Server Y DAC Overflow error

[WagonWheel]

Hello all. Sorry in advance for the long first post. I have a 1996 EZ-Trak DX Series 1 machine that I just bought and have immediately broken. I am an old timer but new to CNC so I come with head bowed and hat in hand hoping for some advice or instruction. Before I got the machine moved into position and setup, I was playing around getting familiar with the software and managed to run the SEM x-axis motor encoder wires and motor into a pile of metal banding that was stacked on the floor. I replaced the Heidenhain X and Y encoders (don’t ask what happened to the Y) with new and still get a Servo Y DAC overflow error. When trying to home the Y jumps a bit (.1030 indicated on display) but the X never moves, and the error shows up. I searched and read all I could find on this error so here are some of the things I tried in no particular order.
Software is v4.5g rev H.

*key with 11235 and pfm.exe through DOS. Results: 1 30 12 0.0 -3048.0 .0004 .0012 0 0.

When I got the machine first thing I did was clone the hard drive to back it up, so I replaced all files from this backup back to the original.

On the BMDC3S when the machine is turned on, the two red leds one is dim the other is on.
After boot, both are out.
After Home do move, both are out.

I switched the X and Y drive cards with no difference.
On the cards at machine turn on, both the red and green leds are on.
After boot, both stay on.

Wiggled, jiggled, pencil erased card contacts and wire connectors. No bent pins.

Continuity checked all fuses I could find.

With power on, neither axis freewheel.

I do get a howl from the x Axis motor sometimes but the machine did this before I wrecked it.

Machintek was kind enough to respond to a similar problem and indicated the L1 on the BMDC can fry. I found the L1 but am an electronic dunce, so not only do I not know how to test it, I don’t even know what it is. I talked to the good folks at EMI and was told if the L1 is bad, whatever caused it to go bad could fry on a new/rebuilt card if the problem is not fixed first. Is there anyway to test for this and is there anyway I can test L1 with a multimeter?

I talked to a nice gentleman at Ajaxcnc so a conversion could be an option.
I really like the canned cycles of the Bridgeport software for a person with my beginning CNC experience level, although I already wish I had a later software version.

Any help, comments, thoughts, recommendations, etc. would be greatly appreciated. It’s even OK to laugh but I don’t think I will join you on that one.
Pittsburgh Area.


Bill

.

[machintek]

A DAC overflow means that the control commanded a move but did not see it. If the encoders are not counting (with the motor moving) then you will still get the DAC overflow.
The encoders need 5 VDC to count which comes from the logic power supply, through the mother board , through the BMDC (through L1 on the BMDC) to the AXSBOB to the encoders. If both do not count, typically it means that the L1 has blown as it acts as a fuse against excessive current such as a direct short when a encoder or wire suffers damage.
Hence, I believe we have cause and effect here with some certainty. You could remove the BMDC and check L1 for continuity with a digital meter.

George

[Gncc50]

Give these guys a call

2SQ Industrial Supply, Inc
1002 Corporate Lane
Export, PA 15632
Phone: (724) 325-7550
Toll Free: (800) 346-3188

[WagonWheel]

Thank you very much for the replies.
Machintek - I removed the BMDC and put a probe on either side of L1. The reading was .224 on the 2k scale. There was no tone for continuity when using the tone setting on the meter. Based on your logic voltage flow, would it also be possible if a person knew which voltage supply pin or wire went to the AXSBOB from the BMDC to test for 5v there to see if the voltage was making it through the BMDC or is this wrong thinking? I tried on the wiring diagram to find which wire this might be but could not figure it out yet.
Gncc50, Thanks! There was a representative from that company at the sale where I bought the machine. He invited me to an open house they were having, but I was not able to make it.
Bill

[machintek]

Yes there is a tantalum capacitor on the AXSBOB in one corner that should have 5VDC on it.

George

[NC Cams]

Thre are some good and some illogical design compromises on the BMDC - some compromises might be OK for industrial use but totally primitive and "what were they thinking" by automotive electronic wiring standards.

