Audible noise from an SSR!

PhilHornby

Senior Member
I've never actually used an SSR before - but I'm pretty sure that 'clicking' noises emanating from them, wouldn't be considered normal :confused:

The noise stops if the load is disconnected. Anyone know what's going on?

(The SSR is being used as part of my 'Fan Heater' control experiments: thread here)

[video=youtube_share;84pORTzafkM]http://youtu.be/84pORTzafkM[/video]
 

inglewoodpete

Senior Member
Does it sound like a pinball machine? All those electrons falling into the holes. I've read that is how semiconductor PN junctions work. :):)

The short answer: "no"
 

premelec

Senior Member
Sounds like an inductive spike / spark - try a small capacitor [high voltage] across load and or relay...
 

hippy

Technical Support
Staff member
Getting all sorts of weird EMI and electo-mechanical noise is not uncommon when switching high mains loads, though that usually ends up as buzzing or hum rather than clicking and there should be far less problem with a mere 14W load.

My experience is with switching more resistive loads (light dimming) than inductive loads, which I guess a motor with a coil in it will be. That might be introducing significant issues.

According to the datasheet ( http://www.fotek.com.hk/solid/SSR-1.htm ) the SSR has zero-crossover switching but I am wondering if it's switching on mid mains cycle and causing problems, or even missing the occasional mains cycle.

Does the clicking stop when you have the load connected and you keep the relay drive signal always low or always high ?

And what happens if you replace the motor with a low wattage mains bulb ?
 

Reloadron

Senior Member
Taken from your other thread:
I just tried an initial experiment, using my cheapo fan heater and an SSR. I connected the SSR to PinC.2 on an 08M2 and knocked together some code, that varies the ON/OFF duty cycle over a period of 100mS. This equates to 5 mains cycles, but there's nothing in my code or hardware to detect the zero-crossing, so I'm out of synch with the SSR. Time periods much greater than 150mS made the fan audibly "woosh" and then free-wheel. A curious, constant-speed knocking noise can be heard at certain settings.
So we can get 1 / 100 ^-3 or a 10 Hz switching rate. My guess is you are hearing the switching rate. This being inline with what Hippy mentions above.

Ron
 

PhilHornby

Senior Member
Presumably then, there's an inductor inside the SSR that's effectively acting as a loudspeaker, in response to my modulated control input? (or the modulated back EMF from the fan motor) Thinking about it, the sound is in sync with the occasional noise emanating from the fan motor.

This sounds a lot less alarming than a 'spark' !

For one awful moment, I thought I'd managed to buy the ultimate fake SSR - one containing a mechanical relay :eek:
 
Last edited:

Reloadron

Senior Member
Presumably then, there's an inductor inside the SSR that's effectively acting as a loudspeaker, in response to my modulated control input? (or the modulated back EMF from the fan motor) Thinking about it, the sound is in sync with the occasional noise emanating from the fan motor.

This sounds a lot less alarming than a 'spark' !

For one awful moment, I thought I'd managed to buy the ultimate fake SSR - one containing a mechanical relay :eek:
Unless the noise is something you can't live with I wouldn't worry about it. :) Just to satisfy your curiosity remove the uC and control the SCR using only a DC from the uC power supply. The noise should be gone. :)

Ron
 

PhilHornby

Senior Member
Does the clicking stop when you have the load connected and you keep the relay drive signal always low or always high ?
And what happens if you replace the motor with a low wattage mains bulb ?
Yes, the clicking was only present when the SSR was being driven by the Picaxe. I didn't have a conventional mains bulb to hand (!) ... but I bit the bullet and switched on the heater elements. That also stopped the noise, but did cause my 'other half to note that the light in her sewing room was flickering! Complicated, this mains control stuff!!

The inimitable "Big Clive" has a tear down of a very similar SSR here: https://www.youtube.com/watch?v=DxEhxjvifyY
It seems "Big Clive" found the same thing that was noted on this 'instructable' ... a Triac rated at half of what it said on the outside label. Presumably, I should treat mine as 20A, not a 40A. Interesting to note that there are no Inductors inside the box though, so I'm still not clear as to what actually causes the clicking.

This is the first time I've used an SSR and I'm coming to the conclusion that discrete components are a better bet.
 
