Facebook
Google
GitHub
Linkedin
Homemade inverter
- Thread starter twenglish1
- Start date
Dr.killjoy
- Joined Apr 28, 2013
- 1,196
LOLOLOLOLOLOLOLOLOLLOLOLOOLSome of those links you guys posted to DIY plasma cutters just scare me.
I know and thats way I decided to buy instead of build ..
That's what I am pushing for as well.
The direct line powered designs and the excuses for doing them that way are a hoot!
To make a real functional and reliable plasma cutter is not a simple job. Sure a sort of functional one can be hacked together from scrap parts but the cost of the parts that do need to be bought plus the time involved will easily buy a person a good used one or a fixer upper that will do what it is intended to do.
I am for the learning as much as anyone but at some point why bother with DIYing a whole machine that will likely be a poor stand in for any basic factory built unit to call it a learning experience when the same can be done with an actual factory built unit that just needs some repair?
The direct line powered designs and the excuses for doing them that way are a hoot!
To make a real functional and reliable plasma cutter is not a simple job. Sure a sort of functional one can be hacked together from scrap parts but the cost of the parts that do need to be bought plus the time involved will easily buy a person a good used one or a fixer upper that will do what it is intended to do.
I am for the learning as much as anyone but at some point why bother with DIYing a whole machine that will likely be a poor stand in for any basic factory built unit to call it a learning experience when the same can be done with an actual factory built unit that just needs some repair?
It's a close call but this seems almost as dangerous as modifying a car especially given the OPs apparent lack of concern for his own safety. Kudos to tcmtech for trying to redirect this thread. To add my $0.02, if you want to build a telescope with a 16 foot mirror, you'll get there faster by building a 2 foot, a 4 foot, and an 8 foot mirror first. That's another way of saying that working with an existing unit is a great way to learn the ropes.
Similar thoughts.
Typically when people pass up on the opportunity to get a complete and good sized repairable machine of any sort off of me for less than parts costs I have to suspect that either they are either in way over their head as to their views of their skill sets and or complexity of their attempted project or more than likely they are too cheap, poor and or ignorant of what they are seeking to know better.
Personally I am leaning on the too poor and too ignorant aspect at this point being I can't imagine anyone passing up a whole complete machine for a fraction of what the most basic parts will cost to buy new or used.
Seriously if anyone thinks they can build a workable 40 amp plasma cutter inverter based or otherwise for under $300 without already having all of the major power components in hand plus considerable prior hands on experience with designing high power SMPS systems they are delusional and in way over their head.
Typically when people pass up on the opportunity to get a complete and good sized repairable machine of any sort off of me for less than parts costs I have to suspect that either they are either in way over their head as to their views of their skill sets and or complexity of their attempted project or more than likely they are too cheap, poor and or ignorant of what they are seeking to know better.
Personally I am leaning on the too poor and too ignorant aspect at this point being I can't imagine anyone passing up a whole complete machine for a fraction of what the most basic parts will cost to buy new or used.
Seriously if anyone thinks they can build a workable 40 amp plasma cutter inverter based or otherwise for under $300 without already having all of the major power components in hand plus considerable prior hands on experience with designing high power SMPS systems they are delusional and in way over their head.
I found a core that can handle up to 7,000 watts, so maybe ill build a 20 amp instead, unless I could put two cores side by side to allow more power? I saw someone do that in some of the research I have done
Yes you can usually stack identical cores side by side to make a larger transformer.
I have the main HF transformers from a number of Miller inverter welders that are built that way. At first look they appear to be one massive ferrite core unit but when dissembled they are in fact three cores stacked side by side.
I have the main HF transformers from a number of Miller inverter welders that are built that way. At first look they appear to be one massive ferrite core unit but when dissembled they are in fact three cores stacked side by side.
How does that effect the total power available? Say I stack two 7,000 watt cores
Alright got it, so say I choose to switch the transformer at 10khz, how should I figure the number of primary turns based on this?
How so? I believe I can do It
Plasmahunt3r
- Joined Feb 6, 2014
- 28
Just do it!!! That is how you become an expert. Run initial test at low voltage and increase the power until you cook the transformer. Then, rebuild it bigger.
I would start in stages.
1. Buy a meter with a frequency scale. You will need it. I got one that does up to 20 MHZ on EBAY (from China, cheap, took a month).
2. Build a power supply for the low voltage side (12-15V). This power is isolated from the voltage used on the power amp.
3. Build your frequency generator and MOSFET gate driver on a breadboard. Buy multiple crystals at different frequencies (they are cheap). I solder a couple solid wire 22 Ga leads to the crystals, so I can swap them out on easily. The MOSFET Gate Drive Transformer will provide isolation from the high voltage side on the power amp.
4. Build the power supply for the Power Amp side. I would start with a low voltage DC (12-30v) for initial testing. Your MOT welder you rewired would be good for this. You get about 1 volt per turn on a MOT. 16ga wire could get you about 20 turns (20v AC and around 28v rectified DC). I would add an AC capacitor (30-36 Uf) to the AC Hot wire to limit the current on the MOT primary (look up Capacitor Reactance). Use this supply to get the bugs out of you Power Amp.
