Rigid Tapping Setup

[RJS100]

When I initially setup the servo motor, it ran rough, so I added the attached snubber, and it immediately ran much smoother (visually).

I do not have an oscilloscope but will try a few of your suggestions to troubleshoot. I did purchase the encoder on ebay and paid about $36.00 buck for it. I suspect this maybe the culprit as I machined a small part and the surface finish was very nice. If the speed was actually unstable, I think it may show up in the surface finish.

[RJS100]

I tested a range of rpm's using the test jog mode and it did indeed run quite a bit smoother. For example, when it was running 2451 rpms, only the "1" was bouncing around a bit.

If I am already using the snubber installed, and the encoder cable is a shielded cable that is grounded on the Acron side, what else can I do to minimize this noise?

[Muzzer]

It's not recommended to put large capacitors (total 200nF in this example) across the output of an opamp. In the case of the LM4324, that max recommended value seems to be around 100pF and certainly below 500pF (where the phase margin becomes zero). It's not going to be possible to define how the device will behave with a load that is 2000 times larger - while it may appear to be stable, you may even be stressing the internals. A better solution would be to put a series resistor between the opamp output and the filter cap, of the order of 1k or more. However, that might affect the voltage, depending what you connect to it.

See fig 5-23 and the bottom of the table in page 9:
https://www.ti.com/lit/ds/symlink/lm324 ... %252FLM324

1 digit of noise in a 4 digit speed is pretty reasonable - I'd be happy at that.

[RJS100]

Thanks for taking the time to look into this. I'm not sure I'm following this completely. I think your note that "1 digit of noise in a 4-digit speed is pretty reasonable" is suggesting that if just the "ones" digit is bouncing, that would be good. I agree. But this was only the case when I ran the spindle in jog mode, I got significantly more bounce when the spindle was controlled by from the analog signal.

With respect to the signal filtering this is a bit over my head as it has been over 40 years since I looked at these diagrams. The filter shown in the previous example was one suggested to me by Uwe while using the same series JMC AC servo JASD controller (his is a 1.8KW as opposed to my 2.6KW).

Are you suggesting that I add an additional series resistor of 1k or more in front of the existing filter? If so, what resistance would you suggest I start with?

Thanks for your time... Richard

[RJS100]

Here is the link that describes the snubber circuit that I used. Perhaps this you will make more sense to you:

https://centroidcncforum.com/viewtopic. ... uit#p46032

Are you suggesting that I try the filter show on figure 4?

Thanks again for your help... Richard

[RJS100]

I tried a different filter as shown below with no change at all. I even tired rerouting the analog signal wire that goes from Acorn to the servo and it made no difference either. I set the servo to display the analog input signal, and the digital display bounced around confirming that the analog signal the servo is receiving is not steady. It even bounces up/down when the spindle is stationary. I once again ran the servo in test mode, and the speed reported by the encoder on the spindle was steady!

Any ideas as to how to better attenuate the noise in this signal would be appreciated. I read a post on this forum suggesting adding a .1uf capacitor across the analog input at the servo end could help. Do you think this makes sense? I'm happy to try it but it is rather difficult to implement as there is a 50 pin connector on the servo end. I did install EMI line filters one the 220V side of the servo. I did not use any ferrite chokes. Do you think these may help?


Please let me know your thoughts...Richard

[Nigelo]

No, the 0.1 UF ceramic cap needs to be fitted between Acorn's analog Vout and Com terminals. It assisted me in bringing RPM under proper control and many other folk in stabilising their analog output. You normally don't need anything else other than proper shielded wiring best practice throughout

[RJS100]

Hello Nigel,
The attached is the snubber circuit that I am currently using that was suggested by Uwe. This is wired across the analog output on the Acorn board.
Are you suggesting that I replace this with just the .1uf capacitor?

[Nigelo]

Hello Nigel,
The attached is the snubber circuit that I am currently using that was suggested by Uwe. This is wired across the analog output on the Acorn board.
Are you suggesting that I replace this with just the .1uf capacitor?

Yes, that's how I have my system but ensure the cap is ceramic

[RJS100]

Thanks. I just verified that I used a ceramic disk capacitor. I will remove the resistor and other capacitor and cross my fingers.

Thanks for your help... Richard