Changed drives from MDS-A to MDS-C1 series and get S01-PR 0018 alarm

Hi everybody,
Long time no see...

I have a problem starting my second 4-axis Meldas 520AM (1998) system since I changed all drives to MDS-C1.
I get S01 PR 0018 Alarm and the contactor shut off power to the power supply unit.
The alarm appears on both my 2-axis servo drives. One blips "F1, 18" and "F2, 18". The other one blips "F3, 18" and "F4, 18".
The Spindle drive only shows "E7" since it got emergency shut down by the contactor.
On the control display I get "Servo Alarm S01 PR 0018 X" Note, it only points out the X-axis here. How come?

Before the swap there was no alarm about any encoder communication problem. I have checked the X-axis encoder cable, changed servo motor/encoder unit to another one. Still the same alarm.

My old drives were MDS-A-SVJ-10 for X, Y, B axis and MDS-A-SVJ-20 for Z axis. Spindle drive was MDS-B-SP-75. Power unit was MDS-A-CR-75.
My new drives are MDS-C1-V2 1010 for X, Y axis and MDS-C1-V2-1020 for B, Z axis. Spindle drive is MDS-C1-SP-75. Power unit is MDS-C1-CV-110 + of course an AC line reactor D-AL-11K.
I use the original servo motors/encoders: HA80NT-E33 for X,Y axis, HA100NT-E33 for Z axis and HA40NT-E33 for B axis.
They are all equipped with incremental motor end encoders OSE253 (25 000p/rev).

I have changed following servo parameters before firing up:
SV025 Mtyp: X,Y=from 0001 to 2201, Z=from 0002 to 2202 B=from 0000 to 2200
SV036 Ptyp: X,Y,Z,B=from 1500 to 0011
SV047 EC1: X,Y,Z,B=from 0 to 100

Since I se "18" flickering on all axis, I tend to think it might be a parameter setting error.
Is there another parameter that can cause the 0018 alrm?

Any guidance would be highly appreciated.
Regards

alarm 18 is the "initial encoder communication error".
i bet it's because the mds-c1-v does not support osa/ose253 encoders. neither do mds-b-v. i think the only other drive (apart from mds-a-svj) which does - is the mds-b-svj2. no idea why, i see nothing special about them...

Thanks for the valuable information. I think it is weird though. Can anyone confirm, elaborate on why C1-V2 drives does not accept OSE 253 encoders?
Another odd thing is that the servo monitor states that all encoders are OSA 105 which they obviously are not!
So now I have to buy OSE 104 or OSE 105 encoders for all axes?
I hope I can still use my HA-N servo motors though?

I am not familiar with your control or your encoders (I have an older M3 machine) but I can give you a few thoughts on encoders in general.... The older Mitsubishi controls actually have some additional tracks. In addition to the normal phase quadrature and index pulse, the encoders have a course U/V/W signals too. These are not used for position feedback to the NC control but, rather, the servo amplifier itself. Specifically, the signals provide the feedback mechanism to properly sequence the windings to energize so that the servo moves. Remember, these are AC servos, not DC, so you need to sequence the coils properly.

So, why is this (potentially) significant? Well, there is nothing unique about a phase quadrature. It is about as exciting as watching honey pour from a bottle. The only variation here is whether or not there are differential amplifiers, but that is easily enough determined. The CPR is very significant, though. Mitsubishi had a tendency, on older control, to provide model number parameter entry and not CPR entry, so there could be some issues there. But, the other place you could easily have an issue is with wiring. I would find, and check, the wiring pinouts carefully between the two models of encoders you are talking about. It is entirely possible that you are not feeding the amps with the correct signals.

Quote:

---

Originally Posted by CNC Viking

Can anyone confirm, elaborate on why C1-V2 drives does not accept OSE 253 encoders?

---

marketing probably? :) i don't have any mds-c1-v drive to confirm, but mds-b-v still support lower resolution quadrature OHE5K. OSA/OSE are serial interfaced, same protocol. not sure about early ones, but later OSA/OSE/253/104/105 are actually using the same hardware/firmware and can be reprogrammed to be either of those. i.e. OSE253S always reports absolute 20bit position just like OSA105, the only difference between them is the encoder ID...

e.g. here's the definite list of encoders supported by the mds-b-v fw version BND-513W000 A6, both motor and load-side:
OHE25K
OHA25K
OSE5K
OHE25KET
OHA25KET
OSA105
OSA104
OSE105
OSE104
OAM-13B
OSA105ET
OSA104ET
OSE105ET
OSE104ET
AT342

Quote:

---

Originally Posted by CNC Viking

Another odd thing is that the servo monitor states that all encoders are OSA 105 which they obviously are not!

