compumill 4000 retrofit
Dynapath parts are way too high for their age on ebay. after dealing parts and smarts we decided they would be the only people we would trust to buy parts from for the dynapath, but they are just as if not more expensive. Their experience with the controllers is invaluable for guaranteed repairs and replacements.The processor itself is worth more than its weight in gold as there is no suitable replacement for it. That was the deciding factor for us to deal with an immediate upgrade.
Please Log in or Create an account to join the conversation.
I take it that your going to keep the power supply, D.C. motors and servos right?
I have no idea how your going to use this P.C motherboard to run your machine.
Details please?
Please Log in or Create an account to join the conversation.
We will not be using the dynapath power supply for the upgrade as it is very expensive to replace and well its with the dynapath control which we returned to the machine seller. There are lots of manufacturers of power supplies that are way cheaper to buy and replace than the deltron on that dynapath which is bad. This power supply has a few pots on it which have been adjusted to lower resistance of the outputs to increase the 5v output ( actually only putting out right at 4v if the others were in spec) which was getting weak as it aged. Thats why the 24v and +-15v are so far out of whack. We will be keeping the existing baldor power supply , servo drives, as well as the existing servos and encoders.
The only thing I will replace of the existing system are the omron relays on the PIC interface board that are 24v with the same type but 12v., which allows us to use a more generic power supply.
All in all its everything the dynapath is, just a bit more advanced, as we can add features such as the spindle VFD which the dynapath cant handle without modifying its software and probably hardware as well as limited memory. We are setting up this machine to run some complex 3d profiling which the dynapath can do by drip feeding, but as we have yet to hear back about our control Im worried they just are not going to return it to us as its problems and replacement parts exceed the price paid to the merchant for the entire machine. At this point we just want the dynapath back working so we can test all mechanical systems. If I have to pull everything loose from the controls and out of the panduit just to test Im not putting a 23 yr old controller back on it. Plus If we find the spindle is shot like on the first machine we purchased its being loaded back on a truck and sent back to the merchant. So I do not want to modify it or really tear into it yet.
I first Installed EMC2 5 months ago on a linux pc and have been working my way through the manuals since. We were going to build a small CNC engraver first to get the hang of it, but thats now on hold even though we need it as our parts must be serial numbered and logged for tracking by the ATF. Id rather be producing parts, but thats not going to happen anytime soon.
I can tell you that this is not as simple as just buying a new control system and connecting it all together, but you will get a better understanding of a machine and how it interfaces with its control, and software. The one thing Ive found is there is better support for EMC2 than a lot of control software because its open source. Its an entire community interacting and working together.
Please Log in or Create an account to join the conversation.
I can tell you that this is not as simple as just buying a new control system and connecting it all together, but you will get a better understanding of a machine and how it interfaces with its control, and software. The one thing Ive found is there is better support for EMC2 than a lot of control software because its open source. Its an entire community interacting and working together.
That is the truth, Rob H from England stayed up real late many nights trying to help me get my Hardinge CHNC original controls up and running via the IRC. Once I made the decision that the only way was to convert to EMC2 the software choice was simple... then the research for the hardware. I screwed up by trying to get a few older computer motherboards to work and wasted a couple of weeks playing with that. All in all I'm glad that I ended up with a good CNC lathe powered by EMC2. I wished I had known about the D510MO at the time as that would have made the PC install much easier.
John
Please Log in or Create an account to join the conversation.
Also if you completely replace the controller and use software only "which is what I would rather do" Then you will also lose to tool cal soft key and many other functions. So I'm thinking that you must have a program to install on a hard drive for that new P.C. board. The other thing is why would you want to use a mother board that is much weaker than other boards that are out there. I would use the best mother board I could. The mother boards from the HP XW8600 are the fastest ones ever made "as far as I know". I do know that they were the fastest in the world not so long ago.
This sounds like a cool project and a very good learning experience.
Please Log in or Create an account to join the conversation.
Raw CPU performance is less important to EMC2 than latency (how quickly the CPU responds to real time requests)
Many very fast systems have poor latency.
the D510 Mini ITX motherboards have known good latency and adequate performance, which is why they are suggested as EMC2 platforms.
Please Log in or Create an account to join the conversation.
It's only software and electronics, it should be perfectly possible to add such things back in to the EMC2 config.Also if you completely replace the controller and use software only "which is what I would rather do" Then you will also lose to tool cal soft key and many other functions.
As it happens, those little, cheap, Atom boards have about the best latency that we have seen. CNC is not at all demanding of raw computing power, they could do it in the 60s and the rate at which you can machine metal has probably no more than doubled since then. What matters most is that when EMC2 wants some CPU time it gets it exactly then. The D510MO has a dither of 5000nS or so on the realtime threads. That's about as good as it gets in a PC board. Some high-performance boards can be 300,000 or more. They are still very fast boards, but they do too much clever stuff that gets in the way of our very basic and simple requirements.The other thing is why would you want to use a mother board that is much weaker than other boards that are out there.
Please Log in or Create an account to join the conversation.
Please Log in or Create an account to join the conversation.
128 GIGs of RAM Memory. He swore I must be wrong on that and thought I was confused with how many GIGs the hard drive would hold. He was wrong.
Anyways the board you guys are talking about is probably good enough.
I guess the real question is how do you connect all the wires that went to the controller to the new P.C.???
Thanks
Please Log in or Create an account to join the conversation.
I guess the real question is how do you connect all the wires that went to the controller to the new P.C.???
Maxxed-out a Mesa 5i22 full of 7i64 boards has 1536 pins of GPIO. Is that enough
Typically you wouldn't do it that way, the 72 IO pins on the 5i23 without all the peripheral cards is plenty.
(Other options exist, from Pico, Motenc, Pluto…..)
Please Log in or Create an account to join the conversation.