[ozzie34231]

I have a Shizuoka CNC built in the mid 70s. I recently adjusted all of the jibs because I had a lot of vibration when making a large cut.
The machine has a 4 3/8" spindle, 3 HP Knee mill.
As a test I have cut some unknown steel with a 7/8" 2 flute HSS mill, .375" deep, full width, 200 RPM, 2 IPM. I don't think the steel very hard.
The chips came bright blue and I had a small amount of vibration.
What should the capability of this machine be ?
Ozzie

[underthetire]

How did you adjust the gibs? Is this a Shizouka ANS ? And why such a large 2 flute? If it's just an old knee Shizouka Bridgeport copy, can't expect machining center cuts.

Sent from my A3-A20FHD using Tapatalk

[ozzie34231]

This is not a copy. This is a quality machine built in Japan and then shipped to Hauppaque NY to be fitted with CNC electronics for a company "Autonumerics". Though these facts are known and documented elsewhere, the machine has no Shizuoka identification on it, and that company has no record of those early machines. The Autonumerics people held some of the very early CNC patents.
I used that particular mill because I thought it would be a tough cut for the machine. I believe the basic mechanics of the machine remained the same for many years; as to weight, size, power.
The Gibs were adjusted in the normal manner.
Ozzie

[Muzzer]

Have you seen Yeltrow's blog? He (like me) has a Shiz CNC and is restoring it.

I got a parts list with my machine although it's not the right one for my machine which is an unusual "square head" version of the AN-S. I scanned it and you can download it from his website.

These seem to be very well built and if they are in good condition it seems they can cut some serious swarf. From what I understand they were shipped as manual machines and converted by a variety of companies. They came from the factory with NSK ball screws and the accuracy report for my machine makes interesting reading.

This is mine (6000 lbs of it)

Murray
Attachment 340768

[underthetire]

This is not a copy. This is a quality machine built in Japan and then shipped to Hauppaque NY to be fitted with CNC electronics for a company "Autonumerics". Though these facts are known and documented elsewhere, the machine has no Shizuoka identification on it, and that company has no record of those early machines. The Autonumerics people held some of the very early CNC patents.
I used that particular mill because I thought it would be a tough cut for the machine. I believe the basic mechanics of the machine remained the same for many years; as to weight, size, power.
The Gibs were adjusted in the normal manner.
Ozzie

I know what a Shizouka is. I worked for the distributor for years. They had a manual only mill, the size and build of a Bridgeport, and yes IT WAS A COPY! Then they came out with the ANS, which was retrofitted by dozens of companies with multiple controls. And what do you consider "normal" and how much indicator movement are you getting on side load.

Sent from my A3-A20FHD using Tapatalk

[ozzie34231]

I remembered that I have the accuracy test sheets for the machine.
The model is an ST-N tested 1976. Machine 56021
Underthetire, the company name is Shizuoka, U before the O. Why would they put 1 3/32" ball screws on a manual machine? Anyway I don't really care about all that, I am just inquiring about the size cuts this machine would be expected to handle, and if under the circumstances I enumerated would a slight shaking be normal, or should I look for additional adjusting.
Muzzer, your machine has more bulk than mine. When I acquired it the dealer said it weighed about 4000 lbs. Do you have any idea what vintage your machine is?
I adjusted the Gibs by first loosening the lock screws, then tightening the adjustment until the table locked, then backed off to a snug setting. Then I locked again. Using dial indicators I cannot see any deflection when I push or pull, and that is true for X,Y, and knee. I attribute that to my just not being strong enough, I'm 78. Under no load, my servos draw just under one amp.
There are two places I'm not sure about. The two knee Gibs appear to only work against movement in and out, i.e. rotation about the X axis. I don't see how rotation about the y axis is adjusted. Perhaps it was expected that the knee always be locked when milling?
The other thing is that the Y axis has box ways and although I adjusted for side to side play, I did not lay under the machine to adjust for up and down play, if there is any.
Thanks,
Ozzie

[underthetire]

Thats why I was trying to get a model number! They didn't just make knee mills, and they did make manual machines as well and were still producing machines till the 2000's.

Yes, on that machine slight shaking would be normal with that size cutter in that material. Try a 4 flute, and i think you'll find the harmonics get better for that machine frame. I think that was a pre- ANS model.

