Controller using IGBTs

[Yummyrum]

EDIT: I split this off the Burstfire ,Auber makes LED lights Flicker Topic
viewtopic.php?p=7810936#p7810936

I’m confused about why you want to go to DC with a big transformer .

My son is Electrical Engineer and does Traction control for Locomotives . He finds it very archaic that we are using Phase control and Burst fire control .

He asked why we aren’t running the mains into a big Bridge rectifier and chopping up the Rectified DC with PWM control at around 10-100kHz using IGBTs .

I asked him why he hadn’t designed one for me

Is that the sort of thing you are looking at ?

[Steve Broady]

He asked why we aren’t running the mains into a big Bridge rectifier and chopping up the Rectified DC with PWM control at around 10-100kHz using IGBTs .

Money, I would say. All that costs money. It’s also more complex and outside the scope of what most people are willing to homebrew, but I expect that a suitable product off the shelf would be a bit niche and therefore expensive.

However, I would LOVE for someone to prove me wrong!

[kennstminet]

However, I would LOVE for someone to prove me wrong!

I'm afraid you are right.

[NormandieStill]

I’m confused about why you want to go to DC with a big transformer .

My son is Electrical Engineer and does Traction control for Locomotives . He finds it very archaic that we are using Phase control and Burst fire control .

He asked why we aren’t running the mains into a big Bridge rectifier and chopping up the Rectified DC with PWM control at around 10-100kHz using IGBTs .

I asked him why he hadn’t designed one for me

Is that the sort of thing you are looking at ?

Exactly! It's what my dimmable LEDs do. The Arduino and the ESP32 have dedicated PWM outputs. Appropriately re-rated my AC element doesn't care that its seeing DC. And ignoring losses the AC feed into the rectifier will see the average power draw and not a choppy all-or-nothing draw.

I think that makes at least three of us who'd like to see the circuit. I'll take a look later but I'm not certain that my electronics-fu is up to the standard. AC-DC power supplies tend to involve some bits going analog and the hand-waving explanations start getting pretty desperate by then.

[Yummyrum]

I’m confused about why you want to go to DC with a big transformer .

My son is Electrical Engineer and does Traction control for Locomotives . He finds it very archaic that we are using Phase control and Burst fire control .

He asked why we aren’t running the mains into a big Bridge rectifier and chopping up the Rectified DC with PWM control at around 10-100kHz using IGBTs .

I asked him why he hadn’t designed one for me

Is that the sort of thing you are looking at ?

Exactly! It's what my dimmable LEDs do. The Arduino and the ESP32 have dedicated PWM outputs. Appropriately re-rated my AC element doesn't care that its seeing DC. And ignoring losses the AC feed into the rectifier will see the average power draw and not a choppy all-or-nothing draw.

I think that makes at least three of us who'd like to see the circuit. I'll take a look later but I'm not certain that my electronics-fu is up to the standard. AC-DC power supplies tend to involve some bits going analog and the hand-waving explanations start getting pretty desperate by then.

There are literally hundreds of IGBT modules around with Voltages up to and exceeding 1000V DC and currents up to several hundred amps . But they need a driver Chip to drive them . There seems to be quite a choice of them as well offering Optocoupled input so could connect opto to ESP32 or Arduinos . But they also need a 20-30v supply ( to run the Driver IC and supply the Gate Voltage ) not that that's a biggy .
So maybe it is doable .Probably need to add a simple filter to keep the noise out of the Mains .


It would be nice to just use a SSR but they seem rare as rocking horse poo when it comes to high voltage at high currents .
I found this unit . It would make DIY pretty easy . Alas it only goes up to 30 Amps . But it can be controlled straight off an Arduino PWM output .
Not the cheapest puppy in the pound but it shows hope .

https://www.power-io.com/products/hdd.htm

SSR.png

And a Bridge rectifier

https://www.ebay.com.au/itm/31583220860 ... sdEALw_wcB

Bridge Rectifier.png

Something like this .

IMG_2793.jpeg

[Steve Broady]

There are literally hundreds of IGBT modules around with Voltages up to and exceeding 1000V DC and currents up to several hundred amps . But they need a driver Chip to drive them .

So, I found this one on Digi-Key
https://www.digikey.com/en/products/det ... 1/21771508

At only $15.35 US, it’s not a bad price… as long as you’re willing to spend $700 on a tray of 40 to meet the minimum order plus shipping.

Price aside, this is where I run into the limitations of my knowledge and understanding. Reading the data sheet, it seems like you can feed this thing DC up to 650V and 50A (70A if you can keep it at 25C) and get three phase AC out if you want. It’ll dissipate 190 watts, and needs a 20V DC control signal.

Have I got that right?

[Yummyrum]

There are literally hundreds of IGBT modules around with Voltages up to and exceeding 1000V DC and currents up to several hundred amps . But they need a driver Chip to drive them .

