Hey, I'll tell you just like I tell her. I fixed the overheating problem. Any other issues with it "not working" aren't my problem and probably unrelated to my solution.
in such cases it is not uncommon for the overheating issue to jump to the human operator of the machine, good thing is that water helps again, as short and long term solution, you lose one wife, but hey - the problem is solved!
Your diagram at 10:40 is actually a pretty good representation of what is going on. But you are missing a stage between the last rectifier and the torch / ground clamp. The first inverter stage you identified is the 40 KHz stage (assuming that is the frequency and you were not just ringing your clamp-on current probe) and is also used to control the current via PWM. This stage is several very large MOSFETs in your machine. After the rectifiers there is another stage before the torch / ground clamp that allows you to change the polarity of the applied voltage with respect to the ground clamp. This stage uses IGBTs and is also a form of "inverter" if you will, as it takes the DC that its rectifier (on the right in your drawing) produces and does the swapping around to put the right polarity on the torch. Now that was a little tiny bit simplistic. There are actually two output rectifiers (and chokes) and the secondary of that transformer is really two secondaries. The two secondaries feed the two rectifiers which feed two different halves of the second inverter which feed the torch. Given that you get good current control in the DC half of the cycle, I am fairly certain your issue is on this board or the board generating the control signals for this board. The good news is that with HUGE devices like these it should be simple to have the kitten tack-solder some flying test leads onto the interesting points and monitor them from the comfort and safety of away from the big scary boards. I would take a look at the unrectified AC coming off both the secondaries, and the control signals to the two banks of IGBTs (one labelled "DC" and one "AC"). If you can grab those control signals both before and after their opto-isolators, all the better. Somewhere in all that I suspect you will find a demon. I would certainly keep your scope running on battery for all this, as the ground reference may be all over the place. You should probably also avoid touching things. Just saying...
Hey Tony. Love the videos. I just started watching. I used to be a machinist before I went into the military as an avionics tech.From what I saw on the scope it looked like a bad capacitor or two. Capacitors, especially the small flat ones usually only last 5-8 years before they dry out and give up the ghost.I know this is a couple years late. But I figured it would be good info for the future fixing projects.Cheers
I really appreciate when you dempen the light from the welding, I know it's not gonna be dangerous at full bright, but it does hurt still. So I appreciate that.
Tony: "I need a new welder. " Wife: "You mean you want a new welder" Tony: "No, I really need one. Look, I made a whole video and the world agreed with me, then I made another one to prove it" Wife: "OK, I watched, looks to me like one of the red caps, but just go and spend the money, I wanna watch Breaking Bad".
Mispronunciation is the hallmark of someone that is self-taught or learns more from books than classes. Miller publishes shop manuals with separate power-off and power-on diags one can perform. I remember this well, because during a power-on diag my test probe slipped a millimeter and I fried one of the six or so $300-dollar boards in my old Dynasty 200 (which is also an inverter). Even with the the manual I eventually got frustrated and sold the thing on eBay for parts. I ended up getting an Everlast, which is a Chinese brand, but, as you say, it's still a better welder than I am. Also it costs about one tenth of what the Miller did. I never would have admitted this online for fear of flak from pro welders, but then I saw how Jamie over at WeldingTipsAndTricks demos Everlast welders all the time.
thzzzt pro welder here, everlast, ahp, and one or two other “off-brand” welders i cant remember the name of, they are actually amazing little machines, they definitely aren’t made for industrial (working construction sites, inside of plants) but small travel jobs that have power supplies (radio towers, sub stations, hydro plants and the like) as long with in a home shop, or a newbi just trying to learn on a “cheap” priced machine, im looking into one of them in the future for personal usage
99.9% of the time electrolytic filter caps are the reason a switching power supply fails. Check out Mr. Carlson's Lab. He does a very thorough video about repairing industrial switching power supplies for CNC machines
Yeah.....agreed...! I made that comment on Part 1, but others said it was too expensive....you can just unsolder and test with DVM and get some idea of the condition.... If this is a half decent ($$$) welder, should be worth the effort.
I'd say it's more like 90 when you specifically look at the filter caps. maybe 95% when you count all possible caps. The rest is diodes, mosfets, support circuity and so on (just had a 40V PoE smps fail on me because the small support cap of the control circuity on the HV primary side failed silently with high ESR and lowered capacity but visually ok) Generally I recommend www.repairfaq.org/sam/smpsfaq.htm (maybe not the best for an inverter welder but the tips and ideas are still great). Oh and testing the Caps is really easy with an ESR meter like the cheap KKmoon Diode/Transistor/Triode/Resistor/ESR testers (based on an OSHW design/concept from here www.mikrocontroller.net/articles/AVR_Transistortester )
I was thinking along the same line. With the machine being a little long in the tooth, in an industrial setting, I’d be tempted to replace as many of the electrolytic caps as I could. They do dry out in time. I think there’s another fault going on too, though no idea what.
I enjoyed this. Your troubleshooting methods are familiar. I know a little, definitely not a lot, about electronics, can identify components, and solder well. So, I take apart, test what I can, look for obvious problems, black things that shouldn't be black, or things that smell bad. It works more often than I'd expect.
The AHP tig welder is an excellent choice for the price . Great duty cycle, lots of bells and whistles, good accessories all for under 800$. Had mine for two years with no problems.
EnlightenedSavage I second this. I have the latest gen bought this spring. I would never have even considered a tig welder at the prices of a blue or red machine. The AHP will weld anything I need for my garage.
