Author Topic: Homemade IGBT MIG welder with ATX PSU parts. [blown yet again][blown count: 5]  (Read 28567 times)

0 Members and 1 Guest are viewing this topic.

Offline RefrigeratorTopic starter

  • Super Contributor
  • ***
  • Posts: 1593
  • Country: lt
Re: Homemade IGBT MIG welder with ATX PSU parts.
« Reply #50 on: June 27, 2019, 09:25:03 pm »
I have a blog at http://brimmingideas.blogspot.com/ . Now less empty than ever before !
An expert of making MOSFETs explode.
 

Offline strawberry

  • Super Contributor
  • ***
  • !
  • Posts: 1199
  • Country: lv
Re: Homemade IGBT MIG welder with ATX PSU parts.
« Reply #51 on: June 29, 2019, 04:27:27 pm »
fan 120x120mm 24V 11.3W 3.7m^3/min
air speed approx 5m/s
SK442-100 0.22K/W @5m/s
5kW HB inverter
igbt 0.6K/W
AlO wafer ~0.3K/W
(0.6+0.3)/2+0.22=0.67K/W
IGBT loss 0.66mJ @400V 20A
IGBT 1.6V 32W@20A
IGBT 43W@65kHz
43*2+32=118W
temperature of silicon at room temperature
118*0.67+25=104C

Diode 0.94V@60A trr25ns
Diode 0.9K/W
Rectifier contains 6pcs in total
Rectifier 200A
Rectifier loss 200*0.94=188W
SK442-100 0.22K/W @5m/s
0.9/6+0.22=0.37K/W
temperature of silicon at room temperature
188*0.37+25=94.5C





 

Offline RefrigeratorTopic starter

  • Super Contributor
  • ***
  • Posts: 1593
  • Country: lt
Re: Homemade IGBT MIG welder with ATX PSU parts.
« Reply #52 on: June 30, 2019, 06:44:01 pm »
I've got my drive circuitry nearly finished now, it's a mega-bodge on some proto-board but it does work very well.
One problem that i've come into is frequency instability, seems that even small changes in RT make massive difference in operating frequency.
Using a 2n2 CT i get 12 kHz with a 5k1 resistor, 100kHz with a 2k2 resistor and about 50kHz with a 4k7 resistor, i might put a 10 turn trimmer in place of RT to set the frequency more accurately.
So far the gate drive transformer was the fiddliest part to make, finding the right core took longer than i'm willing to admit.
In the end i settled for a core from a common mode choke from the same ATX PSU.
I made it a 1:1 transformer with the same construction as in an ATX PSU, bifilar wound secondary sandwiched between two primaries, 10 turns each.
Gate drive signals look super nice, i used the signal on one of the GDT drive transistor bases as a reference to check if my gate drives were correct.
« Last Edit: June 30, 2019, 06:46:18 pm by Refrigerator »
I have a blog at http://brimmingideas.blogspot.com/ . Now less empty than ever before !
An expert of making MOSFETs explode.
 

Offline xavier60

  • Super Contributor
  • ***
  • Posts: 3034
  • Country: au
Re: Homemade IGBT MIG welder with ATX PSU parts.
« Reply #53 on: June 30, 2019, 10:10:47 pm »
Does the 5k1 resistor actually measure 5k1?. It seems to be the odd one out.
HP 54645A dso, Fluke 87V dmm,  Agilent U8002A psu,  FY6600 function gen,  Brymen BM857S, HAKKO FM-204, New! HAKKO FX-971.
 

Offline RefrigeratorTopic starter

  • Super Contributor
  • ***
  • Posts: 1593
  • Country: lt
Re: Homemade IGBT MIG welder with ATX PSU parts.
« Reply #54 on: June 30, 2019, 10:42:37 pm »
Does the 5k1 resistor actually measure 5k1?. It seems to be the odd one out.

I've already tossed it in the bin so i can't measure it, but i replaced RT with a 2k 20 turn trimmer and a 3k3 in series and the lowest frequency i can go to is 55kHz, so the 5k1 resistor might not have been 5k1 afterall.
Checked all of the gates of my IGBTs to make sure they're all being driven properly and everything seems fine at 55kHz.