They seem to use the 12v buss on the computer and regulate it to 5v on the BMDC and then feed it that out to operate some of the BOB cards. This is fine so as to keep the 5v at a common level BUT, it can lead to some bizarre consequences should control or feedback line short out which it seems that you did.

With the inductor partially open, you won't get any or at least adequate control voltage downstream. This lack or inadequatecontrol voltage could "tell" the BMDC that something isn't right which could result in DAC errors.

If you can identify the make, brand and value of the inductor (L), you should be able to find a duplicate and remove and replace it. If not, EMI should be able to replace it and do a bench test.

HOWEVER, before you button up the DIY fix on the inductor, you could TEMPORARILY remove the inductor and solder in a pigtail fuse connector with a 1 amp fuse it its place.

If there still is a downstream short, you'll pop the fuse and not toast the BMDC. There are some real strange places where shorts can occur - ours occurred in the box where the ESTOP button is. We had to go into there and the wires are real tight fit. So we jammed them in, and pinched one between a switch case and the threaded conduit connector in the process.

Real SOB to find - not even EMI and their $150 bench test (at the time) of the BMDC fixed or even identified that one. It took some real judicious study of the Schematic and pinning out each and every control wire on the BOB's to find the damn thing. Especially when the short also took out a diode on on the AUXBOB card (or was it AXISBOb???) card.

If you haven't, now would be a good time to pin out each and every feedback and power line between the servo's and the source bus in the control cabinet. Ditto checking each wire to ground. This time consuming fix is way, way cheaper than a Centroid. If you don't have a schematic, get one. BPT did a very good and detailed job of documenting the wiring and it is pretty easy to understand. Thankfully, all the wires are coded and numbered which makes it MUCH easier than going at it blindly.

WHen you get the gremlin identified that you shorted, and perhaps remains shorted, you'll probably lose interest in retrofitting the 'Trak with anything.

I already tried looking into the retrofit trick on a Trak and found that you can't use ANYTHING on the 'Trak with Centroid - that's what they told me anyway. If you do decide to pull the BMDC et al, don't hack out the wiring or BOB's. Rather, give me a call. I may be interested in buying the take outs for spares.

BTW, If you look up "Eztrak retrofit" on the Zone search engine, I already tried that and got essentially NO replies.

[WagonWheel]

NC Cams, Thanks for the help!
You make some very good points. I had searched and printed out all info from the net that I could find on this problem and I saw your thread with no responses.

From EMI for my particular model board the test cost is approx $350.00 with a replacement board being approx. $1400.00. I am not sure, but it sounded like it was either going to be good or bad; not them doing a repair to mine. In other words I think it was going to cost either $350.00 or $1400.00.

I am going to take your advice on checking the wires. There were only a couple of the encoder wires that were stripped of insulation and were not broken when it hit. The trouble was the wire was shielded with the metal shielding and the stripped wires touching so there had to be a short of something. Those encoder wires are really tiny, especially for my eyesight, to strip and solder so I will double check that I have it correct before powering up.

Tomorrow I will call that company Gncc50 recommended and see what they have to say. It is ironic that a machine that can work marvels on metal like this is in some ways like handling eggs. Short out a wire as tiny as a human hair and the whole thing is out of business. I was hoping the machine would be a little more forgiving of my stupidity.

Bill

[machintek]

I just diagnosed a short in an encoder on a VMC with 4 encoders. He has a BMDC with a breaker protecting the BMDC. It had tripped. I reset the breaker, put my meter on OHMS, and used a spare connector on the logic power supply (red to black) to check for a short in the circuit. Yes the machine was OFF. I unplugged each encoder one at a time watching my meter and found the X axis encoder had shorted. I unplugged it at the encoder pigtail to make sure it was not the cable running to it but the encoder itself that had the short. If your BMDC L1 is "open" this will not work.
If you mess with the BMDC, EMI will not accept it as a trade in and your cost of replacement will double.

George

[NC Cams]

The EMI price is probably $1400 EXCHANGE. That means, they either check your board for $350 and find it to be OK or fix it or exchange it for a tested good/repaired one for the $1400. Perhaps they'll change the inductor (comon fix) for $350 but I wouldn't count on it.