Hi, NIce videos, very informative. Just a quick thought, the bit where you turn on the current probe showed a low battery led lit ( at 5:05) . if you turn round the current probe on the earth clamp it should invert the waveform on the scope. if it doesn't the probe is asymetric due to a flat battery ( or faulty). Keep up the good work. Kevin
Whoa - BATTERY powered scope?!? Now that I've got to get! (Lately I've been working on mains-powered equipment, and been needing to use an isolation transformer so I could measure mains-referenced stuff using my current scope)
What if the "logic board" is actually full of op-amps and comparators cleverly controlling all the parameters of the waveform? Judging by the two small electrolytic capacitors, trim pots and high number of resistors, this might be a fair assumption. You could look up chip part numbers and check whether they are powered correctly. Could be that one of the board's power supply rails is down, which might be where 50% maximum duty cycle number comes from. (assuming dual rail power supply, e.g. +-15V). And yeah, if the assumptions above are correct, the voltages across the two SMD electrolytic capacitors near the ribbon cable on the left at 12:35 should be equal in order for the circuit to operate properly - easy check.
13:10, it is likely that one of the relays is to limit in-rush current when the machine is plugged into mains. Otherwise it will try to charge all capacitors instantly and easily do 50 or 100 A from the mains for maybe half a second. Instead after rectification, with relay off, the rectified mains are connected to capacitors via a current limiting reistor. Once they are charged (based on voltage monitoring or just timer), the relay shorts the resistor and allow fully current to be used.
I would suggest using your IR camera to look for hot capacitors (greater ESR of failing cap causing more dissipation) or other unusually hot components
I've been using my AHP Alphatig that I picked up in 2015 and I've been loving it. They expanded a couple of the setting options for the newer models. Duty cycle is 60% at 150A and 100% at 115A. You should check it out.
I would check the full-wave rectifier module, which is the potted module with 4 spade tabs that appears in the left edge of the screen at about 12:50 elapsed time. Just a thought.
i dunno why there is always such a big rush to get in early, but this is the earliest ive ever been... This old tony, your content is awesome, love it, keep it up!
try switching every black cable with its corresponding red cable, it should just flip the output wave to the opposite, and you'll generate electricity every weld you make. trust me, im a whitecoat
One aspect of using it might make it worse, it might make it worse enough that the magic smoke comes out of the failing part which makes it a LOT easier to diagnose.
If you decide to throw it away please post about it on the forum, I' m 16 years old and I started learning stick welding on my own three years ago, i'm pretty good at it, but i never weld anything thicker than 3 mm so i was looking for something better (aka TIG) but at the moment I can't afford a new machine and most of the used ones you find here are three phase or really over priced Something used, even if has some defects and doesn't run properly would still be great As always a great video !
Thank God. I was running out of entertainment. Dude if were you I would buy an AHP 200. I got one last year. That machine is fantastic....and cheap too.
From my experience working at a well known welder manufacturer, all the boards are tested individually as they are made and them put together in final assembly. And each board powered by a power supply. So you would have to test each board individually under power to find the problem.
On the board with all the orange caps, looks like a cracked MOV (center back, behind the two fat black wires. The case is cracked across the face. see on the left side below the relay you are talking about at 12:48.
you need to be able to look at the triggers to the IGBT ( i presume it has them) which floats on the supply voltage, so care must be taken. There are usually isolated dc-dc converters driving the high side.
lol. Great video. Typically I preplan a failed disassembly as an excuse to purchase a new welder . By the way , I have a "this old tony" shrine i worship and light a candle to every night before I go to bed .
I'd suggest checking the capacitors. They are one of the few solid state components that actually go bad. Also it's only a few minutes to look for bloated caps.not sure how to actually test them.
"So you're getting predictable, Tony," I thought to myself as you 'heard something rattling around.' *Tools and toys fall out...* "Ah hah! I am so sma..." *kitten* "I bow to thee, my Master."
Really graphic video, very useful to know how this stuff works. There is not much more you can do with the tooling available, it was probably time for a machine renewal. I am trying to install on my Oerlikon 1500 ACDC, which has the same control card as this one, the potentiometer to limit the pedal current. But I wonder how you installed one with 2 cables when they usually have 3.
13:06 inrush. When the caps are charged the relays will close to bypass current limiting resistors. Gas solenoid doesnt require a relay, its an electromechanical device in itself.
At 3:22 the region around the 10K resistors shows cold solder joints. The joints to the right of the board look nice and shiny but those look really bad. Before checking anything else, I would reflow/solder those as they look very suspect. Looking at the wave forms from your first video would also lead me to believe the problem lies in the control circuitry.
7:00 God damnit. After the last video, I literally went on a 2 hour long jaunt in to reading articles and watching The Organic Chemistry Teacher videos on how transformers and DC power supplies work (again) to understand how a welder can turn a 20A source in to 150A to weld with, and here you are, explaining it in the very next video. 🤦♂️ Thanks for going over this anyway, lol
I notice 2 surface mounted electrolytic capacitors on the logic board . These '' caps '' are known to cause problems after some time . You Can try to heat ( hair dryer) or cool ( a can of contact 75. freeze ) them , watch what happens with the wafeform on the oscilloscope .
First I thought maybe because of the cat :) then I realized, this is the high frequency oscillator with four legs. Positive, negative, control voltage and the output :)
Look on the bright side, that asymmetric current you have is actually a feature on some much higher priced welders. :D Be glad it's favoring the negative at least. Plus you scored another kitten out of the deal.