---

i think can explain that. the drive negotiates with the encoder asking its ID, and when it doesn't recognize it - it load the default one, which is the OSA105

Quote:

---

Originally Posted by CNC Viking

So now I have to buy OSE 104 or OSE 105 encoders for all axes?
I hope I can still use my HA-N servo motors though?

---

yes, that should help. HA motors are still supported by mds-c1-v according to the manual.

or take the cover off the encoder and post a photo of the PCB. i can probably tell if it's reprogrammable. i have a HA23-NC motor with OSE104S encoder which happily works as OSA105...

MrMetric:
I checked the encoder cable buildups for both mds-a-svj/mds-b-svj2 drives and mds-b-v/mds-c1-v drives and they all seam to have same pinouts so I am not too worried about that.

dm17ry:
I also checked the encoder compatibility list for mds-c1-v drives and it only lists these as motor end encoders: OSE/OSA 104, 104S,104S1,104S2 ; OSE/OSA 105, 105S, 105S1, 105S2; OBA13, OSA14 and OBA17.
Since my current OSE253 encoders are not listed as compatible, I take it that I must replace those with OSE 104 units.
Now, what is the difference between OSE 104, OSE104S, OSE104S1 and OSE 104S2? Very confusing and I cant find any info on that. Anyone knows about this?

Here is a picture of the PCB in my current OSE253 encoders
Attachment 474788

i believe the difference between S, S1, S2 are additional mounting holes or something mechanical like that, if any at all. i don't have enough statistics to say it for sure. but for the drive they all are identical.
the encoder PCB looks the same (except the A1 chip revision vs A2 on my later OSE104) and i'm pretty sure your OSE253's can be reprogrammed into OSE/OSA 104 or 105. but i'm not sure if anyone does that except some clever chinese guys and me :) but i can't do it remotely..

Quote:

---

Originally Posted by dm17ry

ibut i'm not sure if anyone does that except some clever chinese guys and me :) but i can't do it remotely..

---

Being an electronics guy myself, I'm curious about this.... Where did you get the information and/or how did you reverse engineer that? The kiss of death, to me, is that there is a custom chip on there, which is almost certainly what you are programming. But getting the pinouts and programming information for that is tough, so I'm curious how you solved that problem.... One of the things I hate about the Meldas stuff is that there are lots of custom SIP packs and chips, so if there is info out there, I'd love to see it!

Regards,
Alna

Quote:

---

Originally Posted by dm17ry

i believe the difference between S, S1, S2 are additional mounting holes or something mechanical like that, if any at all. i don't have enough statistics to say it for sure. but for the drive they all are identical.
the encoder PCB looks the same (except the A1 chip revision vs A2 on my later OSE104) and i'm pretty sure your OSE253's can be reprogrammed into OSE/OSA 104 or 105. but i'm not sure if anyone does that except some clever chinese guys and me :) but i can't do it remotely..

---

OK If I have OSE253 now, I better chose the OSE104 and none of the "S" versions so I dont run into a physical misfit issue.
Regarding "reprogramming": Wow that would be really neat, being able to use the existing OSE253's!
How do you do it? Do you make a custom cable that fits into the small white terminal on top of my pic? What about the programming equipment and procedure?

MrMetric, i have a long history with mitsubishi servos... :) it started 12 years ago when i bought a few mr-j2s-b sscnet servo drives off ebay with the intention of reverse engineering the protocol and retrofitting my chinese benchtop mill with them under linuxcnc control... i actually did it by snooping at the motion controller-drive traffic and then dived into the rabbit hole... :) since then i've been having fun exploring the inner working of many generations their servos. found a number of development/debug hooks present in the hardware/firmware.. as of encoders, osa/ose/253/104/105 are kinda easy to crack as they have an external eeprom chip which contains the encoder id. you can pluke it off the pcb, compare/reprogram it offline and solder it back. i believe the guys selling those encoders on ebay from china are doing exactly that. but the encoder firmware is capable of writing to the eeprom also, when told to do so via normal drive serial interface. it's kinda non-trivial to figure out how to do it, because the host requests to it are just 1-byte packets. they have something like a finite state machine implemented in the encoder firmware, so you have to send a number of bytes in the correct order to put it into eeprom writing mode. i was interested in it because osa17/18/24 encoders do not have an external eeprom chip, the talking to the firmware is the only way to reprogram them. but they have a way to read back CPU memory via drive interface, so i was able to dump complete cpu address space and disassemble the firmware...
the last trick i was able to do is to build my own encoder speaking mitsubishi protocol, hack a mr-j3-70b drive into mds-d-spj3 and make it drive a generic 24krpm induction motor spindle: https://youtu.be/SIhoIOcmU-c

CNC Viking, i can do it via normal drive connector using my own usb-rs422 adapter and some linux software...