Last edited:

Jeremy Harris

Senior Member
I have several "proper" SSRs in our new house switching stuff. All are completely silent. Having read tales of early failure, poor isolation etc with some cheap Chinese SSRs, on the Open Energy Monitor forum (when I was building my excess PV energy diverter, here: http://www.picaxeforum.co.uk/showthread.php?24286-Photo-Voltaic-Immersion-heater-power-diverter-SAFETY-WARNING! ) I opted to only use well-known brand name devices. Mine are either Cruezet or Crydom, and are proving to be reliable and completely silent in operation.
 

hippy

Technical Support
Staff member
Interesting to note that there are no Inductors inside the box though, so I'm still not clear as to what actually causes the clicking.
I would guess it's either some electro-magnetic effect, something magnetising, or thermal expansion or similar. One would have thought that would require quite large currents and that such a light load would not have much effect, but in-rush currents can often be quite high compared to normal running.

Having lights flicker suggests to me there is a lot of interference, might be a lot of current involved. Are you sure that you have everything wired correctly, haven't done something silly like shorting out the motor and briefly connecting live to neutral ?

That said, a relay energised by a battery through its own N/C contact - so it alternates on-off - will give quite a kick if one touches the coil, so it might not need large currents.

Have you checked to see if the PICAXE is resetting ? It could be that the SSR is activated, noise occurs, the PICAXE resets, starts again, so isn't doing what is expected.

The other thing to try and determine is if the click is when the SSR turns on or off. You might achieve that, and determine the above, by using a slower control signal, one second on, one second off.
 

Jeremy Harris

Senior Member
I've heard noises from snubber capacitors before now. The latter were all a consequence of using a poor choice of snubber capacitor - the decent metallised polypropylene ones, with a low self-inductance, are usually far better than other types.
 

techElder

Well-known member
If you switch an inductive load with a zero-crossing switch, you will cause a maximum current in the inductor's circuit until backemf is developed. This max current can cause strange things to happen.
 

PhilHornby

Senior Member
One would have thought that would require quite large currents and that such a light load would not have much effect, but in-rush currents can often be quite high compared to normal running.

Having lights flicker suggests to me there is a lot of interference, might be a lot of current involved. Are you sure that you have everything wired correctly, haven't done something silly like shorting out the motor and briefly connecting live to neutral ?
My reputation goes before me :p

To my mind, turning on the heater elements generates the flickering next door, just as a consequence of adding and removing a 7A load from the Ring Main every 60mS. Obviously this would be an issue in my finished project, but thought it interesting that it stopped the SSR 'clicking'.

If you switch an inductive load with a zero-crossing switch, you will cause a maximum current in the inductor's circuit until backemf is developed. This max current can cause strange things to happen.
So, would the 'interference' be caused by the fact that the back EMF from the fan motor is varying (every 60mS or so), as it speeds up and slows down in response to the input signal... and that switching on the big heater element load drowns out this effect?

Incidentally, I've got a 'clamp type' ammeter and pickup coil from an energy monitor in the line to the Fan Heater. With just the fan motor activated, I can't detect any appreciable current flow with either device. With the heater elements ON, I can, and can see a very strange run-on effect when the SSR is de-energised - but I'll detail that back in my other thread, since it's probably not related to this strange SSR 'clicking' effect.
 
Last edited:

inglewoodpete

Senior Member
Residential street distribution wiring can easily cause interference when high currents are being switched. Depending on your distance from the local distribution transformer, there will be impedance in the lines, however small it may appear to be. You could check this be measuring your mains voltage several times when the heater is off and when it is drawing a steady full load (8A?).

I once lived next door to a character who did arc welding projects in his back shed during the evenings. It was "only" a 250v 15A max welder, and I would expect that it was seldom being used at capacity, but the lights in my house would flicker as he did his welding.

If you are going to switch the heater element in-and-out, you will need to do it "very often" eg on half cycles' zero crossing or use a triac to turn on part-way through each half cycle like light dimmers do. The second option may create quite a bit of RFI with a nominal 2kW load.
 

techElder

Well-known member
So, John, what magic will an MOV perform to solve the problem?

P. S. Oh, not trying to hurt feelings. Just leading to a conclusion.
 
Last edited:

PhilHornby

Senior Member
Residential street distribution wiring can easily cause interference when high currents are being switched. Depending on your distance from the local distribution transformer, there will be impedance in the lines, however small it may appear to be.
That triggered a pertinent (and worrying thought): the cottage where my Dimplex 'stove' lives is in a terrace of three houses. All three are fed via a single overhead cable which fans out from a pole at the bottom of the garden. Interference from one cottage could easily end up in the others...

him as well said:
If you are going to switch the heater element in-and-out, you will need to do it "very often" eg on half cycles' zero crossing or use a triac to turn on part-way through each half cycle like light dimmers do. The second option may create quite a bit of RFI with a nominal 2kW load.
Choice of RFI or Flickering eh? :)

I've dreamt up a scheme that uses various combinations of the heater elements, which might avoid both. See other thread.
 

johnlong

Senior Member
So, John, what magic will an MOV perform to solve the problem?