Dear Tony. I have been working as an électrotechnicien for 35 years. I might be wrong, but just in case. Looking at the wave form, you have the plus and minus wave forms. So, triacs are good. It is I believe what controls those triacs That is Wong. I would not dig too far deep in the control boards. Electronic and components are just mathematic, It works or it don’t. It does not work halfway. Capacitors especially the electrolytic can loose there capacitance with time. I don’t think it is your case. In my experience, what wears most are the moving parts. In your machine case, the potentiometers. Check those out first.
The click after power-on is probably a soft-start. The primary-caps would be charged initially through a big R so as to not blow up the rectifier. Once they reach a good fraction of working voltage, the relay closes and shorts the resistor. It's a very low-tech approach. PFC would be much better and more efficient but this is cheaper and just good enough to not pop your breaker at power-on.
Remote debugging is always tricky, but here is my 2 cents: Check if the 4 capacitors are coupled in series and if there is a center track going out ( and maybe back to L or N on bridge rectifier ). If they are, there is a chance this serves as a +/- supply for the inverter and could account for the lack of current in one polarity and the duty-cycle problem. Measure capacity, AFTER checking voltage is 0 across all of them. They should all read the same, about 470 uF.
For someone like me, with little to none electrical knowledge, I would de-solder all capacitors and check'em. I even saw a guy here on UA-cam checking them in place. I don't wanna be a tool-blocker here (almost the same, but blocking new tool purchases), try it, and if you manage to fix it, you can still have your new tig. Just send the old one to me :-D
13:08 that's probably something I would describe as soft start or anti inrush current circuit. I guess when u type "what is in rush curent" there should be many more clear descriptions of subject than I would do.
Tony, because you are missing about 1/2 of your positive voltage it might be a Diode after the transformer. By using 2 diodes and 2 caps you can double the voltage of the secondary winding. if 1 fails, you half the voltage.
I'd bet that surface mount logic board is full of op-amps and comparators, given the layout and pin count. Maybe some inverters to create an oscillator for the pulse and AC features.
Not easily. Low heat and a nickel-iron rod will work, but it smokes a lot, so use a respirator and be sure to wear hearing protection to avoid the loud screeches.
weshowe - I've had good results using E7014 with HF stabilization but that was on a long-haired breed. It's hard to keep the arc lit on longhairs without the HF. Fume extractor is a must.
Also, I'm all for replacing caps and mosfets until you find the right one but I'd take a page out of the millwright's book and let it buck. Do all the aluminum and heavy steel you can until it gives up the ghost. The check what let go and replace it, what ever is up stream to be on the safe side and then put it on Craigslist if it starts working again. At this point Mrs. This Old Tony will be so tired of hearing about welders that a nice new Esab, Alpha or Thermal Arc will slide right past the radar.
As somebody else already said, the delayed relay is one to allow full line current AFTER the big capacitors have been charged up. The inrush limiter resistors (likely) would combust if not bypassed by the weld time. On the other hand, without the limiter, the inrush could be 200 to 600 amperes and would definitely pop the breaker. Without seeing your part 1 video, I speculate that your issue is the Variable Polarity pulse duty cycle that remains fixed at about 50%.. The problem must be in the oscillator (analog or digital) section. Don't try to replace the big red capacitors, until they are at least 10 years old. Or even older if they are rated 105 degrees C temperature - see what is printed on them. The original AC welders for aluminum used line frequency and got more or less 50% pulse duty cycles. The inverter generation can have true variable duty cycles, with the Reverse Polarity (Electrode Positive) part typically from 10 to 20 %. In TIG (or officially GTAW for Gas Tungsten Arc Weld) the controlled feature is current. The voltage is set by the shield gas and the arc length. In automated welding the voltage is measured and used to run a servo motor to raise or lower the torch. Your VP welder primary inverter works O.K. There must be another inverter or bridge AFTER the transformer in order to provide the weld current reversals. Moreover, the arc starter may need to be fired on every polarity reversal moment to prevent a loss of the arc. As you get the polarity reversals, even the VP bridge is O.K. That means the problem is in the control section of the VP bridge, close to the potentiometer. With argon as a shield gas, the straight polarity voltage is around 10 V, while the reverse polarity time it is likely to be around 25 V. As the above should suggest, I have electronics control experience and even in welding system design. But I can do only so much diagnosis remotely...
Heck, if you still have it, I'd try replacing the caps in the power supply. Maybe you did a vid on this actually fixed and I just haven't gotten there yet in my search through your older vids 👀
I tell you what.... it's going to be a bad capacitor or a dirty pot! Nothing more! Just use a propriety pot track cleaner and your good to go for another ion! Thumbs up from Spain amigo!
With your penchant for referencing other channels, I was fully expecting an "Electro BOOM" shock sequence at the end, and some dancing eyebrows.... 8^D
you should look into avortec welders, i bought one that does mig, tig, stick, and plasma cuts all in one unit and i love it. and if you ever had an issue with it, you could get the guy that designed and built it on the phone. you can get him on the phone before you buy it as well. i bought it a year ago, and just find myself wanting to use it more than i have time for.
you could have a bad zener diode that it flowing voltage to early I would look for one shorted open all the time of a capacitor that is dry that has not failed all the way yet so you would need to check the ESR so get out the soldering iron and start pulling caps and checking them. the diodes you can check with out power or desoldering. if you need someone else to check it I would be glad to help.
How much do you trust the test equipment, especially that amp meter? Not to cast more doubt and confusion on your pile, but any chance you have a different current meter you could try? The sillyscope seems skookum enough, but hall effect sensors can get ausgeficked surprisingly easily. At least it's an easy double check before tearing apart the entire....oh right. Anyway, best of luck! Love your channel.
I noticed a 8 pin din surface mount component that is inserted into a socket. These are notorious for having “chip creep” that is when the component begins to push it way out of the socket due to vibration. Simply removing the component and reinserting it back in fixes the problem.