The inverter part is basically finished now, i'll just have to put it all in a box tomorrow.
Next will be the wire feed and speed. I wanted to make a closed loop wire feed mechanism to keep the speed constant but i think i'll do it the easy way with a buck converter.
I have a blog at http://brimmingideas.blogspot.com/ . Now less empty than ever before !
An expert of making MOSFETs explode.
 

Offline strawberry

  • Super Contributor
  • ***
  • !
  • Posts: 1199
  • Country: lv
Re: Homemade IGBT MIG welder with ATX PSU parts.
« Reply #55 on: July 01, 2019, 06:23:55 am »
you should set dead time for IGBT. TL494/KA7500 is determined by DT pin voltage
 

Offline xavier60

  • Super Contributor
  • ***
  • Posts: 3034
  • Country: au
Re: Homemade IGBT MIG welder with ATX PSU parts.
« Reply #56 on: July 01, 2019, 10:44:31 am »
you should set dead time for IGBT. TL494/KA7500 is determined by DT pin voltage
The maximum duty cycle would be something less than 50% with only one output used.
What bothers me about the TL494 is not being certain that double pulse suppression  has been added.
HP 54645A dso, Fluke 87V dmm,  Agilent U8002A psu,  FY6600 function gen,  Brymen BM857S, HAKKO FM-204, New! HAKKO FX-971.
 

Offline xavier60

  • Super Contributor
  • ***
  • Posts: 3034
  • Country: au
Re: Homemade IGBT MIG welder with ATX PSU parts.
« Reply #57 on: July 01, 2019, 11:23:47 am »
Gate drive for a one or two switch forward inverter is complicated by  the changing symmetry of the drive waveform with change of duty cycle.
I remember with one welder that I repaid, the drive transformer had a large inductance and was driven by a single MOSFET in flyback mode. The flyback voltage was clamped by Zener diode across the primary.
I wonder if the same drive that's normally used for half-bridge forward inverters can be used. The negative pulse will just be ignored by the IGBTs.
HP 54645A dso, Fluke 87V dmm,  Agilent U8002A psu,  FY6600 function gen,  Brymen BM857S, HAKKO FM-204, New! HAKKO FX-971.
 

Offline RefrigeratorTopic starter

  • Super Contributor
  • ***
  • Posts: 1593
  • Country: lt
Re: Homemade IGBT MIG welder with ATX PSU parts.
« Reply #58 on: July 01, 2019, 12:40:41 pm »
you should set dead time for IGBT. TL494/KA7500 is determined by DT pin voltage
The maximum duty cycle would be something less than 50% with only one output used.
What bothers me about the TL494 is not being certain that double pulse suppression  has been added.

One thing i noticed when testing the circuit out is that the voltage on the open collectors of the TL494 does not go down to 0V but rather to something like 0.7V, which explains why there are almost always one or two diodes between the base drive transistor emitters and ground in the ATX PSU schematics.
« Last Edit: July 01, 2019, 04:07:35 pm by Refrigerator »
I have a blog at http://brimmingideas.blogspot.com/ . Now less empty than ever before !
An expert of making MOSFETs explode.
 

Offline RefrigeratorTopic starter

  • Super Contributor
  • ***
  • Posts: 1593
  • Country: lt
Re: Homemade IGBT MIG welder with ATX PSU parts.
« Reply #59 on: July 01, 2019, 04:14:50 pm »
Well i seem to have run out of solder for today but i finished the bridge rectifier and while i was at it i added a simple relay circuit to bypass the PTC resistor after two seconds, this way i won't cook the PTC.
Pretty much the same thing done as in big old CRT TVs.
I have a blog at http://brimmingideas.blogspot.com/ . Now less empty than ever before !
An expert of making MOSFETs explode.
 