So, I found this one on Digi-Key
https://www.digikey.com/en/products/det ... 1/21771508

At only $15.35 US, it’s not a bad price… as long as you’re willing to spend $700 on a tray of 40 to meet the minimum order plus shipping.

Price aside, this is where I run into the limitations of my knowledge and understanding. Reading the data sheet, it seems like you can feed this thing DC up to 650V and 50A (70A if you can keep it at 25C) and get three phase AC out if you want. It’ll dissipate 190 watts, and needs a 20V DC control signal.

Have I got that right?

Yes that is a set of 6 power switches that are commonly used in Variable Frequency Drives ( for running 3 phase Induction motors off single phase ) and Three phase Inverters etc . It is a bunch of dumb switches . Needed , is a controller that drives each switch with a PWM signal in such a way as to synthesise a three phase power at the U,V, W outputs . Each output isn’t a smooth sine wave but a high frequency PWM so it is usually run through a low pass filter ( Coil and capacitor)to remove most of the PWM artefacts leaving a smooth sinewave . However , if it is driving an induction motor for example , the motor inductance will do the same job .

All we need to control an element is a single IGBT module not one with multiple IGBTs

This sort of thing

https://au.mouser.com/datasheet/2/268/A ... 444492.pdf

IMG_2796.png

And an opto -coupled driver IC . These are plentiful and about $5 .

Like this one .
https://www.onsemi.com/download/data-sh ... 8321-d.pdf

or this one .
https://au.mouser.com/datasheet/2/408/T ... 915943.pdf

The IGBT needs a voltage of around 20v to switch it hard on . The driver chip is designed to very quickly charge and discharge the Gate of the IGBT .IGBTs need to switch very quickly so they are either fully on or fully off . The transition period is where they can quickly over heat .
The driver also has an inbuilt Opto coupler so that it provides electrical isolation between the IGBT circuit ( which is connected to the Mains ) and the controller , IE Arduino or the likes .
However , you can see in that schematic that the Driver IC needs a 20v supply . (Vee and Vcc).Note that this is also on the live side .

IMG_2797.jpeg

The input signal could also be a PWM signal from a simple 555 timer. It doesn’t need to be an Arduino .


LOL , the more I think about this , the easier it looks .

IMG_2798.jpeg

[Yummyrum]

So combining all that , adding a low pass filter but still not showing 20v supply for driver .
Note that the low pass filter is not your standard one with large electrolytic capacitors . It simply has to stop the 10-20KHz ripple getting back into the Mains . the waveform will look like a fullwave rectified Mains .

Controller.png

Going to run it by my son and see what he thinks and what drivers and IGBTs he recons will match and what filter values to try .

[NormandieStill]

I like the "just use a 555" attitude. I'd want to stick with my arduino because the controller needs to double up as a mashing controller for the RIMS brewery.

I guess the reason that this seems more complex is that we are outside the realm of hobbyist electronics so there are no neat, prebuilt, mass-produced modules available.

I'd have loved to be able to adapt my existing controller but I see another build project in my future!

[Steve Broady]

I like the idea that this design should work equally well with any mains voltage without the need to adjust potentiometer values, etc. It looks like a very nice tidy design. Like Normandie, I see another project coming.

[Yummyrum]

I like the "just use a 555" attitude. I'd want to stick with my arduino because the controller needs to double up as a mashing controller for the RIMS brewery.

I guess the reason that this seems more complex is that we are outside the realm of hobbyist electronics so there are no neat, prebuilt, mass-produced modules available.

I'd have loved to be able to adapt my existing controller but I see another build project in my future!

Damn that 555 been around for over 40 years and still has it’s place . 1/2 dozen components and it’s gonna give 0-100% PWM power control .

Thats the thing , anyone that can follow this topic can adapt what they feel comfortable with .Want a pot that goes 0-100% , use the 555 circuit .

Want extra control and know how to program ESP32 , Arduino , go that route .

It’s just the interface that needs to be nailed and then we have a whole new avenue .
I agree that this is a bit above entry level electronics , but once we have a working model that can be copied , that is the goal
.

I like the idea that this design should work equally well with any mains voltage without the need to adjust potentiometer values, etc. It looks like a very nice tidy design. Like Normandie, I see another project coming.

Yes , Steve , and the Pot should be very Linear .

Gezz I hope I am not hyping this up and it turns to shit . But I’ve seen this design used elsewhere ( only up to 10amps ) but I see no reason why it can’t work up to 100amps using modern IGBT modules .

[Steve Broady]

Thats the thing , anyone that can follow this topic can adapt what they feel comfortable with .Want a pot that goes 0-100% , use the 555 circuit .

Want extra control and know how to program ESP32 , Arduino , go that route .

I promise, this is not just me being mentally lazy and wanting someone else to do the design work for me!

Well… mostly…

If at some point this thread ends up with a solid parts list and basic circuit design, it would open the door for a lot of folks who know how to assemble basic circuits to build their own. I think a lot of the hesitation (certainly on my part, at least) is in breaking new ground and trying things that might work beautifully, but might just as easily go zap, poof, kablooie, or even kaboom. Having a tried and true design in hand, or at least some solid design guidance, is all a little folks need.

