Ever been so mad at a project you wanted to just smash it into the concrete?

[mendip_discovery]

I have a old ametek joffra in the lab, it had a busted blade on the fan so I thought. Easy just undo the big screws that hold the frame on, but it doesnt budge and I dare not hit it with a hammer as it does work at the moment and it may break more trying to repair it. I put it back together ready for a day when I am in a better mood.

[coppercone2]

Dynasty 200dx I have a useless service manual I need like component level lol, I replaced maybe 50 parts on it bc of various burns n stuff. Milled and glue charred pcb too. It’s why man invented the hammer like 2001 space odessey

I have this manual on my NAS, it might be what you have already though? Once I remember the Miller partner portal login (it's been 10 years...) I'll see if I can find more.

https://drive.google.com/drive/folders/1AbmCZZFmvnaZQ04w5Bh-TJpSg_JAkeDf?usp=sharing

yeah I got that one for a long time, the failure I have is not there.

[coppercone2]

replaced that stupid PCB, it looks like it works now. Nothing getting too hot. Need to diassemble and clean the magnetics section now that I know it might work, then get some accessories to test it with, and add thermal grease to a few things, but damn, its that stupid inverter control board, it must have had some kind of short inside of the PCB

18 months I think, but I suspect if I replaced that board first it might have had other problems

[coppercone2]

what conformal coating do you recommend to recoat miller boards?

I ordered mg chemicals silicone coating, so I can redo the coat on the mainboard.

Right now I am disassembling all the front end power electronics parts and I am ultrasonic cleaning all the connectors, plan to put them back together with deoxit grease. Some of the contact surfaces don't look too great, but nothing is horrible, and doing some cleaning on the big air inductor, and I might take one of the transformers off and clean the rust off of it and laquer it a little with a brush.


Maybe a little OD but I wanna take care of this thing if it works, store it inside, etc. Lots of black dust inside of the coils. I guess this thing works like some kinda magnetic sieve for fine grinder dust

[coppercone2]

also it looks like I found part of the original problem, there is a distribution bolt attached to some thick wires near the magnetics, there is a adapter to connect a fast-on there (like a eyelet to faston) and it looks blued and there is arc damage on the end of it. Probobly should replace that crimp too and check that wire.

[coppercone2]

It looks like the heavy coils come out OK, the entire magnetic front end of the welder is pretty serviceable, you need to bend some bus bars to get stuff out, or is it heavy square wire, whatever... maybe miller could cut some slots that are reinforced with plates so that stuff comes out with less bending though. Like unscrew a plate over a slot so that the heavy wires can lift out instead of having to be pulled out, you could probobly do it with 1 bent tab and a single screw or even some spring in place crap



Lots of crap still in there, tried to get it off with spray alcohol and a air hose before disassembling but I see that the greasy dust aint gonna come out of there easily. And the plate choke could use some more laquer or a better paint process, it should not go off this quick with surface rust, but I guess it probobly gets hot down there at high currents so its probobly hard to prevent that

And they gotta do a better job with that honey comb air vent thing they got going on, its clever for air flow but cleaning that is a total pain in the ass, they should have screwed in a structural vent there (that is both structural support but also a vent) that you can remove and wash.

I do like that the large diameter air coil is on a bent bracket that screws in with only 2 big bolts, thats a really nice feature for cleaning.

I would give the magnetic section a 4/10* for build quality, and the air intake system a 3/10 for clean ability, FUCK that caked up mill scale shit in the honey comb!! If you could take that vent out like a slot it would be REAL nice. Put a stainless pipe latterally to stabilize it and stop making god damn monolithic vents

Bet this thing would have not exploded if it was easier to clean....

*I would have given it 8/10 but I renembered that the transformer is built mad bootleg. They stuffed a teflon tube into a winding and then jammed a wire thats loosely wrapped around a ferrite rod in there. Its like a gimmick capacitor style. That needs to be made a real part because its ridiculous! That component looks like it came off Hindu-hacktube! I dunno if that was an attempt at a adjustable core or something but holy crap if your just jamming wire into a tube thats super bootleg. Dude musta been a pump designer that transferred to electronics because its like hydraulic wadding or something lmfao, I don't want high power blunt wraps. Irma say that's 20/10 for prototyping power electronics but for premium retails GET REAL. I felt like I found a knife stuck in the patient 

[coppercone2]

lots of cleaning, every component resulted in a dirty ultrasonic bath (I put fresh liquid after each subassembly).

Now I will bake them out, clean up some of the bigger connectors and conformal coat the reworked board, varnish the transformer exterior after a bit of rust removal, change the rusty bolts on the spark gap arc start, and hopefully it will turn back on again looking fresh inside . Even the inside of the relay was filthy, I carefully dipped it in the cleaner as not to get the coil wet though, I don't trust those relay coils for getting wet.


still cleaner then that one harrison power supply I got on ebay though, that think was RANK²

[helius]

what conformal coating do you recommend to recoat miller boards?
I ordered mg chemicals silicone coating, so I can redo the coat on the mainboard.

I've had good results with it, but the fumes are noxious (ethylbenzene) and it takes a fairly long time to harden, like 6 hrs. It does flow and level very easily.

[TERRA Operative]

We used to use MG Chemicals stuff in the workshop, but I'm not sure now what conformal coating we used exactly. It was a high voltage rated one in a spray can we bought from RS Components.
If you buy the highest voltage rated one you can find in the RS catalogue, it was probably that.

