EEVblog Electronics Community Forum
Electronics => Repair => Topic started by: NetworKing on June 10, 2020, 06:16:19 pm
-
I ended up with a free Snapon YA205 mig welder, said did not work. I tested all of the welder power supply transformer, choke, diodes, solid state relay. I determined that there was no voltage being output from the control board to the solid state relay or gas solenoid valve.
I unsoldered at least one lead on all the through hole parts on the main control board and tested them they all seem fine. I even tested the transistor i suspected was faulty but seems fine so i can only assume trigger to it is problem. The motor does change speed, but it does not seem to work that well fast does not seem fast, and about 15-20% on low end it does not even seem to run but just whine.
Tested the dial board too all the parts seemed fine has 3 10K potentiometers. There is one unmarked SMD chip on main board, the dial board has what looks like 2 555 ICs but no idea how to test those. I did try the spot, stitch, continus settings they seemed to work about what i would expect. Stitch like pulses off, on and two dials adjust weld time and frequency. Spot weld just makes spot for X time when trigger is hit. Continuous makes it go constant. So i think that PCB is good there is just some fault in the main board i think.
I do not know would work but i did have the thought of just replace the main board with an arduino, or something like it. Wire in the dial knobs for the settings to that, and then find a 12V DC motor controller module that could interface to the arduino (or similar).
Attached are some images of the parts, and a page of notes i made of connections. The pic named main-bot-num has numbers marked near the headers in gray pen, header 1 is on lower left, header 2 is above that, header3 is at the top left, header 4 is at top right. All headers pin 1 is on the left of the connector as viewed from the bottom. I am kind of lost as all the parts seem fine and dials seem to work, just seems like the output transistor(s) don't trigger. So could be faulty transistor but i am thinking fault in the driver for that transistor. Note-2 has some notes of parts i tested, mostly caps, resistors, these corelate to numbers by the part on the top of the board.
A new PCB is like $280 and only a few places have them. I would think an arduino, etc i would be looking at like $100, i have tons of relays, etc that could be used too. Seems like this PCB should be simple to find issue, i almost feel i missed something, but tested everything i could, and reflowed all the solders i am not sure.
-
These controllers usually use back emf to regulate the motor speed, so it is usually a bit more complicated than just a pwm controller.
I had a crappy noname MIG welder that made abysmal welds. I ended up building a motor/feed assembly and control unit from scratch. I used optical feedback on the motor speed control. That thing makes better welds than my Miller 250 (still has a crappy duty cycle though).
-
Most welders live in very dirty environments. From my own experience, I have found that dirt gets into the pots and makes them intermittent or worse. The openings in the pot cases you have pictured certainly are suspect as well as the open rotary switch. Another thing to look for are fusible resistors.
Just because it has a snap-on logo doesn't mean it is of the highest quality but probably not the worst, either.
Edit: I know that Century made some welders and other equipment like battery chargers for snap-on. If you can find a large welding supply that repairs welders in house they could tell you exactly who made it and the equivalent model from the real manufacturer probably from a few pictures.
-
I've come across a couple of design schemes in MIG's that might help here.
Sometimes wire feed motors supplies are linked to the welders output as in the Automig attachment (below) so for it to perform properly the welding transformer needs be ON.
However I'm not sure of it's physical control for welding current, is it the 4 position switch or one of the 10K pots ?
Looking through your pics it might seem this MIG is TRIAC controlled but more accurately an Alternistor, a specialized type of TRIAC used for inductive loads using ordinary phase control techniques but I can't spot such a device. :-//
Instead it seems maybe this uses switched rectified mains into the welding transformer primary via the NPN MJ15015 on the separate Ali heatsink.