[amh71]

Dammit, I just talked myself out of building a burst fire controller because of the led flickering, plus the fact there is nothing actually wrong with my current one.

Then along comes something even newer and even more shiny to build!

[shadylane]

My son is Electrical Engineer and does Traction control for Locomotives . He finds it very archaic that we are using Phase control and Burst fire control .

Archaic is a plug and play, variable auto transformer that's adjustable from zero to 120%
Try getting that range with any other option
If running a still off DC, I'd use IGBT's and PWM for control.
With AC, a cheap, simple and readily available phase angle control works best.

[Yummyrum]

Dammit, I just talked myself out of building a burst fire controller because of the led flickering, plus the fact there is nothing actually wrong with my current one.

Then along comes something even newer and even more shiny to build!

amh71, don’t be put off by the LED flicker on the Burstfire controllers . It is a thing but not everyone suffers from it .

This topic may go a few ways , but I believe it is the way forward .

Archaic is a plug and play, variable auto transformer that's adjustable from zero to 120%
Try getting that range with any other option
If running a still off DC, I'd use IGBT's and PWM for control.
With AC, a cheap, simple and readily available phase angle control works best.

You are correct on both accounts Shady .
If I could get my hands on …. And lift a Variac that could handle 10kw , I would . And be proud to call it archaic

Also agree that phase angle controller ( SCR or SSVR ) is the simplest solution for AC .

But , they are crude .
The linearity of control sux .
The range of the pot sux .
The metering accuracy due to the chopped wave sux.
What they do to meters sux.

And they are so simple that there is no challenge ….that really sux

Run your AC through a Bridge Rectifier , then you have DC . What are the choices ……IGBT with PWM control .

[Yummyrum]

Speaking of meters .
I haven’t really thought too much about it .

As Normandie said , using high frequency PWM on a rectified sinewave means that at least a current meter will give an accurate reading compared to a Phase angle controller .

But sticking an off the shelf meter that reads AC volts or power will not work when connected across the element as the internals need AC to power the meter . So we are back at square one . The meter hacks , like Stevebroady did still won’t work because the AC to power the meter is not the same as the rectified DC being measured .Unless an isolation transformer was used to power the meter .So it gets messy .

However an analog Voltmeter across the element would give 100% accurate readings

More complications , but an Arduino sitting on the live side could measure the element voltage and send it via serial through an opto-coupler to the safe side where another Arduino can read the serial and display it or use feedback control ti the PWM controller to maintain regulation with changing mains voltage .

Just gets more and more chaotic .

[kennstminet]

However an analog Voltmeter across the element would give 100% accurate readings

I'm afraid, a moving coil analog Voltmeter in AC Mode is calibrated for sine wave and will not be accurate with square wave.

[NormandieStill]

More complications , but an Arduino sitting on the live side could measure the element voltage and send it via serial through an opto-coupler to the safe side where another Arduino can read the serial and display it or use feedback control ti the PWM controller to maintain regulation with changing mains voltage .

My thinking was more to monitor the supply side voltage. The same transformer that we're using for arduino power and zero crossing detection could be used to measure amplitude of the sine wave. A calibration step with a multimeter can then get you an accurate read of voltage in. Add a CT and you've got apparent power. You can use this to adjust for a target power.

So rather than targeting a specific percentage, you give the nominal value of the element in watts (to set a maximum) and then dial in a target power (I.e. 900W). The monitoring side of that is what the Emon TX Shield from the Open Energy Monitor project does.

[Yummyrum]

However an analog Voltmeter across the element would give 100% accurate readings

I'm afraid, a moving coil analog Voltmeter in AC Mode is calibrated for sine wave and will not be accurate with square wave.

Doh , yes you are totally correct .

[Yummyrum]

More complications , but an Arduino sitting on the live side could measure the element voltage and send it via serial through an opto-coupler to the safe side where another Arduino can read the serial and display it or use feedback control ti the PWM controller to maintain regulation with changing mains voltage .

My thinking was more to monitor the supply side voltage. The same transformer that we're using for arduino power and zero crossing detection could be used to measure amplitude of the sine wave. A calibration step with a multimeter can then get you an accurate read of voltage in. Add a CT and you've got apparent power. You can use this to adjust for a target power.

So rather than targeting a specific percentage, you give the nominal value of the element in watts (to set a maximum) and then dial in a target power (I.e. 900W). The monitoring side of that is what the Emon TX Shield from the Open Energy Monitor project does.

I shouldn’t have drunk and posted last night

Totally over complicated it in my head . Yes , I agree , by measuring the input voltage and knowing the % ( from the pwm ) , should be able to calculate power from known element . This would take into account voltage drop as the power is turned up .

And yes , adding a CT , could measure current so would not need to be set to a know element .