Samsung-Quad QSA30A servo amps

[jmelson]

So, adding to the long saga of keeping a Quad QSA30A P&P running for light duty, I ran some boards last month and the machine ran fine.  I fired it up last weekend, and had some PC hard drive boot-up issues, and then both of the XY servo amps failed to come up.  They get an error code on the single 7-segment LED of 3 horizontal bars.  Each of these amps failed with the same symptom about 5 months apart, about 2 years ago.  I got replacement used amps and got the machine running again.  I am guessing that the EPROM in the microcontroller has lost its contents, or at least failed the checksum.  When this last happened, at first the machine would not home properly, with jerky moves, but then after resetting the fault, it would run fine.  Later, it went completely dead with the 3 line symbol on the LED.  That does sound like a fading EPROM issue.

So, now the real issue:  Is it crazy to try to keep a 20+ year old machine running by replacing parts with 20+ year old used parts?
And, is it possible to replace just the Sanyo Denki servo amps with something newer?  I fear that new Sanyo Denki amps are going to cost more than I have invested in the whole machine.

I'm kind of disappointed with this whole mess.  This machine has been quite unreliable.  I first chalked that up to poor storage conditions after the machine was taken out of service.  But, now two more servo amps have died with the same symptom.  The outfits that sold me the spare servo amps don't seem to want to help, other than sell me two more at full (used) price.  I ran an even older Philips CSM84 (made by Yamaha) and it was insanely reliable (just not so accurate on placement).  I had one sensor die and an accumulation of commutator dust in one motor in 13+ years of use.  I don't know why this Samsung-built machine is just falling apart.
Suggestions welcome!
Jon

[jmelson]

Ugh!  And it gets worse.  The spares from a broker are both defective.  One shows the same failure as my previous bad units, the other shows some different failure code.
Jon

[MR]

how long has the spare been on the shelf 20 years, and where was that shelf - near the coastline in a humid area? Some ICs are specified for 10 years on my Mitsubishi Amplifier, they're 15 years overdue already. Certainly many parts exceed 10 years but if you want to build a business on it you need reliability of course...

When I recovered my CNC machine I also had to replace an amplifier, also one encoder and clean 2 encoders.
My fallback if more will break will be that I will rip out everything and replace it with Delta Servos.

[Kean]

That does sound like a fading EPROM issue.

I don't know anything about these particular servo amps, but the fault you describe sounds more like an analog circuit failure which the microcontroller is detecting.
(Actually no - see below, this is just the idle state display - again meaning the micro is starting properly, but maybe with lost settings?)

Have you tried any reverse engineering of them?  What inputs and outputs do they have?  Model number?  I'm guessing these are not the modern Ethernet connected style.

Maybe you can retro fit a more modern servo driver like the Gecko G320X
https://www.geckodrive.com/product/g320x-digital-servo-drive/

Those take the common step pulse and direction logic inputs used by DIY CNCs.  You'd obviously need to check motor and encoder compatibility.
If the Sanyo Denki units have a proprietary communications protocol, you'd need to hack something like grbl to simulate the protocol and provide the step/dir outputs.
And you'd fake any other outputs, like fault or ready signals.

A quick search and I found a manual for the Sanyo Denki PE series, which has a 7-seg display and uses pulse inputs (one for CW, one for CCW).
The 3 horizontal bars means "The bus power (R, S, and, T) is present and the start ready complete (SRDY) signal is ON."
So this isn't an error state, more of a idle state.

See page 92 and onwards for troubleshooting info.  Even if yours are a different model, they might be similar.
http://static6.arrow.com/aropdfconversion/92a3e2036951025af81eb72e9bbeb02c384191cc/molmanualpe1_05.pdf

[MR]

can you take some photos of the servos (possibly motor and encoder part numbers) and amplifiers? So we know what you're writing about?

[jmelson]

how long has the spare been on the shelf 20 years, and where was that shelf - near the coastline in a humid area? Some ICs are specified for 10 years on my Mitsubishi Amplifier, they're 15 years overdue already. Certainly many parts exceed 10 years but if you want to build a business on it you need reliability of course...

When I recovered my CNC machine I also had to replace an amplifier, also one encoder and clean 2 encoders.
My fallback if more will break will be that I will rip out everything and replace it with Delta Servos.

