what drive and power supply?maybe its not the motors
I'm getting more & more issues with our slightly aged router, and am keen to try another servo to see if that helps eliminate the spasmodic positioning error messages we frequently get in the Y axis.
Does anyone know of a Qld or eastern states supplier of 2nd hand servos?
(1/4" shaft, 1000 cpr, with encoder of course, what's currently on is a Pittman, 2.24" diam, about 7" long altogether.)
(The router used to be an Esab in 1990 modified to become a Tekcel in 2001, but with variations added to the theme)
(PS we're N-W of Toowoomba)
what drive and power supply?maybe its not the motors
It's about 30 vDC...
well the power supply is a toroidal 240:30 volt AC, plus whatever 30 vAC rectifies down to- maybe 48 vDC.
Drives are Tekcel printed circuit boards.
Perhaps I can prise the cover off the encoder & maybe blow a bit of dust out, or something.
Hi Stewey, it could be the encoder or ground noise rather than the servo itself, I use these from Digikey on my router :
You can select the CPR you want and they also come with several shaft mounting size option's, it may be cheaper than getting a new (ish) servo, and if it doesn't make any difference it would be handy for a spare. They have an index pulse but no differential driver, so you may have to add that yourself if you have long encoder cable runs.
Other than that I would go looking for ground loops in the controller and servo wiring.
Does the servo fault out when doing slower than rapid moves?
I eliminated a few other issues, and the router's running well most of the time.
I still get spasmodic Y axis position errors. I found out the encoder is a 500 cpr optical one- perhaps it just has dust in there, but I'm unsure how to take to cover off without risking damaging it irrepairably.
It's an HEDS 5500-A06 encoder by Argent or HP.
It tends to fault out when going fast, not slowly.
Would you explain what you mean by 'ground loops' please? (faulty earthing???)
I have replaced the wiring cables from the controller board to the servo board, and from the servo board to the encoder. No difference.
I have swapped encoder/servo boards, and I have swapped controller boards- no difference- the problem still is in the Y axis.
The servo itself seems fine.
I appreciate the help-thanks!
I'll look into another encoder via your link. I may just have to be more creative about the 'snapping of the cover down onto the base plate' in reverse- trying to work out how & where to release clips to see if I can de-dust it.
Hi Stewey, sounds like you have tried pretty methodically to find the problem, how hard is it to swap the Y axis servo with the X or Z servo ? Sounds a bit painful I know but this will narrow it down to the servo, and not the Y axis wiring. If they are the same it may be worth a shot.
By ground loops I mean two paths for earth or induced earth current to flow between two points, say the encoder cabling has its shield earthed when it leaves the drive, then you also earth it to the servo, which is also earthed from the motor cable shield or frame earth, this can generate noise and/or unwanted voltage on the encoder signals that may be enough to fault out the drive.
Don't get me wrong the servo could just be worn out and not giving enough power so it cannot keep up with its original top speed, slowing down the maximum rapid speed for that axis may stop the errors but I wouldn't call it a fix, more of a workaround
I had a source of servo's from another zone member that lived in Korea but unfortunately he has since moved, I managed to get a small stash of Electrocraft E543's, if you are really stuck I could loan you one, I would offer one to keep but I have only just enough for my remaining machines on the "To Do" list and like you have found it difficult to find a supplier that I don't have to sell one of my kids to buy from (as tempting as that is sometimes )
Here's the actual encoder that's on the X & Y servos. (The Z one is a whisker different)
From Farnell- $58ish
The lead from the encoder to the little encoder board cable-tied to the servo is a flat ribbon. The lead from that board, to the main Y axis servo controller board, is about 4 metres long, and is shielded. However, the shielding is not connected, but sealed & heatshrink-insulated at each end.
It would be slow, but possible to disconnect the two servos (X & Y) and see if that changed things. I'd guess it at the most a couple of hours' tedium to be neat & tidy. That's why the thought of maybe cleaning an encoder- or replacing one even at $50ish is tempting.
ON the other hand, now you've suggested it, maybe I will swap them over.
(There were three E544s or 588s on ebay last week in USA which went for about $110 for the lot, with 1000 cpr encoders on 1/4" shafts. A seller was clearing out new unsold stock)
Hang onto your E543s - you'll need them as a set, I'm sure, but thanks for the thought!
Can you put a pic of the encoder and little board ? (the link didn't work for me but my internet is being temperamental tonight) It sounds like it has a differential line driver connected to the encoder, this is a good thing as it helps with noise immunity.
I would be tempted to connect the shielding to ground at the driver end before trying anything else, leaving it (the shield) unconnected at the servo end. This may be enough to be causing your problems. It is probably the easiest thing to try as well, assuming you can re-terminate the cable at the drive and it is long enough.