[coppercone2]

I was a little worried about temperature and flex because the PCB is kind of one of those bugs bunny things that wobbles like a diving board and there is some power resistors that over heat, so I got silicone

But, if you compare the dielectric strength, its 1000V/mil, vs 3000V/mil for the dielectric/insulting one


Do you need to get the higher value one? When I see a 300% difference I got a little worried

i.e. silicone conformal coat vs super corona dope (1kV / mil vs 4kV/mil)

This welder coating discussion has a new thread now, so I don't hijack the stories thread
https://www.eevblog.com/forum/repair/conformal-coating-for-tig-welder-pcb-(milller)/new/#new

[helius]

Super Corona Dope for high tension wires or points, 10kV and up. Long ago I used it on a voltage multiplier for a nitrogen laser.

[Zero999]

Yes, but it's people, often managers who make be feel this way. Quite a few times, I've had to walk away, then rant about how much of an expletive a certain person is. Politics, egos and power trips are the main problems I encounter. I find it infuriating.

[jmelson]

I had my first pick and place machine up and running in one week, making boards.  I had a lot to learn about the process.
After 13 years, that machine was not accurate enough for the stuff I was doing, it had some 0.5mm lead pitch flat packs and SSOP parts that were not placed precisely enough.  So, I decided to upgrade.  I bought a Quad QSA30A from an auction for the starting bid ($500.)  What I didn't know was that it had been sitting in hot humid space for SIX years after monkeys had messed with it.  The heat/humidity soak caused a LOT of failures of components shortly after I got it installed. I spent WAYYY too much time trying to debug those issues, then I found this model had large parts commonality with a very popular Samsung model so parts are readily available cheap from the far East.  The other thing I found out was that the manuals are AWFUL, some of the worst I have ever seen.  It is just a guide through the GUI pages with VERY little info on how various parameters interact, etc.  I was totally tearing my hair out.  Crazy stuff like they don't even tell you how the board coordinate system works.  +Y is to the rear of the machine, but plus X is to the LEFT!  However, the machine coordinates have +X to the right, as is normal.  Completely crazy!  It does make a tiny bit of sense, as the conveyor stop is at the lower right corner of the board, so coordinates grow from there.  Fortunately, another user put me in touch with the guy who did factory training at Quad.  He straightened me out on a lot of these issues.  But, at one point I was REALLY ready to chop the machine into little pieces.
Jon

[EEVblog]

[TERRA Operative]

I had my first pick and place machine up and running in one week, making boards.  I had a lot to learn about the process.
After 13 years, that machine was not accurate enough for the stuff I was doing, it had some 0.5mm lead pitch flat packs and SSOP parts that were not placed precisely enough.  So, I decided to upgrade.  I bought a Quad QSA30A from an auction for the starting bid ($500.)  What I didn't know was that it had been sitting in hot humid space for SIX years after monkeys had messed with it.  The heat/humidity soak caused a LOT of failures of components shortly after I got it installed. I spent WAYYY too much time trying to debug those issues, then I found this model had large parts commonality with a very popular Samsung model so parts are readily available cheap from the far East.  The other thing I found out was that the manuals are AWFUL, some of the worst I have ever seen.  It is just a guide through the GUI pages with VERY little info on how various parameters interact, etc.  I was totally tearing my hair out.  Crazy stuff like they don't even tell you how the board coordinate system works.  +Y is to the rear of the machine, but plus X is to the LEFT!  However, the machine coordinates have +X to the right, as is normal.  Completely crazy!  It does make a tiny bit of sense, as the conveyor stop is at the lower right corner of the board, so coordinates grow from there.  Fortunately, another user put me in touch with the guy who did factory training at Quad.  He straightened me out on a lot of these issues.  But, at one point I was REALLY ready to chop the machine into little pieces.
Jon

Maybe you now know why it was sitting unused for six years....

[jmelson]

Maybe you now know why it was sitting unused for six years....

No, the machine works quite well, once you know how to use it.  A previous owner replaced one of the mounting heads.  The earler version of the machine used 2-phase stepper motors for nozzle rotation, the later version  used 5-phase motors.  The previous owner didn't have a CLUE why that head would not home.  All you had to do was look at the part number on the motor and see they didn't match!
That's the monkey part.
Then, I had some issues with the fiducial camera, sometimes it would detect the fiducial perfectly, sometimes it would fail terribly, being a couple mm off.  It would show a one pixel wide crosshair where it thought the centroid was, so you could tell it was wrong.  I pulled the DSP board for the camera out and put it under a microscope and found seven FPGA leads had bad soldering!  Fixed that, and it worked like a CHARM!!!  Amazing it left the factory like that.
Then, there's a program called the "optimizer".  It does a HUGE job in the setup for a placement job.  You import a file with the part designator (R103), part descriptor(10K 0805) and XY and rotation.  You then select all the required parts out of the library.  Then, you run the optimizer and it matches the part descriptors, pulls the part info from the part library and selects nozzles and the order to place each part for minimum nozzle changes and travel time.  This program would occasionally screw up and scramble the mounting coordinates to different parts.  A guy sent me all the programs from his machine and I just swapped out the optimizer, as it was a separate program from the main mounting program.  I was not sure you could swap different versions of main program and optimizer, but it worked!  And, so far no scrambling of the mounting data.
I do not know if this was "bit rot" or some damage to the program image.  I kind of doubt the machine was delivered like that, as the screw-ups of the optimizer were frequent enough that they couldn't have gotten through testing without it happening at the factory.
I don't run this machine in what people would call full production, but it sure works well once you have it set up.  Setup IS complicated, as you have to tell it all the XYZ dimensions of every part "footprint".  So, you don't have to separately describe 10 Ohm vs. 10K Ohm 0805 resistors, those are all listed under 0805 resistor, but you do have to have footprints describing every resistor size, every capacitor style, every IC package and so on.  Once you have the setup done, it will place about 4000 components/hour quite reliably, and plant 0.5mm lead pitch parts perfectly!
Jon