Tree Journeyman / Dynapath Servo Tuning

I'm working on a really old Tree Journeyman 325. It has a Dynapath Delta 20 control. The issue it has is that the axes will not always be "in position" at the end of a rapid move and the control will just hang there forever waiting for the axes to be "in position".

The machine has a Servo Dynamics DC servo drive that's original to the machine. It uses Baldor DC servos with encoders. I think it may also have tachs on the motors.

I have gone through the servo tuning procedure provided by Tree, but I am not able to really reduce the servo lag. I tried to swap the servo drives with those from another machine, but I still have a lot of servo lag.

The only way I can get around the issue is to change the parameters and reduce the rapid speed. The Z axis is the worst offender, and I have reduced it all the way to 40 IPM. It's OK for the work this machine does, but I would like to find a permanent fix.

Any thoughts? I have not inspected the brushes, but I will. I suspect that the encoders are working OK because the machine does run fine at reduced rapids.

30 yr old caps on servo dynamic board can be problematic.
Check encoders they may need cleaning. I think baldor encoders of that era used a small light bulb in them and maybe they are too weak with age to put out a good enough amount of light at higher speeds for positive signal output.

The electronics being so old is a crap shoot.

Dirty corroded servo driver board also no good. Loose connectors or cables. Bad worn out brushes. Excessive friction or binding. Built up sludge on the box ways.

Also while your checking the brushes, look at the commutator for dirt or excessive wear. That can result in a weak servo motor.

What are the exact symptoms when it's waiting for "in position"?

Is the servo hunting for position, or overshooting?

Does it always fall short or always stop past position? By similar amounts?

Does it whine or squeal?

It might not even be the drives that are at fault. Old analogue servo systems have a DAC on each axis controller on the control that generates the analogue voltage to the drive. These can drift over time, becoming offset + or -, or low or high.

Even if the balance is perfect on the drives the DAC output might be sufficiently skewed that the position loop can't compensate. Some controls have trimmers on the output DACs, others rely entirely on the balance offset on the drive.

On the Dynapath control, there is a "*" next to the position display when the axis is in position. if you don't get the "*", the control just waits. The axes all move and appear to be in the right position, but the encoder count must not be exactly correct.

ewlsey said:

On the Dynapath control, there is a "*" next to the position display when the axis is in position. if you don't get the "*", the control just waits. The axes all move and appear to be in the right position, but the encoder count must not be exactly correct.

Click to expand...

So do the position readouts display the actual position, or the programmed position? If the readouts show the correct position, but it's not "in position" then either they are not showing the actual position, or there is something wrong with way the control calculates "in position". Another possibility might be that the servo is oscillating faster than the position readouts can display. Just guessing here, not familiar with Dynapath. I can't imagine that it would be wait for it to settle down to the last encoder count to say that it's in position. Every control I've ever seen has a "window" for in position.

Have you tried a physical test? Zero an indicator against a block when it is "in position", then move away and rapid back.

Post more on youtube. Sick of the Hobbyist stuff

The control is set for .001 tolerance for in-position on all axes.

m98custom1212 said:

Post more on youtube. Sick of the Hobbyist stuff

Click to expand...

Yeah I need to do that. My work has really picked up in the last 3 weeks. It should level off once the weather finally turns cold. Then I can get back to videos.

What is your youtube name... searched you user name here and came up with nothing.

ewlsey said:

I'm working on a really old Tree Journeyman 325. It has a Dynapath Delta 20 control. The issue it has is that the axes will not always be "in position" at the end of a rapid move and the control will just hang there forever waiting for the axes to be "in position".

I have gone through the servo tuning procedure provided by Tree, but I am not able to really reduce the servo lag. I tried to swap the servo drives with those from another machine, but I still have a lot of servo lag.

Click to expand...

On the 5 (very similar system) you can adjust the counts of error allowed on the in-position indicator. So you could just open this way up, but that's not going to solve the problem.

You say you can't reduce the following error, mmm, as I remember following error wasn't static, at lower feeds it was less and at high feedrates it would get bigger, but as the drive gets close to the end of the move, it drops off. Then at rest "in position" it's supposed to be just a few counts.

Are you watching this on the maintenance screen ? That will tell you how many counts out of position you are, both at rest and in motion. In maintenance mode I think you have to drive the axis with a battery, cuz you can't enter any commands because you are in the wrong mode, but that may be faulty brain cells.

When the thing is just sitting there unmoving but not in position, do you have an error signal to the drive, telling it to move ? Seems to me you can adjust the allowable following error, too - but that may be on the drive. If you allow too much following error on the drive but not enough on the control, then it's never telling the axis to go, so it will never get into "in position" distance.