As a pixie wrangler (bachelor in electronics, and studying for a master's in Electronics Engineering) my professional opinion is that you won't be able to fix this thing. You gave it a good shot, but with the horrible physical layout, the lack of a service manual and the real danger that come with testing this thing under power: you're not going to get this fixed. The only way you're fixing this, is by getting some extra help and a lot of luck. You got two options: give it to someone who can fix it or run it till you can buy a new one.
Balance Adjust Pot...hmmmmm...but I agree with the post below as regards probably an electrolytic capacitor going bad in a circuit that controls switching times. The top bulges as they go bad (sometimes). It's really kind of a time thing - how much time is it worth spending if you want a larger/newer machine anyway? If you can't find a schematic or service manual, the complexity of troubleshooting tends toward infinity - not impossible, just time-consuming.
The bridge rectifier on the input side is probably rated for 35 Amps and up to 600 Volts. It can be this small because the current from your outlet is probably about 20 amps at 120 volts, this is before it gets stepped down, the diodes on the output side are rated for a lot more, probably hundreds of Amps and to handle that kind of current they need to be much larger. You said earlier in the video you didn't know anything about switch mode power supplies but you know a lot, it sounds like you understand the theory behind it, which is more than most people.
I was scanning the daughter board and I am fairly certain that there is a poor connection between your temporal flux capacitor and the half-nut on your lathe. Probably intertemporal oxidation. Try a 1/4 teaspoon of baking soda in your next batch of marinara . It makes a huge difference in the sweetness of the sauce.
It looks like it can't equalize the bias between the two periods of the square wave. This might happen when a capacitor in the main supply shorts internally when it warms up (actually the internal resistance falls). It might be also the same thing but on the IGBT board, such as one of the filter/bypass capacitors. If those capacitors are not rated for the temperature they work at (typical of Chinese machines), they start to give up in a similar way.
electrolytic capacitors over time tend to dry out and lose capacity especially at hot environment it usually cause similar problems at the beginning of that.
Tony this is great, fantastic, amazing, unbelievable how much you know about it.......... I did not understand a word of it. I know nothing about electronics. so my question is: why are you waisting your time with mechanical stuff when this is far above it, it's a higher stage of the technical world. I hope I will enjoy more of your presentation which I won't understand at all but people with more brain I have will greatly appreciate. I only have a Lincoln 175 and after years I already know how to fuse steel, never mind I live in in last centufy. That Old Bob
12:40 Is there any way that the knobs were turned during re-assembly into the pots when you first had it apart? I can't tell, but if the center of the pots are hexagonal, it wouldn't be difficult to have it off by 60° which might give a bad value to the control board.
Lots of interesting stuff fell outta there when you shook it! Lol. If you shook it harder the factory worker that built it might have fallen out as well!
The very flat top on the waveform suggests to me that the confuser thinks that's where it's supposed to be. I think the problem's on the board that's taking in all the voltage & current monitors and doing the A to D.
When it overheats you should just pour water on it. That's what I did to my wife's laptop and it hasn't overheated since.
sure, it hasn't overheated, but does it still turn on?
Cant overheat of it doesnt turn on.
Hey, I'll tell you just like I tell her. I fixed the overheating problem. Any other issues with it "not working" aren't my problem and probably unrelated to my solution.
in such cases it is not uncommon for the overheating issue to jump to the human operator of the machine, good thing is that water helps again, as short and long term solution, you lose one wife, but hey - the problem is solved!
Oh, Al!
This old Tony, now with 100% more cats!
That's a move that I certainly approve!
Rumors say that Tonys cat is called "Schrödinger".
well this is the internet so cats are to be expected...
nice!
So, he had half a cat in other videos??? :-O
Your diagram at 10:40 is actually a pretty good representation of what is going on. But you are missing a stage between the last rectifier and the torch / ground clamp. The first inverter stage you identified is the 40 KHz stage (assuming that is the frequency and you were not just ringing your clamp-on current probe) and is also used to control the current via PWM. This stage is several very large MOSFETs in your machine. After the rectifiers there is another stage before the torch / ground clamp that allows you to change the polarity of the applied voltage with respect to the ground clamp. This stage uses IGBTs and is also a form of "inverter" if you will, as it takes the DC that its rectifier (on the right in your drawing) produces and does the swapping around to put the right polarity on the torch.
Now that was a little tiny bit simplistic. There are actually two output rectifiers (and chokes) and the secondary of that transformer is really two secondaries. The two secondaries feed the two rectifiers which feed two different halves of the second inverter which feed the torch. Given that you get good current control in the DC half of the cycle, I am fairly certain your issue is on this board or the board generating the control signals for this board.
The good news is that with HUGE devices like these it should be simple to have the kitten tack-solder some flying test leads onto the interesting points and monitor them from the comfort and safety of away from the big scary boards. I would take a look at the unrectified AC coming off both the secondaries, and the control signals to the two banks of IGBTs (one labelled "DC" and one "AC"). If you can grab those control signals both before and after their opto-isolators, all the better. Somewhere in all that I suspect you will find a demon. I would certainly keep your scope running on battery for all this, as the ground reference may be all over the place. You should probably also avoid touching things. Just saying...
Great post! Tack-soldering on some flying test leads is a brilliant idea; saved for future reference...
Hey Tony. Love the videos. I just started watching. I used to be a machinist before I went into the military as an avionics tech.From what I saw on the scope it looked like a bad capacitor or two. Capacitors, especially the small flat ones usually only last 5-8 years before they dry out and give up the ghost.I know this is a couple years late. But I figured it would be good info for the future fixing projects.Cheers
The cgi scene is like a 90s intel commercial... :D
It's all about the Pentiums.
