220V to 110V step down questions

[okw]

Hello!

I have a resistive load 110V which I want to power from 220V. I found many designs, but they seem to not handle a lot of current. I found one promising design (see attachment). However, after I wrote this post, I found an almost identical design (drawn by someone else). They differ with one trace (separating the diode bridge) I assume the black design is correct, and the diode bridge should be separating the two "sides".

The black design says max 200W, but my resistive load is 1000W and I wanna make sure the design can handle up to 1400W (for some other elements I have) - so about 7A. It will go full on for maybe a minute or two until the water is hot, then maintain water temperature (so on/off very briefly).
I want to have the most effective, low cost, safe design.

I will exchange the TIC226 with a BTA41 (for better thermals). Is this a good choice for this application? I like designing things with common parts.
Is C106 (TH1) a good suited thyristor, or should I also replace this with a "newer"/better/more common type?

Does this design need a zero-cross detection (to prolong the life of the triac/thyristor/load)?

[Zero999]

Phase control will do, just use the standard lamp dimmer circuit. You could also use a half wave rectifier, but the power distribution network won't like it.

Why not use a transformer? An autotransformer would be half the size of an isolating transformer.

This really isn't the best way to do this, which would be to get the correct voltage rating elements in the first place.

[okw]

I have limited space (about 3.5 cm height). So I need something smart

[Terry Bites]

Zero crossing keeps the radiated and conducted noise down
If its a simple heater then burst control set to 50% can also be used to set the power dissipation in the element.
Drive the input of a MOC3041 with a 50:50 square wave. Add an external triac and other external components as shown in the data sheet.
All you'd need to add is a thermistor and a comparator to hold it at a set point.

[Zero999]

Look for heating elements rated for 220V or connect them in series?

I was talking nonsense when I mentioned half wave rectification. It would cut the RMS voltage by a factor of 1/√2, leading to twice the power being dissipated by the element, as its rating.

Phase control will work. Here's an example. I just played with the value of R1 until it gave around 50% power dissipation in the load.

C3 and R3 are just for the purposes of calculating the average power in the load, using the simulator.

Here's an application note on phase control.
http://educypedia.karadimov.info/library/an1003.pdf

[Zero999]

Zero crossing keeps the radiated and conducted noise down
If its a simple heater then burst control set to 50% can also be used to set the power dissipation in the element.
Drive the input of a MOC3041 with a 50:50 square wave. Add an external triac and other external components as shown in the data sheet.
All you'd need to add is a thermistor and a comparator to hold it at a set point.

Burst control, a type of PWM, will work, but you appear to have made a similar error to me in my original post. A 50% duty cycle will not reduce the RMS voltage to 120V, but by a factor of 1/√2. The duty cycle needs to 30% to give an equivalent RMS voltage of 120V from 220V.

[okw]

Thanks for excellent answers
Would phase or burst control be my best choice?
I assume the triac will dissapate a lot of heat, so keeping things cold would be a priority as I don't have much space for heat sinks.

[Zero999]

Thanks for excellent answers
Would phase or burst control be my best choice?
I assume the triac will dissapate a lot of heat, so keeping things cold would be a priority as I don't have much space for heat sinks.

Phase control is easier, because it needs no microcontroller, or logic circuitry, but it can be a source of radio interference and will require filtering, which isn't shown on the schematic I posted above.

Burst control requires logic, such as a microcontroller, to turn the TRIAC on for 3 cycles and off for 7, but no filtering.

I forgot to say, you need protection against fire. If the TRIAC fails short circuit, which is normally how they fail, the heater will turn on at full power and overheat.

[okw]

What kind of filtering would be required? A parallell X2 cap? Or a Y2 cap to ground for both lines? A series inductor or a choke? The two last ones I fear will be big and expensive to push through 7A.

[Zero999]

What kind of filtering would be required? A parallell X2 cap? Or a Y2 cap to ground for both lines? A series inductor or a choke? The two last ones I fear will be big and expensive to push through 7A.

An inductor and capacitor will do. Refer to the application note linked to in my previous post.

[okw]

Thanks. I suppose Figure AN1003.15 is the one? Or is a single-time-constant sufficient for a simple heating element?
If I go with the double, I see C2 is 100V in the schematic, but 400V in the table.
Which voltage rating should I choose for R1, R2 , R3, C2, C3? I guess R1, R2, C1, C2 must be over 310V?
And wattage (i see R3 is 1/2w but what about the others)?

I guess the potentiometer is to adjust the output voltage. But will the output vary with load current?

