[dubble]

The different sound is nothing to do with the different kernel frequencies. The kernel frequency is the maximal frequency of the step signals and the OP has the same settings, same kernel freq. in Mach3 and the UCCNC and hears different sounds of the motors. Also different kernel frequencies will not cause different sounds if the running feeds are the same.

The reason for the different sounds is that Mach3 has a defective CV motion planner where the accelearions and deccelerations are often not optimal and especially the deccelerations are often higher than what you have setup in the configuration.
In the UCCNC the CV motion planner always obeys your settings and the profile is always perfectly symmetrical, this can be hear on the sound of the motors since the trajectories generated by the two system are different and therefor the sound.

An example logic analyser print of the X axis step signal with a simple G1 X0 ... X1 ... X2 movements

With Mach3:

Attachment 295098

With UCCNC:

Attachment 295100

Notice the different trajectories for the accel and deccel part in the Mach3 measurement, this has to be symmetrical if it was optimal.
With more complex paths the situation is often worse than this as this was a very simple one axis movement.

[mactec54]

The different sound is nothing to do with the different kernel frequencies. The kernel frequency is the maximal frequency of the step signals and the OP has the same settings, same kernel freq. in

Really, and the higher the step signals ( output Frequency ) the smoother the motors will run, with the motors running smoother, the sound they make also changes, this is not rocket science

[ger21]

Kernel frequency (max frequency) has nothing to do with the "output frequency", which constantly varies with velocity.

[dubble]

Really, and the higher the step signals ( output Frequency ) the smoother the motors will run, with the motors running smoother, the sound they make also changes, this is not rocket science

Nope, the output frequency varies the running speed or feedrate in other words, has nothing to do with smoothness. The kernel frequency is the maximum output frequency.
There could be one thing which could vary the smoothness about kernel freq. is the max. jitter (maybe you thought about this), but there is no exact relation, so again the kernel freq. does not directly influence anything about motor running smoothness.
Moreover high kernel frequency like 4Mhz will make the system noise sensitive since 4Mhz frequency means 125 nanoseconds pulse width at 50% duty cycle, I would not run anything with this short pulses using step/dir interface, not even a hobby machine, but sure not a machine in an electrically noisy industrial environment.

[mactec54]

dubble

I understand what you are saying, so what is the reason for this to be happening, I know what you said in the first reply, that is normal for any good motion control, like I said that's not rocket science, we have known about Mach3 motion planner for 10 years, I would hope that your's is much improved

so if someone runs your UC100 with Mach 3 why do the motors sound better and run smoother, no other changes just adding the UC100

[curiosity22]

Hey G'day,

Thank you for the explanation, I get the basic gist of whats going on. So pretty much due to a better CV control the steppers are making a sweeter sound. Which I can only assume must be better for the motors.

Cheers,

Steve

[daniellyall]

a smooth sounds is good (but you can play music with a stepper) if it sounds smooth it may well be running with a nice ramp up and ramp down, running a machine comparing m3 and m4 through the parallel port you get the same affected as using a UC100 it`s just newer and runs everything better same with m3 to uccnc software, thing just run better so it sounds better.
I don't want to know why I just want something that run`s nice a smooth 100% of the time also if it starts sounding ruff something could be wrong

[dubble]

dubble

I understand what you are saying, so what is the reason for this to be happening, I know what you said in the first reply, that is normal for any good motion control, like I said that's not rocket science, we have known about Mach3 motion planner for 10 years, I would hope that your's is much improved

so if someone runs your UC100 with Mach 3 why do the motors sound better and run smoother, no other changes just adding the UC100

It runs smoother when adding the UC100 and still running Mach3 is because the jitter is in avarage 100 times lower with the UC100 than it is with the LPT port. Tens of microseconds range compared to hundred nanoseconds.
A stepper motor will run much smoother, because it runs in open loop and is a high pole count syncronous motor, so if the syncronisation is better timed, the jitter is lower means a smoother running.
Imagine a simple example of a fixed steps frequency running, e.g. 100Hz. The motor will run smoother when this 100Hz in realism is always between 99.9 and 100.1 and will run less smooth when it fluctates between 99 and 101. The same thing is what happens when you add the UC100 to Mach3, the signals will be better timed or in other words it will have less jitter.

[mactec54]

It runs smoother when adding the UC100 and still running Mach3 is because the jitter is in average 100 times lower with the UC100 than it is with the LPT port. Tens of microseconds range compared to hundred nanoseconds.
A stepper motor will run much smoother, because it runs in open loop and is a high pole count synchronous motor, so if the synchronization is better timed, the jitter is lower means a smoother running.
Imagine a simple example of a fixed steps frequency running, e.g. 100Hz. The motor will run smoother when this 100Hz in realism is always between 99.9 and 100.1 and will run less smooth when it fluctuates between 99 and 101. The same thing is what happens when you add the UC100 to Mach3, the signals will be better timed or in other words it will have less jitter.