5. Now, start on the Power Amp. Decide which topology you want. I like the Class E, but Half Bridge is simpler. Class E has a lot of variables. I would start using low voltage DC on the Amp and gradually increase it.
Remember to post your results (success and failures) and schematics. We all want to know what tweaks and mods you made. What worked and what didn't.
I like to think of the cost as training cost. It doesn't do me any good to be a consumer and buy something already made. The pleasure is in learning how it works.
I would start in stages.
1. Buy a meter with a frequency scale. You will need it. I got one that does up to 20 MHZ on EBAY (from China, cheap, took a month).
2. Build a power supply for the low voltage side (12-15V). This power is isolated from the voltage used on the power amp.
3. Build your frequency generator and MOSFET gate driver on a breadboard. Buy multiple crystals at different frequencies (they are cheap). I solder a couple solid wire 22 Ga leads to the crystals, so I can swap them out on easily. The MOSFET Gate Drive Transformer will provide isolation from the high voltage side on the power amp.
4. Build the power supply for the Power Amp side. I would start with a low voltage DC (12-30v) for initial testing. Your MOT welder you rewired would be good for this. You get about 1 volt per turn on a MOT. 16ga wire could get you about 20 turns (20v AC and around 28v rectified DC). I would add an AC capacitor (30-36 Uf) to the AC Hot wire to limit the current on the MOT primary (look up Capacitor Reactance). Use this supply to get the bugs out of you Power Amp.
5. Now, start on the Power Amp. Decide which topology you want. I like the Class E, but Half Bridge is simpler. Class E has a lot of variables. I would start using low voltage DC on the Amp and gradually increase it.
Remember to post your results (success and failures) and schematics. We all want to know what tweaks and mods you made. What worked and what didn't.
I like to think of the cost as training cost. It doesn't do me any good to be a consumer and buy something already made. The pleasure is in learning how it works.
Lots of good info! Thank you.Any pointers on making the switching and gate transformers? I know the switching transformer is gonna need a large ferrite core rated for about 10,000 watts, I'm guessing both can be a 1:1 ratio that way I will get my mains voltage on the switching transformer, and the gate transformer is just for isolation. For the switching transformer how do I determine the number of windings for the given frequency I switch it at?
Plasmahunt3r
- Joined Feb 6, 2014
- 28
After experimenting with different chokes and autotransformers, I came up with some general rules.
1. Use inductance for figuring out your coils. (buy an inductance meter)
2. The current drawn on your power supply is directly proportional to the inductance value of your ferrite transformer.
3. The higher the inductance value (more turns), the less current the ferrite transformer requires.
4. The lower the inductance value (less turns), the more current the ferrite core transformer requires (Less turns is closer to a short, so higher current draw).
5. For KHZ circuits, use values in the mH range. At 32.8 KHZ; I have drawn 7.5A with 1.54mH and have drawn 3.5A with 6.26 mH (see how the inductance value and current are proportional).
6. For MHZ circuits, use values in the uH range. Closer to a short, but "ON" for a shorter time period, so the circuit can handle it.
7. Use a Clamp Ampmeter on your Power Supply Secondary to see how much current your Power Amp and Ferrite Transformer is using.
Three essential tools:
1. Inductance Meter (Make sure it has uH & mH range)
2. Multimeter with Frequency scale (KHZ to MHZ range desirable)
3. Clamp Ampmeter
As for Gate Transformers, I use a Common Mode Choke 1:1 with a Capacitor/Diode to create Pulsed DC Voltage Doubler. A MOSFET requires at least 10V to be fully "ON". Lots of Common Mode Chokes on the market. I am using a Panasonic 3A Common Mode Choke because I had one. A 1A Common Mode Choke would have been enough. You can see it in the schematics in the Class E Magnetic Induction Heater thread.
If you are going to control 2 MOSFETs with one Gate Transformer, you have to wire your own using a small toroid.
1. Use inductance for figuring out your coils. (buy an inductance meter)
2. The current drawn on your power supply is directly proportional to the inductance value of your ferrite transformer.
3. The higher the inductance value (more turns), the less current the ferrite transformer requires.
4. The lower the inductance value (less turns), the more current the ferrite core transformer requires (Less turns is closer to a short, so higher current draw).
5. For KHZ circuits, use values in the mH range. At 32.8 KHZ; I have drawn 7.5A with 1.54mH and have drawn 3.5A with 6.26 mH (see how the inductance value and current are proportional).
6. For MHZ circuits, use values in the uH range. Closer to a short, but "ON" for a shorter time period, so the circuit can handle it.
7. Use a Clamp Ampmeter on your Power Supply Secondary to see how much current your Power Amp and Ferrite Transformer is using.