If you aren't getting in-position, then either the axis card in the control is not doing its job (there's test points to check for output voltage) or the drive is not doing its job. Either one may be bad but I'd tend to think axis card ...

Dirty encoders sounds reasonable but pretty sure if it's missing counts, it would just go to the wrong place, not refuse to get where it belongs. That is, it thinks it is somewhere it isn't because it dropped some counts, but it should still finish up the move properly even though it's in the wrong place.

On the 5 you can swap axis cards around to give a better idea where the problem is. Don't know about the 20 but if you can, that might give some clues. But check for output first. If it's not in-position you should get an error signal to the drive.

edit: hang on a sec ... two more brain cells showed up. Did you check the following error in both directions ? It's supposed to be the same. And it's supposed to be the same across all the axes, otherwise you don't get round circles. I think there is a balance pot on the axis card to get + and - even. If they aren't even, then you can get the situation you describe; somewhere in the back grey regions the word "balance pot" is floundering around.

I'm sure there's other causes too but this is an easy one to check.

SeaMoss said:

You say you can't reduce the following error, mmm, as I remember following error wasn't static, at lower feeds it was less and at high feedrates it would get bigger, but as the drive gets close to the end of the move, it drops off. Then at rest "in position" it's supposed to be just a few counts.

Click to expand...

Does the Dynapath have a display for following error? I have only ever had one control that did (old Gildemeister EPL). I have never used a Dynapath, am just going by my experience of other old analogue servo systems.

From a purely tuning perspective (physical faults and mechanical issues ruled out) excessive following error is usually a symptom if insufficient gain. Excessive following error in one direction can be attributed to balance offset.

A good indicator of balance offset is the ballscrew twitching when it's holding position, as the servo tries to drift and the position loop compensates.

Ewlsey, I'd try just turning the gain up and see where it gets you. The servo will start to whine if you go too far.

Yes, the control has a servo tuning page that shows lag. I have a factory procedure to tune the servo drives, but I am not able to achieve the lag described in the manual.

Another service guy suggested to me that the tuning pots may be bad. That makes sense to me since adjusting them doesn't seem to do much. However, I swapped drives from a working machine and didn't notice any improvement.

ewlsey said:

Another service guy suggested to me that the tuning pots may be bad. That makes sense to me since adjusting them doesn't seem to do much. However, I swapped drives from a working machine and didn't notice any improvement.

Click to expand...

Check out the balance thing for sure. Some of this is coming back, it really has to be the same in both directions. There was also a guy in Forest Lake CA with a service that specialized in Bendix, he knew these things inside-out. George Something. That'll probably come back some time later, too.

gregor - most of the controls I've had will display following error and a lot of other diagnostics. Maybe newer stuff figuers you don't need to know those things ? Can you watch the ladder on a Fanuc ?

SeaMoss said:

gregor - most of the controls I've had will display following error and a lot of other diagnostics. Maybe newer stuff figuers you don't need to know those things ? Can you watch the ladder on a Fanuc ?

Click to expand...

I spoke out of turn about that. Thinking about it more, the only controls I have used that didn't show following error anywhere in the diagnostics were old Hurcos - those being the most prominent in my mind when thinking about this stuff...

My last DC servo machine was retired about 4 years ago and I haven't had to deal with any of this since, thankfully.

Ok. Generally there is two types of following error. Running and standing. I have never used the baldor system but have 100s of dynapath w/gettys drives. On these are two pots. Offset for standing error and sensitivity for running error. Anyway. Based on the age I would think you have dirty wipers on your adjustment pots. Try carefully the turn the pots back and forth to help clean the wipers and the surface. Then see if your lag adjusts. Good luck.


Sent from my iPad using Tapatalk

ewsley - did you get it balanced ? Betcha fifty cents that will solve your problem.

No, I have not messed with it. I'm going to look for a spare drive board and pull the drives one at a time to see if I can improve things like the adjustment pots.

ewlsey said:

No, I have not messed with it. I'm going to look for a spare drive board and pull the drives one at a time to see if I can improve things like the adjustment pots.

Click to expand...

Drive or axis cards ? I'll up the bet to seventy-five cents, if you balance + and - the in-position problem should go away ...

SeaMoss said:

Drive or axis cards ? I'll up the bet to seventy-five cents, if you balance + and - the in-position problem should go away ...

Click to expand...

I had the same problem of not getting lag to adjust on a delta 20 I called Dynapath and they told me that the servo drive card was bad. I called Parts and Smarts as I had just got the control back from them not working. Parts and Smarts sent me a different card claiming that wouldn't fix it. The different card allowed me to get it adjusted

Are you talking about the servo drive card in the control chassis?