actually smoking, but the machine made so much dust that nobody could tell. I smelled the burning rubber, nobody else did but once I pointed them to it, it was very obvious. Turned out I had grossly over estimated the HP conversion from engine to electric even though my math was correct, because an electric motor develops torque instantly, an engine has to build it slowly. The belts that had worked fine for 50+ years couldn't handle that.

The point is, it's NOT automatically a control loop problem. By the way, I've experienced the fluid coupling issue too. I also had that on a magnetic particle clutch where voltage drop from the main motor drive was allowing the clutch to slip for years, then when I put in the VFD, I also changed the transformer and fixed the voltage drop, so the particle clutch fully engaged and the machine ran faster.

Yep. The electrics are usually the first to be blamed!

I had a similar issue with belts on winch drives. Quite an interesting application.
A little bit of background. These winches were to be operated in pairs on a tin mining dredge. They worked by pulling the dredge back and forth across what was called a paddock. The pairs of winches were mounted on opposite sides of the vessel. To travel to the starboard the winch on that side winched in and the port side one winched out or rendered. They had our 4-quadrant DC drives on them.

The engineering company who built the winches wanted an endurance test before they were shipped overseas. Well, not everyone has a paddock or a dredger in their back yard so a test rig with two of the winches on it was built. They worked by one winching in the rope and the other rendering until the rope got to its end then the roles were reversed and so on. This was continually repeated 24/7 over a period of two weeks.

One Saturday morning, I got a call. One of the drives was making a thumping noise. So, of course, I went to look. Sure enough, there was a distinct thumping noise every now and again. It sounded like one of the DC drives misfiring. I checked and it wasn't. But obviously something was causing it.

The transmission between the motor and the winch was a Morse belt. This, as you probably know, is an internally toothed belt. Each of the shafts had a cog with the same tooth pitch as the Morse belt.
One of the belts simply wasn't tight enough and intermittently jumped a cog and that's what caused the thump.

Interestingly or otherwise, the guy who called me out was the winch designer, a mechanical engineer................

 

[junkhound]

In a lighter 'vain':

The wonderful internet has allowed a jump to the logical conclusions:
given the OP's (who has not posted since?) location is in (small town) MA, which has the highest per capita number of GOLF courses of any town in the USA and ALSO is the home of PUBLIC SCHOOL philosophy.

Thus, the obvious cause of the speedup is that the local public school educated electricians cannot count, simply are mis-stating the speed -

OR, there are any number of stray golf balls jamming the conveyor belts and as they fall out the speed increases;

Either is MORE plausible than a control loop problem IF the system is open loop instead of closed, which we do not know !

OR, even more of a potential root cause,

Maybe the meter is being held upside down by a PS educated tech and an L1 error code seen at startup, or the speed is de-creasing to 12 fps? ? or a a 5d3 error readout on a VFD ?

The OP did try to post a time domain pic, would have been interesting to C if the rate of increase was exponential or linear or ?

 

[Besoeker]

In a lighter 'vain':

The wonderful internet has allowed a jump to the logical conclusions:
given the OP's (who has not posted since?) location is in (small town) MA, which has the highest per capita number of GOLF courses of any town in the USA and ALSO is the home of PUBLIC SCHOOL philosophy.

Thus, the obvious cause of the speedup is that the local public school educated electricians cannot count, simply are mis-stating the speed -

OR, there are any number of stray golf balls jamming the conveyor belts and as they fall out the speed increases;

Either is MORE plausible than a control loop problem IF the system is open loop instead of closed, which we do not know !

OR, even more of a potential root cause,

Maybe the meter is being held upside down by a PS educated tech and an L1 error code seen at startup, or the speed is de-creasing to 12 fps? ? or a a 5d3 error readout on a VFD ?

The OP did try to post a time domain pic, would have been interesting to C if the rate of increase was exponential or linear or ?

You have a way with words..........