This is the best reply yet, and is what It was meant to be in the very first post, this happens to anyone that has added UC100 or Smooth Stepper to there system, with the motors running smoother,this also changes the sound the motors make as well

As for the Smooth Stepper it defaults to 256 Hz and is adjustable down or up to it's max, so does not run at 4 Mhz unless someone was to try/set it at its max

[dubble]

This is the best reply yet, and is what It was meant to be in the very first post, this happens to anyone that has added UC100 or Smooth Stepper to there system, with the motors running smoother,this also changes the sound the motors make as well

Well, IMO this shoud not be the first post, because this is not an answer to the OP's question. This was an answer to your question only which was a different question than what the OP asked about.
He used Mach3 + UC100 and then he used the UCCNC + UC100 and comparing these 2 setups, therefor the jitter is the same and is not the answer to his question, the answer in this case is the different CV planner as I have described previously.

[mactec54]

He used Mach3 + UC100 and then he used the UCCNC + UC100 and comparing these 2 setups, therefore the jitter is the same and is not the answer to his question, the answer in this case is the different CV planner as I have described previously.

That makes sense now, of course both will have the same jitter with using the same output device, I did not know he was comparing systems

[Bad Wolf]

I haven't been back to Mach 3 since early in this discussion. It was fairly evident from early conversation that it was plain inferior to UCCNC. It is old technology and really isn't the benchmark to be evaluating best of breed against in the product class, if anything the only real rival UCCNC has is Mach 4 and seeing that it has no CV planner, it will be sometime before it makes any kind of headway on UCCNC. Hence my decision to stay with UCCNC.

[ger21]

if anything the only real rival UCCNC has is Mach 4

Mach4 does have a CV mode, just no control over it at the current time. But Mach4 has many other issues.

UCCNC has several other competitors:

Planet CNC
EdingCNC
LinuxCNC

[Bad Wolf]

OK sorry I should have put it more in my CNC dummy speak... Mach 4 isn't as smooth. Gerry are any of the options you listed serious contenders?

[ger21]

Yes, all of them are. They're arguably more polished than UCCNC, which is still early in it's development.

[Bad Wolf]

Thanks I may look into them at a later date, currently I am still sorting out mechanical issues with my CNC, so until I have it right there is little point to seek better software.

[battwell]

honest appraisal
i own 7 routers.- all work horses- from small to huge ex industrial machines. all capable of very high speed machining
most have been retrofitted with mach3 and galil motion controller. and have worked ok
ive only been using uccnc since october- but using it every day cutting parts for my business.
im using uc300 with mistubishi servos.
all i can say is wow! they have done an amazing job on the trajectory planner! (read as accel/deceleration planner if you wish)
not once have i had to switch to g61 exact stop mode. which for cabinet work with mach3 is a must if you want perfect results of parts that fit (machining at high speed)
yes there are old code things it cant do- which is a bit of a bind- as i like to program some parts parametrically - g41 42 are not available
hardly a big thing these days as most of my drawings are done in vcarve pro etc.
i actually cut at up to 20 metres/minute when doing cabinet work which uccnc handles superbly using the stock cv settings - slowing down for corners etc really nicely and accelerating away as id like it to.
for carving work (vcarving) i only cut at max 2m/minute to get desired work finish . 3d carving 6metres/minute.
im amazed how much more detail is in all my work. (and for cabinet work- everything fits together perfectly first time)
only a couple of issues ive had- probably caused by me, but machine was put straight to work as soon as it was running
1 a few usb dropouts
2 cant for the life of me get soft limits to work running a program- but work perfectly jogging to maximum extents on my machine
i have noted this to balaz
not really worried about soft limits as i have bomb proof hard limits set up (double stage via servos and if all else fails servo estop)
waiting for them to release the 5 port version of ethernet controller and il refit another 3 of my machines. (shame they wont run old analogue servos as buying new derives/motors for the others is way out of my budget!)
i have scripted a really nice tool height probe and just ready to add my 9hp hsd atc spindle and rotary carousel stolen off a biesse. hopefully get time to do this over next couple of weeks
happy cutting to everyone in 2016 . benny

[vmax549]

I think A new Step/Dir to analog interface board would be a great Idea and would be a good seller. I would not build it into the UC300 but as an outside option to retrofit older machines with analog drives. There have been several addon boards over the years so we know the idea is doable. Seeing how CNCdrives builds servo drives I am SURE they understand the process involved.

(;-) TP

[ger21]

There's already one available. The Step2Linear
Viper Servo motor products

[battwell]

thanks ger. hadnt seen that one before!
know anybody thats tried it?