DIY phase controller

If it plugs in, post it here.

Moderator: Site Moderator

Post Reply
User avatar
zed255
Distiller
Posts: 1021
Joined: Thu Jan 05, 2017 7:06 pm
Location: Ontario, Canada

DIY phase controller

Post by zed255 »

Just a few pics of a bog-standard application note circuit I made PC boards for. This one has an 88A 1200V TRIAC and heavy duty screw terminals. It can be set up for either 120V or 240V mains. Made a (recoverable) error in the layout requiring a trace be cut and a jumper installed because I mixed up the TRIAC MT1 and MT2 terminals.
20210430_102431.jpg
20210430_102533.jpg
20210430_102448.jpg
Didn't really cost much to get 10 boards printed so I have extra bare boards if anyone is interested, just pay postage. I'll provide parts selection guidance and outline required changes to anyone wanting to build one. Send a PM if interested.
----------
Zed

When the Student is ready, the Master will appear.
If you can't explain it simply, you don't understand it well enough.
User avatar
zed255
Distiller
Posts: 1021
Joined: Thu Jan 05, 2017 7:06 pm
Location: Ontario, Canada

Re: DIY phase controller

Post by zed255 »

Here's the schematic, taken from a Littlefuse application note:
Phase.png
Parts I chose, many substitutions possible depending on application:

TR1 - IXYS CLA80MT1200NHB *** Any TO-3P/TO-247 TRIAC with the same pin-out, properly rated for the voltage and current required
D1 - STMicroelectronics DB3TG *** Trigger DIAC 30-34V, 2A device
D2 - ON Semiconductor DF06M-ND *** 4-pin DIP 600V, 1.5A bridge diode
VR1 - Bourns 91A1A-B28-A23L *** 2W potentiometer, may be another type mounted off board via wires, 500K/470K for 240V, 250K for 120V
C1-C2 - Kemet R474I31005001K *** 0.1uF 1kV X2 class, IMPORTANT! This MUST be an X2 class capacitor!
R1-R2 - Stackpole RSF2JT33K0 *** 2W 33K
R3 - Stackpole RSMF3JT1K00 *** 3W 1K
Terminals - Keystone Electronics 7799 ***6-pin 30A M4 snap-in PCB terminals

One note on the TRIAC: The one I chose DOES NOT have an isolated tab and requires and insulating pad / mica washer between the device and the heat sink. Others are available with an isolated tab that will be easier to mount without any insulating hardware.

I also suggest using nylon stand-offs to mount the board.
----------
Zed

When the Student is ready, the Master will appear.
If you can't explain it simply, you don't understand it well enough.
User avatar
Yummyrum
Global moderator
Posts: 7653
Joined: Sat Jul 06, 2013 2:23 am
Location: Fraser Coast QLD Aussie

Re: DIY phase controller

Post by Yummyrum »

Good onya Zed .Good to see people still constructing stuff with real components :thumbup:

I agree hat some caution should be taken and that a triac with insulated tab should be used . Sure , insulation kit could be used by experienced constructors but regardless , I’d highly recommend that the heatsink is grounded ( Earthed) as a precaution .

Incidentally ,if anyone is wondering how it works , the 500k resistor in that schematic is the pot although they have not used a Variable resistor symbol .

This is what is inside one of those Chinese SCR controllers or inside a SSRV controller. It’s a classic simple time constant phase controller circuit thats been around since the 50’s .

What intrigues me is the four diodes and two 33k resistors . I’ve never seen that used before and I’m buggered if I can see what they do . That circuit would work just the same without them AFAIK .

