Does acceleration make a difference?

Yes, to show the difference I have just taken a video, same Velocity (20m/min) in both sections of video.
 First section is 1000mm/s/s
 Second section doubled to  2000mm/s/s.

The whole cabin was shaking with the faster acceleration but it is not on level ground and needs shifted anyway. Hopefully it will be a bit more steady once moved

https://youtu.be/pVwKnkPx2YA

Will be interesting to see what you can do with the closed loop  steppers.

Hood

Thanks Hood, pretty snappy I had to de-tune my plasma a fair bit as was shaking itself enough  to give slat-wobble which wrecked the cut.

Don't forget though that this new project is a heavy old milling machine, not plasma, I have no idea what acceleration does to a mill but i'm fairly certain it's not as important as a plasma.

There is always going to be a max acceleration that is possible no matter what, it will depend on the rigidity of the machine. There is also the fact that Mach is basically bang bang with the acceleration, full accel, full decell. If it had S Curve then you could likely get much higher accel and things would stay smooth.

Regarding milling, well accel is as important in milling as anything else, the fastest you can get, whilst still being smooth, the better. Most  parts will run quicker with the higher accel and no problems with corner rounding or such like.

On my Bridgeport, when it had steppers I actually got a boost in performance by lowering the Velocity as it allowed me to increase the accelerations, parts ran much faster.
I think it was something like dropping the Velocity from 2500mm/min down to 2400mm/min allowed me to increase the Accel from something like 40mm/s/s up to 120mm/s/s. although I cant be certain of these numbers as it was probably getting on for 8 or 10 years ago.

As previously said my Beaver NC5 runs servos with Vel of 8m/min and accel of 800mm/s/s (just looked to confirm that, first time it has been fired up for a long time )
The 8m/min is just what I limited it to as I reckoned for old iron it was fast enough, I did have it cranked up to 20m/min (same as the  Chiron) but it was way too fast for that style of machine.


Hood

Thanks Hood, Interesting.

It looks like a series 1 BOSS would fit in, trouble is they are rare now

Hmmm...

Don't think there is much difference in height between series 1 and 2. Seemingly you got Series 1 and 2 both with sliding ram head and rigid heads, I have only ever seen the rigid ones here in the UK though so possibly a USA only thing with the sliding ram heads.

Out of curiosity I just set the plasma to old Bridgeport stepper Vel and tried different Accel, funny to think I used to think it was fast. Anyway here is a vid of  it.
https://youtu.be/I3JPAHYCxMk

Hood

I'm having trouble finding the correct heights for the various models, so many different series, models and mk's!

Well, nothing wrong with 2500mm/min, provided you only want to cut 12mm thick slabs all day Surprising how much sharper it looks as the acceleration went up.

Had a look at my manual for the Series 1 Boss I had, says 86" high.

Yes, the faster accel would make a big difference on a job, especially if it had a lot of smallish moves.
Seeing  and hearing it one after the other really shows the difference.

Hood

Thanks, even that sounds a bit tight, will have a measure tomorrow but its looking like I'll be converting the manual machine I think.

2:1 or 3:1 reduction on X&Y?

Let me answer that this way as it's all a trade off with steppers.

A reduction will increase the steps per unit value so resolution goes
up,torque delivered to the screw will be increased, but the motor must
turn faster for some desired feedrate / rapid move and that can affect
motor acceleration and max velocity. So you need to look at it as a system and
find what is required to accomplish what you want.

Also note that stepper motor curves are done at voltages that may be less or more
than what your power supply may deliver. Higher voltage will increase motor speed
but remember that as speed increases the torque will drop.

That's why knowing what smeone else did with a similar machine is very helpfull.
Even if all is calculated you don't know until the machine is run what you have.
But if you did do some basic design you know things should be satisfactory.

RICH

Story so far based on data collected...

Servo's seem to have won, 750W is recommended size. Ball-screws have 5.08mm pitch so with motors flat out, I think i will go for 3:1 reduction to give me loads of torque, good resolution and 5000mm/min rapids - this still sounds high for an old machine though. The 750W option was recommended buy a 30+ conversion guy so seems sound.

Resolution would be around 0.0006mm/step? Based on 2500 line encoders and 3:1 reduction and 5.08mm screws.

Control options is the nest step, Ethernet smooth-stepper and use step/dir signals OR CSMIO/A and use 10-0-10v analogue - this also offers me actual DRO read-outs for manual knob twiddling. Cost is the factor - £150 vs £480. Both will work with Mach3??

I would prefer the CSMIO option I think as I really want the loop to go right back to the Mach DRO so I get full position read-out in real-time. I can the recoup some of the cost by stripping and selling the new DRO i fitted to the mill a year ago, plus the working but redundant power-feeds and all other bits removed.

Have I missed anything out??