Well, any spares are likely over 20 years old (pulled from scrapped machines.)  Whole story, my machine was bought from auction traces in the hard drive indicate it was used up to 2013 or so, and then left off for 6-7 years before going into auction.  I got it in 2020, and it fired right up for a couple days.  I was still trying to get manuals.  Then, on day 3, it would not come out of E-stop.  I spent a month tracing signals all over the machine and finally decided the board that handled the Estop was bad and replaced it.  The machine had a BUNCH of other issues (wrong head replaced from another machine type) bad cameras, etc.  I got all that fixed and the machine ran well for 2 years, then the main XY servo amps started flaking out.  I replaced them, and all was well for another 2 years.  On April 28, I ran a batch of boards and the machine ran fine.  Then, I tried to start it up last weekend, there were computer hard drive issues, and then I got to coming out of E-stop and both servo amps were dead again.  I got replacements installed, but they were both dead.
My theory is that the EPROMS in the DSP micro in the servo amps have lost their contents.  Tomorrow I will call the US rep for these servo amps and see if there is a way to get them fixed.  If my theory is correct, they just need to reflash the CPU and it will work for another 20 years, but I don't know if they will do this on such old gear.
Jon

[jmelson]

can you take some photos of the servos (possibly motor and encoder part numbers) and amplifiers? So we know what you're writing about?

The motor labels are hidden inside castings.  The servo amps are Sanyo Denki PZ0A030A and PZ0A015A, made about 1999 or so.
Jon

[jmelson]

[Maybe you can retro fit a more modern servo driver like the Gecko G320X

These drive large 3-phase brushless motors.  I'm thinking they are "sensorless", meaning no Hall signals.  I believe the controller sends CW or CCW pulses, not step/direction.
The drives run off 200+ VAC, so I think anything from Gecko would be insufficient.

A quick search and I found a manual for the Sanyo Denki PE series, which has a 7-seg display and uses pulse inputs (one for CW, one for CCW).
The 3 horizontal bars means "The bus power (R, S, and, T) is present and the start ready complete (SRDY) signal is ON."
So this isn't an error state, more of a idle state.

I have never seen the 3 bars on a properly-working drive.  But, this is VERY interesting - thank for finding this manual!
When the drive is powered-on, the 7-seg display has a pattern rippling through all 7 segments going around a figure 8 about once a second.
Now, in fact, this 3 bars display MIGHT exist for a moment right after drive power comes on, I'd have to be in 2 places at once to see it.  And, MAYBE what this means is that the drive is not detecting the SRDY signal or refusing it due to some detected issue.  I DO know that the issue is in the controller board, not the power board, as I have swapped those parts from good to bad units.  The last time I had this issue, I got used drives from brokers and it FIXED the problem.  At that point I stopped investigating things.
I do note these drives are set up for 3-phase input, and have R and T tied together on single phase power at 238 V.  The label says 200 V ac input.  Maybe I have the machine's main transformer strapped for the wrong voltage and I have been frying some component on these drives.  Yes, it IS ODD that I have had these drives fail here every TWO years!
Thanks for the comments!
Jon

[Kean]

OK, I don't know much of anything about these Sanyo Denki parts, but here are another couple of manuals to look over and some thoughts.

General Datasheet: https://www.artisantg.com/info/SanyoDenki_P_Series_Motors_Amplifiers_Datasheet.pdf
PY Series manual: https://www.artisantg.com/info/SanyoDenki_RP001_Manual2.pdf
This second file is called the RP-001 manual, but it is actually the PY series manual with info on the RP-001.

From the first document, the PE and PZ seem to be very closely related - so the previous doc is very likely to apply for basic usage.  I couldn't spot the differences other than some beefier models in the PZ range.

Even the PY series seems quite similar based on that second manual, lots in common with the previous PE series manual.  Same status indicators on page 56.

The pages in the operation section discuss the start sequence with SRDY and SON signals.  SRDY is an output, and if I read it right the application of SON input on CN1 should change it from the 3 bars to the rotating figure 8.

If there isn't some issue with the upstream controls, the the RP-001 remote programmer may be what you need... but they don't seem to be cheap.  Maybe you can borrow one from somewhere, but you may not have the correct settings documented for your system...

Out of interest, have you seen any of the other error codes displayed?  See the troubleshooting section.

[jmelson]

OK, I don't know much of anything about these Sanyo Denki parts, but here are another couple of manuals to look over and some thoughts.

General Datasheet: https://www.artisantg.com/info/SanyoDenki_P_Series_Motors_Amplifiers_Datasheet.pdf
PY Series manual: https://www.artisantg.com/info/SanyoDenki_RP001_Manual2.pdf
This second file is called the RP-001 manual, but it is actually the PY series manual with info on the RP-001.

From the first document, the PE and PZ seem to be very closely related - so the previous doc is very likely to apply for basic usage.  I couldn't spot the differences other than some beefier models in the PZ range.