Offline strawberry

  • Super Contributor
  • ***
  • !
  • Posts: 1199
  • Country: lv
Re: Homemade IGBT MIG welder with ATX PSU parts.
« Reply #60 on: July 01, 2019, 06:14:18 pm »
thermal compound is not used with silicone alike thermal pads
did you mean NTC ? PTC wont limit inrush current
 

Offline RefrigeratorTopic starter

  • Super Contributor
  • ***
  • Posts: 1593
  • Country: lt
Re: Homemade IGBT MIG welder with ATX PSU parts.
« Reply #61 on: July 01, 2019, 08:23:11 pm »
thermal compound is not used with silicone alike thermal pads
did you mean NTC ? PTC wont limit inrush current
Yes, i meant NTC.
The pads are reused from the ATX PSU and it looked like most of the contact was made right around the screw hole and not lower down on the TO-220 so i added some thermal paste to help the diodes out a little bit.
I have a blog at http://brimmingideas.blogspot.com/ . Now less empty than ever before !
An expert of making MOSFETs explode.
 

Offline techman-001

  • Frequent Contributor
  • **
  • !
  • Posts: 748
  • Country: au
  • Electronics technician for the last 50 years
    • Mecrisp Stellaris Unofficial UserDoc
Re: Homemade IGBT MIG welder with ATX PSU parts.
« Reply #62 on: July 01, 2019, 08:40:15 pm »

I learned to MIG weld when i was 13, it's basically a hot glue gun for metal.
Assuming the welder is correctly set up it takes about 30 minutes to learn MIG.

Just started a blog at http://brimmingideas.blogspot.com/ . Not much in it as of now but more is sure to come

You have a very interesting blog, exactly the kind of thing I love to read because you engineer your projects using materials that you have, to make what you want, without buying everything of the shelf.

Kudos on your moving coil meter project, I was amazed it even worked! I spent 3 years repairing and calibrating moving coil meters in instrumentation and automotive. One rugged automotive method used for a meter bearing is a small tube around the moving coil shaft that has a viscous liquid such glycerin in it. This method adds a lot of damping but seemed to be low friction because the shaft 'floated' in the bearing. It only had one longish bearing of this type with the other end floating free.

Keep up the excellent work, you have a long and interesting career in front of you in my opinion.
 
The following users thanked this post: Refrigerator

Offline strawberry

  • Super Contributor
  • ***
  • !
  • Posts: 1199
  • Country: lv
Re: Homemade IGBT MIG welder with ATX PSU parts.
« Reply #63 on: July 02, 2019, 06:29:53 am »
For beginners it is nice to have parts you can blow without worries..
downside is large space filled with PCB boards and staff
 

Offline RefrigeratorTopic starter

  • Super Contributor
  • ***
  • Posts: 1593
  • Country: lt
Re: Homemade IGBT MIG welder with ATX PSU parts.
« Reply #64 on: July 02, 2019, 03:36:46 pm »

I learned to MIG weld when i was 13, it's basically a hot glue gun for metal.
Assuming the welder is correctly set up it takes about 30 minutes to learn MIG.

Just started a blog at http://brimmingideas.blogspot.com/ . Not much in it as of now but more is sure to come

You have a very interesting blog, exactly the kind of thing I love to read because you engineer your projects using materials that you have, to make what you want, without buying everything of the shelf.

Kudos on your moving coil meter project, I was amazed it even worked! I spent 3 years repairing and calibrating moving coil meters in instrumentation and automotive. One rugged automotive method used for a meter bearing is a small tube around the moving coil shaft that has a viscous liquid such glycerin in it. This method adds a lot of damping but seemed to be low friction because the shaft 'floated' in the bearing. It only had one longish bearing of this type with the other end floating free.

Keep up the excellent work, you have a long and interesting career in front of you in my opinion.

I actually started the blog because i couldn't find anyone else doing the kind of stuff i was doing, and i thought surely there are more guys like me out there.
So now if anyone's interested in the kind of projects that i'm into they can find my blog.  :)
I have a blog at http://brimmingideas.blogspot.com/ . Now less empty than ever before !
An expert of making MOSFETs explode.
 

