The Sequel - Oh Blimey I bought a CNC Lathe (Beaver TC 20)

[awemawson]

This afternoon I reverted to my original SMCC card (the servo subsystem card that drives the turret via Siemens power amplifiers)

Initially the turret would unlock by driving the tool disk forwards, but when rotated wouldn't re-lock when driven back due to angular error. Driving it to the correct position by manipulating the lock / unlock solenoids and physically turning it, it then would settle back and I was able to select tools by program.

I get the impression that the positioning is not spot on as the re-coupling of the curvic coupling is a bit hesitant. I came to the conclusion then that the output stage of the SMCC controller had a small error offset due to a hardware issue.

OK this where I was a couple of years back before shingles destroyed my brain. I have sourced another SMCC card that was in another Beaver TC20 lathe but that one was controlled by a Fanuc PLC. Reading it's internal program  it is virtually the same as my one only differing in a couple of lines but these are to do with addressing the card so probably crucial.

My attempts at programming these cards in the past have drawn a blank - more research needed.

[tom osselton]

It’s coming along nicely

[awemawson]

Well it WAS Tom but now I'm in the 'one step forwards, three back' scenario !

So with the progress shown in my last post I left it for the evening and returned the next day to find the servo system totally refused to function with the PLC reporting "6022 Axis Drive Fault".  Now the Axis Drive Fault is reported to line 404 of socket X121 on the 'Measuring Card' and is derived from the Siemens Servo drive box control card 6SC6100-06B00 in the servo crate. It just so happens that I have a known good spare so I swapped it. No change in the symptoms.

  [ You are not allowed to view attachments ]  

Maybe it's the "Measuring Card" in the 820T? This card drives the axis servos, has all the various encoder interfaces and is buried deep in the 820T Siemens Control - access to change cards is possible but by no mean straight forwards due to physical constraints. I changed it. OH heck now nothing on the monitor screen - replaced original card - still nothing - argh !!!!

  [ You are not allowed to view attachments ]  

At this stage with nothing to loose I have pulled the entire Siemens 820T control from it's rack for bench diagnosis and am retiring to lick my wounds. Looking far back in this saga I've confirmed that I could power it up and get data on the monitor with nothing else in the way of interfaces connected so my next approach will at least physically far easier!

  [ You are not allowed to view attachments ]  

  [ You are not allowed to view attachments ]  

  [ You are not allowed to view attachments ]  

[awemawson]

So having enjoyed a smoked mackerel salad  with home grown tomatoes for lunch (very nice ) I threw caution to the winds and connected up a local mains supply so that I could continue diagnosis.

Much to my surprise the monitor came to life showing what it should (don't worry about the low battery warning - it's disconnected at the moment.

  [ You are not allowed to view attachments ]  

So now I suppose it's a case of tapping and wriggling to see what had come loose - assuming some sort of mechanical issue.

. . . . the plot thickens 

[vtsteam]

Andrew, some days problems are strictly additive. I think it's the moon phase or something.   

[AdeV]

More importantly.... did you manage to fix the tractor shed door, and how many deaf crows are there in the vicinity? 

[awemawson]

Ade aludes to a problem that I had when he recently visited (good to see you Ade). It may well be linked to the lathe issue - seems we've had a few significant power surges on our overhead power lines. During this period a friend has lost two roller shutter controllers, I've lost mine and the lathe packed up 

So no - the shutter isn't fixed but I have a replacement (expensive) controller arriving on Monday.

And Steve yes these things definitely ARE additive - ah well just get on with it I tell myself and stop moaning !

[AdeV]

It was good to catch up with you too, Andrew - It was lovely to chat in person.

What a pain about the controller! I don't suppose there's any reasonably priced anti-surge gear you can add where you power comes in, not for 3-phase anyway.

I still haven't got my new Boxford CNC out of the container.... I'm girding my loins to do it on Saturday (or maybe Sunday); it's a bit of a step down from the size of gear I'm used to... but at least I stand half a chance of moving it around on my lonesome!

[awemawson]

Three phase surge arresters do seem enormously expensive Ade - it must be said though that since I fitted a single phase one to my external PIR LED floodlight it's been fine, but before it didn't last more than a few months (if that) before dying so certainly worth investigation as the shutter is fed from the same supply albeit on all three phases.

So apart from an oddity whereby the controller refused to power up with the battery connected but was ok without it I couldn't find any faults on the bench. There are two ram retention batteries in this 820T controller - one mounted on the RAM card that I think holds the PLC program and was the original cause of all that corrosion back in the beginning. And another mounted in the rear of the CRT monitor that I think backs up other parameters on the CPU card. This later one is now remotely mounted in a battery box. I did also re-locate the PLC battery in the same box to prevent anther corrosion event but the card failed and the replacement still has the battery on the card. Sorry if that is a confusing explanation but necessary for completeness !

Nothing for it but to return the 820T controller to it's rack and carry on. Relatively simple exercise only complicated by very poor access.

  [ You are not allowed to view attachments ]  

However it does require a full reload which is quite an involved and time consuming process that sadly I'm getting too familiar with.

With it all connected I'm back to where I was with the Axis Enable line in the fault state ie not asserted. This is derived from the "Measuring Card" (which I had already exchanged for a known good unit) but presumably ultimately from the CPU card - it was changing this card that had left me unable to bring up displays on the monitor - I suppose reluctantly I'm going to have to repeat the exercise.