This guy:
Datasheet: http://www.svntc.com/TPDF/625.pdf (http://www.svntc.com/TPDF/625.pdf)
(https://www.eevblog.com/forum/repair/snapon-ya-205-mig-welder-control-repair/?action=dlattach;attach=1001890)(https://www.eevblog.com/forum/repair/snapon-ya-205-mig-welder-control-repair/?action=dlattach;attach=1001892)
If so it seems controlled by the 16 pin IC and output from pin 3 which just might indicate it's just a PWM controller like those used in SMPS.
This guy, top right:
(https://www.eevblog.com/forum/repair/snapon-ya-205-mig-welder-control-repair/?action=dlattach;attach=1001884)
Flip the PCB over and the beefy TO3 MJ900 is probably for the wire motor speed control.
Datasheet: http://electols.com/pdfs/mj1001.pdf (http://electols.com/pdfs/mj1001.pdf)
Just what is the device top left ? Voltage regulator, some LM78** ?
Looks like it's split but is that just a heatsink ?
(https://www.eevblog.com/forum/repair/snapon-ya-205-mig-welder-control-repair/?action=dlattach;attach=1001886)
Does the welding transformer hum at all when you pull the hand-piece trigger ? Does local lighting flicker like it's drawing mains current ?
Any voltage on the output ?
Can you lay all the bits out and show interconnects in a single photo that we can enlarge it ?
Also a pic of the transformer with something to scale its size would help us estimate if it's powered at high frequency or at mains frequency.
-
A few details: The main transformer i am fairly sure is mains (60HZ) driven. When I connected the two leads on the contactor side of the solid state relay the main transformer hummed and i got voltage from gun to ground. I tested the solid state relay by putting a solenoid volt meter on it and spliced that into an old mains power cord, i connected the "coil" of the solid state relay to a small bench supply it calls for 3-32V DC I gave it 12V and the solenoid voltmeter activated when powered on the bench power supply, when removed power to the bench supply it shut off. With solid state relay removed i probed voltage and hit the trigger got like .035 volts. Solenoid seems to be same issue like they use the same drive transistor or they share the same trigger mechanism.
The part that was under the heatsink is an LM317T i had voltage on both sides of that. 12V DC and like 7V. I did read some places that these run on 24V but the power transformer seems to put out 12V. I can't imagine a transformer failing in a way that it just suddenly puts out half the voltage. This transformer is in the unit where the PCBs are, i could see it it looked basically like a doorbell or thermostat transformer, but getting it out looked like a ton of work and is spliced into the tether cord between the base unit and the unit with the PCB/ wire feed.
Pics attached showing interconnections, the LM317 with heatsink off, the main transformer next to a 1 quart can of paint.
Conn 4 pin 1,2 are to solenoid valve. Conn 3 pin 5,6 go to solid state relay. These both seem to go to the two 2n5193 transistors that tested ok. The base of the solenoid transistor goes to an SMD resistor marked 470, then goes to the trigger connection on Conn3 pin3. and would complete circuit when this wire connects to the negative line (conn3 pin4). The relay transistor base seems to link to the small transistor, the unmarked IC, and a resistor that connects to what seems like the output of the voltage regulator.
Links to youtube videos:
Main unit video: https://youtu.be/im1QWGmbdsQ
PCB video: https://youtu.be/2KoRRcakFws
-
Quick further question.
The 2 controls on the right panel in this pic are for what ?
(https://www.eevblog.com/forum/repair/snapon-ya-205-mig-welder-control-repair/?action=dlattach;attach=1002310)
-
The switch type control is the amperage settings and taps into the main transformer. The potentiometer is the wire motor feed speed control.
-
The switch type control is the amperage settings and taps into the main transformer. The potentiometer is the wire motor feed speed control.
Ah, OK. Sorta like the Automig then, then it will be mains straight to the primary not some SMPS design.
What test gear do you have ?
-
I have a klein DMM with capacitance measurement, i used that to test all the parts i could. I have a little tester kit from electronic goldmine i sued to test the transistors and all seemed fine. I have a fluke PM3094 scope but i am very new to scopes.