I was thinking How It's Made. Either way yes.
I started rendering it in the 90s, only finished a few days ago. :)
@@ThisOldTony just about fell out of my chair when I read this message .. lol
I really appreciate when you dempen the light from the welding, I know it's not gonna be dangerous at full bright, but it does hurt still. So I appreciate that.
Your videos are getting better every damn time... Congrats man...
Thanks FCA!
Tony: "I need a new welder. "
Wife: "You mean you want a new welder"
Tony: "No, I really need one. Look, I made a whole video and the world agreed with me, then I made another one to prove it"
Wife: "OK, I watched, looks to me like one of the red caps, but just go and spend the money, I wanna watch Breaking Bad".
thisnicklldo n
Mispronunciation is the hallmark of someone that is self-taught or learns more from books than classes.
Miller publishes shop manuals with separate power-off and power-on diags one can perform. I remember this well, because during a power-on diag my test probe slipped a millimeter and I fried one of the six or so $300-dollar boards in my old Dynasty 200 (which is also an inverter). Even with the the manual I eventually got frustrated and sold the thing on eBay for parts. I ended up getting an Everlast, which is a Chinese brand, but, as you say, it's still a better welder than I am. Also it costs about one tenth of what the Miller did. I never would have admitted this online for fear of flak from pro welders, but then I saw how Jamie over at WeldingTipsAndTricks demos Everlast welders all the time.
thzzzt pro welder here, everlast, ahp, and one or two other “off-brand” welders i cant remember the name of, they are actually amazing little machines, they definitely aren’t made for industrial (working construction sites, inside of plants) but small travel jobs that have power supplies (radio towers, sub stations, hydro plants and the like) as long with in a home shop, or a newbi just trying to learn on a “cheap” priced machine, im looking into one of them in the future for personal usage
@@Lazybiker60 Thanks, this helps me get an idea of what I should be looking at!
This Old Tony, you made my Saturday evening complete. It didn't take much, but you did it. Keep posting, I enjoy everything you make!
If the kitten couldn't figure it out there's no hope.
99.9% of the time electrolytic filter caps are the reason a switching power supply fails. Check out Mr. Carlson's Lab. He does a very thorough video about repairing industrial switching power supplies for CNC machines
Yeah.....agreed...! I made that comment on Part 1, but others said it was too expensive....you can just unsolder and test with DVM and get some idea of the condition....
If this is a half decent ($$$) welder, should be worth the effort.
Doug Ankrum lol I did too. Easy fix if that's all it is
I'd say it's more like 90 when you specifically look at the filter caps.
maybe 95% when you count all possible caps.
The rest is diodes, mosfets, support circuity and so on (just had a 40V PoE smps fail on me because the small support cap of the control circuity on the HV primary side failed silently with high ESR and lowered capacity but visually ok)
Generally I recommend www.repairfaq.org/sam/smpsfaq.htm (maybe not the best for an inverter welder but the tips and ideas are still great).
Oh and testing the Caps is really easy with an ESR meter like the cheap KKmoon Diode/Transistor/Triode/Resistor/ESR testers (based on an OSHW design/concept from here www.mikrocontroller.net/articles/AVR_Transistortester )
I was thinking along the same line. With the machine being a little long in the tooth, in an industrial setting, I’d be tempted to replace as many of the electrolytic caps as I could. They do dry out in time. I think there’s another fault going on too, though no idea what.
I enjoyed this. Your troubleshooting methods are familiar. I know a little, definitely not a lot, about electronics, can identify components, and solder well. So, I take apart, test what I can, look for obvious problems, black things that shouldn't be black, or things that smell bad. It works more often than I'd expect.
paused the video and tried the knob-rectifier-trick before you revealed the joke. electrifying experience
no, seriously, I love your humor
this channel is pretty much a bunch of really well filmed dad jokes.
The AHP tig welder is an excellent choice for the price . Great duty cycle, lots of bells and whistles, good accessories all for under 800$. Had mine for two years with no problems.
EnlightenedSavage I second this. I have the latest gen bought this spring. I would never have even considered a tig welder at the prices of a blue or red machine. The AHP will weld anything I need for my garage.
New $4k machine: Justified.
LOL @ 5:50.
Great writing & production as always.
Hi,
NIce videos, very informative.
Just a quick thought, the bit where you turn on the current probe showed a low battery led lit ( at 5:05) .
if you turn round the current probe on the earth clamp it should invert the waveform on the scope. if it doesn't the probe is asymetric due to a flat battery ( or faulty).
Keep up the good work.
Kevin
12:07 top board is inverter drive, then the logic board. The board with heatsink is primary inverter.
If you haven't subscribed yet, you need to go to 3:45, freeze, and read carefully...
Whoa - BATTERY powered scope?!? Now that I've got to get! (Lately I've been working on mains-powered equipment, and been needing to use an isolation transformer so I could measure mains-referenced stuff using my current scope)
I know right!
Tony's version of BOLTR, with less swearing and half-arsed guesses.
the half-arsed guess ratio seemed pretty equal to me
Thumbs up....seems about right about BOLTR...
I LOVE BOLTR, but I don't have to shut off Tony when the kids come into the room. I really appreciate that.
Even AvE turns off BOLTR when the kids walk into the room.
Thank god... I thought I was the only one who watches ave and tot
Can't wait for TIG part 3, making out with the new machine.
Love your videos. Educational and entertaining.
