Ok all you old DC drive guys... I need help on an "S21" regulator 1

[jraef]

Today's challenge for the graybeards:

I must come up with a way to interface a new AC vector drive system (for a hoist) to an existing control cab for a GE DC M-G set drive system that was 50+ years old. The customer is insisting on retaining the existing operator control system that he says is using something called an "S21 regulator" on the Master Switch for speed control. It apparently takes a 120VAC source and splits it with a special transformer that has a moving coil with a variable air gap(?) so that if you push the handle forward from center, it drops the 120V as the air gap increases by 0-30VAC in-phase with the source, then if you pull the handle back past the center point, it drops the AC voltage by 0-30V, but at 180 degrees OUT of phase with the source. The phase difference becomes the sign of the signal, the 0-30V is the amplitude of the signal. This was then somehow fed into the M-G set power control to provide DC voltage control to the hoist motor.

I had, in my misspent youth, yanked out several Ward Leonard systems and replaced them with Solid State DC drives, but we NEVER retained any of the old operator interface systems so I never bothered to see what they were. I therefore have zero experience with this, but the guy I'm doing the project for insists these are (were) very common on old Ward Leonard and GE DC drive systems. Google is ignorant of the term "S21" with regard to anything speed control related, other than to confirm that is was indeed used on old crane controls. All I'm really looking for is a more accurate description of what it is/was and does so that I can come up with a way to transform it to something a modern drive can deal with, such as +-10VDC, but with a degree of accuracy one would expect for a hoist drive. A drawing would be fantastic.

I have by the way already offered to just replace the Master Switch with a more modern version, but he is insisting that anything that changes the operator's perception of how it works, real or not, is unacceptable. I MUST use the exiting system, although a transducer is acceptable once it leaves the control cab. Needless to say, I cannot find any sort of 0-30V In-Out of Phase to +-10VDC transducer and before I start cobbling together components to make one, I'd like a more accurate picture of what it is.

"Will work for (the memory of) salami"

 

[waross]

Hi Jeff. How about a small transformer to generate 30 Volts or more and place it in series with your signal. That will transform your 30-0-30 VAC signal into a 0-60 VAC signal. That you can rectify and scale to something that the drive will accept as an input.
Pity the next service tech. grin

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[danw2]

Shouldn't anyone so insistent on keeping a 50 year old MMI be expected to put some effort into supporting his fantasy?

To that end, 50 years ago electrical controls were pretty well documented with theory of operation, schematics, parts lists, troubleshooting, even an alphabetical list in the back called an 'index'.

Somewhere in that plant, there's a manual for these controls because back then capital jobs always required at least 4 copies of the operations and service manuals and at least one has survived, somewhere.

It's probably a non-descript 'report cover' titled with a typed adhesive label, the kind with the metal strips that go through the punch holes and then fold over. It's somewhere, he just hasn't put any effort into finding it.

You might encourage him to participate by painting him a picture of the alternative: dealing with a series of 26 year old kid salesmen with 4 months of 'drives' experience under their belt promising the world because everybody's line is 'world class' and their service guys 'really know their stuff'. Give you the shivers? It will for him, too.

 

[MikeHalloran]

Star for danw2; gave me the shivers, too.


Mike Halloran
Pembroke Pines, FL, USA

 

[Skogsgurra]

Jeff
Agree that anything else would be better than that original HMI.
Agree that there are documents describing it, somwhere.

The thing you describe is nothing but an ordinary differential transformer and all you need to make it useful with a modern drive is to put a synchronous rectifier and some smoothing between that device and your drive's analogue +/-10 V input.

If you dig into it, there is probably already a synchronous rectifier in place and that rectifier was used to feed the excitation winding of the generator. Tell the guy that wants to keep that HMI that the synch rectifier is part of the HMI and that he can't have one without the other. Adapting it is easy, just put an RC filter (no electrolytics) and a voltage divider so your +/- ca 40 V are reduced to +/- 10 V.

