RYE QUANTUM - Omicron 3000 Controller SWAPOUT OPTIONS

[BEACHBLANKET]

OLD THREAD - http://www.mycncuk.com/threads/13855-Rye-Quantum-First-Start-Drive-Fault/page6

Ok having finally been able to confirm all servos and hardware are working I'm looking for advice on whether or not its possible to just swap out the ageing controller on this thing or will I need to replace drives/electronics and controller.
I look forward to your comments...(I think!!!)...Thanks in advance...:encouragement:

Spec:- 3 PHASE

Servos - SOELMA MP48 DC12.3V

Servo drives - AXOR - MiniSpeed 200

Encoders - SOELMA CK58-L-1000ZCV18

Inverter - Control Techniques Commander CDE AC DRIVE

Spindle - PERSKE KRS51 14-2-D

Controller - AMC OMICRON 3000

Pics to follow...
28681
286722867328674286752867628677286782867928680

video...

https://photos.app.goo.gl/VMT1Y9Wx2Nmg9bnT6

[BEACHBLANKET]

I was thinking maybe something along the lines of UCCNC?

Sent from my Mi A1 using Tapatalk

[JAZZCNC]

I was thinking maybe something along the lines of UCCNC?

Sent from my Mi A1 using Tapatalk

It depends on the drives and if they have Step & Dir control inputs. If not then you can't use any of the UC line of controllers because they supply Step & Dir signals only.
Most older machines which use DC motors are +/-10v analog control inputs and often they will have Tacho's or resolvers on the motors, thou in your case it looks like you have encoders which makes things easier because worst case scenario you could change the drives for DC drives with Step & Dir signals. Thou it does depend on the encoders as well so you'll need to check them out.
CNC Drive who make the UC cards also make DC drives with Step & Dir.

You can get Motion controllers which provide +/-10V Analog output but they tend to be expensive and can be a bare to setup compared to step & dir.!

If the existing drives do have Step & Dir Inputs then it's a Doddle to convert with just a UC300 and Decent BOB that provides Spindle control and away you go.

Edit: But I'll say it again because it shouldn't be underestimated how much work is involved.!! . . . . You will have to setup and write your own ATC code or code to control dual spindles plus pneumatics etc. And this machine seems to have a few bells n whistles on it.!

[BEACHBLANKET]

Thanks JAZZCNC.

I'll check out the encoders and drives in more detail.

The one thing I would probably change is the pneumatic head change.

One head will more than suffice so it could be fixed directly to the Z negating the pneumatics etc.

[BEACHBLANKET]

Existing AXOR - MiniSpeed 200 Drives spec

Tachogenerator feedback ✔
Armature feedback ✔
Torque mode feedback ✔
SMT technology ✔
IGBT or POWER MOSFET bridge ✔
✔ Encoder feedback
✔ PWM + Dir. feedback

So the drives are PWM and not step and direct... there is literally nothing I can find on the encoders...

Due to my lack of CNC and machine code experience...Do you think I should really be looking at a full swap out i.e Controller/drives/motors?

[JAZZCNC]

So the drives are PWM and not step and direct... there is literally nothing I can find on the encoders...

Due to my lack of CNC and machine code experience...Do you think I should really be looking at a full swap out i.e Controller/drives/motors?

Ok well, I don't have any experience of controlling drives with PWM but my limited understanding is it's very much like +/-0-10v and while many controllers have PWM outputs I don't think they will be accurate or fast enough for this application.
However, I do know MESA who make cards that work with LinuxCNC do a card that will output PWM to do this but trust me it's a massive learning curve to undertake and I wouldn't advise it for a new user.

Regards full swap out then I suppose it depends on your budget and what your plans are for the machine.?
If I was using this machine as part of my business then it would be a full swap out for reliability reasons and easy access to replacement parts if it went down.

If it was for semi business that isn't reliant on the machine then I'd probably just swap out the controller and drives for Step + Dir DC drives and reuse the old motors.

If it was solely for playing around with then I'd probably look at LinuxCNC and MESA cards and reuse everything but the old controller.

Regards the Encoders the numbers point to it being 1000 line encoder but whether or not it'ss single-ended or Differential or not I couldn't say but it's easy to find out. Just count the number of wires coming out of the encoder. Single-ended will have 5 wires, Differential will have 8 wires.

Edit: You could also always go with a controller that outputs +/-0-10v of which there are a few. Cslabs IP-A or the ET10 from PV-automation and there are others.
Also again MESA makes +/-10v cards that will work nicely but again you'll have a learning curve because while LinuxCNC and MESA are super powerful they are unforgiving beasts and not very intuitive to set up for a beginner.
To give an example I'm very experienced but I've just built a plasma table that uses MESA - Linux with servos and I don't mind admitting I struggled massively at first and needed a lot of encouragement from an experienced Linux user (Clives S) to stop me from smashing the F@~#ing thing to pieces. Great when it's working but a B'##'rd to get working correctly.