What if the "logic board" is actually full of op-amps and comparators cleverly controlling all the parameters of the waveform? Judging by the two small electrolytic capacitors, trim pots and high number of resistors, this might be a fair assumption. You could look up chip part numbers and check whether they are powered correctly. Could be that one of the board's power supply rails is down, which might be where 50% maximum duty cycle number comes from. (assuming dual rail power supply, e.g. +-15V). And yeah, if the assumptions above are correct, the voltages across the two SMD electrolytic capacitors near the ribbon cable on the left at 12:35 should be equal in order for the circuit to operate properly - easy check.
13:10, it is likely that one of the relays is to limit in-rush current when the machine is plugged into mains. Otherwise it will try to charge all capacitors instantly and easily do 50 or 100 A from the mains for maybe half a second. Instead after rectification, with relay off, the rectified mains are connected to capacitors via a current limiting reistor. Once they are charged (based on voltage monitoring or just timer), the relay shorts the resistor and allow fully current to be used.
I would suggest using your IR camera to look for hot capacitors (greater ESR of failing cap causing more dissipation) or other unusually hot components
I've always called them silly scopes, so your pronounciation is fine by me.
I've been using my AHP Alphatig that I picked up in 2015 and I've been loving it. They expanded a couple of the setting options for the newer models. Duty cycle is 60% at 150A and 100% at 115A. You should check it out.
I would check the full-wave rectifier module, which is the potted module with 4 spade tabs that appears in the left edge of the screen at about 12:50 elapsed time. Just a thought.
Wow! this is my new favorite channel. How in the world do you not have a million subscribers?
i dunno why there is always such a big rush to get in early, but this is the earliest ive ever been...
This old tony, your content is awesome, love it, keep it up!
Thanks Woggin!
try switching every black cable with its corresponding red cable, it should just flip the output wave to the opposite, and you'll generate electricity every weld you make. trust me, im a whitecoat
One aspect of using it might make it worse, it might make it worse enough that the magic smoke comes out of the failing part which makes it a LOT easier to diagnose.
If you decide to throw it away please post about it on the forum,
I' m 16 years old and I started learning stick welding on my own three years ago, i'm pretty good at it, but i never weld anything thicker than 3 mm so i was looking for something better (aka TIG) but at the moment I can't afford a new machine and most of the used ones you find here are three phase or really over priced
Something used, even if has some defects and doesn't run properly would still be great
As always a great video !
Thank God. I was running out of entertainment. Dude if were you I would buy an AHP 200. I got one last year. That machine is fantastic....and cheap too.
From my experience working at a well known welder manufacturer, all the boards are tested individually as they are made and them put together in final assembly. And each board powered by a power supply. So you would have to test each board individually under power to find the problem.
Check the ESR (equivalent series resistance) of the capacitors that you found - maybe one of them is on the way out?
Wild guess: a bad cap somewhere in the control circuit for the negative half of the cycle... perhaps on the boost converter for the mosfet driver? 🤔
Yeah, definitely looks like something bad in the mosfet/igbt bridge that switches the welding current polarity.
dark winter,, is that called clipping when part of waveform Peak, is down like that?
Soooo... on the board with the high current connections going off it, under the big heatsink Tony didn't remove, then?
Yeah, What these guys ^ said !!!
Hippie Engineering Works its always the one thing you dont check
2 very informative and interesting videos on TIG! Brilliant!
On the board with all the orange caps, looks like a cracked MOV (center back, behind the two fat black wires. The case is cracked across the face. see on the left side below the relay you are talking about at 12:48.
you need to be able to look at the triggers to the IGBT ( i presume it has them) which floats on the supply voltage, so care must be taken. There are usually isolated dc-dc converters driving the high side.
Check the solder joints on all power devices. Thermal cycling usually causes bad joints.
lol. Great video. Typically I preplan a failed disassembly as an excuse to purchase a new welder . By the way , I have a "this old tony" shrine i worship and light a candle to every night before I go to bed .
I'd suggest checking the capacitors. They are one of the few solid state components that actually go bad. Also it's only a few minutes to look for bloated caps.not sure how to actually test them.
"So you're getting predictable, Tony," I thought to myself as you 'heard something rattling around.' *Tools and toys fall out...* "Ah hah! I am so sma..." *kitten* "I bow to thee, my Master."
loving the post production! keep up the good work tot.
Thanks Tom!
Really graphic video, very useful to know how this stuff works. There is not much more you can do with the tooling available, it was probably time for a machine renewal.
I am trying to install on my Oerlikon 1500 ACDC, which has the same control card as this one, the potentiometer to limit the pedal current. But I wonder how you installed one with 2 cables when they usually have 3.
13:06 inrush. When the caps are charged the relays will close to bypass current limiting resistors. Gas solenoid doesnt require a relay, its an electromechanical device in itself.
At 3:22 the region around the 10K resistors shows cold solder joints. The joints to the right of the board look nice and shiny but those look really bad. Before checking anything else, I would reflow/solder those as they look very suspect. Looking at the wave forms from your first video would also lead me to believe the problem lies in the control circuitry.
7:00 God damnit. After the last video, I literally went on a 2 hour long jaunt in to reading articles and watching The Organic Chemistry Teacher videos on how transformers and DC power supplies work (again) to understand how a welder can turn a 20A source in to 150A to weld with, and here you are, explaining it in the very next video. 🤦♂️ Thanks for going over this anyway, lol
I notice 2 surface mounted electrolytic capacitors on the logic board . These '' caps '' are known to cause problems after some time . You Can try to heat ( hair dryer) or cool ( a can of contact 75. freeze ) them , watch what happens with the wafeform on the oscilloscope .
