Pleasant (tentative) surprise with Bridgeport Customer Service

any, and I mean ANY VGA monitor will support EGA resolution - it was (is) part of the spec - there is an adapter available so you can connect a 15 pin cable to the 9 pin ega/cga monitor output, and the wiring for the adapter is available on line.

As far as I know, the original XT did not support EGA, the choice was either a CGA card that had a 9 pin and composite video with a resolution that was really pathetic (640×200), ega went a bit higher (640×350 ) and VGA higher still (64-X480).

wikipedia provides the pinouts

formatting link
The other choice on XT machines was the monochrome (MDA) monitor which was a much more useful monitor and much higher resolution - it also used a 9 pin connector, pinouts below - note that the MDA display was frequently destroyed by connecting it to a CGA card - but the repair is easy, replace a $3 horizontal drive transistor - I fixed many.

formatting link
> Ignoramus16651 wrote:

Reply to
Bill Noble
Loading thread data ...

[ ... ]

You *don't* want to use the contactors for the motor if you are building a VFD into it to give three phase.

Instead -- send TTL commands to the forward and reverse inputs (assuming the voltage will work -- otherwise try CMOS which will go as high as 15 V), and set aside a spare axis D/A converter in the controller to send speed voltage commands to the VFD. Just set up the old contactor to drop power from the VFD's *input* when the computer is turned off.

Good Luck, DoN.

Reply to
DoN. Nichols

Encoder 101:

You have a Heidenhain control. They used a great feedback idea - linear scales. Some had quadrature feedback but most older ones had some sort of sinewave system kind of like a resolver. That's all I know about them. But I have read of folks converting these to quadrature. Linear scales with quadrature feedback would be top-of-the-line. If you go this route, any backlash will kill you (servo oscillation) but its more accurate. Heidenhain scales bring good money on eBay. My control has a system to use two feedback devices, I'm almost certain EMC can only use one. So you use this or a shaft encoder, not both.

Encoders on the servo motors give feedback position also. No backlash, they are on the motor shaft. USdigital.com makes an inexpensive kit that bolts on the shaft extension on the back of your servos. I've not worked with Heidenhain, make sure you have a spot and motor shaft to work with.

It helps to choose encoder counts carefully. My control counts "edges" with four counts per encoder pulse - I think EMC does the same. I won't explain this unless you need, Anyway, the best way to eliminate rounding errors is to have 100,000 or 50,000 or some very even counts per inch. For example, if your servo turns five revolutions to go one inch get 2500 pulse encoders.

Also double check that EMC can use differential encoders. These are MUCH more noise immune. I pulled all my hair out (I'm totally bald now) over what turned out to be encoder noise during tool change.

Karl

Reply to
Karl Townsend

Right, you have to look it up by model #, but most of the ancient models are on somebody's database, on the net.

I know you have an oscilloscope from your work on the TIG welder project, so finding the sweep rates of the CNC control should be easy, once you get it powered up.

Jon

Reply to
Jon Elson

If the motors have Harowe Controls resolvers in the NEMA size 11 package, it may be hard to fit encoders in place, as they are 1.1" (28mm) diameter. I have a resolver converter that makes them look just like encoders to the CNC control.

Resolvers are not obsolete. They were the top choice when cost was no object and optical encoders had light BULBS in them, as you got a servo runaway when the bulb burned out! Now they are just more expensive, but will run underwater, fully filled with oil and coolant, and at 150 C. They are nearly indestructible.

If you mean linear encoders from DROs, the problem with them is the general DRO-class encoders have low resolution, typically .01mm ~= .0004" This is really not good enough for CNC motion control. You generally want 10X the encoder resolution compared to displayed resolution. I have 50 uInch resolution on my X and Y, and 25 uInch on Z on my Bridgeport mill retrofit. You can get higher resolution linear encoders, and now 1 micron resolution is becoming pretty affordable, but they will still cost several hundred $ each. If your machine has resolvers and good ballscrews, just keeping that system and using the resolver-quadrature converter may be the easiest way to go.

Jon

Reply to
Jon Elson

Karl Townsend wrote: My control has a system to use two feedback

Nope, not true. EMC2 out of the box doesn't do dual-loop feedback by default, but a number of people have certainly done it. It does require additional encoder counter inputs and a little bit of special HAL code to link it all up. From the

formatting link
case studies page
formatting link
's a lot of unrelated stuff about Stuart Stevenson's awesome shop in Wichita, but there is also a couple paragraphs about his dual-loop servo setup on the G&L HBM under "the Project". (By the way, he is using my PPMC analog servo interface for this.) The hal and ini files are on line, but I don't want to post the URL here. I could send you these files.

Well, as long as the resolution is high enough, there really is no "roundoff" problem, at least with EMC. All calculations are done in floating point format. Now, a low-res (.01mm) scale is just low, and non-inch, resolution. That isn't really "roundoff" it just has position points that don't fall on even .001" boundaries. You get .0000. .0004, .0008, .0012, etc. readings, and nothing in between. That is why it is good practice to spec a higher res encoder. Much easier to get there with shaft encoders and ballscrews than linear encoders.

This is entirely a function of the encoder counter board used. If it doesn't natively support differential inputs, you can always use a US Digital diff. receiver board, they are $12, I think.

Jon

Reply to
Jon Elson

Yes, except DC SSRs work really well, and safely, for controlling the VFD command inputs. Otherwise, I completely agree. The command inputs are set up for contact closure type controls.

Jon

Reply to
Jon Elson

PolyTech Forum website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.