Power Transformer buzzing when fed via Solid State Relay-Why?

I have a 550VA Toroidal 220v AC transformer fed from the mains via a 50Amp Solid state relay. The relay requires a 3 to 32V DC input to turn it on. It is a zero switching type device. The transformer buzzes ( no load) when powered via the relay, and does not when placed directly across the mains. The enable voltage applied is a clean 12vDC - I even tried it with a 12v battery. I tried a 2nd ( and a 3rd!) relay to be sure its not the relay. It is NOT the transformer core or windings buzzing due to loose elements - the windings are vacuum impregnated with varnish.

The relay is a Carlo Gavazzi type P/N - RM1A23D50 - datasheet here if interested:

**LINK**

SSR's are supposed to handle this kind of application with no problems...

Any Ideas?

Joe

Does it buzz on load?

It sounds like you are getting a switching transient on the output which may be an artefact of the fact that there is no load on the triacs. The spec sheets call for a minimum current of 250ma.

If you have an oscilloscope available have a look at the output of the relay.

If it doesn't buzz on load, try wiring an incandescent lamp (60w will probably do) across the output to provide a baseload.

Edited By Peter Cook 6 on 25/03/2021 18:51:52

Odd. I notice the Datasheet quotes a minimum operating current of 150mA, and I guess it's because the switching electronics are non-linear below that value. As the transformer is off-load, the current draw is probably below minimum and the transformer rattles due to the sine wave going clickity clack near zero.

Without engaging brain or looking up Thyristors, I visualise something akin to the non-linear conductivity of diodes, where silicon doesn't conduct below 0.6V, causing a little clunk near the zero point, and the effect is swamped under full load. A scope on the output would show if the waveform is distorted or not. If the waveform is clean, I'm baffled!

Dave

Peter/Dave - good thinking! I will attach the load tomorrow and try - also get the scope on it - Will let you know what results.

Thank You.

Joe

It shouldn't be a problem with a continuous gate voltage applied, but triacs have a holding current. If the current through the main terminals drops below the holding current then the device turns off. But that usually only applies if the gate voltage has been removed.

My best guess is that the output waveform is badly distorted. It's going some to make a toroid buzz. I agree with using a 'scope to look at all signals, input and output.

Andrew

You should not use a zero crossing SSR with a highly inductive load like a transformer. Turning on at zero causes very high peak currents. See:
https://www.te.com/commerce/DocumentDelivery/DDEController?Action=srchrtrv&DocNm=13C3206_AppNote&DocType=CS&DocLang=EN

Robert G8RPI.

Edited By Robert Atkinson 2 on 25/03/2021 20:57:19

As I read the data sheet the SSR has two back to back thyristors rather than a single device triac but as Andrew said the spec shows a 150mA holding current so they are probably dropping out and taking chunks out of the waveform on both half cycles. An SSR with a Triac output stage may perform better.

regards Martin

If so, why doesn't Joe's Datasheet mention inductive limitations, and it does say the SSR is OK for electric motors? I'm wondering if Robert's Application Note, quoting a 1974 reference, is describing a problem now fixed? Joe's SSR claims to have 65kA short-circuit protection, so maybe it doesn't matter if the transformer core saturates briefly?

I know nothing about SSRs!

Dave

Robert is correct. We discussed this problem with our SSR supplier when I was working. They supplied us with some sample SSRs specifically for use with transformers. I think some of them ended up in my workshop. We never used them in production though. Most of the transformers we used were low voltage high current secondaries and we used phase angle control on the primary side.

Peter

The laws of physics don't change. While the references may be old the application note metadata indicates it was authored in 2000 and last updated in 2013

Data sheets hardly ever highlight the shortcomings of the device.

Thyristor SSRs tend to have short off periods at low load current excerbating this problem.

Robert G8RPI.

Well, I must admit I am having difficulty getting my head around this...

Robert's reference is baffling - I find one other reference to the same concept and yet in NONE of the datasheets from Gavazzi, Craydom, etc, etc, do they refer to this issue. These are two very prominent manufacturers of SSR's and they do not mention this issue in relation to the use of their products with inductive loads at all. yet they have a zero switching and a peak switching product...

mentioned in the text of that reference is the residual magnetism in the core - surely after a few mains cycle that would stabilise in a symmetrical fashion? Especially in a good quality transformer core?

I spoke with the Factory Electrical Engineer at Rossing Uranium Mine ( close by to our town..) - they use SSR's by the hundred to control pumps and crushers - He claims never to have heard of this being an issue! And when he dug in stores, found the SSR's they have are all zero crossing. Interesting..

Anyway, I did some tests - the 'scope measurement I need to re-do - there was significant secondary AC waveform distortion, but the secondary is a CT 140V winding, feeding a full wave rectifier and 40,000uf of filter cap - and then driving the load. The diodes and filter cap is part of the waveform distortion problem, so I need to disconnect the rectifiers and load the windings directly.

The buzz did not change intensity with the transformer load varying from no load to 400 watts. The 'scope waveform also did not change much, measured on the AC side of the rectifier.

Joe

This paper 'Solid State Relays Common Precautions' is interesting, including their inputs being sensitive to inductive noise.

My favourite: SSRs may occasionally explode.

Dave

Saw that paper as well, Dave, and they also mention SSR use with motors and inductive loads, almost in passing - page C-343, par 3 and 4 - NO mention anywhere of the need for zero switching, peak switching or random switching in those situations...All they talk about is sizing for inrush currents. And OMRON is also a reputable mnfr in this field.

I am going to ignore the slight buzzing..

Joe

Electronics isn't my field, and I know approx. nothing about SSRs, but I'm aware that the audio folk sometimes have trouble with buzzing or humming toroidal transformers. It seems that even well-made toroids can hum, given half a chance, and common causes seem to be a DC bias on the input, dirty AC input, and sharp secondary current pulses as a result of large filter caps. An isolating transformer, or a 'DC trap' can deal with the first cause and some R before the filter cap can deal with the last cause, and both are easy diagnostic lash-up tests. Dunno about input waveform though - presumably fairly horrible from SSRs, and, if not symmetrical, can a DC bias result? Hope this helps, but it's probably far too naive...