First I thought maybe because of the cat :) then I realized, this is the high frequency oscillator with four legs. Positive, negative, control voltage and the output :)
Look on the bright side, that asymmetric current you have is actually a feature on some much higher priced welders. :D
Be glad it's favoring the negative at least. Plus you scored another kitten out of the deal.
ha! exactly what I thought when I saw it. "Wait this welder does asymmetry?!"
im here for the humor. I learn a little bit too.
Dear Tony. I have been working as an électrotechnicien for 35 years. I might be wrong, but just in case. Looking at the wave form, you have the plus and minus wave forms. So, triacs are good. It is I believe what controls those triacs That is Wong. I would not dig too far deep in the control boards. Electronic and components are just mathematic, It works or it don’t. It does not work halfway. Capacitors especially the electrolytic can loose there capacitance with time. I don’t think it is your case. In my experience, what wears most are the moving parts. In your machine case, the potentiometers. Check those out first.
The click after power-on is probably a soft-start. The primary-caps would be charged initially through a big R so as to not blow up the rectifier. Once they reach a good fraction of working voltage, the relay closes and shorts the resistor.
It's a very low-tech approach. PFC would be much better and more efficient but this is cheaper and just good enough to not pop your breaker at power-on.
Remote debugging is always tricky, but here is my 2 cents:
Check if the 4 capacitors are coupled in series and if there is a center track going out ( and maybe back to L or N on bridge rectifier ).
If they are, there is a chance this serves as a +/- supply for the inverter and could account for the lack of current in one polarity and the duty-cycle problem.
Measure capacity, AFTER checking voltage is 0 across all of them. They should all read the same, about 470 uF.
For someone like me, with little to none electrical knowledge, I would de-solder all capacitors and check'em. I even saw a guy here on UA-cam checking them in place. I don't wanna be a tool-blocker here (almost the same, but blocking new tool purchases), try it, and if you manage to fix it, you can still have your new tig. Just send the old one to me :-D
that dark spot on the back of that circuit board at 13:45 at the bottom near that copper strap.... is that a shadow, or is it a burn
13:08 that's probably something I would describe as soft start or anti inrush current circuit. I guess when u type "what is in rush curent" there should be many more clear descriptions of subject than I would do.
Tony, because you are missing about 1/2 of your positive voltage it might be a Diode after the transformer. By using 2 diodes and 2 caps you can double the voltage of the secondary winding. if 1 fails, you half the voltage.
I'd bet that surface mount logic board is full of op-amps and comparators, given the layout and pin count. Maybe some inverters to create an oscillator for the pulse and AC features.
and the constant current control probably
I'm getting biased, I click like before watching the video
Your just expressing the joy of seeing another This Old Tony video.
That's how Stalkers begin....
Thanks!
This Old Tony is my new favorite channel, up there with Cody's Lab And Matthias Wandell
Does the kitten weld?
Now that he fell into the Black Hole chip, he probably doesn't do much of anything ... endlessly.
Robert Pinder. typical cat
Not easily. Low heat and a nickel-iron rod will work, but it smokes a lot, so use a respirator and be sure to wear hearing protection to avoid the loud screeches.
Does the kitten weld?
The psi-function of the system contains both the kitten welding, and the kitten not welding.
weshowe - I've had good results using E7014 with HF stabilization but that was on a long-haired breed. It's hard to keep the arc lit on longhairs without the HF. Fume extractor is a must.
Also, I'm all for replacing caps and mosfets until you find the right one but I'd take a page out of the millwright's book and let it buck. Do all the aluminum and heavy steel you can until it gives up the ghost. The check what let go and replace it, what ever is up stream to be on the safe side and then put it on Craigslist if it starts working again. At this point Mrs. This Old Tony will be so tired of hearing about welders that a nice new Esab, Alpha or Thermal Arc will slide right past the radar.
As somebody else already said, the delayed relay is one to allow full line current AFTER the big capacitors have been charged up. The inrush limiter resistors (likely) would combust if not bypassed by the weld time. On the other hand, without the limiter, the inrush could be 200 to 600 amperes and would definitely pop the breaker.
Without seeing your part 1 video, I speculate that your issue is the Variable Polarity pulse duty cycle that remains fixed at about 50%.. The problem must be in the oscillator (analog or digital) section. Don't try to replace the big red capacitors, until they are at least 10 years old. Or even older if they are rated 105 degrees C temperature - see what is printed on them.
The original AC welders for aluminum used line frequency and got more or less 50% pulse duty cycles. The inverter generation can have true variable duty cycles, with the Reverse Polarity (Electrode Positive) part typically from 10 to 20 %.
In TIG (or officially GTAW for Gas Tungsten Arc Weld) the controlled feature is current. The voltage is set by the shield gas and the arc length. In automated welding the voltage is measured and used to run a servo motor to raise or lower the torch.
Your VP welder primary inverter works O.K. There must be another inverter or bridge AFTER the transformer in order to provide the weld current reversals. Moreover, the arc starter may need to be fired on every polarity reversal moment to prevent a loss of the arc. As you get the polarity reversals, even the VP bridge is O.K. That means the problem is in the control section of the VP bridge, close to the potentiometer.
With argon as a shield gas, the straight polarity voltage is around 10 V, while the reverse polarity time it is likely to be around 25 V.
As the above should suggest, I have electronics control experience and even in welding system design. But I can do only so much diagnosis remotely...
Heck, if you still have it, I'd try replacing the caps in the power supply. Maybe you did a vid on this actually fixed and I just haven't gotten there yet in my search through your older vids 👀
I tell you what.... it's going to be a bad capacitor or a dirty pot! Nothing more! Just use a propriety pot track cleaner and your good to go for another ion!