[Muzzer]

I remembered that I have the accuracy test sheets for the machine.
The model is an ST-N tested 1976. Machine 56021
Underthetire, the company name is Shizuoka, U before the O. Why would they put 1 3/32" ball screws on a manual machine? Anyway I don't really care about all that, I am just inquiring about the size cuts this machine would be expected to handle, and if under the circumstances I enumerated would a slight shaking be normal, or should I look for additional adjusting.
Muzzer, your machine has more bulk than mine. When I acquired it the dealer said it weighed about 4000 lbs. Do you have any idea what vintage your machine is?
I adjusted the Gibs by first loosening the lock screws, then tightening the adjustment until the table locked, then backed off to a snug setting. Then I locked again. Using dial indicators I cannot see any deflection when I push or pull, and that is true for X,Y, and knee. I attribute that to my just not being strong enough, I'm 78. Under no load, my servos draw just under one amp.
There are two places I'm not sure about. The two knee Gibs appear to only work against movement in and out, i.e. rotation about the X axis. I don't see how rotation about the y axis is adjusted. Perhaps it was expected that the knee always be locked when milling?
The other thing is that the Y axis has box ways and although I adjusted for side to side play, I did not lay under the machine to adjust for up and down play, if there is any.
Thanks,
Ozzie

My machine left the factory in Shizuoka (a city in Japan) in May 1982, according to the inspection report. It's a model AN-SB, s/n 63008. My ballscrews are metric (for the UK market).

Have you looked at the parts list on Yeltrow's site? Although it may not be the right one for your machine, the machines seem to share a lot of common construction features.

One key difference between my AN-SB and the other AN-S machines is that my ram is bolted solidly to the main body of the machine. That must make it more rigid but is less adjustable which may be why they stopped doing it. Another is that it has fixed ratio pulleys (2 ratios, solenoid controlled) - different speeds are provided by a Hitachi VFD and a motor designed for proper variable speed operation giving 10-6000rpm. It also has a deeper table with 4 slots. The original sales invoice for the finished CNC machine shows a price of $125k in today's money. I have scoured the entire internet inside out and I have only found one other machine like mine (also in the UK). It's been converted back to manual, so mine may now be unique.

Shizuoka still manufacture machine tools, both manual and CNC. Beyond my budget but still very desirable looking machines.

Murray

[ozzie34231]

Yes I've looked at all the stuff on the Yeltrow site.
My machine had been used manually, stripped of all its electronics. The Z was controlled only at the knee, there was no way to move the quill. It certainly was built to be CNC, there are no rack teeth machined into the quill. I have returned it to CNC, servos, 40,000 steps per inch, DuGong type drivers, ESS motion device, Mach 4, VFD.
Ozzie

[Muzzer]

That's interesting. My machine has the original 1980s controller and was apparently in working condition when last used 10 years ago before going into storage when the guy retired. If I can persuade it to work, I'd like to use the existing DC servo drives but replace the controller with a Chinese Fanuc sort-of-clone (about £500 / $600). The G codes in the existing controller are quite different to those we use today, despite the fact that G codes were defined some years before (1960s / 70s?). If that fails and I have to scrap the original electronics, I will need to replace the servo drives ("servo amps") with something more modern (retaining the existing DC motors if possible) although most servo drives these days are designed for AC motors. The DuGong system was one of the few available (and affordable) DC servo drives I was able to find. Sound Chinese although the company is based in Hungary?

Are you using the braking module as well?

Yesterday I finally managed to track down the previous owner who actually bought it new in 1983. He's retired now but was able to tell me all about it. It's only ever been used with neat mineral oil, so there is almost no signs of rust and he was a precision engineer / die maker who took great care with it. I think I've been very lucky. I'd been hoping he might have had some documentation for the electronics but it seems I have all the info he ever got for it.

Murray

[ozzie34231]

Hi Murray,
Are you sure the motors are servos, not steppers? I think steppers would have been more common for the time.
If you get the old electronics running, how will you generate G-code for an obsolete system?
I'm not using braking modules.
It sounds like your machine was a super find.
Ozzie

[Muzzer]

No they are definitely servos - "SEM" DC brushed motors with both tacho and encoder. Seems the US conversions were mostly stepper based but the UK Matchmaker systems were servo. Runs an 80V bus, so the DuGong system sounds compatible.

Any g-code would be read straight from the manual! I'm not going to learn an obsolete version and there's no chance of a post processor. This would be a very short term solution just to show it works before the controller is swapped out.

If I get the old system running it is just to show that the electronics work. But the controller itself will have to go, no matter what happens. That should be easy enough though if it still works, as the servo system actually uses step / dir between the controller box and the servo system. Resolution is 10um (0.4 thousandths), so should be a fine beast when it's working.

[m_c]

I've got an ST-N which has turned into a bit of a long term retrofit project.

I bought mine from somebody who had bought it for the sole purpose of getting the Heidenhain TNC? controller as a spare from it, along with the SEM servos, however the Heindehain had been retrofitted by a UK company who's name I've forgotten.
As I've been working on mine, I've found remnants/evidence that it originally had a stepper motor system fitted. Things like with the X-axis servo mount covers removed, what would of originally been a spacer/mounting block for a stepper motor/flexible coupling, now has an extra bearing and shaft extension bolted to it, with an unpainted square behind the belt pulley where a stepper motor would of been fitted. And the additional barrier terminal board, that was buried in the rear of the machine behind a control box that had been added to house the Heidenhain controller, while the original large side mounted control cabinet was left to only contain the servo drives/transformers.