Three essential tools:
1. Inductance Meter (Make sure it has uH & mH range)
2. Multimeter with Frequency scale (KHZ to MHZ range desirable)
3. Clamp Ampmeter
As for Gate Transformers, I use a Common Mode Choke 1:1 with a Capacitor/Diode to create Pulsed DC Voltage Doubler. A MOSFET requires at least 10V to be fully "ON". Lots of Common Mode Chokes on the market. I am using a Panasonic 3A Common Mode Choke because I had one. A 1A Common Mode Choke would have been enough. You can see it in the schematics in the Class E Magnetic Induction Heater thread.
If you are going to control 2 MOSFETs with one Gate Transformer, you have to wire your own using a small toroid.
Last edited:
Ok thanks, so if I understand correctly, I will need to wind the primary turns of switching transformer to draw 40 amps from the igbts to get the 40 amps output, if the switching transformer is wound in a 1:1 ratio, correct?
Plasmahunt3r
- Joined Feb 6, 2014
- 28
You have to try different inductance values and check the current draw on your power supply secondary, until you get it right.
Just try one value, see current drawn, and then rebuild Ferrite Transformer with more/less turns as desired.
I didn't say to use 1:1 for the ferrite transformer. If you have a 2:1 ratio, you will have twice the current on the secondary but half the voltage.
EX: With a 2:1 ratio, if your the Ferrite Transformer primary is drawing 20A (get this reading from the power supply secondary), your Ferrite Secondary is getting twice the Amps (40A), but half the voltage. READ Transformer Tutorials on the internet. This Amps/Voltage works the same way if you wire your Ferrite as an AutoTransformer.
Where I was suggesting a 1:1 ratio was on the MOSFET Gate Transformer so you can use standard Common Mode Chokes for this purpose.
You will have 3 different Transformers, each with a different function.
1. Power Supply Transformer for Power Amp. Get Amp reading from the secondary. Rewiring your MOT would be good for this transformer. Could use two primaries (from two different MOTS) on one iron core to create an isolation transformer. You could use a Step Up/Down Autotransformer (3000 Watt) but no isolation from the mains (Dangerous). Could use Capacitor Reactance to limit current draw on the power transformer. Many options available. Study AC Power Supply Circuits on the Internet; It's important.
2. MOSFET Gate Transformer. Used to isolate low voltage Frequency Controller from Power Amp.
3. Switching Ferrite Transformer. Goal of your circuits.
Repeat: Study Transformer Tutorials, AC Power Supply Circuits, and Capacitor Reactance. This is important info to grasp. Over time, you will build many specialized power supplies, for different uses. Your Inverter Plasma cutter will have two distinct power supplies (one for Low voltage circuits and one for the Power Amp). People can spend an entire career on Power Electronics.
Just try one value, see current drawn, and then rebuild Ferrite Transformer with more/less turns as desired.
I didn't say to use 1:1 for the ferrite transformer. If you have a 2:1 ratio, you will have twice the current on the secondary but half the voltage.
EX: With a 2:1 ratio, if your the Ferrite Transformer primary is drawing 20A (get this reading from the power supply secondary), your Ferrite Secondary is getting twice the Amps (40A), but half the voltage. READ Transformer Tutorials on the internet. This Amps/Voltage works the same way if you wire your Ferrite as an AutoTransformer.
Where I was suggesting a 1:1 ratio was on the MOSFET Gate Transformer so you can use standard Common Mode Chokes for this purpose.
You will have 3 different Transformers, each with a different function.
1. Power Supply Transformer for Power Amp. Get Amp reading from the secondary. Rewiring your MOT would be good for this transformer. Could use two primaries (from two different MOTS) on one iron core to create an isolation transformer. You could use a Step Up/Down Autotransformer (3000 Watt) but no isolation from the mains (Dangerous). Could use Capacitor Reactance to limit current draw on the power transformer. Many options available. Study AC Power Supply Circuits on the Internet; It's important.
2. MOSFET Gate Transformer. Used to isolate low voltage Frequency Controller from Power Amp.
3. Switching Ferrite Transformer. Goal of your circuits.
Repeat: Study Transformer Tutorials, AC Power Supply Circuits, and Capacitor Reactance. This is important info to grasp. Over time, you will build many specialized power supplies, for different uses. Your Inverter Plasma cutter will have two distinct power supplies (one for Low voltage circuits and one for the Power Amp). People can spend an entire career on Power Electronics.
I understand how transformers work, and my reasoning behind the ferrite transformer being 1:1 is because I need the voltage to be at least 240v on the output, and I will have more than that from the rectified mains, so I figured draw 40 amps on the primary side and output 1:1 to get the 40 amps out and the same voltage from the mains.
Alot of the circuits I have looked at don't have the initial isolation transformer, is it required? Or are you just suggesting it as a way to work with lower voltages while I am developing the main power supply? Or is the power supply transformer for the power supply that powers the control circuit and mosfet gates?
Alot of the circuits I have looked at don't have the initial isolation transformer, is it required? Or are you just suggesting it as a way to work with lower voltages while I am developing the main power supply? Or is the power supply transformer for the power supply that powers the control circuit and mosfet gates?