Don't really mean to hijack the thread, so I'll keep this short... But, interesting.... I have an M3 with that 50 pin ribbon IDC bus that goes between the control and drives. I've been mulling over reverse engineering that too! If you have done it, then maybe it isn't necessary.... But, yeah, I've been getting "deep" into the servo amplifiers. It has been interesting. The frustrating thing is when you start dealing with the custom chips because there is no information out there on them and, even if there were information, there are rarely and sources for replacements. I have a bad Medas SIP that I'm actually removing the epoxy on right now. Fortunately, in that case, I was able to locate the appropriate SIP. Now I'm just curious what was potted.

CNC Viking... Often if there are 4 pins it can be a potential JTAG interface point. That can help. The connector in this case is 6 signals. That doesn't mean it isn't JTAG, but....

yeah, sorry CNC Viking. but in a nutshell, you need to get compatible encoders and your new drives should be functional again. or contact me at yurtaev at gmail dot com if you wanna try to convince your existing humble ose253's that they were really born to be full-blown osa105's ;)

MrMetric, nah, i put a line for myself that i don't mess with that ancient parallel bus stuff from the 80's :) mds-e and cc-link ie tsn are waiting..
i've found the pinout of that bus in some manual a while ago. if you have a working controller - those signal names and a logic analyzer should give you a good start

:) Yeah, I have that pinout.... And, yes, it is ancient. I've been meaning to stick a logic analyzer on there but I don't have one wide enough right now. Plus, it is a PITA to go through it all, but I"m guessing the pattern will show up fairly quickly.

exactly, that's why i don't like parallel interfaces :) but 16 (or even 8) channel LA should be enough to get started. when it's clear what's going on in there, i would make an fpga board with a number of diff transceivers, try to decode the traffic first, then talk to drives myself... shouldn't be complicated

Quote:

---

Originally Posted by dm17ry

yeah, sorry CNC Viking. but in a nutshell, you need to get compatible encoders and your new drives should be functional again. or contact me at yurtaev at gmail dot com if you wanna try to convince your existing humble ose253's that they were really born to be full-blown osa105's ;)

---

So it is doable with a custom drive cable, a USB-RS422 adapter (like Exsys EX-1309-T) and a laptop with Linux on it?
One question though: The OSE253 has a resolution of 25 000p/rev while OSE104 has 100 000p/rev. Surely that must be physical capacities, not possible to "re-program"? Will the drive accept an "OSE104" with only 25 000p/rev?
Attachment 474816

that's my point.. regardless of the programmed encoder id and what the nameplate says those encoders are physically the same and always report 1,048,576 counts per revolution (20 bit) and are all absolute (hence the supercapacitor on the PCB). but when identified as 104 the drive rounds the position down to 100,000 cpr, when 253 - to 25,000. and if set being an "OSE" - doesn't use the absolute capability.

This gets better and better...:):) I will definitively try to re-program my OSE 253's now!
I have no ABS battery module and is used to always do home-search on start-up anyway, so I will go for OSE type. Which do you suggest though, OSE104 (100 000p/rev) or OSE105 (1 000 000p/rev)?

i use a single 18650 li-ion cell cannibalized from a dead laptop battery for encoder backup :) it keeps 4 motors alive for about a year with the machine off most of the time. then recharge and good to go for another year...
if the controller is ok with OSA/OSE105 - of course the higher the encoder resolution - the better... SHG control probably needs 1M resolution. but that may be a payed option in control...

I checked the SV019 and SV020 parameter ranges for the Meldas 520 AM Controller:
The setting range is 1 to 1000kp/rev, so i believe we should be OK with type OSE105 then.:)

Just a little update,

Right now dm17ry evaluates some possible USB-to-RS422 adapters in order to recommend one to build
my cable with. (Thanks dm17ry).

I will be back later when I have some hardware and configured a COM-port on my Laptop.