Some photos of mine can be found at https://www.flickr.com/photos/mc_mtb...57632396902054

[Muzzer]

M_C Yes, I saw a post you made about yours some time back. Although a Matchmaker, it's clear your system is quite different to mine. This is what mine currently looks like:


I managed to get in touch with one of the 2 guys who built most of these machines at Matchmaker in the 1980s. He's currently running his own CNC machine tool business in Oz, having emigrated there a few years back. We was able to send me a user manual for the "Digiplan PWM4" servo amps (but no circuit diagrams). I found another manual for the "servo interface module" from the Parker website. Although not my actual model, it is pretty much identical. It converts the digital signals from the controller (step / dir) and the encoder and outputs an analogue /-10V signal to the servo.

Seems the later Matchmaker machines used Digiplan (Parker) drives and a Micon controller with SEM servos. Later they changed to another system I believe. Dave in Oz told me "The square head version was a later model than the rounded version. At best I’m guessing there were twice as many of the rounded version as these were originally fitted with POSIDATA 1800 or 2800 control systems and MICON retrofitted a lot of the rounded version with the MICON controller and later to the square head version". I don't know anything about the Posidata systems - were they stepper based?

Some of these Matchmaker machines were subsequently retrofitted by Eagland Machine Tools with the ah-ha systems. They are no longer in business but the website has some files and documents. I contacted Ian Eagland (now retired) and he told me that he actually has a machine identical to mine (perhaps the only one!) that he converted to an AC servo system, starting with a bare machine. "As a retrofit company we ended up removing Micon controls. We were Ahha agents but they are no longer in business."

[Muzzer]

.

[m_c]

That control cabinet looks familiar, but the contents certainly differ.

I can't remember the name of the company that retrofitted mine. I want to say Pendle, but I know it's not. It's certainly something that sounds like a Cumbrian town name, and I have in the past looked at their website.

[Muzzer]

You can see from my photo that the old Hitachi VFD (black), braking module (white, next to it), servo amps (above the braking module) and the servo interface module (left of that) take up much of the space. The 3-phase transformer is massive, partly because the VFD takes its power through it (220V, down from 415V) and partly because the servos come through it (3 x 1kW plus losses). If you then remove the 2 boards at the top (opto isolators, relays etc for the IO), there's not a lot left.

Running a modern compact VFD (I have a nice new Yaskawa V1000 with braking resistor) and a switched mode PSU for the servos (ditto some Meanwell servo PSUs) would free up much of the space. Modern controllers have much of the IO built in already and the likes of the DuGong servos need very little to support them. Shouldn't be too much of a job to fit it all into a much smaller cabinet. The original cabinet is almost big enough to fit a hammock - which may be handy if my wife catches me buying stuff like servos, controllers etc.

Hope to have it running by the summer, although I haven't narrowed down which year yet .

[ozzie34231]

My back gear is stripped, (plastic), and I did the cut I described by slowing the spindle from 650 RPM to 200 RPM using my very cheap Chinese VFD, worked great.
However if I try to slow the spindle too quickly, the VFD faults out. Any idea of which of the dozens of settings I should look at.
Ozzie

[Muzzer]

Seems to be a thread here on this forum.

The variable you need to change will be called "deceleration time". That seems to be variable PD15 and this guy suggests setting it to 5 seconds.

The optimum time will depend on the moment of inertia of your spindle. You want it to come to a stop reasonably quickly but you want to avoid tripping. Experimentation may be required.

Better quality VFDs will have usable DC injection (the Huanyang only works below 10Hz) and provision for a braking resistor. Depends if you can live with the decel time you end up with.

Doesn't your Shiz have a friction brake? Most of the examples I have seem seem to have one operated by air or solenoid(?).

[ozzie34231]

Yes, it has a brake. The problem is not when stopping but decelerating from 100% down to maybe 25%. If I turn the pot too quickly the unit faults. I then must shut it down, wait for the display to go away and then start up. There is a reset terminal but I can't get that to work.
I got this a couple years ago and was happy to just get it running, the total Chinglish is tough. I suppose it's time to get back to the manual and see what I can find.
I'm thinking about raising the frequency limit from 60 to perhaps 75-90. I've read conflicting advise on over-speeding the spindle, some saying that running at 150% should be no problem, bearing in mind that this is a hobby application, no long runs, and that it doesn't cause any heat. I've run the spindle at 3800 for as much as 30 minutes and it was cool to touch. I'm thinking to try higher speeds gradually.
Ozzie