[BEACHBLANKET]

Yes I looked at LinuxCNC after Routerdriver mentioned it...daunting!

I bought the machine for £500 and went 50/50 with my brother in law, my initial idea being that it provided an excellent mechanical base to start from even if everything electrical needed replacing.

We are looking to start a new business and this machine will form some small part of that business, mainly 2d cutting sheet ply, along with hobby stuff (small 3d stuff, photo carving, sculpture etc. hence the interest in the Vectric Software)

I'm a bit of a recycle reuse nut so the idea of being able to upgrade as little as possible is the most appealing. However given that it's taken me nearly 100 hours to figure out the existing controller/head/software/post processor setup I need to be realistic with time in order to get this up and running efficiently.

I think budget wise, at this point in time we could probably put in another £600 or so what would that get us?

I can't find anything on the encoders in the schematic drawings I have so I'll check wires tomorrow when I'm back with machine.

[BEACHBLANKET]

Just reading your conversion thread using the SZGH controller etc...

I'm guessing realistically we are looking at changing the encoders, drives and controller...

[routerdriver]

I hope it doesn't come as too much of a shock but the comments about the amount of work involved in post #3 will apply no matter what control system you change to.As it is you have a machine with a serious,but dated,integral controller and it sort of works.The alternatives involve running a PC front end to achieve the same performance and a mountain of work to get to where you are now in terms of producing parts.

You also need to get a lot more familiar with both CAD and CAM software in order to generate the Gcode.If there is any kind of business imperative requiring parts to go out of the door and bring in an income it may all take quite a while.Unless you get going with whats there and learn more about the upgrade path in the background i.e. evenings and weekends so a quick and effective swap can take place while still producing during the day,it could be s long time before you make any money.

All CAM software has the same requirements,a series of connected elements for outlines holes located where required and a list of tools with which to cut the parts.You may have to set the tools accurately in their collets and model them precisely in a tool library.Then you tell the program which of the tools will be required for each operation,which side of the line it has to go and how deep it has to cut.At which point you post process the information and send it to the machine.I don't know if you have to draw within AlphaCAM or if you are importing from other software,to an extent it doesn't matter,just as long as complete information is present.The difficulty you seem to be having is setting the part datum relative to the home position,otherwise your dated software seems to be working.Have you compared an AlphaCAM upgrade to the cost of the alternative hardware and the extensive learning and development process?

[JAZZCNC]

Ok well you have bought it right and it would be a shame to spoil it by cutting corners because you'll never build anything like it for £500, it wouldn't even buy the steel.!

Now regards your intention to use it for a business then I can tell you now that you'll use this machine far more than you ever think you will and in ways you probably haven't imagined yet.!
I see this happen all the time with machines I build, people start off with an intention to do this and that on a limited basis and very quickly they are going off at tangents cutting things they never dreamt about then next thing the machine is flat out busy and they want another or bigger machine.
So what I'm saying is invest now and get it running properly because it will cost more later when it's holding you back or breaking down which is the worst thing you need when orders are stacking up.!

£600 will just about get you an Analog controller which would get you going with the existing drives etc.

For a complete refit if you buy Servo Drives and motors from China then you'll just about do it for £2k for a full refit with controller etc.

But even then if you had £3k in the machine fully fitted you have a very cheap machine and it's the route I would take if using for business, even a small business.

[BEACHBLANKET]

Ok chaps points noted.

In its current setup the main problems I've had are...
1 - understanding the controller
2 - getting a post processor for the machine/software combo

Now after much help from this forum...

3 - I now understand the controller
4 - I've now modified a post processor from which I'm going to trial tomorrow.

As this is my first CNC machine (and was a punt tbh) I think Routerdriver is right in that I should use this as a starter/learning platform and see where it goes.

Software wise I can generate all of my design/CAD info in software I already own, it is literally just the tooling/toolpathing and NC code that I need Alphacam for.

Yes Jazzcnc it is built as they say!!
It's nearly 3 ton that's what appealed to me initially, a solid foundation.

Then longer term if and when a business need does start to develop I can press the button on upgrades which have been carefully researched and considered.

There's no pressing need to be fully operational at the moment, but I would like to get a plan or direction in place should the need arise.





Sent from my Mi A1 using Tapatalk

[BEACHBLANKET]

Thank you both, for your opinions...

Sent from my Mi A1 using Tapatalk

[BEACHBLANKET]

Just a quick update...

After modifying the post processor it now starts and stops and cuts where it should at the correct height. :toot:

https://photos.app.goo.gl/GfUPKpiU7RNKK4fXA

(Only test cutting at -1mm so extraction was removed)

So for now, I'm a happy chappy.

I've also worked out how to drip feed code to it, so nearly 10 minutes of code streamed consistently no problem.

Obviously, taking onboard JAZZCNC's comments I'm aware that this may be a ticking time bomb so I would still really appreciate any thoughts/comments on controller/motor/electronics options.

