My Troyke CNC rotary table is actually working!

A while ago I mentioned this Troyke CNC rotary table that I bought off ebay.
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item 0587743949 http://igor.chudov.com/projects/Bridgeport-Series-II-Interact-2-CNC-Mill/24-Troyke-Rotary-Table-U12PNC/
It was described as "untested, as is", so I called the seller and spoke on the phone. My impression from him was that he was an honest person, so I bought this table despite it being AS IS.
When I received it and applied power to the motor (30v, no more than 5A limited power supply) it would not budge. Called Troyke, they have NOTHING on it.
I set it aside and continued to work on my CNC mill.
Now that I have only a few odds and ends to do on the mill, like buttons and switches. They are complicated due to HALUI etc, so I wanted to do soemthing easy. I looked at the table again.
I took the motor off. Without the table, the motor spun happily.
I stuck a bar into the table's motor connection (star type plastic force transmitter thingy) and tried to turn the table. After a little effort with a wrench, it moved. I found one more screw that might (not sure) have stopped the table, backed it out, and in any case, after a little while of rocking around the table felt much easier to move. It might have sat for 10 years or whatever and just got all stuck.
Put the motor back and again, the motor would not move it. At this point I realized that 5A is just not enough to get it going.
I called in in heavy artillery (military battery charger going up to 20A, 150V). Set it to 30 volts. When connected to that, the motor MOVED, and kept going.
More voltage, more speed. I kept it running for a while (20 minutes or so). The running amps were approximately 3 amps. From that, the motor became warm, but barely, had a body level temperature.
I explain this by assuming that the screw is preloaded or some such, and is hard to turn because Troyke tried to eliminate backlash.
What I know at this point is that it is a DC servo driven table and it works.
What I do not know is what is the operating voltage of this table. How would I find out? I am thinking, to take the motor out again (or take the rear encoder cover), and increase voltage until I get approximately 2,000 RPM from the motor. That should be appx. the max voltage. I do not really care with this table, to get max speed from it, as rotary tables do not often need to turn fast. I just do not want to go beyond the parameters of the motor.
My second question is how do I decipher encoder wires. I will make a separate post about it.

Ignoramus32079 wrote:
  Quoted Text Here  
http://igor.chudov.com/projects/Bridgeport-Series-II-Interact-2-CNC-Mill/24-Troyke-Rotary-Table-U12PNC/
So you broke the table brake???
I'd have done some careful exploratory surgery to see where the wires all went (table home switch, etc.) and look for brake mechanisms before forcing anything.

  Quoted Text Here  
I do not think so. There was no feeling of anything "breaking" as I unstuck the table.
There is a table lock nut, so labeled, that I fully loosened before starting everything.
  Quoted Text Here  
The wires are all contained inside the motor. No wires going to the table. I believe that there is no home switch on the table and nothing electrical or pneumatic in the table body.
I believe the wires are: motor ground, motor power, tachometer, encoder. There are no unaccounted-for wires and no pneumatic connections anywhere.
The table does not really need a brake, as it is probably a worm drive with a high ratio.
i

Ignoramus18921 wrote:
  Quoted Text Here  
If it was intended as a light duty low precision table, perhaps it doesn't have a brake. All the heavy duty precision tables have brakes to improve rigidity and protect the worm gear.
No table home switch??? If it really doesn't have one, you're going to need to add one.

  Quoted Text Here  
I would not know. But it does not look like there is a brake besides the locking nut.
  Quoted Text Here  
Probably, yes. I want to first figure out if I can somehow deduce which encoder wires are which. I will make a separate post about it.
i

  Quoted Text Here  
Pete, this model is U12PNC. This table is the same as the widespread model Troyke U12, like this one:
http://auctions.machinesused.com/listings/details/index.cfm?itemnum „6975105#DESC
I sold a U12 once as well, but for some reason cannot find the auction listing in the archive.
Anyway, what this U12PNC is, is a U12, with the manual crank handle removed, and a servo motor with what I think is an extra reduction (extra worm?), installed.
This was probably Troyke's first attempt at making CNC rotary tables, so they reused a successful manual design and their existing production line.
This U12 table is a decently accurate table, at least when it was new, and lists accuracy of 30 seconds, or 1/43200th of a circle.
Here's a home retrofitted Troyke U12 with a servomechanism.
http://ef.algebra.com/e/330456487538
Mine seems to be a much better design, taking much less space and overall more professional.
i

On Mon, 02 Aug 2010 08:55:07 -0500, Ignoramus18921
  Quoted Text Here  
Same one I have..but I think mine is horizontal only.
Ill have to check. Ive got another one that is horizontal/vertical that is the same size, but has the 4 dividing plates to go with it.
Cant remember the brand of the second one, if it was marked at all. Probably chinese.
I sold a nice Yuasa INDEXING rotary table last year, which went on a CNC mill after I installed a trigger switch and whatnot on the rig. Send the mill to a given location, it would step the table around to the next location, drill, mill, return it to the location, rotate, move to the next location and so forth.
Air driven with pins for the number of locations per full rotation, up to 36 IRRC
Gunner
Gunner
"

wrote:
  Quoted Text Here  
A lot of those tables didnt have a home switch. And in most cases, they really arent needed. When one drills/mills a bolt pattern for example..all the holes tend to be symetrical.
I see a few of these in my clients shops, and they are generally set up as a fixture that holds an asymetrical part that is indexed via control the same way each time and are set to go "home" via counts.
A home switch IS nice, but not always required.
Gunner
"