Offline RefrigeratorTopic starter

  • Super Contributor
  • ***
  • Posts: 1593
  • Country: lt
Re: Homemade IGBT MIG welder with ATX PSU parts.
« Reply #65 on: July 02, 2019, 05:35:44 pm »
Finished the wire speed controller, I had some 555's laying around somewhere but couldn't find them.
No big deal because i found a neat little LM393 PWM controller schematic on google and made the speed controller with that.
Drew the traces with a permanent marker and etched the board, soldered everything on top because i couldn't be bothered to mirror the layout on the other side and drill the holes.
I put dots around the circuit with my marker, those dots are there in case i ever wanted to experiment with feedback so that i'd have pads to solder to.
Planned to use a TO-220 MOSFET but the only ones i had were IRF640 and IRF614, which would get way too hot, so i bodged in a NTD70N03R that i pulled from a dell motherboard.
My PWM is tuned to 30kHz and the MOSFET stays ice cold.
For my wire feed i'm using a VW windshield wiper motor, which seems to have some good torque so the open loop speed control might not be as bad as i had anticipated.
I have a blog at http://brimmingideas.blogspot.com/ . Now less empty than ever before !
An expert of making MOSFETs explode.
 

Offline strawberry

  • Super Contributor
  • ***
  • !
  • Posts: 1199
  • Country: lv
Re: Homemade IGBT MIG welder with ATX PSU parts.
« Reply #66 on: July 02, 2019, 06:07:20 pm »
is it posible to use TL494 as PWM controller
more detailed info for TL494:http://www.ti.com/lit/an/slva001e/slva001e.pdf
 

Offline mikerj

  • Super Contributor
  • ***
  • Posts: 3382
  • Country: gb
Re: Homemade IGBT MIG welder with ATX PSU parts.
« Reply #67 on: July 02, 2019, 06:59:45 pm »
For my wire feed i'm using a VW windshield wiper motor, which seems to have some good torque so the open loop speed control might not be as bad as i had anticipated.

That going to be a very heavy duty wire feed, but it's a nice choice for an easily found/cheap part to repurpose.  Getting/making suitable rollers and assembling then in an accurate, mechanically rigid assembly is key to a decent wire feed, and typically is where all the cheap MIG welders have problems (flexible plastic rubbish held together with self tapping screws).
 

Offline RefrigeratorTopic starter

  • Super Contributor
  • ***
  • Posts: 1593
  • Country: lt
Re: Homemade IGBT MIG welder with ATX PSU parts.
« Reply #68 on: July 02, 2019, 08:27:55 pm »
is it posible to use TL494 as PWM controller
more detailed info for TL494:http://www.ti.com/lit/an/slva001e/slva001e.pdf

Yeah i saw that but the LM393 circuit is much simpler.

For my wire feed i'm using a VW windshield wiper motor, which seems to have some good torque so the open loop speed control might not be as bad as i had anticipated.

That going to be a very heavy duty wire feed, but it's a nice choice for an easily found/cheap part to repurpose.  Getting/making suitable rollers and assembling then in an accurate, mechanically rigid assembly is key to a decent wire feed, and typically is where all the cheap MIG welders have problems (flexible plastic rubbish held together with self tapping screws).

While i was looking up wire feed mechanisms on the internet i found quite a few that use a 24V windshield wiper motors in them.
The roller problem was sorted with the help of a cheap skateboard bearing that i welded shut to keep it from spinning, this one will be mounted to the windshield wiper motor and will function as the drive roller.
One problem that i came into though is that the bearing is so cheap that the races are not even hard, my dull file bites pretty well into it.
I might try quench hardening to see if it helps because as i assume a soft metal drive roller won't last long.

« Last Edit: July 02, 2019, 08:29:30 pm by Refrigerator »
I have a blog at http://brimmingideas.blogspot.com/ . Now less empty than ever before !
An expert of making MOSFETs explode.
 