Even the PY series seems quite similar based on that second manual, lots in common with the previous PE series manual.  Same status indicators on page 56.

The pages in the operation section discuss the start sequence with SRDY and SON signals.  SRDY is an output, and if I read it right the application of SON input on CN1 should change it from the 3 bars to the rotating figure 8.

If there isn't some issue with the upstream controls, the the RP-001 remote programmer may be what you need... but they don't seem to be cheap.  Maybe you can borrow one from somewhere, but you may not have the correct settings documented for your system...

Out of interest, have you seen any of the other error codes displayed?  See the troubleshooting section.

The PE looks VERY close to the PZ, all the wire terminals and the high density connector pinouts look the same.  The PY is quite a bit different.  Anyway, I hacked in a replacement optocoupler on the servo enable (SON) input, but I still get the same indication.  After reading the PE manual, it appears that the drive is not getting the enable, although the controller is sending the right signal.
I've reached the end of what I can do, and am going to see how much Sanyo Denki will charge to repair them, if they even can.
I haven't been able to figure out if the drive is supposed to supply the +24 V or the controller is supposed to.  If that supply is kaput, then there will be NO optoisolator inputs.
Thanks,
Jon

[Kean]

The 24V supply would be external, else you don't get isolation from the optos.
This is shown in the example wiring diagrams - see Note 8 on page "76" (74 of the PDF) on the Amplifiers Datasheet I linked earlier.

[jmelson]

The 24V supply would be external, else you don't get isolation from the optos.
This is shown in the example wiring diagrams - see Note 8 on page "76" (74 of the PDF) on the Amplifiers Datasheet I linked earlier.

Right, I saw that after posting.  I should check that it is working, but it is real hard to get into those connectors to probe them.
Anyway, last time I had this exact issue, replacing the servo amps fixed it.
Thanks,
Jon

[SMdude]

The 24V supply would be external, else you don't get isolation from the optos.
This is shown in the example wiring diagrams - see Note 8 on page "76" (74 of the PDF) on the Amplifiers Datasheet I linked earlier.

Right, I saw that after posting.  I should check that it is working, but it is real hard to get into those connectors to probe them.
Anyway, last time I had this exact issue, replacing the servo amps fixed it.
Thanks,
Jon

Are you sure this is not an io issue with the pc?
You mentioned hdd issues and I think I read last time you also had hdd issues?

My Juki would very occasionally come up with some connectivity errors after not being used in a while and I found some slightly tarnished/corroded contacts in the some of the PC card slots. Since cleaning that up properly I have not had this issue resurface.

Might have been luck that the new servo drives worked last time?

[jmelson]

The 24V supply would be external, else you don't get isolation from the optos.
This is shown in the example wiring diagrams - see Note 8 on page "76" (74 of the PDF) on the Amplifiers Datasheet I linked earlier.

Right, I saw that after posting.  I should check that it is working, but it is real hard to get into those connectors to probe them.
Anyway, last time I had this exact issue, replacing the servo amps fixed it.
Thanks,
Jon

Are you sure this is not an io issue with the pc?
You mentioned hdd issues and I think I read last time you also had hdd issues?

My Juki would very occasionally come up with some connectivity errors after not being used in a while and I found some slightly tarnished/corroded contacts in the some of the PC card slots. Since cleaning that up properly I have not had this issue resurface.

Well, the PC loads the operation program to a 68040 in the VME crate, and then acts as the man-machine interface.  There is a lot of self-checking, so I don't really know how much I can test before the machine is homed.  But, a lot of stuff seems to be working normally.  It senses the safety doors and the button sequence to come out of E-stop, and then reports that the X and Y servo amps are not on line.  I did "hot wire" the servo enable opto coupler and the amp STILL doesn't go live. 
I have sent two units to Sanyo Denki for evaluation.
Jon

[jmelson]

PROBLEM SOLVED!  ALL MY FAULT!
Well, when I first tried to start the machine up over a week ago, the hard drive would not spin up.  I reseated connectors.  Then, the PC would not communicate with the VME computer.  I reseated socketed chips in the interface board, and then that came up.  I had unplugged a few cables while pulling VME boards, and failed to reconnect the the one to the XY axis board!  There are so many cables in the front of the VME box that it was not obvious!  I spent several days probing and testing, trying to determine what was wrong.  And, of course, a big dose of confirmation bias, I assumed that the same indicator display meant the same problem!  I should have been suspicious that two servo amps failed at the same time with the same issue, and that would imply a common cause, NOT in the servo amps.
Well, good to have the machine back running!
Jon

[Kean]

Great news!
We've all been there and done that.
And you learnt a bit more about how the motor control works... for next time?