Actual limit on encoder frequency

[H.O]

Hi,
Got the SmoothStepper, nice piece of equipment, I literally had it running my single axis test setup in about 5 minutes. Anyway....

With the 3600 lines encoder, M=0 and without the 2200pF caps in place I found that 580rpm was OK but 590 and above wasn't. I then tried with M=1 and found that 290rpm worked but 300 and above didn't so it's basically the exact same frequency as with M=0. (580*14400/60=139.2kHz)

So, just as with the LPT-port, ~135-140kHz seems to be the limit on my setup. I'd like to point out that I have the recalculation interval set to 75 which relaxes the strain on the UHU-chip a little, I tried with 66 (which is default) but then got false transitions at 590rpm's too - YMMV.


Then I installed the 2200pF caps, set M=0 and started to test. I immediately got O-counts at 590rpm, I tried 400rpm but got O-counts there too so I stopped. The pulse-stream from the SmoothStepper is very clean, so why I get O-counts with it and not with the signal generator beats me. I'm missing a few parts so I can't complete the remaining two HP-UHU's and next week I'm going away for a few days so it'll be a while before I can test with another drive.

/Henrik.

[contactirfu]

hmmmm interesting

[jcdammeyer]

Originally Posted by contactirfu hmmmm interesting

I'll call your "hmmmm interesting" and raise you a "hmmmmmmmmmm interesting".

[contactirfu]

well sort of my way to show helplessness

[jcdammeyer]

Originally Posted by contactirfu well sort of my way to show helplessness

Just trying to inject some bad humour.

Hopefully this weekend I can get to my HP UHUs and do some testing too. It's just I've been so busy with writing up this blog for assembling the ELS

http://www.autoartisans.com/ELS/packaging_the_els.htm Did lots yesterday but haven't updated the blog with that info yet. Now I'm off to go get a keypad overlay laminated. I always find packaging anything the hardest part of a project.

John

[H.O]

Hi,
Just a quick note. I've assembled my two remaining HP-UHU's and I've managed to put one of them thru some initial testing. Seems to work fine so far - as long as I don't install the 2200pF capacitors. As soon as I put them onboard I get O-counts, just like with the first drive. If I remove them it's fine again.

Tomorrow I'll test the third one but my gut-feeling tells me it will behave the same.

Strange one this.....

/Henrik.

[H.O]

Hi,
So, the third and final drive is now complete and tested. It behaves exactly as the other two does. With the 2200pF caps installed I constantly get O-counts and as soon as I remove them I can run four hours without getting a single one.

I have no idea why, and I will definitely keep an eye on it once the machine is running but right now it's running better than ever.


Although a bit OT I thought you might be intersted to see what I'm working on, so I'm attaching three photos.

First one shows the machine that is being converted. It's an Abene VHF-3 that originally came as a manual machine but converted to CNC by a previous owner so it had ballscrews, motors and old SCR drives but missed the control.

Second shot shows work in progess on the electrical system. I pretty much scrapped everything there was and started from scratch, originally fitted VSD-A drives but they seem to be too small for my motors so I'll give the HP-UHU a proper chance. The console features a 17" touch-screen a MPG and a couple of hardware buttons that will interface to Mach3 thru the CUBLOC board and MODBUS - some day....

Third photo shows the first HP-UHU mounted as well as a place for the remaining two. I have rather long cables from the power supply to the drives so I made a small PCB with a 1000uF/250V cap, a fuse and anti-parallel diode, and a 200V MOV that will hopfully aid in killing any inductive spike if/when the fuse blows. I don't know if the extra cap is actually needed but it can't hurt and it also gave me a place for the fuse and diode.

On top of that is another board that takes 18-24V in and produces 15V for the HP-UHU. These are being fed from a separate 24VDC supply that is dialed down to around 20V. I already had this supply and couldn't find an afordable 15V switcher for DIN-rail mounting so I ended up doing it this way.

/Henrik.

[kreutz]

Originally Posted by H.O Hi, So, the third and final drive is now complete and tested. It behaves exactly as the other two does. With the 2200pF caps installed I constantly get O-counts and as soon as I remove them I can run four hours without getting a single one.................................... /Henrik.

Congratulations!! Beautiful job!!

Regards,

Kreutz.

[kreutz]

I have been looking around on the Internet for complains about the filter capacitor on the SN75115 chips, but so far, apparently nobody had noticed this strange behavior...
My advise is to get rid of the 2200 pf capacitors, as Henrik's tests proved that they are the main cause of the O counts.

[H.O]

Thanks Kreutz,
Regarding the caps, apparently they are supposed to be there so something strange is going on. My advise would be if you don't see any O-counts then leave them on-board. If you DO get O-counts try without them and see.

I have three new USDigital encoders on order, when I get those mounted I will do another test with and without the caps just to make 100% sure.

Thanks!
/Henrik.

PS. Kreutz, have you had a chance to look on the trace-cut-mod I PM'd you about? We don't need to discuss it in this thread, just a little reminder in case you have a minute over to look at it.

[kreutz]

Originally Posted by H.O Thanks Kreutz, Regarding the caps, apparently they are supposed to be there so something strange is going on. My advise would be if you don't see any O-counts then leave them on-board. If you DO get O-counts try without them and see. I have three new USDigital encoders on order, when I get those mounted I will do another test with and without the caps just to make 100% sure. Thanks! /Henrik. PS. Kreutz, have you had a chance to look on the trace-cut-mod I PM'd you about? We don't need to discuss it in this thread, just a little reminder in case you have a minute over to look at it.

Hello;

I keep it in mind, but have been too busy lately..., as soon as I have some free time I will check it.

Regards,

Kreutz.

[BobWarfield]

Very interesting thread!

I find myself wondering what the different pulse sources look like AFTER the R/C circuit as that circuit seems to be the differentiator. It's supposed to filter high-freqs, but the edge of a square wave is a high freq. Maybe that edge looks enough different from Mach 3 versus the sig gen that after the R/C smooths it to a lower freq the transition is no longer detected as reliably? Or perhaps it just changes the pulse in some other way that differs between the two sources.

I make this stuff up as I go along, so don't take it too seriously. The good news is you got it working.

Cheers,

BW