Problems with triac control on IR heaters

[jimscot]

We have recently (Oct) fitted 7 IR heaters which use triac based controllers. There have been any number of problems with the controllers and IR halogen elements failing. The supplier has now suggested that we may have interference or spikes on our mains supply which they claim may affect or even cause the controllers to fail! The heaters are on a separate new ring main. We do have some older fans for an oil heater on a separate ring main and a oil pump and water pumps on systems fitted in 1989 and which are on yet another ring main
Is this a reasonable possibility as an explanation? And if it is a problem what can be done?

 

[NascentOxygen]

If the controllers are known to be sensitive to spikes on the mains, then there should be appropriate filtering built into the controller circuit to render them relatively immune to ordinary levels of interference.

I sense an opportunity for the manufacturer to revise his circuitry and come up with a proper reliable design, using your home as a test bed. Presumably the manufacturer doesn't want to earn a reputation for having his controllers fail left, right, and centre due to design oversights and deficiencies?

 

[jimscot]

Dear nascent, it does seem obvious that should be the case, but we have bought the equipment through an agent who has been very helpful but we seem to be grasping at straws with unsuppressed noise etc. It strikes me as very unlikely but the alternative is a poor batch of heaters, which still seems more likely.
I just wondered if the forum had anyone with more experience that could say, it certainly is likely or very very unlikely!

 

[FOIWATER]

it is certainly possible for a triac to experience nuisance triggering under voltage spikes.

It is one of three conditions for a thyristor, or triac to fire when not wanted (it is dubbed dv/dt triggering, the others are triggered by exceeding the temperature rating for the device, or by exceeding the forward (or reverse, in the event of as triac) breakover voltage rating.

I have not experienced this problem directly with triacs, but I have with "hockey puck" style thyristors.

We kept getting a DC voltage at our motors even when the operator of the machine was not requesting forward movement (gating the scr's) The issue, was dv/dt spikes causing unwanted triggering of the scr.

How it is rectified, no pun intended, is for a resistor and capacitor combo filter to be added in parallel with each thyristor of the bridge.

Now I don't know the specifics of your device, how susceptible they will be to noise, or even if what I am saying is relevant in your case, but it is certainly possible for these devices to trigger due to spikes and it is certainly detrimental to equipment... and the way to stop it from happening is to suppress the spikes with a filter.

 

[jim hardy]

I just wondered if the forum had anyone with more experience that could say, it certainly is likely or very very unlikely!

Are there any big switching devices on same transformer? Variable speed motor drives for example?

We had a situation not unlike yours but on a 130 volt DC battery bus.
Some 7.5kva inverters, about sixty amps draw, modulated the bus voltage with 100 volt spikes at 120 hz.

Before we got that fixed, one unfortunate supplier redesigned his equipment (much smaller inverters, ~ 100 watts) away from thyristors to transistors to make it tolerate the horrid supply voltage we handed him. I always felt he'd got abused...
We later added input filters to the big inverters so they were less disruptive to our battery bus.
I'm sure his original design would have been okay after we cleaned up the supply to it.

So i would say you need to observe your power source with an oscilloscope for a while.
That'll let you know which you need to filter - your heater controllers or something else connected to the bus.
Observe near the devices that are failing. At switching frequencies the inductance of a loop in supply wires to them can give trouble. Supply and return lines should be routed together to minimize area of loop.

just 2 cents from an old guy