As I have no clue what's good/bad general brand/manufacturer guidance would be a massive help...I can then research in more depth... Many thanks.

[routerdriver]

It seems like you have cleared a major hurdle.If it cuts what it should and nothing else,you have cracked a large part of the operation.I'm a bit puzzled about the need to drip feed the program.Is there insufficient memory within the controller or do you not have the ideal comms package?I'm wracking my brain for the name of some of the programs that were used for sending large plot files to pen plotters in the early nineties.I also expect you may be finding the baud rate verging on neanderthal.None of which really matters if the machine produces what you intend it to.I have mentioned on this forum and elsewhere that I have a German friend who programs a very rugged vintage CNC mill using DOS and it will probably function for the rest of his working life unless a lightning strike fries the electronics.Older industrial quality controls can live very long lives but depending on this happening without some kind of contingency plan could be very risky.

I watched the linked video and it looks like you now have a spoilboard with several grooves in it.If there is an integral vacuum system you will need a new spoilboard before doing production work.For the time being could I suggest you simply(!) cut a few basic shapes to become more familiar with all aspects of the operation-from generating the code to setting datums to creating parts.Something like a small rack for the workshop to hold the stuff related to the machine might be useful.Meanwhile,keep doing your homework on modern alternatives but don't lose sight of the amount of work in successfully implementing them.

[BEACHBLANKET]

Hi Routerdriver, The memory on the controller is only 1mb... The file I sent was a photo vcarve so the machine had lots to do!!
I drip fed it from Alphaedit which comes with Alphacam, managed it without fail.

I'm using a usb to rs232 which to be fair for £12 I'm really pleased with as it allows me to use a modern laptop without searching for one with a serial port.

Unfortunately the vacuum pump was not supplied with the machine, it had been removed prior to sale... I may look at getting another...early days yet...

I'm a long way off trying to make any money with this, but Iam enjoying seeing what it can do and working out what it can't!!


Sent from my Mi A1 using Tapatalk

[routerdriver]

Great to learn that yet more progress is taking place.Have you tried a job with a tool change yet?Once you can clear that hurdle a lot of possibilities open up,although I have used an initial program and called up a series of sub-routines using different tools and the original part location offset and it worked very well as long as the machine was homed before each run.

Drip feeding longer sequences gets you past the limitation of the modest amount of memory and it might be worth keeping an eye on ebay in case any memory modules turn up.I wouldn't touch the controller at this stage but having a bit more memory on the shelf sort of guarantees that it won't be needed.Likewise a high flow vacuum pump might turn up and is a very useful thing to have on hand.

Please keep the progress reports flowing.

[JAZZCNC]

While it's all very well while playing and learning the machine with this controller I'm not sure I'd want to be using it for any real work where time is money or mission-critical work using expense sheets of material because drip feeding is going to bite you sooner or later.!
It doesn't take many dropouts and wrecked sheets at £60-100 a pop and that's not to mention wasted time to recover the costs of an upgrade. Then when you factor in the potential upgrade in performance I'm not sure I'd be wasting any effort struggling to learn this controller or its ancient software.!

Biting the bullet and paying for upgrading sooner rather than later will pay for its self within a very short time if you have the work for it.

[BEACHBLANKET]

I've not used the machine for a week or so I'm waiting on a guide to arrive.
Having now greased everything I found one of the guide dust covers had failed/was damaged so still waiting for delivery before doing attempting anything else.

Jazzcnc I'm a long way off using this for 'paid' work... It's purely me experimenting/playing around and getting to grips with the CNC environment, Fusion 360 next!

I wholly take on board what your saying and would really appreciate any suggestions you have with regards to possible upgrades.

I don't have any paid work to feed the machine at the moment, I'm looking at December at the earliest before that happens and then it will only be part time.





Sent from my Mi A1 using Tapatalk

[BEACHBLANKET]

Routerdriver...
Yes more memory would be great I'll keep my eye open.

I've successfully used fusion 360 to design then output NC code into Alphacam/Alphaedit and post processed it to the machine with no issues at the moment.

Obviously I'm still doing very basic stuff... But I'm making progress[emoji106]

Sent from my Mi A1 using Tapatalk

[finski84]

Just popping my head in to see how your getting on?

[BEACHBLANKET]

Just popping my head in to see how your getting on?Hi Finski,
Thanks for the message. Unfortunately not had much time to progress things due to illness... Hopefully things will get back on track in the next couple of weeks and I'll post some progress pics/info. How's sleep going??? All good with you??


Sent from my Mi A1 using Tapatalk

[finski84]

Sorry to hear you've not been well. I hope your on the mend though.

Sleep is a rare thing but getting better :thumsup: Have been busy 2 other machines at the moment as our 3 axis router is inactive due to reduced work load from this covid-19.

28958
Finally got the thermwood up and running. Been a lengthy 6 months sorting this one out.

28959
Tinkering with the HAAS Mill is more fun though i will admit!