Gettys 360 Drive

We got run-aways on the three axis that we have sent spark to on our retrofit.

Machine has not seen spark in 19 yrs, so quite possible that there have been some caps or whatnot that have dried out and gone bad.

Would like to get a schematic if possible.

I have a full set of spare drives, but those haven't seen spark in about 10 yrs and are a days drive away at that.

I have loads of Reliance books on these machines, but we aint seen nothing Gettys related at all.


"Type" 11-1060-01
Machine build date 4/80


We did see some amount of inteligence at one point when the one axis changed direction once, but now only goes that direction now.

Another axis reacts diff if we short the input V to zero as compared to just unhooked. With shorted it slowly drops - seems very much like gravity, but as soon as E-stopped - it stops dead. I see no sign of a brake on that unit either. Just fine pitch screws, so we ass_u_me that it is actual/gradual movement of the drive/motor.

Other than that - it's Balls to the Wall for the O/T switch on all axis!
(Have not hooked up the B yet.)


FWIW - the resolvers have been replaced with encoders to the new PC control, but the tachs have never been unhooked. (other than for rigging purposses.)


--------------

Think Snow Eh!
Ox

I would find the analog ground and short all the velocity inputs to them first, then adjust the drift until the thing stops. Clearly the drives are out of tune to some degree, but until you have them all shorted to ground, you won't know just how much. Depending on the drive input design, it may not be capable of floating, so it'd best to always have it set. Another option which is pretty common, is to install 1meg ohm resistors between the analog ground and both of the inputs. This puts a slight bias towards ground, so they won't have any stray voltages on the input when floating.

Ox- I have manuals for the 360 drives- at least the material that Gettys gave to customers. drop me an e-mail and I'll dig up the info. I also think I have the Bendix servo setup procedure if that would help. I think Perry is pretty close about grounding the input to set the Gettys end if memory serves me. Its been about 7-8 years since I worked on these.
dmcgearsATyahoo.com
Dan

Isn't incorrect phasing of the encoders the classic reason for runaway? Even reversing those did not help?

Not sure about encoders, but unless Ox has changed things, these use resolvers. I know for a fact that reversing the tach connections will make the servo run away.
Dan

Dan from Oakland said:

Ox- I have manuals for the 360 drives- at least the material that Gettys gave to customers. drop me an e-mail and I'll dig up the info. I also think I have the Bendix servo setup procedure if that would help. I think Perry is pretty close about grounding the input to set the Gettys end if memory serves me. Its been about 7-8 years since I worked on these.
dmcgearsATyahoo.com
Dan

Click to expand...

Howdy! How about this, reviving a thread from 2009!

I have N350 manual but not N360, even tho we have at least 1 N360 still on the warehouse shelf. A friend asked me for a manual to help him with a problem and I am embarassed to not have one.

Might you still have a N360 manual you could send?

Mike- send me your e-mail- dmcgearsatyahoodotcom

Dan

Dan from Oakland said:

Not sure about encoders, but unless Ox has changed things, these use resolvers. I know for a fact that reversing the tach connections will make the servo run away.
Dan

Click to expand...

Yes, that's right, these sweet old controls have been using resolver signals for feedback. The runaway happened mostly when the resolvers haven't been exited in all or when the output wave was lost by electronics. Sometimes they used to have additional resolver on the way, with double winding and brushes, used for grid shift. These brushes have been nightmare, frequently causing loss of signal. This is as far as position loop is concerned. Less likely, but still possible, is loss of tachometer signal in velocity loop.

Stefan

Dan from Oakland said:

Mike- send me your e-mail- dmcgearotcom

Dan

Click to expand...

Thanks; email sent.

PROBE said:

These brushes have been nightmare, frequently causing loss of signal.

Click to expand...

In your dreams, maybe. Have had many machines with resolvers and never a problem with them. They are extremely reliable.

Everything else I can think of has broken, but not one resolver. In fact the H40 had three per axis, geared together so the machine could never lose position, and even that setup was reliably running on the originals twenty-five years later.

Whoop encoders any day of the week

At the last place I werked (>30 yrs ago) we had a Brown and Sharp Hydra-Tape.
I'm sure that it was from the 60's.

The machine would just stop at times.
Had tech come in and adjust, or whatever you doo to resolvers, and yet the issue pursisted.

We found the "fix" was that we kept a 5# rubber mallet on the corner of the table, and if you saw it stopped, you just went over and gave the table a smack, and off i'd go again.

This had the equiv of a 50 taper spindle, and a 3' x 5' table or bigger.
She was a big heifer.


----------------------

Think Snow Eh!
Ox

Ox said:

....
We found the "fix" was that we kept a 5# rubber mallet on the corner of the table, and if you saw it stopped, you just went over and gave the table a smack, and off i'd go again....Ox

Click to expand...

Hahahaha! Knowing WHERE to smack it was worth its weight in gold!

Friend and I developed an encoder to resolver converter a year ago; we use it to use encoders now to run into older controls and drives that required resolvers. No one else has such a thing.

OK, well .... where in the hierarchy does the Inductosyne to digi convertor fit?