Of course, I could also build you one, using modern stuff like an Arduino and communicate with your drive over USB or WiFi - perhaps even via a 5 V analogue signal plus FWD/BWD logic signals. There may even be a +/-10 V signal - but that will cost extra.

But do look for that synch rectifier first. It is there. Promis.



Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[mikekilroy]

unless u saw it and it indeed is a iron wad going linear in/out of a coil, it sure sounds like a center tapped variac?

 

[CREED01]

This "S21 Regulator" sounds a lot like the "Induction Master" used on P&H Static Stepless Crane controls,
but I don't remember any phase shifting being involved.

 

[jraef]

I have never seen it myself, I was going off of his description only and he was not entirely sure exactly how it worked either, hence my query.

But I did manage to find info on old GE Crane Control Master Switch catalog for variable speed DC crane drives, and I think I see why he wants to keep it. The reference output is non-linear, the signal voltage changes in something like an S curve in either direction. So when you first come off of center, the change is very small with relation to handle movement, but the further away you get from center, the higher the rate of change. So operators will be 'used to" the non-linearity of it and if it is replaced without that consideration, there may be training issues and risks of error. There cranes are on a hydro dam for servicing the generators and turbines, so making a slight mistake as you lower a rotor into a generator can be very costly, if not dangerous.

The differential transformer consists
of a stator with three coils and a uniquely
shaped rotor. As the handle position is
changed, the rotor changes the amount of
coupling between the three coils of the stator
and produces an output. Fig. 61 illustrates
this output voltage and clearly shows
that very fine control may be obtained at
creeping speeds. In the low-speed range
where fineness of control is desired, a large
change in handle position provides a small
change in output. At high speeds where a
fast response is required, a small change in
handle position provides a large change in
output.
The neutral, or OFF position, is notched
to provide positive position ‘‘feel.’’ All other
points are un-notched to provide the operator
with a truly smooth, stepless ‘‘feel.’’
The switch produces no torque to oppose
movement of the handle, minimizing operator
fatigue. Contacts are used for directional
purposes to select the reversing contactors.
The primary coil has a maximum rating
of 125 volts. Any change in the input voltage
will directly affect the output characteristics.
The output curve shown is based on a
load resistance of 1000 ohms. With this
load, the effective power output is 0.25
watts.

His description seems to indicate a different version where instead of using contacts to select Fwd or Rev contactors, it is using the signal change from this "in-phase / out-of-phase" condition. He had mentioned that there were no other contacts involved.

I think now that I see this and understand his concerns, it appears that he is most likely unaware how easy it is now to pop in a simple algorithm to create that non-linear response to a signal change on analog inputs. The only tricky part may be in determining, most likely empirically, what the shape of that curve looks like.



"Will work for (the memory of) salami"

 

[iop995]

If there are acces to reference signal to use for phase / direction change, you need a rectifier and scale down to 0-10V. Need care to have good response for snall signal values. If there are not acces to reference signal need to add a sensing solution for FWR/REV. Check also if AC drive have this "S" function for input signal so you can setup a new system with same feeling for user. If it's not included in AC drive this function, may be added by a separate circuit.

 

[Skogsgurra]

The old differential transformers had another quality that may be difficult to achieve with a joystick - they were virtually indestructible. And they had lots of "authority" in their robustness and mass.

I have worked with crane operators (container cranes mostly) and they are very sensitive to how the control's feel. They get nervous if it feels too delicate. One can understand that. Loading and unloading containers is more art than craft - at least if you are going to get it done quickly and safely.

As I said, look for the old synchronous rectifier and use it. All you need is to smooth it and scale down the voltage. Put Si diodes there instead of the selenium rectifiers that probably were used. Synch rectifiers were usually built with switching rectifier bridges without any voltage drop, so there will not be much difference. I think that your customer can like the idea.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.