Phase angle control of vibration pump with opto triac

[rsatchel]

Hi everyone,
I am working on a project that uses a microcontroller PWM pin to control the pressure of a vibration pump. I have the microcontroller part working, but I am a bit stuck with the opto triac part.
The opto I am using is the VO2223A from Vishay. Its AC side is rated to 0.9A, which is well in excess of the 330mA drawn by the pump (40W @ 120V)
Its datasheet is here - https://www.vishay.com/docs/81166/vo2223.pdf

While I have used DC optocouplers & zero crossing SSR's before, I am not very familiar with random fire opto triacs. Would I be correct to drive the pump from pins 6 and 8 and ignore the gate on pin 5?

On the DC side of the VO2223A I am using a 370 ohm current limiting resistor. I calculated this from the datasheet's Trigger Current of 10mA and Forward Voltage of 1.3V.
(5-1.3) - 0.010 = 370.
Closest value without exceeding the max values from the datasheet is 390 ohm.

When I connect AC power to the pump via the VO2223A (I have tried connecting via pins 6 & 8, as well as pins 5 & 8), the pump begins to hum, even when the microcontroller is disconnected from power. Can anyone tell me why this is happening?

My research leads me to believe that I will require a RC snubber, however all of the application notes and schematics I have found are for 240V inductive loads, while the pump is only 120V. How do I calculate the values of R and C for this?

Any advise is most welcome.
Thank you in advance

 

[jnnewton]

Do you have a pull up / down on the led side signal? i'm getting that the microcontroller is still connected, and if so, you could have some current flowing somewhere that you're not aware of on the input side (or you've damaged the output by doing something with that gate (i'm pretty sure it should be left open). Try hooking up between 6 and 8, and just short the input pins. That should tell you if it is actually turning off properly. If you post a schematic, there may be someone who can see a wiriing / hookup issue. You could also add a resistor to that gate to pull it to a known state if you need too (i think that's why it's available)

 

[Mr RB]

A vibration pump needs the full mains power to operate and can't be phase angle controlled, if given less than mains voltage it will sit and hum (and burn out).

You need to change to a zero crossing type solid state relay, and switch the pump on and off every 10 seconds or so, to maintain an average pressure.

You will also need a pressure accumulator (air tank like on an air compressor), to average out the pump cycling to become an even and average pressure.

 

[rsatchel]

Do you have a pull up / down on the led side signal? i'm getting that the microcontroller is still connected, and if so, you could have some current flowing somewhere that you're not aware of on the input side (or you've damaged the output by doing something with that gate (i'm pretty sure it should be left open). Try hooking up between 6 and 8, and just short the input pins. That should tell you if it is actually turning off properly. If you post a schematic, there may be someone who can see a wiriing / hookup issue. You could also add a resistor to that gate to pull it to a known state if you need too (i think that's why it's available)

I am using a current limiting resistor from the mcu to the optotriac.

 

[KeepItSimpleStupid]

That is a very odd device. I would not say that 0.9 A is overkill. It may very well be underrated.

This datasheet doesnt tell all if you ave never worked with triacs before. You ave to provide for this device, 25 mA for a minimum load, otherwise the triac will always think it's ON.

Now, I'm not sure exactly what the fancyness is doing, yet, but you can turn on a triac by keeping the LED circuit powered.

But, you CANNOT just turn te triac on an expect the zero crosiing to turn it off, like you might with a light dimmer.

Why? Because the voltage and the current are not in phase. In te simplest case, let's assume that the voltage is sinusoidal and the current is sinusoidal, but not in phase.

So, you turn it on and start counting when the voltage is zero. At some point the current will become zero, but the voltage will not be and the motor will turn off. So, you don't get what you expect.

To pase angle fire any such load, you have to do it periodically or continuously, between the zero voltage crossings. When the device goes through a current =0, it will turn off. With a resistive load V(t),I(t)=0 at t=0, but not for an inductive load.

Now, since you are creating this pulse train for on, you have to monitor the current in the load. Thus if the peak current is exceeded, the pulse should be removed.

One system I worked on was able to limit I this way. Max I becomes a settable paramter. In the systems I used they used back to back SCR's for th AC line. One reason is that both could be fired togeter to mediate surges. The adjustment range was 10%-100%, so a 10 A triac could be limited from 1 to 10 Amps.

During my tenure, I convinced them to always use pase angle fired with current limit and always use 25 A SCR units as they were called for 10-15 Amp loads. These devices had integral semiconductor fuses at $30.00 a pop and they were popping them left and right due to the nature of the system. Adding a properly sized 3AG fuse cut the number of semiconductor fuse blows significantly.

Each of these triac thingy were operating into a transformer to get like 30 V at 10A out.

Later, I conviced the powers that Be, that A DC power supply would save money in the long run, be smaller and would not pop fuses. As an added benefit, the power calculations of V * I would be real.