Kind'a sounds like you just need to back-feed that?

(apparently not)


----------------

Think Snow Eh!
Ox

EmanuelGoldstein said:

In your dreams, maybe. Have had many machines with resolvers and never a problem with them. They are extremely reliable.

Everything else I can think of has broken, but not one resolver. In fact the H40 had three per axis, geared together so the machine could never lose position, and even that setup was reliably running on the originals twenty-five years later.

Whoop encoders any day of the week

Click to expand...

My dear Emanuel. Although maybe God is with you ("God with us" is the meaning of name Emanuel in Hebrew), but you doubtedly can have any idea about my dreams. As to the matter:

1. I would strongly recommend you, before being so sarcastic, to spend some time and have a look on this small paper https://www.moog.com/literature/MCG/synchrohbook.pdf. Maybe this will enrich your knowledge.
2. From reliability point of view brushed encoders have been nightmare.
3. I have to second you: From reliability point of view brushless resolvers outperform the encoders.

Stefan

Ox said:

OK, well .... where in the hierarchy does the Inductosyne to digi convertor fit?

Kind'a sounds like you just need to back-feed that?

(apparently not)


----------------

Think Snow Eh!
Ox

Click to expand...

Yes, wouild have been nice if it Farrand to encoder unit could be backfed! But alas, not in the cards. Besides, the Inductosyn is ua or milli amp level signals, not like resolvers. No one has ever needed a encoder or analog signal replacement for a resolver in the past I guess. We did for retrofitting the big Cincinnati (and others) aerospace gantrys and profilers. We used to supply motors with TWO resolvers inside - one for the drive and one for the CNC. We got tired of having to build special motor back endbells that provided the second resolver, and in addition, most of those big machines needed the position resolver to be physically geared to the motor shaft too! We geared them so 0.200" made 1 CNC resolver feedback rotation. A lot of the CNCs used yet today on these machines cannot do the math so needed the resolver to be exactly .200" per rotation for scaling. So my buddy and I decided to use an arduino engine and design electronics to accept the drive or CNCs resolver reference and modify it into the required syncronized sine & cosine outputs - and make it scaleable so we could eliminate the 2nd motor resolver AND the gearing. Our geared resolver emulator outputs any level of resolver signals required (5V, 20V, whatever). We use either an incremental encoder and/or a 0-10V position signal to generate the gearerd resolver emulated sin/cos. Both if we need it to be absolute over more than a single resolver revolution (like for X & X' dual axis slides with auto anti-skew software that requires keeping track of the position even if moved during power off. Lots of fun. So it can be dialed down to the low outputs to replace Farrand inductosyns too...

PROBE said:

1. I would strongly recommend you, before being so sarcastic, to spend some time and have a look on this small paper https://www.moog.com/literature/MCG/synchrohbook.pdf. Maybe this will enrich your knowledge.

Click to expand...

Thanks, but after keeping a whole bunch of resolver-driven machines running, I'll trust experience over anything Moog has to say. Moog, fer chrissake, of all places ...

mike_kilroy said:

... Both if we need it to be absolute over more than a single resolver revolution (like for X & X' dual axis slides with auto anti-skew software that requires keeping track of the position even if moved during power off. Lots of fun.

Click to expand...

Cool device ! Also cool that Cincinnati remembered their older systems ... the Accramatic machines had three resolvers geared together, one for inches, one for hundredths, and one for thousandths, so the machine never lost its position. Both the horizontal and verticals did that. Was a spiffy deal, and that control was more reliable than twenty years newer ones. A little dumb, but who needs interpolation anyhow

EmanuelGoldstein said:

Thanks, but after keeping a whole bunch of resolver-driven machines running, I'll trust experience over anything Moog has to say. Moog, fer chrissake, of all places ...

Click to expand...

Quite arrogant approach, but this is your choice.
Good luck in any case.

Stefan

PROBE said:

Quite arrogant approach, but this is your choice.

Click to expand...

If you think ignoring Moog is arrogant, then I guess you've never seen a Moog machine ! But I guess if you want an authority on unreliable, they'd be way up the list ...

PROBE said:

My dear Emanuel. Although maybe God is with you ("God with us" is the meaning of name Emanuel in Hebrew), but you doubtedly can have any idea about my dreams. As to the matter:

1. I would strongly recommend you, before being so sarcastic, to spend some time and have a look on this small paper https://www.moog.com/literature/MCG/synchrohbook.pdf. Maybe this will enrich your knowledge.
2. From reliability point of view brushed encoders have been nightmare.
3. I have to second you: From reliability point of view brushless resolvers outperform the encoders.

Stefan

Click to expand...

I never heard of brushed encoders before. Is that a thing?

Thanks for the link to the manual. Just read a bit of it.

Heheh.......the box of spare and used/ broken parts that came with one of my Lathes..,,that contained no less than 5 bad resolvers and about 4 new-in-box ones makes me LOL at your words here.

EmanuelGoldstein said:

In your dreams, maybe. Have had many machines with resolvers and never a problem with them. They are extremely reliable.

Click to expand...