I have a water heater rated at 1200 W. Typically, the heater operates using AC power, which can be controlled with a thyristor. However, I now require a setup where the AC power is first converted to DC and then used to switch the heater on and off using PWM control. My query pertains to designing an H-bridge configuration that can provide a voltage around 150 V and handle a current of 8 A. Are there any readily available plug-and-play H-bridge solutions that meet these voltage and current requirements?

 

Why do you think you need an H-bridge, when all you need is a single MOSFET or IGBT?
if it is unsmoothed DC, then you can still use the thyristor.

 

After the bridge rectification, if I only use a MOSFET or IGBT, will this work? Ideally, when using a microcontroller, we should have a smoothed DC.

 

First of all, be very careful with these high voltages. They can be lethal!
Of course, a single Mosfet or IGBT will work. Practically, You can put it into the negative line, using an N-channel fet.
The H bridge is only in cases necessary, when the output current must flow in both directions, for example when You drive a motor. In case of a water heater it is not the case. Even the DC is not necessary, the water will heat up from AC as well. The thyristor can also be controlled by a microcontroller.
A smoothed DC is necessary for the microcontroller, but it requires 5V or 3.3V, not 150V.

 

After the bridge rectification, if I only use a MOSFET or IGBT, will this work? Ideally, when using a microcontroller, we should have a smoothed DC.

But you are not running your microprocessor off 150V
Do you realise how big a capacitor you would need to smooth a 1200W supply?

 

I understand your point, and I share your perspective. My intention in shifting to DC is to overcome the challenge of dealing with varying frequencies in AC power, which might not always be at 50Hz. By using DC, I aim to simplify the control process and eliminate the need for frequent calibration to determine the zero-crossing point.

 

But you are not running your microprocessor off 150V
Do you realise how big a capacitor you would need to smooth a 1200W supply?

No, the 1200W refers to the heater element. After smoothing, I will have approximately 330V DC RMS from the 230V AC input. The microprocessor will generate a gate pulse to initiate the switching. Once the system is activated, I expect to have sufficient voltage and current passing through. Am I correct in this understanding?

 

First of all, be very careful with these high voltages. They can be lethal!
Of course, a single Mosfet or IGBT will work. Practically, You can put it into the negative line, using an N-channel fet.
The H bridge is only in cases necessary, when the output current must flow in both directions, for example when You drive a motor. In case of a water heater it is not the case. Even the DC is not necessary, the water will heat up from AC as well. The thyristor can also be controlled by a microcontroller.
A smoothed DC is necessary for the microcontroller, but it requires 5V or 3.3V, not 150V.

understand your point, and I share your perspective. My intention in shifting to DC is to overcome the challenge of dealing with varying frequencies in AC power, which might not always be at 50Hz. By using DC, I aim to simplify the control process and eliminate the need for frequent calibration to determine the zero-crossing point.

 

No, the 1200W refers to the heater element. After smoothing, I will have approximately 330V DC RMS from the 230V AC input. The microprocessor will generate a gate pulse to initiate the switching. Once the system is activated, I expect to have sufficient voltage and current passing through. Am I correct in this understanding?

No. If you want to smooth the supply, you need a capacitor.
To get it down to 10% ripple (23V) the capacitor needs to be C=It/V, which gives 2200uF.
Price: about £33.

What is the big problem in dealing with a variable frequency? In most countries it doesn’t vary very much. If you use a zero-crossing opto triac such as MOC3041, you don’t even need to make a zero-crossing detector circuit.

 

understand your point, and I share your perspective. My intention in shifting to DC is to overcome the challenge of dealing with varying frequencies in AC power, which might not always be at 50Hz. By using DC, I aim to simplify the control process and eliminate the need for frequent calibration to determine the zero-crossing point.

You don't need calibration at all. What You need is: Synchronisation.
You mentioned, You're planning to use a microcontroller. Just interface the AC signal into one input of this micro*, and the firmware can synchronize itself in every half-period, in every single zero-crossing whether it is 50Hz or 60Hz or something. This is the only reliable method to find the zero-crossing.

*: I mean: interface it. Don't wire the 230V direct to the microcontroller, it will not survive.

 

No. If you want to smooth the supply, you need a capacitor.
To get it down to 10% ripple (23V) the capacitor needs to be C=It/V, which gives 2200uF.
Price: about £33.

What is the big problem in dealing with a variable frequency? In most countries it doesn’t vary very much. If you use a zero-crossing opto triac such as MOC3041, you don’t even need to make a zero-crossing detector circuit.

Initially I was going to use a solid state relay to switch on and off the AC so If I use this opto triac and will fire into this relay this should work am I correct ?

 

You don't need calibration at all. What You need is: Synchronisation.
You mentioned, You're planning to use a microcontroller. Just interface the AC signal into one input of this micro*, and the firmware can synchronize itself in every half-period, in every single zero-crossing whether it is 50Hz or 60Hz or something. This is the only reliable method to find the zero-crossing.

*: I mean: interface it. Don't wire the 230V direct to the microcontroller, it will not survive.

Can you please elaborate more what is meant by synchronisation. Thanks in advance

 

I have a water heater rated at 1200 W. Typically, the heater operates using AC power, which can be controlled with a thyristor.

I'm kind of dumb I guess. What is wrong with using the temperature controls made for the water heater? If your thinking that you will be able to make the temperature control more exact by using a micro I'm pretty sure your wrong, the mass of the water in and out won't allow it.

 

I'm kind of dumb I guess. What is wrong with using the temperature controls made for the water heater? If your thinking that you will be able to make the temperature control more exact by using a micro I'm pretty sure your wrong, the mass of the water in and out won't allow it.

This is a project and I am going to control the unit by using implement a fuzzification system.

 

Synchronization means you sample the AC input to find the zero crossings. It would never work to try use a separate frequency source to match the frequency. No matter how close you matched the frequency, it would soon be off.

What advantage do you hope to get by doing this instead of a simple thermostatic control?

 

Synchronization means you sample the AC input to find the zero crossings. It would never work to try use a separate frequency source to match the frequency. No matter how close you matched the frequency, it would soon be off.

What advantage do you hope to get by doing this instead of a simple thermostatic control?

A thermostatic control is like having one already on the water heater. I need to control it myself, depending on how this will be set by the end user through software interference.

 

Initially I was going to use a solid state relay to switch on and off the AC so If I use this opto triac and will fire into this relay this should work am I correct ?

Triac/opto-triac and solid state relay are the same thing.
Forget fuzzy logic, just make a proportional circuit, with a 555 and a comparator.

 

Take a look at Fairchild AP-3006 fig 3. Uses a 555.
Also Microchip has an app note specifically aimed at AC heater, power triacs . I don't have the # off hand.
The Microchip app replaces the old style bimetal strip control.

 

Take a look at Fairchild AP-3006 fig 3. Uses a 555.
Also Microchip has an app note specifically aimed at AC heater, power triacs . I don't have the # off hand.

https://www.onsemi.com/pub/collateral/an-3006.pdf
http://ww1.microchip.com/downloads/en/appnotes/00958a.pdf
but don’t use a pot with a metal shaft!

Here’s the 555 and comparator circuit. The slow triangle wave on the capacitor is buffered and used to vary the comparator threshold slightly over a period of a second or so.

 

https://www.onsemi.com/pub/collateral/an-3006.pdf

Didn't realize that Onsemi had taken them over!