New to me CNC toy mill - Taig

Just picked up a little Taig off Craigslist which was outfitted with the old IMService/Globe servo setup.
I did a couple Mach3 projects before and they worked OK but even though the interface looked a little bit like "real" machines, I always thought it was distractingly busy and just plain ugly. Plus the technical issues...
Anyway, going with LinuxCNC this time and I've got three axes moving. Just have to refine configuration a bit and start wrapping my head around it all. I'm not 'fraid of lernin' sumethin' new. I figger I'll be digging in fairly deep so I actually know what's going in inside and that will be a ways into the future, but looking forward to it.

I'm a maintenance machinist now (so it's all manual) but back in the day I spent thousands of hours using CNC machines of all sorts in production and I kind of see what they are referring by Mach looking "professional", but only inasmuch as it's referring to the types of things a screen might show on the available tabs. But Artsoft could have done a vastly better job of it - it's atrocious really in my opinion, but oh well. Actual "professional" machine UI's are generally MUCH more user-friendly and easy on the eyes comparatively. Even the old text-only screens of old at least showed pertinent information with a logical layout and were easier to deal with than Mach, which looks like shuffled playing cards splattered on a cafeteria tray to me. Blaeughh!

I am really looking forward to the UI Lcvette et cie. are working on becoming a usable thing. It looks entirely awesome. It's my opinion that something like that might be so widely appreciated that Linux CNC might gain huge adoption compared to before. It's a hypothesis.

I've been scouring the net for possibilities on a replacement spindle motor I can apply PC speed control. Nothing jumped out. Current motor is a basic single phase AC motor.
Then I got to thinking about the huge pile of things I already have on shelves and boxes and realized that I have a decent sized DC motor with encoder which might do the trick in style! I picked up a Chinese counterfeit servo controller (DCS810) several years back which might run the thing.
Does this seem like a feasible possibility? I've seen folks using pretty high-count encoders for spindles and it seems like this might be a pretty slick arrangement if it works. The Taig doesn't require much spindle power of course and the motor I dug up would appear to be close to the size and power of the existing motor I think. I don't have any documentation on the motor but I just got done hooking up some gear and reverse engineering the wiring. Haven't worked out the PPR yet, but the encoder is just a 2-wire/2-channel quadrature but that's pretty standard stuff so probably just perfect... if everything else works.
I'm pretty new to all things servo though and it's common for me to be missing important parts of any given puzzle.
Any advice is appreciated.

One thing I'm not sure of is the rated voltage for the motor. Is there a procedure I can work through to determine how high I can go with the voltage on it? It's just a brushed DC, though it does appear to have 4 brushes instead of just the 2 like I'm used to seeing (but still just 2 wires), in case that matters.
Thanks.

Well I hooked the encoder up to the oscilloscope and put some power to the motor leads. The power supply I have spins the motor at 1285rpm and I'm getting 74µs per pulse (rising edge to next rising edge) at that rpm. I don't know what tolerance these measurements fall into, but it should be halfway close.
My math puts the encoder at somewhere around 600ppr - specifically 631ppr but that exact figure doesn't seem likely, does it? I'm gonna go with assuming it's 600ppr unless I think of a more reliable way to measure things. But dang, the oscilloscope ought to be within a RCH and the rpm is off one of those optical reflective rpm meter things which seems like ought to be pretty decent as well.
I couldn't think of a stupid easy way to just count the pulses through one revolution. My scope might even be able to do that but I'm not as familiar with it as I wished I was.
Any thoughts?
Thanks.

An easy way to do this is wire the encoder up to linuxcnc so you can count pulses,
then rotate the shaft multiple times (say 25 times) referencing a mark on the shaft/pully

Baahh!
First, thanks for the suggestion.
The despair comes from the fact that I had a NIB driver for the servo and software to tune it, but of course it takes an RJ11 on one end and serial on the other. Of course I don't have that nor the stuff to make one, even though I have piles and piles of components, wire, collections of.... oh just absolute tons of stuff around here and I do not have any RJ11 cables or ends.
Oh well. Maybe I'll just wire it up to the LinuxCNC breakout and just see if I can run the thing (the special cable is just for tuning purposes).
Been dorking around with this thing off and on all day and I'm a paper's dust away from trying the motor out.
Send it!

Well I punted on a proper tune for the spindle servo motor and just wired it up & tried it.
NOT BAD !!
The default settings in the driver actually seemed pretty smooth at least. I do need a higher voltage supply as I can get about 1000 rpm or so out of it with a 24V supply. I have an unused higher voltage one at work so maybe I'll go in & grab it tomorrow.
At any rate, it looks like it's gonna be a beautiful replacement solution. Plenty of torque for this little mill.

I do have to figure out how to set LinuxCNC up for it though. The stepconf wizard has spindle options for CW/CCW and PWM but not step/direction which is what the driver presumably wants. The driver manual shows no other mode options than pulse/dir. Maybe it would work with PWM pulses? Or maybe I need to do some text editing in a configuration file? Any help on this would be greatly appreciated.

I think I'm close though. And didn't need to buy anything new. Woohoo!

edit to add: Well I just realized PWM won't help any since it essentially varies duty cycle and not frequency, so I need to figure out a way to control the spindle with frequency and the stepconf wizard doesn't offer that option (unless I'm missing something). If anybody can either help me directly or at least point me to where I can find the info I need for this I would greatly appreciate it.
One thing that seems possible is that if I'm driving my spindle with this servo motor, then I oughtta be able to have the only Taig toy mill I've ever seen that can do rigid tapping!
The motor seems to operate beautifully if I configure it as the A-axis, but that won't quite do for spindle since it just needs to run until it gets switched off rather than move a determined number of steps.

Yes you can use pwm with a trick. Have an oscillator signal on one side of an and gate and the pwm signal on the other side. This will generate a controllable amount of pulses. Last you have to tune the servo in a way that it smoothes the pwm base frequency. A bit fiddling, but chepo and good chance to get a fast result.
P. S. I'm not sure whether u can use hal components for that. My guts feeling tents towards hw. A cd4093 could do the job.

That's already over my head. I wish it wasn't, but it is.
Does that do some kind of feedback loop that varies frequency based on duty cycle or something?
The frustrating thing is that LinuxCNC CAN output a completely appropriate signal for what I want to do, but only on an axis. I verified this by setting it up on a 4th axis just to test whether the stuff worked, and it was a beautiful thing! I got it spinning at around 1100rpm with plenty of torque as well as around 4rpm (not a typo - I'm guessing though, maybe it was 10rpm) and it was rock solid. I have a gorilla grip and I couldn't phase it as far as I could tell. Tons of torque.
Seems like it would be a doable thing for the software to send out the appropriate signal, but do it for the spindle output pins instead of on an axis. I wanna cry that I feel like I'm so close to success but just can't get there from here.