Values observed followed by rating
R1 1.012K, 1K 5%
R2 1.009K, 1K5%
R3 1.192K, 1.2K 5%
R4 1.485K, 1.2K 5%
R5 2.194K, 2.2K 5%
6-10= diodes diode test mode worked only one way as should registered about .5 on all of them.
11 56UF rated 47UF
12 3773 UF rated 3300 UF
SMD resistors value then numbers written on it.
20 47.1 470
21 .473K 471
25 2.719 272
There are some other white or brown parts near the unmarked IC but i can't tell what they are with no PCB markings and no markings on the parts.
Tested all of the pots, selector switch all registered about 10K ohm from both ends. I also did not notice any sign of loose contact when i connected the wiper terminal and turned the shaft.
To me the fact that even the gas valve does not activate and is tied directly to the base of the transistor tells me there is some issue with that transistor even though tested fine or there is some issue with the voltage regulator or power supply for this PCB.
I am partly tempted to just see what i have or find some new transistors and change those. Maybe even the caps, resistors but they do all test fine. out of circuit so not sure that would get me anywhere. I have not quite figured out what the MJ900 in the heatsink on the main PCB and the other MJ5105 seem to be wired in series and used to control the wire feed motor.
-
Most welders live in very dirty environments. From my own experience, I have found that dirt gets into the pots and makes them intermittent or worse. The openings in the pot cases you have pictured certainly are suspect as well as the open rotary switch. Another thing to look for are fusible resistors.
Just because it has a snap-on logo doesn't mean it is of the highest quality but probably not the worst, either.
Edit: I know that Century made some welders and other equipment like battery chargers for snap-on. If you can find a large welding supply that repairs welders in house they could tell you exactly who made it and the equivalent model from the real manufacturer probably from a few pictures.
According to this post (on a different forum) the YA 205 was made by Century.
https://www.pirate4x4.com/threads/snap-on-ya-205-welder.642716/#post-7789441 (https://www.pirate4x4.com/threads/snap-on-ya-205-welder.642716/#post-7789441)
"I have a YA205A that I bought new in '88. It's a 130 amp, 220V machine that was manufactured by Century for Snap on. And yes, it was insanely overpriced for what it was.
It's a great unit for light materials. I used it for bodywork mostly, but did some big stuff with it too. Runs awesome on the low end of the scale."
Maybe this will motivate the OP to get his free welder working.
:)
-
The unmarked white and brown components are multi layer ceramic caps not normally marked or troublesome.
-
The unmarked white and brown components are multi layer ceramic caps not normally marked or troublesome.
Ceramic caps do have a tendency to crack in unfriendly environments like welders may experience. Dave has done a video on this: https://www.youtube.com/watch?v=QgKY5QWehME (https://www.youtube.com/watch?v=QgKY5QWehME)
Thank you Tautech for running with this thread, you obviously have had more experience fixing welders than have I.
-
The unmarked white and brown components are multi layer ceramic caps not normally marked or troublesome.
Ceramic caps do have a tendency to crack in unfriendly environments like welders may experience. Dave has done a video on this: link removed.
Thank you Tautech for running with this thread, you obviously have had more experience fixing welders than have I.
Not heck of a lot but some all the same.
Your first comment reminds me of a small MIG caddy I real trouble fixing some years back after it had been dropped in the workshop. ::)
The PCB PSU transformer, a small PCB mount jobby with pins through the board had pulled one pin and broken the wire to it but as it was a potted transformer I couldn't reattach the wire although I did try. After replacing it the MIG apparently worked fine but came back to me in short order as it was intermittent.
