Whoops! Smoked all 4 channels?

Well, today was a good day. NOT. Accidentally pinched 2 of the wires on the Y channel of my 4-channel Xylotex board (cut through the insulation and shorted to the machine table), now all channels appear dead... The Y stepper is "locked" with the board turned on *or* off (IE: feels like you're shorting the stepper wires together), the other channels are just plain unresponsive - motors don't lock with power on and don't respond to the PC. I do get the red light when power's on. Anyone know if it's possible to blow all the channels at once if one motor shorts? I bought the board as a kit with the Taig from someone other than Xylotex so repair by them isn't possible, so I just ordered 6 driver chips to replace the 4 on the board (and have spares), hopefully that takes care of it. Also ordered the 1 octal buffer on the board just in case. Besides a voltage regulator (that seems OK), those appear to be the only active components on the board. I don't see any burned traces and didn't smell anything, it just died. :( A few minutes later I was digging for some parts in a box and cut my hand open, woo! :) I'm using the 269oz motors from Xylotex, anyone have a good reccomended reference voltage to run? I had them pretty high before but it seemed to run faster when I cut them down to 2.5v or so.. Anyway, hopefully the new chips will fix it!

If you're lucky, only one channel is blown. It might be shorting out the power supply so there's not enough voltage for the others to run.

Try removing the locked motor and see if the others work. Warning: if you keep the locked motor powered too long, you will likely burn it out also.

Unfortunately I tried all the channels individually (with the other "known good" motors) and they each appear dead. The one that actually shorted turns out to be the "A" channel, with the board completely removed from the chassis if you turn the motor it feels like you're shorting the wires together on the motor. I removed the "A" channel driver chip, and the motor turns by hand now. (so at least that chip was toast) I just now tried the other channels, and they don't lock up the motor when you turn the power on.. :( I should probably do some more investigating with the power before I replace the other chips -- it could be possible similar to what you said that the power supply circuit is being / was hosed by the blown chip. I don't see a fuse of any sort, but it's possible 1 common part blew and I just haven't seen it. (what's that big "block" component over by the jumpers / headers? It only has 2 pins and in one picture I saw it looked like someone replaced it with a resistor or an inductor...)

Well, I replaced the 1 "definitely" blown chip, that motor now moves by hand with or without power applied. Buut, all 4 channels are still dead including the one with the new chip. (they don't even lock when power turns on..) Has to be something common to all 4, but I haven't found it yet. 24v is good, 5v is good, nothing seems burned. Lesson learned - don't work on the machine with the power on! Ugh.

Well, last post I guess! Good thing I'm not a debug tech any more! I double, triple checked everything after I replaced the 1 blown driver chip.. Or so I thought. I ended up taking off the other 3 "just in case" they were causing any problems. Still dead. Looked at the chipspec for quite a while, couldn't find anything wrong. It turns out that in the Block diagram / pinout of the chipspec, they show the "RESET" pin as active HIGH. But in the text, it has a bar over it and they say it's active LOW. I missed that little detail even though I probably read it 5 times. On the Xylotex 4 channel board, there's a small SOT-23 NPN transistor just above the X channel current adjustment potentiometer. It's apparently being used as a pullup diode (base is connected to GND, not any signals..) to pull up the RESET* signal to all the driver chips. Apparently a current overload on 1 channel can feed back through and blow out that diode, causing the RESET* to be stuck low. On all the other boards I've seen that use these driver chips they pull RESET* up with a 1k or so resistor, so that's what I replaced that Xistor with and now ALL IS FINE. :) Well, after the PITA of soldering on (4) 44-pin chips with screw terminals really close by anyway. I now have 4 spare driver chips just in case I blow out another one. :) If anyone needs the #, it's Digikey 620-1076-ND for the "industrial" temp range @ $6.14, and 620-1078-ND for the "consumer" temp range (that's what Xylo uses) @ $5.54. Saved me a couple hundred $$! As a side note, while I had everything apart and measuring I noticed that my 24v supply dips down to 17 or so if I do a rapid (54ipm) movement and then abruptly stop. Could explain why my Y axis locks up sometimes when I'm doing an X-Y rapid..... I think it's a 2.5A @ 24v. Time for an upgrade. Thanks for the great form guys! Edit: SCORE! Forgot I had a 24v @ 12A IPD power supply in the parts bin! It can be adjusted up to 27v, internally fan cooled, and made in the USA. NICE. :) Think I paid like $5 at the MIT Flea market years back..

I believe I just smoked a channel myself... and I was wondering how do you go about checking individual channels are alright?

Set the channel to full step, no connected motor and measure the 4 motor outputs dc voltage. They should approximately alternate between the supply voltage and 0V as you step.

The PS is switching on and off, I tried disabling all the axis but no change....

Probably at least one of the 3977's is fried then, maybe more. If you know which channel is suspect you can remove that 3977. Simplest method is to cut the pins where they enter the plastic package with a sharp exacto knife, when the pastic body is free then remove each individual pin with a soldering iron. Clean the pads with solder wick.
Or the easy route, have Xylotex repair it.

I replied to your PM, but to keep it short, what he said ^^^ :) According to Xylotex's website, they will only repair the board if you bought it directly from them though.. You're supposed to contact the person you bought it from for repairs if you didn't get it from them...

I pulled of the heat sink and removed the chip, and the other three axis are working perfectly now.

Thanks :D

I only need three axis at the moment for the current project, but how did you go about soldering the replacement chip on....

Basically just clean off the existing solder / any leftover pins, place the new chip on there correctly, then carefully solder down all 44 pins of the new chip. Helps to have a small soldering iron tip and a magnifier / microscope... :)

I have the same problem you originally had now... Would you mind sharing a picture of the location of the chip that needs to be replaced?

I looked at ordering some replacement chips from digikey, and the exact part number on the chip is discontinued? Could you shed any light on that?

The same part is a3977KED. The PN after the KED is just how the parts are packed for delivery, t=tube TR or T&R means tape and reel for surface mount machines.

http://img353.imageshack.us/img353/1065/dsc00068pf1.jpg

There is the general location I believe...

The transistor I blew is the little 3-pin surface mount guy to the right of that 20-pin chip in the top left corner of your picture. I soldered a (I believe) 1k resistor diagonally across the pads, but not sure if it was the top or bottom pin at the moment. If I remember right only 2 pins were used, so I would just take a 1k resistor (surface mount if you have one) and place it between the single pin on the transistor, to one of the other pins. (try each one) If you look up the chip datasheet for the driver chips (digikey should have a little link for "data sheet" on the page), it'll tell you which pin is *RESET, you can check that pin with a multimeter to see if it's being pulled high. (+5v, which that transistor / diode does, or in my case, the 1k resistor) If that pin is held low (0v) then your driver chips will be disabled.

Removed the transistor, testing the pads, one does nothing and the other shuts down the entire board, I think this one just might be screwed...

Hmm... with the transistor removed, the board should be disabled. (signal is being pulled low internally to the driver chips I believe) You need to pull the signal high to enable the chips, which is what that transistor (or resistor) would do. With the transistor removed, is the board disabled? (all motor output lines = +24v or whatever your supply is) What are the symptoms you're having now, before the transistor was removed?

The lines are all 0v with the transistor removed. The status light is on, but when I cross the pads the board just turns off.

That's odd... I'm not sure what would cause the lines to be 0v... I know "disable" is all lines = 24v (at least on my board), but that sounds odd.. Have you tried contacting Xylotex at all yet? Unfortunately that's about the extent of my knowledge with the board... Once I got it working I made a new box for it and haven't touched it since...