Retrofitting a 3-axis VMC with DC servos - guidance needed

Is the noise on the analog wiring side or the encoder side?
Are the original analog cables in use? If yes, those will have an analog ground and a general machine ground, should not be wired together, although i had a single case when it was a must.

From what I've measured, the issue seems to be on the encoder side.

I initially thought that analog ground and power/system ground were connected. The original cables are currently in use. According to the diagrams, the analog signal for the axis is connected to the machine's frame/ground.

Is it better to isolate the analog ground from the system ground? Under what conditions is it strictly necessary to tie both grounds together?

Japoo_Ness wrote:

From what I've measured, the issue seems to be on the encoder side.

Shielded cable, shielding grounded on the Mesa side.

Japoo_Ness wrote:

According to the diagrams, the analog signal for the axis is connected to the machine's frame/ground.

Analog cable has 2 wires plus shielding, shield must be grounded on the Mesa side, and one of those 2 wires is the analog ground.

Japoo_Ness wrote:

Is it better to isolate the analog ground from the system ground? Under what conditions is it strictly necessary to tie both grounds together?

There is no exact reference for it, see above, use a DVM to measure between analog and machine ground both in AC and DC mode > if any of those read more than a Volt or two = do not connect. If that is all correct but still does not work, you can try grounding the analog ground to machine ground through a 1K resistor for a single line, test with only a single line/analog drive/connection to mesa.
-
As a general rule, use a very good power supply for the Mesa as the encoders are powered from it.

Then, I need to connect the shielding for the encoder signals and the driver signals. Is it okay to connect the chassis from the 7I97T to a bus bar with these shields, or is it better another way?

I also want to ask: doesn't this keep this ground somewhat floating, since it doesn't have a discharge path, or am I wrong? Or when you mean ground, do you mean power ground?

All shielding should be wired to a terminal block to machine ground.
Search for HyperTherm grounding best practices, they had a very good pdf about it.
And since we are here, grounding and shielding is never simple, so requires a bit of time, but it is very well worth it for the time it saves when working with such machines.

Thanks for the link and recommendation, that helped a lot. The machine is now moving on all axes, homing correctly, and the noise disappeared.

A quick question about the 7i97T bitfile: the firmware I flashed doesn't include dither support. Is there an alternate build available with pwm mode 1? I couldn't locate it in the repository.