Long story short, the PCB was flexed so badly by the mass of the transformer when originally dropped that the PCB when flexed would click indicating a broken component or solder joint yet despite careful inspection with magnification no cracks were found. :-//
Finally I reflowed every darn component on the PCB manually with an iron and the welder gave no further trouble. :phew:
Another, similar to the topic of this thread, but not a caddy but a reasonable sized wheeled jobby had cooked its Alternistor and the DIP package that controlled it. ::)
Its cooling fan was mounted just above a wire netting grille at the bottom of the unit and someone had dragged the welder over some junk on the floor pressing the grille into the fan which of course killed any forced cooling. ::)
The IC could not be sourced so I shoehorned a 8 pin DIP U2008B phase control IC into the 16 pin footprint well enough to connect with a few of the nets and some bodge wires did the rest. :phew:
Worked like a charm and still does. ;D
The things we do for mates. ::)
For NetworKing
I want you to download an LM317 datasheet and identify the resistors used in your PCB to set its output value.
Then if we measure it's Vout and it doesn't match set value, either something's dragging its output down or the LM317 or its setting resistors are faulty.
-
So i did a little testing;
I get 2-2.1V when i probe from the main positive to the base of the transistors with the trigger activated. From the datasheet for the 2N5193 this seems to be the low end of the trigger on voltage, but seems should still work.
The Vreg i could tell has one SMD resistor marked 392 and a couple other white caps? but my DMM could not measure them.
Here are some measurements though, i tried to look at the datasheet but the schematics they use seem backwards to what this board has.
Rectified PCB main input 14.2V DC.
I got 12.74V from - rail to the center pin
11.92V from - line to pin1 which ties to the - main with the 392 resistor
2.35v between pin 2 and 3
I don't know that much about Vregs to me it seems like it is ok, i tried to measure everything i could think to try that would help. To me it is now seeming like issue with the transistors as base voltage is there but they don't activate although is almost like voltage could be low too.
-
317 datasheet attached.
You can throw some numbers in this and see if it matches what you measure:
https://circuitdigest.com/calculators/lm317-resistor-voltage-calculator
-
Ok did a little more testing looks like the other (white) SMD part by the vreg is a 211.7 ohm resistor. The PCB has a .961K resistor between the main negative bus and pin 1on the vreg. There is a 211.7 ohm between pin1 and pin 2 on the vreg. Pin3 is the positive bus at about 14V DC. There is a small SMD cap between pin3 and the negative bus, i tried to test that cap i briefly got 4UF then never got a good reading on it again.
When I probe the vreg output i get 12.87V when probe the main negative bus and the output. When I probe the main positive bus to the vreg out i get 1.546V.
I put these values in the calculator site and first it seemed right said about 1.5V but i had R1 and R2 flipped. When i fixed this and put the 211,7 in for R1 and the .961K in for R2 I get 6.924V this is way off what i observed. I ams wondering though if maybe i did something backwards in calculation or if they are using a negative regulator as a positive or the other way around. Or maybe the Vreg is just toast.
-
~14V V+ in
12.9V V+ out
~1.55V dropout
Sounds OK to me.
R measurements could be wrong and measuring multiple paths so lets assume the Vreg is configured for ~12V circuit operation.
Little steps. ;)
-
Ok in that case what should i do next. If vreg is fine then change those transistors? Transistor seems like only remaining suspect at this point having established voltages are ok, the transistors get ~2V to the base. To me seems like that is my only culprit left even though tested ok however my transistor tester is made more for the little NPN/PNP than these things i just have some wire clip leads i use to connect these as the socket does not work on these. So totally possible they are actually bad and my tested did not identify it. I have seen a few ways a DMM could test transistors but there were so many ways and people saying did not work, etc. IU have no idea best option for that.
If i do replace them should i use same part or could i interchange with something the same spec i have or is there a better more rugged part i should use that may take shop abuse better?
-
Ok in that case what should i do next. If vreg is fine then change those transistors? Transistor seems like only remaining suspect at this point having established voltages are ok, the transistors get ~2V to the base. To me seems like that is my only culprit left even though tested ok however my transistor tester is made more for the little NPN/PNP than these things i just have some wire clip leads i use to connect these as the socket does not work on these. So totally possible they are actually bad and my tested did not identify it. I have seen a few ways a DMM could test transistors but there were so many ways and people saying did not work, etc. IU have no idea best option for that.
