Electronics > Projects, Designs, and Technical Stuff

IGBT Dimmer - PWM AC Power Control for an Immersion Heater

<< < (5/7) > >>

drussell:
It just occurred to me that since the heating element doesn't care whether it is fed AC or DC, another possibility, in theory, is to use a "power factor correction" type circuit with a regulator controller designed for PFC, some of which have a way to vary the output voltage via a control signal.

It is basically the same idea as the OP already has but it might be more easily implemented using components typically intended for PFC rather than rolling your own.  PFC is designed to basically just make your load look resistive, even when it isn't, which is essentially what the OP is trying to do, in a general sense.

Just a thought.  :)

jbb:
I’m curious about what might be best for this...

From the perspective of ‘don’t be mean to the inverter,’ a TRIAC is out (generates lots of harmonics in phase control or ‘lumpy’ waveforms in full wave mode).

So it’s MOSFETs / IGBTs. For a given rating they behave in somewhat similar ways.

For maximum efficiency, I would look at back-to-back connection (large MOSFETs would have lower losses that IGBTs and will make cooling easier and cheaper). An incoming line filter (e.g. 2x Y caps, common mode choke, filter inductor and X cap) will be required to keep the PWM out of your house wiring (and away from the inverter). A snubber will be required to eat voltage spikes across the power devices - and any inductance in the heater circuit will push more and more energy into the snubber.

Is there a thermostat in the heater wiring? Or an over-temperature cutoff? These devices may not be able to switch off properly (I.e. arc over and catch fire) in some circuits.

I don’t suppose you can get high voltage DC straight out of your inverter?

Richard Crowley:
I have an old reel-to-reel tape machine (Sony TC-850) that can operate at the user's selection of 3.75 or 7.5 or 15 inches per second (IPS)  It uses a servo to maintain the selected speed. (Toothed gear and magnetic reluctance pickup to sense motor speed.) And a split-phase AC motor with the capacitor to generate the alternate "phase".  Technology from the 1960s.

But for the control element for the motor, I was amused that they just put an AC "resistor" in series with the motor. The "AC resistor" consisted of a bridge rectifier with a TO-3 transistor (2N3055 or equivalent?) on the DC side.  Rather clever, I thought. With the pass transistor bolted to the center of a beefy aluminum plate around 100mm square.  Certainly much quieter (electrically and acoustically) than some other alternatives.  Of course, that configuration was only possible because the big power transformer had a dedicated secondary winding to power the capstan motor.

willz1200:

--- Quote from: drussell on August 25, 2018, 08:36:49 pm ---It just occurred to me that since the heating element doesn't care whether it is fed AC or DC, another possibility, in theory, is to use a "power factor correction" type circuit with a regulator controller designed for PFC, some of which have a way to vary the output voltage via a control signal.
--- End quote ---
I'm not very familiar with power factor correction controllers and haven't had much luck in finding a controller that accepts a control signal, can you point me in the right direction? Its worth considering if it will also allow for fine-grained control :)


--- Quote from: jbb on August 25, 2018, 09:35:31 pm ---I’m curious about what might be best for this...

From the perspective of ‘don’t be mean to the inverter,’ a TRIAC is out (generates lots of harmonics in phase control or ‘lumpy’ waveforms in full wave mode).

So it’s MOSFETs / IGBTs. For a given rating they behave in somewhat similar ways.

For maximum efficiency, I would look at back-to-back connection (large MOSFETs would have lower losses that IGBTs and will make cooling easier and cheaper). An incoming line filter (e.g. 2x Y caps, common mode choke, filter inductor and X cap) will be required to keep the PWM out of your house wiring (and away from the inverter). A snubber will be required to eat voltage spikes across the power devices - and any inductance in the heater circuit will push more and more energy into the snubber.

Is there a thermostat in the heater wiring? Or an over-temperature cutoff? These devices may not be able to switch off properly (I.e. arc over and catch fire) in some circuits.

I don’t suppose you can get high voltage DC straight out of your inverter?

--- End quote ---

Yes I think the extra money I spend on the MOSFET will save money on the heatsink, So there's probably not a huge difference in the price point... Worth going for MOSFETs. Excellent and just RC snubbers will be sufficient? In some cases I've seen snubbers dedicated to the optocoupler is this necessary? I think the ball park recommendation is 100nF 100ohm but I've got an app note somewhere to find the proper values. Yes the immersion heater has a builtin thermostat. High voltage DC would be ideal my inverter is an ABB PVI-3.6-TL-OUTD-S unfortunately it looks like theres only AC out.


--- Quote from: Richard Crowley on August 25, 2018, 10:35:25 pm ---I have an old reel-to-reel tape machine (Sony TC-850) that can operate at the user's selection of 3.75 or 7.5 or 15 inches per second (IPS)  It uses a servo to maintain the selected speed. (Toothed gear and magnetic reluctance pickup to sense motor speed.) And a split-phase AC motor with the capacitor to generate the alternate "phase".  Technology from the 1960s.

But for the control element for the motor, I was amused that they just put an AC "resistor" in series with the motor. The "AC resistor" consisted of a bridge rectifier with a TO-3 transistor (2N3055 or equivalent?) on the DC side.  Rather clever, I thought. With the pass transistor bolted to the center of a beefy aluminum plate around 100mm square.  Certainly much quieter (electrically and acoustically) than some other alternatives.  Of course, that configuration was only possible because the big power transformer had a dedicated secondary winding to power the capstan motor.

--- End quote ---
So was the transistor just operating in its linear region to limit the speed? I bet the block of aluminium kept it cool, finding nice big PCB mounted heatsinks (For a decent price) has proven to be quite tricky, cpu heatsinks always work but they're not very nice to mount.   

Richard Crowley:

--- Quote from: willz1200 on August 25, 2018, 10:59:58 pm ---So was the transistor just operating in its linear region to limit the speed?
--- End quote ---
Yes. It seems rather crude in the 21st century, but it had some elegance to the basic concept.


--- Quote ---I bet the block of aluminium kept it cool, 
--- End quote ---
Yeah, the heat-sink was larger than the motor.

Navigation

[0] Message Index

[#] Next page

[*] Previous page

There was an error while thanking
Thanking...
Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod