Electronics > Projects, Designs, and Technical Stuff

IGBT Dimmer - PWM AC Power Control for an Immersion Heater

<< < (4/7) > >>

David Hess:

--- Quote from: willz1200 on August 25, 2018, 03:46:43 pm ---
--- Quote ---IGBTs like bipolar transistors have an advantage in cost at high voltages because MOSFETs have a die size which is proportional to the square of the voltage.
--- End quote ---

When you say higher voltages do you mean in the kV range? Because briefly looking around 400V appear to be a similar price point.
--- End quote ---

Very few IGBTs are made for use below 600 volts because MOSFETs will generally be more economical.  600 volts and higher is where IGBTs become competitive.

I actually did not know that they made any IGBTs below 600 volts until I just looked but there are a few for special applications.


--- Quote ---So do you think MOSFETs are the way to go? I was thinking of trying the FDA24N40F which is 400V, 23A maybe a slightly higher voltage rating would be better? I'll pull some ideas together in LTspice.
--- End quote ---

Pick a IGBT and then compare the price to that of a MOSFET with an on resistance which produces the same voltage drop.  For greater cost the MOSFET will always deliver lower static losses which will dominate in a low frequency switching application.


--- Quote ---
--- Quote ---The gate drive has to come from somewhere.
--- End quote ---

Of course, but what I mean is one of the schematics uses a zener diode to clamp the gate voltage and the other uses a drive IC.
--- End quote ---

AC applications like this would usually use a TRIAC or MOSFET based solid state relay which do not support PWM faster than the line frequency.  Higher frequency PWM requires more gate drive power so a different circuit is used to provide it.


--- Quote ---
--- Quote ---Sometimes two or more common and differential mode suppression networks are used in series to provide enough suppression.
--- End quote ---


Okay, so is each individual suppression network tuned for different harmonics that can be expected?
--- End quote ---

Something like that.  The lower frequency suppression network will not perform well at high frequencies so a separate high frequency suppression network is also used.

Zero999:

--- Quote from: Richard Crowley on August 25, 2018, 04:21:38 pm ---
--- Quote from: Hero999 on August 25, 2018, 04:06:12 pm ---I don't understand the question.
--- End quote ---
Look at the ENTIRE system (from solar energy to the temperature of the water).  The water is a THERMAL "capacitor". It doesn't care whether you are powering the heater from a varying DC current, or from a PWM AC current, or from a slow on-off power (whether AC or DC).  The "slow on-off" is the method used by the overwhelming majority of temperature control systems.  Because of the massive integration of the energy sink (whether the water in your hot water tank, or the air temp in your house.)

--- End quote ---
I understood your point. What I didn't get was the original poster's question.


--- Quote from: willz1200 on August 25, 2018, 03:54:47 pm ---
--- Quote ---The thermal mass of the water is like a humongous capacitor that is integrating your nice PWM back into a running average.
--- End quote ---
But won't this result in a running average of a lower voltage which will cause less power to be dissipated in the heater?

--- End quote ---


--- Quote from: willz1200 on August 25, 2018, 04:28:18 pm ---
--- Quote ---A TRIAC is probably the most efficient option.

--- End quote ---
Thanks for clearing that up, doesn't the unsymmetric nature of triac phase cutting, generate more noise than PWM?
--- End quote ---
If the PWM is at a much higher frequency, than the mains, then it will generate much more RFI, than a TRIAC with phase control.



--- Quote ---
--- Quote ---The additional complexity can be avoided by using an opto-coupler TRIAC with a built-in zero crossing circuit, such as the NTE3097 or MOC3042.
--- End quote ---
Aren't these drivers only suitable for burst mode control? Which would give me a small dimming window? Meaning less dimming levels.

--- End quote ---
The range of power control is the same with burst control as with phase control, the difference is it occurs over a much longer timescale.

willz1200:

--- Quote from: David Hess on August 25, 2018, 04:59:56 pm ---For greater cost the MOSFET will always deliver lower static losses which will dominate in a low frequency switching application.
--- End quote ---
Yes after looking around I found these:
FMH30N60S1 - 106 mOhms (£1 each on aliexpress)
FDL100N50F  - 43 mOhms  (£2.50 each on aliexpress, £13.19 each on RS)
The only issue is that they could be counterfeit parts on aliexpress but probably worth a try. It might be best to just get a big heat sink and cheap out. After all it won't be running at high current for a great deal of the day.

Thanks for the information :)

Zero999:

--- Quote from: willz1200 on August 25, 2018, 04:41:55 pm ---
--- Quote ---The fine-grained control is needed for the supply side, not the load, to be able to smoothly use only the excess generated power.
--- End quote ---
Yes that's my main aim, slowly increase power using a pid controller until the grid tie is on the edge of import/export.


--- Quote ---But then what difference does the efficiency of the switching methodology make when the ultimate point to the exercise is to simply waste power as heat energy?
--- End quote ---
Point taken maybe I should bolt the FETs to the hot water tank :-DD I was just wondering how the use of IGBTs and MOSFETs compared and each PWM topology. But yes the main idea was to put a more uniform load on my inverter. The triac method causes some odd audible noises from my inverter, compared to boiling the kettle which doesn't produce the same noises.

--- End quote ---
Oh I see what you're trying to do now: make a variable resistor. Yes PWM can probably be used to to that, but you'll nee a fairly large filter to get rid of the EMI.

willz1200:

--- Quote from: Hero999 on August 25, 2018, 06:19:53 pm ---If the PWM is at a much higher frequency, than the mains, then it will generate much more RFI, than a TRIAC with phase control.
--- End quote ---
Okay, Would the PWM noise be easier to suppress due to the uniform change in current?


--- Quote ---The range of power control is the same with burst control as with phase control, the difference is it occurs over a much longer timescale.
--- End quote ---
Wouldn't that result in more difficulty in modulating the instantaneous power?


--- Quote from: Hero999 on August 25, 2018, 06:22:17 pm ---Oh I see what you're trying to do now: make a variable resistor. Yes PWM can probably be used to to that, but you'll nee a fairly large filter to get rid of the EMI.
--- End quote ---
Super, Do you have any recommendations on what suppression network I should use?  :) 

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