[awemawson]

Well wasn't that fun 

Dismantle, replace CPU card, full re-load, and . . . . . . . no change 

I'm obviously missing something here as I'm getting dizzy going round in circles - need to find a quiet corner and have a major think.

[vtsteam]

This kind of thing seems all to familiar! My thoughts are with you, Andrew. You'll crack it, I'm sure.

[awemawson]

Thanks Steve for the kind words.

At last I can report a little positive progress. I've spent the last three days going round a logical loop regarding the Axis Enable Line not being asserted. This involved 'fault' outputs from the internal Simodrive servo drive supervisory card reporting the error (yes I know it's not enabled - is that your error?) feeding into the 810M CNC controller 'PLC' logic where the signal to enable the drive unit emerges. So it's a circular argument and rather reluctantly I decided that somehow I had to break the circle.

It is fortunate that I spent months on end decoding and mapping out the PLC program several years ago - it was a big help.

The method that I chose was slightly risky (*) - remove the 'fault signal' from the PLC and drive it to a clear state to see what happens. Sure enough the 'Enable line' goes to the correct state (24v) and I can jog in X & Z axis's . A slight oddity was that the fault line had to start at 0 v then go to 24 v - presumably to satisfy some sequence needed by the PLC.

(*)  risky in the sense that these are powerful servos reporting a fault so may not behave benignly.

So I deduce that the error is within the Simodrive not the 820T - most likely one of the two dual axis pre-drive cards. There is the supervisory card that I've mentioned (and changed!) that sources the error signal but it has monitoring circuitry looking at the pre-drive cards, the power drive cards and the power supply. Sadly card changing is pretty involved - not only loads of plug connectors but also a fair number of bolted high power connections for the servo motors themselves @ 40 amps. as well as numerous option settings on daughter boards.

  [ You are not allowed to view attachments ]  

   [ You are not allowed to view attachments ]  

Just need to get down to it

[awemawson]

Well that's two cards exchanged in then back out with no change 

  [ You are not allowed to view attachments ]  

Ah well I suppose it's now down to the power drivers

[awemawson]

Well if I wasn't confused before I certainly am now!

I went through exchanging each of the analogue servos cards in turn using spares from my 'Christmas Tree' spare controller and in each case when the symptoms stayed unchanged re-installed the original

  [ You are not allowed to view attachments ]  

Then I went through exchanging the power driver cards - again one by one and returning the originals when symptoms were unchanged.

  [ You are not allowed to view attachments ]  

  [ You are not allowed to view attachments ]  


At this point I thought that I might as well return the supervisory card that I'd swapped out on the 13th. This is when things got very confussing. The green LED saying that the servo system wasn't enabled went out, and initially I could jog in X&Z however the machine started giving  D to A error in X&Z and the X axis started moving the turret upwards but the motor was growling and the machine reported X clamping issue and went into eStop.

Had enough for today - yet again going to lick my wounds indoors!

[vtsteam]

My guess is it's a Gremlin. Do you have rat trap you can put in the main cabinet at night? Bait it with a working Android cellphone, or a string of series connected incandescent Christmas lights. Bound to be some action overnight. 

[awemawson]

You are probably right Steve - whatever it is it's doing my head in !

[awemawson]

Well excellent news today. Logically I thought reverse the last step and re-replace the 6SC 6100-6B00 card. No - very similar symptoms - X axis going upwards and growling - more head banging. I then took a VERY close look at the wiring and the various connectors that bring signals on and off the board. I then realised that there are two cable forms terminating in 20 pin headers that are both labelled X311 and are bundled together.

Surely not - no - did I put them back swapped - OH YES I DID 

  [ You are not allowed to view attachments ]  

Reversing the error I'm back to "no unwanted moves upwards" on reset - phew. These cables are from the feedback encoders from the X and Turret servo motors no wonder it didn't know what it was doing. Fortunately their pinouts and voltage levels are identical so no lasting damage done.

Still got the green error light on the Go card 6SC 6100 6B00 BUT using my bodge on the error input line 404 I can jog X & Z in a controlled fashion.

OK lets put the original Go card 6SC 6100 6B00 back - oh ho - green light out system ready and I can jog X & Z without doing the 404 line bodge

Hoo blooming ray 

Back indoor to absorb the relief with a nice cup of tea and quit while I'm ahead !

[vtsteam]

Yay!!!

(I'm guessing the labeling was not your doing, Andrew, nor that it will stay as such!)

 

[awemawson]

Too true Steve on both counts!

More definitive labels will be applied tomorrow, meanwhile I've progressed from the "nice cup of tea" to a glass of "Famous Grouse" scotch with a modicum of ice so I'd best not do it now 

[tom osselton]

Glad to see you found the problem

[awemawson]

Thanks Tom.

Mind you that's just THIS problem - there are a few more that need knocking over down the line 

[vtsteam]

Famous Grouse, the Wild Turkey of the highlands.

Famous Grouse, "Waiter, what's this fly doing in my soup?"
Waiter, "The backstroke?"

Famous Grouse: Humphrey Bogart: "When I athk for a grilled cheese thandwich, I don't ekspekt muethelidge."