[Zero999]

Thanks. I suppose Figure AN1003.15 is the one? Or is a single-time-constant sufficient for a simple heating element?
If I go with the double, I see C2 is 100V in the schematic, but 400V in the table.
Which voltage rating should I choose for R1, R2 , R3, C2, C3? I guess R1, R2, C1, C2 must be over 310V?
And wattage (i see R3 is 1/2w but what about the others)?

I guess the potentiometer is to adjust the output voltage. But will the output vary with load current?

The voltage ratings are for the peak mains voltage, plus a margin.

The potentiometer is to adjust the firing angle of the TRIAC, which will alter the RMS voltage. I can be skipped and replaced with a fixed resistor. Refer to the schematic I posted above.

[okw]

Thanks. Will 1/10w resistors do? Except R3 which is 1/2w.
Will BTA41 and DB3 work fine together?

[Zero999]

Thanks. Will 1/10w resistors do? Except R3 which is 1/2w.
Will BTA41 and DB3 work fine together?

Why not use ¹/₂W for all resistors? That would make more sense.

The TRIAC and DIAC you've mentioned should be fine.

[okw]

Just asking, I think i have some 1/4 watts. But if needed, I'll order 1/2w.

[richard.cs]

Burst control requires logic, such as a microcontroller, to turn the TRIAC on for 3 cycles and off for 7, but no filtering.

As an alternative to a microcontroller for burst control, there are dedicated chips for this. e.g. Telefunken/Temic/whoever-owns-it-nowadays U217B. http://www.griederbauteile.ch/download/Datasheet/Temic/U217B_B.pdf (AKA T2117 and others)

It is a bit clunky by modern standards, but it does work. Hunting around a little it looks like most (maybe all) variants are no longer manufactured, but for a one-off for a home project they look quite obtainable. Onsemi UAA2016PG looks similar but a little more sophisticated.

[okw]

I want to test the step down before connecting it to a heating element. What resistor value and rating can I use? I do a few 5w and 10w, but a lot of 2w and 3w.

[okw]

To adjust firing angle, can I use a 220V 35W halogen bulb and adjust the potmeter until I measure 110V over the bulb?
Or will the firing angle vary depending on the load. Ultimately the 110V load will be around 1100-1300W (resistive), but I don't wanna start out with the actual load, as it's expensive to replace, in contrast to the bulb. And I choose a 230V bulb so it won't burn if I start out with completely wrong resistance on the potmeter.

[Zero999]

Using a 35W bulb won't affect the firing angle in the circuit I posted. The component values given should give an RMS voltage close to 110VAC, with 220VAC in.

Do you have a true RMS meter capable of accurately measuring non-sinusoidal voltages? If so, that would work.

[okw]

Yes, I have a Fluke 183 (true RMS).

I tried now, but it only lets out a few volts. I duplicated "Figure AN1003.15" in the application note you attached (load is on the other side of your R1, and your R2 is 100R instead of the 15k in the application note). The PCB is already produced, so I'd like to tweak this to work. Schematic is attached. Is my R2 value too large? Or what could be wrong?

[NiHaoMike]

Try changing R1 to 1k, then test again.

[okw]

(I'm cautious when working with high voltage): Before I do, why did the application note say 15k (and I have the identical circuit)?

[okw]

Still 0.72VAC (52Hz) over the lamp. If I disconnect the lamp (46W 240V halogen), the output goes to 242V (same as input).
I removed the the RC filter (L3 inductor and CX4 cap) and then got 45mVAC (50Hz).
I then went on to replicate the circuit Zero999 simulated (but kept my R1 1K (he: 100R), R2 3.3k and R3 470k (he: total 390k) from my original design.
Now it was 0.3mVAC. Very strange.
Could it be the load is just too low to trigger the circuit?

[NiHaoMike]

Adjust R3 down until the circuit starts to trigger, then continue turning it down until the voltage is what you expect. It's basically just a standard dimmer circuit.

[okw]

Thanks. Worked now
Had to go down to 132k on the pot.
What is the effect of using 15k and not 100R (which Zero99 did) as R1? Is it some kind of hysteresis adjustment?

I did the LTSpice simulation, and see the tops are (of course) still at 310V, but cuts shortly after (so the RMS will be 120V).
Since it's a dimmer circuit, it's probably designed for (and used with) e.g. light bulbs and motors which is already rated to run at 220-250V (and of course tolerates the ~310V tops).
Will the 120V heating element be damaged over time by the high 310V spikes? Or will the short spikes be absorbed by the heating element before creating such heat it will do any harm?