The cables are pretty sturdy in their design and construction. Our machine is a 1995 model, sees relatively light service and we still have the original cables. You must have struggled really hard to hurt the cables. Regardless, your options are to pin out and carefully repair/test yours or, again, pay EMI for new pre-made ones.

I do believe that they have taken over all service ot the BMDC based systems via a deal with Hardinge - your options to the contrary are relatively limited. Fix it yourself or rely on EMI or a local tech who will probably refer any BMDC service to EMI.

Regarding the identification and "matching" of the inductor: if you can find the exact same one (most of the stuff on the BMDC was garden variuety) it would be real hard for anyone to tell if it had been replace if a good job or R and R'ing were done. The inductor is essentially acting as a/the "fuse" that BPT should have put on the BMDC when they made the darn thing.

Interestingly, at the time, there were "automotive" grade voltage regulator IC's that serve as both inductors and overcurrent devices for automotive circuits. Why sutomotive in an industrial device? Simple, momentary shorts and voltage dumps and inductive spikes are QUITE common in cars which is what is going on due to the systems that are interacting in a CNC

These "automotive grade" IC's were available and could have been designed into the BMDC instead of an inductor. But they were a bit more costly. Too bad BPT opted to save $0.75 or so instead of using a MUCH more robust regulator/protection circuit.

At this point, you can test the L and see if it has voltage on either side of it. There should be voltage there. Better yet, go downstream and start looking for encoder feed voltage at the servo plug - you should be able to figure out which pins are which from the wiring diagram which you hopefully do have for the machine. YOu can also use a fine wire and plug into the terminals on AXISBOB card - that's how we found our short.

Any pinout work is a real PITA but it saved us $1500 replacement of the BMDC which EMI claimed was our "only cure" at the time. I think not.

All this was learned via some study of the wiring diagram we got with the machine. If the machine wiring diagram is 11x17, do yourself a favor and go to an office supply/copy store and have them blow it up to D or E size - MUCH easier to read and trace the wires.

George's trick with unplugging the encoders is worth trying. If the motors and or encoders need service and they are SEM's, send them to Clarkson. Pricey but good.

The wiring of the Extrak is somewhat complicated but traceable and understandable. It takes time and study and the careful use of a voltmeter.

Some issues are a process of elimination like those advocated by Machinetek. Being an engineer, I go about things by following the route of the enable/disable signals which should or should not be at a particular point. BOth ways work, it all depends on your frame of reference.

[WagonWheel]

Machintek - Yesterday I tried your testing method from the power supply Red (5v) and Black (Ground). I was not able to work it out and since L1 is a little black around it I will have to assume it is open. Can you tell me what amp value the person used as a breaker, or the fuse value you have used in the past for protection? Or another way, would you have any idea what the maximum volts and amps L1 sees before frying?
In addition I called 2SQ but their tech is out of town.

NC Cams
Thanks for the tips. I especially like the idea of oversizing the wiring diagram printout. My cousin is an electronic whiz and might be able to figure out the ID out that inductor. Unfortunately, he is going through one of life's rough patches so I don't want to bother him with this.

Bill

[machintek]

Sounds like you have the surface mount type of L1.
I have saved many customers a pile of money by pulling 5VDC from the logic power supply through an automotive type in line fuse (2 amps), running it to the prior mentioned capacitor on the AXSBOB 9soldering the wire to the + leg of the capacitor. This powers the encoders, is fused and back feeds the BMDC. I vary the size of the fuse based on application.
This is what I do. I do not suggest this to anyone that cannot follow a schematic well. I assume no liability for what you do.
The customer above has a later control. Bridgeport figured out that customers were not happy replacing BMDC boards and started protecting them with breakers.

George

[WagonWheel]

Machintek
Thank you so much for sharing your knowledge and for your patience with me. I realize you have talked about this problem several times in the past and I really appreciate you going through it one more time. I'm back to making chips now. Everything seems fine. Somehow when you said back feeds everything became clear to me.
I would also like to thank all those that took the time and were good enough to respond. I only hope that I will be able to contribute something in the future.

Bill