Thumbs up from Spain amigo!
Thanks Guillermo!
With your penchant for referencing other channels, I was fully expecting an "Electro BOOM" shock sequence at the end, and some dancing eyebrows.... 8^D
you should look into avortec welders, i bought one that does mig, tig, stick, and plasma cuts all in one unit and i love it. and if you ever had an issue with it, you could get the guy that designed and built it on the phone. you can get him on the phone before you buy it as well. i bought it a year ago, and just find myself wanting to use it more than i have time for.
Oh my gosh that kitty is adorable!
Tony, you are messing with my belief in creationism. You have evolved so much!! (LOVE IT!!). Master.
you could have a bad zener diode that it
flowing voltage to early I would look for one shorted open all the time of a capacitor that is dry that has not failed all the way yet so you would need to check the ESR so get out the soldering iron and start pulling caps and checking them. the diodes you can check with out power or desoldering. if you need someone else to check it I would be glad to help.
How much do you trust the test equipment, especially that amp meter?
Not to cast more doubt and confusion on your pile, but any chance you have a different current meter you could try? The sillyscope seems skookum enough, but hall effect sensors can get ausgeficked surprisingly easily.
At least it's an easy double check before tearing apart the entire....oh right. Anyway, best of luck! Love your channel.
I noticed a 8 pin din surface mount component that is inserted into a socket. These are notorious for having “chip creep” that is when the component begins to push it way out of the socket due to vibration. Simply removing the component and reinserting it back in fixes the problem.
At least its still doing what it was doing and it still welds, I wonder if the capacitors are failing? just a thought.
As a pixie wrangler (bachelor in electronics, and studying for a master's in Electronics Engineering) my professional opinion is that you won't be able to fix this thing. You gave it a good shot, but with the horrible physical layout, the lack of a service manual and the real danger that come with testing this thing under power: you're not going to get this fixed. The only way you're fixing this, is by getting some extra help and a lot of luck.
You got two options: give it to someone who can fix it or run it till you can buy a new one.
Just hilarious a kitten falling out of your welder
at least it wasn't a cat-astrophe!
Balance Adjust Pot...hmmmmm...but I agree with the post below as regards probably an electrolytic capacitor going bad in a circuit that controls switching times. The top bulges as they go bad (sometimes).
It's really kind of a time thing - how much time is it worth spending if you want a larger/newer machine anyway? If you can't find a schematic or service manual, the complexity of troubleshooting tends toward infinity - not impossible, just time-consuming.
The bridge rectifier on the input side is probably rated for 35 Amps and up to 600 Volts. It can be this small because the current from your outlet is probably about 20 amps at 120 volts, this is before it gets stepped down, the diodes on the output side are rated for a lot more, probably hundreds of Amps and to handle that kind of current they need to be much larger. You said earlier in the video you didn't know anything about switch mode power supplies but you know a lot, it sounds like you understand the theory behind it, which is more than most people.
I was scanning the daughter board and I am fairly certain that there is a poor connection between your temporal flux capacitor and the half-nut on your lathe. Probably intertemporal oxidation. Try a 1/4 teaspoon of baking soda in your next batch of marinara . It makes a huge difference in the sweetness of the sauce.
you don't think the baking soda would throw off the AC balance?
It looks like it can't equalize the bias between the two periods of the square wave. This might happen when a capacitor in the main supply shorts internally when it warms up (actually the internal resistance falls). It might be also the same thing but on the IGBT board, such as one of the filter/bypass capacitors. If those capacitors are not rated for the temperature they work at (typical of Chinese machines), they start to give up in a similar way.
I loved it: "The good news is...; the bad news is..." The story of my life.
electrolytic capacitors over time tend to dry out and lose capacity especially at hot environment it usually cause similar problems at the beginning of that.
very entertaining, well done.
awww :3 the kitten just gave you sooo many internet-points :P
Tony this is great, fantastic, amazing, unbelievable how much you know about it.......... I did not understand a word of it. I know nothing about electronics. so my question is: why are you waisting your time with mechanical stuff when this is far above it, it's a higher stage of the technical world. I hope I will enjoy more of your presentation which I won't understand at all but people with more brain I have will greatly appreciate. I only have a Lincoln 175 and after years I already know how to fuse steel, never mind I live in in last centufy. That Old Bob
ahah thanks Old Bob!
I really enjoy your stuff what ever it is. I enjoy the presentation, just about the best on you tube.Old Bob
Did you check the flux capacitor?
Funny, good prank call for calling your buddies that work in parts houses.
"The potentiometer is quite simply connected across the bridge rectifier." haha I lost my head laughing at that
Welding project!! thanks for uploading
Love your humor !!!
Thumbs up for the kitten
"[the pot] quite simply is connected over the bridge rectifier", genius.
12:40
Is there any way that the knobs were turned during re-assembly into the pots when you first had it apart?
I can't tell, but if the center of the pots are hexagonal, it wouldn't be difficult to have it off by 60° which might give a bad value to the control board.
Good catch, but they're D-shaped and have molded in stops. If I put them in wrong the knobs have like.. 2* of motion :)
Check da capacitors. It might be a partial short internally. I've seen it many times.
I have that Hantek. Works about as good as an indicator LED for sub 1A currents. Surprisingly decent beyond that though.
Lots of interesting stuff fell outta there when you shook it! Lol. If you shook it harder the factory worker that built it might have fallen out as well!
The very flat top on the waveform suggests to me that the confuser thinks that's where it's supposed to be. I think the problem's on the board that's taking in all the voltage & current monitors and doing the A to D.