  Quoted Text Here  
Good point.
Gunner, a home switch should be a piece of cake to install, no? I have a bunch of oil tight limit type switches for this purposes (most used, some new). All I need is to add a dimple to the side of the table and a switch to touch the dimple, right?
i

Ignoramus18921 wrote:
  Quoted Text Here  
That's pretty much it. Mill up a little ramp piece (double ramp) and attach it with a couple countersunk screws. Also consider that you can do this with the switch mounted on the RT body and pointing up towards the table surface and use a ramp piece milled so the ramp faces down towards the switch. This arrangement will likely allow a more compact home switch setup.
FYI: Early RTs generally operated as indexers, connected to a dedicated control which received triggers from the mill's CNC control via one or two "general purpose outputs". You programmed your index steps into the dedicated control and then the CNC control would trigger the indexer control to move to the next programmed position and then to reset to home at the end of the cycle. This allowed CNC controls that didn't support a 4th axis to do limited 4 axis work. Since you CNC control supports a full 4th axis, it's really nice to have a proper home switch for it.

wrote:
  Quoted Text Here  
Good post.
Gunner
"

  Quoted Text Here  
This is a great idea. Thanks!
  Quoted Text Here  
Yeah. I think that I have a chance, with this rotary table, encoder on the spindle, etc, to have a really great machine that could make almost anything I would consider doing. Considering that the table only cost $149, I may as well do it right.
So, at this point, I should strive to do a good job on all fronts. On the rotary table, specifically, I will look into the encoder and probably install a new one, and I would also add a limit switch to it as well.
Then I will have a true 4 axis CNC machine, also capable of rigid tapping, pretty much a dream machine, at a relatively low overall cost. I am estimating that the total cost will be $2,350, based on my spreadsheet.
Tonight I will open up encoder housing to see what options I have on this RT.
i

On Mon, 02 Aug 2010 10:19:40 -0500, Ignoramus18921
  Quoted Text Here  
Normally..good used rotary tables for say...Fadal....cnc capable..run about $3000 used.
You are doing very well indeed
<G>
Gunner
"

  Quoted Text Here  
Thanks... This was a bit of good luck with that rotary table. Even if it is a little worn and is only 1 arc minute accurate (as opposed to factory accuracy of 30 arc seconds), it will be good.
I will try to open it up to clean and photograph. It should be easy. The nut holding the slotted table is on the bottom most likely.
I have a Haas 5C indexer that I will sell, and it will pay for the whole new CNC mill.
i

Ignoramus18921 wrote:
  Quoted Text Here  
You don't need limits on a RT, since they are continuous rotation devices and you can't "crash" them. The home switch does need to be a double ramp so that you can overrun it in either direction. There are some cases where you may want to rotate more than 360 degrees, such as running out a milled groove so you get a clean finish and no start/stop burrs.
  Quoted Text Here  
Very good.
  Quoted Text Here  
Post some good pics of it. I wouldn't bother replacing the encoder unless it's an incompatible type or really low line count. Count the turns/rev for the table so you know the drive ratio or course.

  Quoted Text Here  
I am sorry, I meant a home switch. Yes, it will be a double ramp type.
  Quoted Text Here  
Sure.
  Quoted Text Here  
I will do so tonight for sure. I am not yet sure how to figure out which encoder wires are which. I am very tempted to just replace it with a US Digital one, just for uniformity's sake. But I will find out facts first. I would like an encoder on this to be fully covered due to its future location on the machine.
i

On Mon, 02 Aug 2010 09:44:07 -0500, Ignoramus18921
  Quoted Text Here  
Something to hit depress the home switch, yes. Now if you rotate at high speeds..you may have an issue...it may well overrun the home switch by a few arc seconds or so. Depends on the motor mass. Unless you set up a run to home, off, coast and then reverse to home again. This may be simple, or a bitch. It will require a seperate axis board to do it right.
Gunner
"

  Quoted Text Here  
I am not worried about overrunning home switch. I have some experience with this. It is a non-issue in this case with this servo driven worm drive and EMC. The table will stop instantaneously in relation to its speed before stopping. Much less than the length of the "dimple" that I would make.
i

Gunner Asch wrote:
  Quoted Text Here  
Absolutely he'll need another servo drive for the 4th axis.
The home switch needs to be a double ramp so it can be approached from either direction and overrun without issues. For the actual homing routine, you'll always home in the same direction, and at a speed that allows you to not overrun too far. You'll use the encoder index pulse for homing, just as with the other axes. Home at modest speed until the home switch trips, then either continue or reverse until the next index pulse and you'll be spot on.

  Quoted Text Here  
I have a shipload of servo drives that should work for this application (30A8T). At just $25-30 on ebay it was a no brainer. I actually have eight 30A8T's.
I have a space to mount the fourth drive alongside the three, I did think about 4th axis from the beginning, and left space for the fourth drive on the mounting bracket.
  Quoted Text Here  
Come on guys. This rotary table will stop almost instantaneously under more or less any speed that is safe for its motor.
i