Hello Guest it is June 21, 2024, 08:15:20 AM

Author Topic: Opinions on use of ancient computers to run Mach3  (Read 9513 times)

0 Members and 1 Guest are viewing this topic.

Re: Opinions on use of ancient computers to run Mach3
« Reply #10 on: December 21, 2008, 11:03:58 AM »
Two points.

1/ It is stupid to even consider power saving on a PC that is controlling a machine tool complete with spindle motors and steppers / servos and associated power supplies.

2/ The modern Core 2 Quad, eg Q6600 series, are remarkable for anyone coming from Pentium4, in every way, though completely unnecessary unless your workshop PC also does duty as your CAD/CAM pc.

The thing that we are really talking about here is the actual electronic component and PCB construction quality of older pukka kit versus modern consumer grade kit, and there really is very little comparison.

On a closely related subject
this is the guy who tests actual PC PSUs...

I can't tell you how many transient computer errors we cured over the years by simply swapping out apparently working within parameters at first glace cheapo PSU's.

Offline simpson36

  •  1,369 1,369
Re: Opinions on use of ancient computers to run Mach3
« Reply #11 on: December 21, 2008, 01:51:43 PM »
Sooner or later, all CPUs, even low level ones suitable for running Mach (if indeed it stays in its current incarnation) will be 'green' as will probably everything that runs on electrons.

The question is not the stupidity of buying a fast compter for Mach, although I suspect that many (including myself) use the same computer for CAD/CAM and then also run the output thru LazyCam, or some post processor and then test the resulting Gcode on a copy of Mach running on the same machine.

I only send the finished and tested Gcode to whatever computer is actually running the machine tool.

It is not a matter of power supplies, component cost or anyone's definition of stupidity. I used Core 2 only as an example of the new 'green' technology that is finding its way into processors. The question is if the new CPUs 'throttle' and if so do they do it internally? Do they do it automatically? In such an instance, would it disrupt a running low level Mach process? If this feature is built in, can it be externally controlled or disabled?

Anyone with the knowledge to shed light on those questions could make a useful contribution to the community.
Re: Opinions on use of ancient computers to run Mach3
« Reply #12 on: December 21, 2008, 04:02:14 PM »
I don't know the ins and outs of Mach3 code, but I do know it isn't "just another app" sat on top of the OS, instead it uses in effect fake realtime code, which is going to bypass any BIOS aware OS code that is for example stepping the CPU clock.

But, all these "green" features are simple to disable in BIOS anyway in every desktop I've ever seen.

I just can't get my head around why people would (given the effective free cost of an old desktop or server box) insist on using a laptop though, which would have a far greater level of such tricks to be disabled, even assuming all of them could....

for example good luck downloading and flashing the "OEM" BIOS onto a dell laptop (instead of the dell BIOS)

Mind you, I have had this (I/O flexibility) conversations with vaxen / PDP-11 owners, who just couldn't understand how our megahertz cpus with astonishing 4 megs of RAM couldn't hold a candle to what their PDPs could do in a few k....

Re: Opinions on use of ancient computers to run Mach3
« Reply #13 on: December 21, 2008, 09:26:19 PM »
There is an answer to these issues.

1. Yes most modern CPUs have speed/power throttling built in.
2. there is a fix for many (if not most) boards/Cpus.
3. I found this solution by searchig this forum but the pointer was old and the program and name have changed. Thanks to the original poster whose name I don't recall.

the solution is;
"There is a fix that works well for many Laptop users and some desktop users.

This is available here;
download the "RMClock Utility".
This enables all the background power save and CPU slowing etc etc functions to be turned off."

I used this on a 1 year old dual core desktop which had a messy driver test and varying speeds. This cleaned the driver test right up!

« Last Edit: December 21, 2008, 09:28:42 PM by tech.tron »