Whats your take on them Zed.
User avatar
zed255
Distiller
Posts: 1021
Joined: Thu Jan 05, 2017 7:06 pm
Location: Ontario, Canada

Re: DIY phase controller

Post by zed255 »

Per the application note the use of either a second time constant for additional phase shift or a diode network ensure there is no hysteresis in the circuit operation. It ensures the timing capacitor is brought to within a diode drop of 0V for each half cycle for reliably symmetrical firing.
There's nothing wrong with a non-isolated tab, so long as you are aware of it and mount it accordingly. And yes, I don't bother with a custom potentiometer symbol and model in LTSpice, so yes the 500K is noted as a VR / potentiometer.
----------
Zed

When the Student is ready, the Master will appear.
If you can't explain it simply, you don't understand it well enough.
User avatar
Yummyrum
Global moderator
Posts: 7653
Joined: Sat Jul 06, 2013 2:23 am
Location: Fraser Coast QLD Aussie

Re: DIY phase controller

Post by Yummyrum »

Cheers Zed .Great stuff in that link .Learned something today :thumbup: Have you tried it with and without and noticed any real improvement .

I agree , nothing wrong with non isolated tabs if done right . Nearly all of the stuff I used to repair in the 80,90 and early 2000 used them . But I have seen quite a few badly done ones .

Not trying to labour the point but if a electronics newby is going to use them , by us talking about it in a few posts , something might sink in that they need a bit of care . :ewink: Its amazing the bleeding obvious things that can get overlooked .
User avatar
zed255
Distiller
Posts: 1021
Joined: Thu Jan 05, 2017 7:06 pm
Location: Ontario, Canada

Re: DIY phase controller

Post by zed255 »

Yummyrum wrote: Sat May 01, 2021 4:37 pm Cheers Zed ... Its amazing the bleeding obvious things that can get overlooked .
Not wanting to sound too hard, but folks who don't know shit from peanut butter shouldn't be playing with mains. I wouldn't consider this for the absolute novice with respect to at least handling mains powered circuits. I did not point it out, but you did and thank you, that the heatsink should be grounded along with the rest of the box.
----------
Zed

When the Student is ready, the Master will appear.
If you can't explain it simply, you don't understand it well enough.
User avatar
zed255
Distiller
Posts: 1021
Joined: Thu Jan 05, 2017 7:06 pm
Location: Ontario, Canada

Re: DIY phase controller

Post by zed255 »

So, noticed an problem...

I ordered up some better mounting hardware and was giving my TRIAC phase controller a real test and noticed the 1K (R3) was getting ridiculously hot when setting the potentiometer for full conduction (Max Power: Strap yourself in and feel the G's!). I've seen application notes with this value anywhere from 1K to 3.6K at 1/4W usually. It is just supposed to limit the wiper current of the potentiometer so the track / contact don't burn out as the potentiometer is dialed towards minimum resistance. For some reason at or very close to maximum conduction the dissipation in this part increases dramatically.

I get to looking at my simulation and come to realize that I did not look at the current flow through this part, as it is almost universally spec'd as a 1/4W part. I gave it no thought and just went big (2W in my case) with the intent on having a cool running uber reliable controller. I get under worst case something stupid like 46W dissipation in R3! No wonder it was begging for mercy. Aside from the timing capacitor reset network it is a bog standard application note circuit, and I've verified that the 'reset' network is not causing the problem. Make R3 too high in value and you affect the timing circuit, make it too low and the potentiometer is not long for this world. My only difference was I wanted to make mine a 3-wire type rather than the old-school 2-wire type, where the main difference is the location of the load. I simulated it with the load in the traditional location and operating it as a 2-wire and there was no issue with the R3 power dissipation. Curious, so I thought.

It looks like the gate draws a lot of current when the timing circuit resistances get low. I compared my parts' specifications against other popular parts and don't find my gate current to be much higher than other parts in the same playing field which makes me wonder why other folks have built similar circuits and gotten away without issue. Turns out the location of the load is the key. My initial testing was with a bulb limiter, effectively operating as a 2-wire phase controller. When I went off the bulb limiter and attached a real load to operate as a 3-wire phase controller is where the difference became apparent. The load effectively shunts the timing circuit in 2-wire mode, reducing the dissipation in the timing circuit.

Project works fine, just not wired as originally intended :oops:. I feel a little dumb posting this prematurely. As I said it works fine, with changes to its implementation.
----------
Zed

When the Student is ready, the Master will appear.
If you can't explain it simply, you don't understand it well enough.
Post Reply