Online tautech

  • Super Contributor
  • ***
  • Posts: 29810
  • Country: nz
  • Taupaki Technologies Ltd. Siglent Distributor NZ.
    • Taupaki Technologies Ltd.
Re: Homemade IGBT MIG welder with ATX PSU parts.
« Reply #69 on: July 02, 2019, 08:53:05 pm »
While i was looking up wire feed mechanisms on the internet i found quite a few that use a 24V windshield wiper motors in them.
Yes because the 24V wiper motor is a good match for MIG welding voltages and simple speed control circuits are needed as the motor is only energised when the handpiece trigger is pulled.
One problem that i came into though is that the bearing is so cheap that the races are not even hard, my dull file bites pretty well into it.
More likely you annealed the hardening from it by welding it.
« Last Edit: July 02, 2019, 08:58:17 pm by tautech »
Avid Rabid Hobbyist.
Some stuff seen @ Siglent HQ cannot be shared.
 

Offline RefrigeratorTopic starter

  • Super Contributor
  • ***
  • Posts: 1593
  • Country: lt
Re: Homemade IGBT MIG welder with ATX PSU parts.
« Reply #70 on: July 02, 2019, 09:03:22 pm »
While i was looking up wire feed mechanisms on the internet i found quite a few that use a 24V windshield wiper motors in them.
Yes because the 24V wiper motor is a good match for MIG welding voltages and simple speed control circuits are needed as the motor is only energised when the handpiece trigger is pulled.
One problem that i came into though is that the bearing is so cheap that the races are not even hard, my dull file bites pretty well into it.
More likely you annealed the hardening from it by welding it.

I dumped it into water right after welding but i guess it wasn't hot enough to harden.
I have a blog at http://brimmingideas.blogspot.com/ . Now less empty than ever before !
An expert of making MOSFETs explode.
 

Offline techman-001

  • Frequent Contributor
  • **
  • !
  • Posts: 748
  • Country: au
  • Electronics technician for the last 50 years
    • Mecrisp Stellaris Unofficial UserDoc
Re: Homemade IGBT MIG welder with ATX PSU parts.
« Reply #71 on: July 02, 2019, 10:30:41 pm »
<snip>
 All the cheap MIG welders have problems (flexible plastic rubbish held together with self tapping screws).

Really? what utter junk. Chinese I assume ?
 

Offline RefrigeratorTopic starter

  • Super Contributor
  • ***
  • Posts: 1593
  • Country: lt
Re: Homemade IGBT MIG welder with ATX PSU parts.
« Reply #72 on: July 03, 2019, 01:21:35 pm »
Wire feed is done and i'm getting really excited, seems my MIG welder might be finished sooner than expected. (knock on wood)
Also the copper pads i left around the speed controller circuitry came in handy as i had to add an enable transistor.
For that i changed the inverting and non inverting inputs of the second comparator so that my duty cycle went up as control voltage went down.
Next i added an NPN transistor that would pull one of the potentiometer leads to ground this way enabling the speed controller as soon as a high signal is received.
Last thing i added was a 47µF cap between the collector of the aforementioned transistor and the control voltage, this way there's a period of high duty cycle on startup, which made a big improvement on the motor speedup.
I have a blog at http://brimmingideas.blogspot.com/ . Now less empty than ever before !
An expert of making MOSFETs explode.
 

Offline strawberry

  • Super Contributor
  • ***
  • !
  • Posts: 1199
  • Country: lv
Re: Homemade IGBT MIG welder with ATX PSU parts.
« Reply #73 on: July 03, 2019, 04:33:22 pm »
I am not sure about post flow gas for MIG
 

Offline RefrigeratorTopic starter

  • Super Contributor
  • ***
  • Posts: 1593
  • Country: lt
Re: Homemade IGBT MIG welder with ATX PSU parts.
« Reply #74 on: July 03, 2019, 05:34:22 pm »
Construction begins on none other than good old plywood, since wood can absorb moisture and become slightly conductive i separated any wood from high voltage with a sheet of plastic.
Now i'll have to find the PS3 power supply that i had laying around somewhere and bodge*ahem* i mean add a variable DTC feature into my TL494 circuit board.
Feeling a bit nervous as i inevitably start thinking about all the things that might go wrong.

I am not sure about post flow gas for MIG

Not necessary as far as i can tell, many MIGs don't even have it.
I have a blog at http://brimmingideas.blogspot.com/ . Now less empty than ever before !
An expert of making MOSFETs explode.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf