Servo amp repair

On Mon, 05 Mar 2012 12:10:59 -0800, snipped-for-privacy@whidbey.com put finger to keyboard and composed:

Bastards!

Electrolytic capacitors often dry out and produce thermal symptoms. Otherwise you could have a dry solder joint.

Servo problems are some of the most difficult to troubleshoot, particularly when they involve motors. You need to open the control loop and very carefully test each stage. When one component fails, there can be catastrophic failures in other components.

Could you upload a photo of the PCB? Maybe someone will recognise it, or at least be able to guess at the identities of the components from the layout.

- Franc Zabkar

Greetings Frank, I should have said that the CNC control decides when the servo amp is working. What happens is that the control says there is a servo amp error, not what the error is, until the amp is working properly. So it's basically binary as far as the info the control gives me. Either the amp is working or it's not. I don't know, and the troubleshooting section of the CNC manual doesn't tell me, what the problem is or may be. I just wait until the amp is warm enough and the control says it's OK. Once warm enough it works perfectly and has been for a year at least. I can post pictures, but where? This isn't a binary newsgroup. I do post pictures for rcm on Metalworking.com. Eric

On Mon, 05 Mar 2012 17:17:26 -0800, snipped-for-privacy@whidbey.com put finger to keyboard and composed:

Some people use file hosting/sharing sites such as ImageShack. My ISP gives me my own free web space, as do most ISPs in Australia. Is this not the norm in the rest of the world? Just curious ...

- Franc Zabkar

I agree with Frank. With limited experience, it mught be easier and cheaper to 'shotgun' replace all the electrolytics and resolder the whole board and see if that solves the problem. I am assuming that you have checked to power supply section of the unit to see that the internal power voltages are stable. A simple drift of the 5V [?] rail during 'warm up' could easily be the culprit.

Neil S.

The problem with the amp is that

Very likely there is a bad connection, either a solder joint on a component or a connector. Look over all the solder joints and see if you can find one that is cracked. You may need to use a jeweler's loupe or other magnifier to find it. It would be best to find this by inspection, but if you can't, you can try to find it by applying force to the board while running. (Of course, watch out for the high voltages!)

Otherwise, if there are plug-in boards or modules, check the connectors there for dirt or corrosion.

It obviously is not a totally failed component or it wouldn't work at all. While it is possible there is a temperature-sensitive component, that kind of thing is fairly rare, and usually they fail when hot, not cold.

Jon

I would second the resoldering of all solder joints. It might well take you less time than it would take to post a reply to this newsgroup, and even if it does not cure the problem, you can then be sure it is not down to a dry joint.

Having reliable information on what is definitely NOT the problem can be very valuable.

Gareth.

I think many don't know to FTP, and use a picture program to generate thumbnails. Comcast gives you 10 gb. Never had a ISP not provide some kind of space.

Greg

That, however, is not the case for electrolytics. They fail more regularly than any other component in modern electronics, and a failure to operate correctly when cold is a very common thing with them. If that was the problem with the OP's board, an ESR meter would almost certainly find the offender in very short order on a cold board.

Otherwise, I concur with everyone else that a bad joint is the next most likely failure area, followed by any semiconductor on the board. It is normally fairly easy to find such problems with the help of a hairdryer, a can of freezer spray, and a little bit of cardboard to confine where the freezer gets to.

As to the tops being ground off the semis, this used to be very common a few years ago. It often indicates that the devices used are actually very cheap and common types, and that they just don't want you to know this, ensuring that defective boards are returned to them for repair. An experienced service person can often work backwards to determine what a device is, by reverse engineering the area of board that the IC or whatever is located in, and checking waveforms and voltages on its pins.

Arfa

Thanks everyone for the replies so far. I will post a picture in the metalworking.com dropbox later today. I think I'll try replacing any electrolytics first and then try the solder joint thing. I did talk to another Fadal owner who has these servo amps in his mills and they exhibit the same behavior. He has been using these machines over 40 hours a week for at least a couple of years by using a hair dryer to warm the boards in the morning. Once warmed this way the machines will run fine all day. Eric

jeweler's

to

component,

regularly

few

cheap

ensuring

in,

If localised cooling/heating ellicits nothing, then ssuming it consists of main series TTL/4000 desolder one IC at a time, place on an IC logic tester in search mode that will then show what device type. If a generic fault , not even a batch fault, then unlikely to be a problem with the IC/straight logic, more like an R or C or something drifting and upsetting timing of a monostable or something. Start with the ICs with most Rs and Cs around them

Has anyone ever used a place like this for support?

of

tester

IC/straight

I've always assumed its like those "sourcing" companies with big bucks upfront, flaff about a bit, and then come up with nothing

r. I will post a picture in the

Well, that's a RIOT! hair dryer? OK, with this additional info, it is beginning to sound like a systematic problem, so the electrolytic problem may be it. Hope you find it.

Jon

Don't know anything about them, but if they are one of the vultures selling used gear, this might sell you more servo amps with the same fault. Anything they have will likely be the same vintage as your gear.

Jon

??? All I saw was repair services.

As well as electrolytics, just have a look to see if there are any crystals, resonators, or hybrid clock generators on the board. These are also all components which can, and often do, exhibit a similar cold / warm failure mode.

Arfa

I took pictures today but forgot to bring the camera home so I will post pictures tomorrow in the Metalworking dot com dropbox. There are two or three electrolytics along one edge of the board and as these are the only electrolytics (as far as I can tell) on the board I'll replace them. I did look for any suspect solder joints with a 10x loupe and didn't see any bad looking joints. Perhaps when the pictures are posted someone will be able to give me an idea of what some of the components may be. Eric

alt.binaries.schematics.electronic