Sometimes transistors check out OK but their Hfe is down but you don't know this unless you compare against datasheets. If you have some known good new ones look up their gain spec and check you tester measures them correctly.
If i do replace them should i use same part or could i interchange with something the same spec i have or is there a better more rugged part i should use that may take shop abuse better?
No I would stick with OEM parts unless they're hard to get in which case you'll need to sub them with an equivalent or better device.
The SMD IC still has me puzzled as to what it's for and why its scrubbed clean. :-//
Looking at the PCB and its surrounds with wide open eyes we only need it for tack, stitch and continuous which makes me think it might be a monostable being used for timer duties and the rest of the PCB is assigned to wire feed variable motor drive and an output to both the gas valve and SS relay.
As much as it's a shit of a job it's time to reverse engineer the whole welder with a focus on the PCB but showing the cable loom pick offs and where they all go.
Write down all the cable colors too when you draw it out. Sorry, it isn't much fun however it can produce the goods when problems aren't obvious.
-
So i gave some thought what to do next and i am going to try and order a few 4-6 of those questionable transistors. That way ia have spare if blow one out. Only place i could really find with decent price for a few was newark. But shipping is like $8 figured what the heck i may as well order every part i think could be bad i don't have and throw it in too. So they have the MJ15015 part (i think this is part of motor drive) but few $ may be worth taking chance than not have and spend more shipping, etc. Anyhow They are showing 3 versions of it the one match to this is out of stock, the other 2 are more $ and higher HFE but i have no idea what that means. I know HFE is for measuring transistors, but besides that means nothing to me. Would it be problem to get a higher HFE one or should i wait for one to come in stock or just not bother?
I was also not able to find that little 2222 transistor on newark, so wondering if would be ok to swap with some similar substitute. I don't see issue in that but ideally it would be the same part. I can search some and see if can find equivalent, unless someone has a suggestion already for this. I'm inclined to say because the timing dials, etc work that the 555 ICs on that PCB are good so probably not worth hassle changing unless someone sees reason i should. Sure they are cheap enough to replace just a harder job to redo those.
-
I know HFE is for measuring transistors, but besides that means nothing to me. Would it be problem to get a higher HFE one or should i wait for one to come in stock or just not bother?
Transistor gain is specified within a range for each type and further some may have a suffix code indicating what sub-range they fall within.
Some transistors have 3 or more sub-ranges of gain listed in the datasheet.
When testing it's important they return a measurement that falls within its specified range.
A little reading while you're waiting for parts:
https://www.electronics-notes.com/articles/electronic_components/transistor/current-gain-hfe-beta.php (https://www.electronics-notes.com/articles/electronic_components/transistor/current-gain-hfe-beta.php)
Of course Google is your friend so do some further hunting for additional understanding if required.
-
I meant to add this in prior post, the part i need seems out of stock could i substitute one of those others in or not?
-
I meant to add this in prior post, the part i need seems out of stock could i substitute one of those others in or not?
Download the datasheet for the first one and compare its specs with the datasheet I posted earlier:
http://www.svntc.com/TPDF/625.pdf (http://www.svntc.com/TPDF/625.pdf)
Without looking I suspect it will be fine.
-
I ordered transistors, a few caps, i could not find the 2222 transistor on newark, the voltage regulator was only in stock in UK figured this one seems fine so not worth the $8 postage for that. Decided to skip the MJU5105 too as did not have exact match and was a bit of $ part and i think this one is fine. I could not find the part inside the tall heatsink on the mainboard anywhere. Says 931-s MJ900 has a big F on it so maybe a fairchild part? Again i think this is part of motor driver so not worth messing with i am thinking.
-
Did you ever get this fixed? I'm working on a few of these boards now.