P. S. Oh, not trying to hurt feelings. Just leading to a conclusion.
Hi Tex
no feeling hurt
all the spec sheets from SSR manufactures recomend them across the AC in there wiring diagams
See Omroms users considerations it explains the need for them
sorry lost my link
regards
john
 

Goeytex

Senior Member
I have used hundreds of both random fire and zero cross AC SSR's in 250 - 1000 watt heater applications. I have never observed one making audible noise. All my SSR's have been respected brand names, usually sourced from Ebay as surplus or used. OPTO, Crouzet, Crydom, et al.

Audible noise implies movement of some kind. As SSRS have no moving parts the noise is disconcerting, since it is most likely due to internal arcing or component movement/expansion. This theory is supported by the fact that adding a cap (Snubber) across the Relay's AC terminals reduces the noise. I further guess is that this is caused by cheap components and/or poor construction used by Fotek. The problem may also be exacerbated due to the rather large turn on current related to heater wattage and non zero cross switching. On the other hand, it could just be a defective relay.

I would suggest using a Name Brand relay instead of this cheap $5 Fotek device. With the large heater element I would further suggest using a zero cross type.

If you must use this cheap relay then use a proper Snubber by adding some series resistance along with the cap. Or just get a QuenchArc device. But understand when SSR's fail they almost always fail shorted to ON.

What I would not do is spend a week or two scratching my head wondering why a $5 POC SSR relay makes noises.
 
Last edited:

premelec

Senior Member
I have read that some SMD multilayer capacitors are quite microphonic so it seems possible they would flex slightly with changing voltage and make a click - especially if mounted on a thin circuit board... another possible component oddity... of no immediate use that I can think of... ;-0
 

PhilHornby

Senior Member
All my SSR's have been respected brand names, usually sourced from Ebay as surplus or used. OPTO, Crouzet, Crydom, et al.
Those names mean absolutely nothing to me. I'm more than happy to accept them as recommendations for the future though.
Him as well said:
The problem may also be exacerbated due to the rather large turn on current related to heater wattage and non zero cross switching...
...I would suggest using a Name Brand relay instead of this cheap $5 Fotek device.
The noise is only present when running the fan alone; adding the Heater load makes the noise disappear.

"Fotek" is a name brand ;-).
him as well said:
If you must use this cheap relay then use a proper Snubber by adding some series resistance along with the cap.
I'm not sure my finished project will use an SSR at all - I just happened to have bought this one recently, and thought I'd have a play.

I have read that some SMD multilayer capacitors are quite microphonic so it seems possible they would flex slightly with changing voltage and make a click - especially if mounted on a thin circuit board... another possible component oddity...
Maybe I'm just going to have to crack the thing open and have a look :)
 

PhilHornby

Senior Member
Maybe I'm just going to have to crack the thing open and have a look :)
OK - the device has deteriorated somewhat - it's turning on with no input signal, and clicking like crazy ... and then it works for a while!

I flipped the top off, but I couldn't locate the source of the noise ... it's coming from somewhere deeper inside.

IMG_0274a.jpg

You're looking at a PCB, seemingly supported by the legs of the TRIAC soldered to it. Maybe the whole PCB is resonating?
 

Jeremy Harris

Senior Member
Looks like a fairly typical cheap Chinese SSR. It's not that well potted at all, and something is probably moving or even breaking down under load.

As for Fotek being a brand name, then that's possibly true, but it has nowhere near the good reputation of the well-known, safe and high reliability brands like Crydom, Cruezet etc. A genuine decent brand name SSR won't make any noise and will probably operate safely and reliably for years.
 

PhilHornby

Senior Member
More dismantling

It actually comes apart non-destructively.

This is the other side of the PCB (with the TRIAC bent up away from the board, so I could see its markings) :-

IMG_0278.JPG (The background is a piece of anti-static bag, in case you can't figure out what you're looking at!)

and this is a close-up of the TRIAC:-

IMG_0277.JPG

Marked (very faintly) BTA16 1000B A1 139 ?

This is supposedly a 40A SSR ... BTA16 implies a 16A TRIAC. But I can't find a datasheet that matches this device exactly. And who's logo is that?
 

PhilHornby

Senior Member
Problem found

IMG_0274a.jpg

The 'clicking' noise appeared to be emanating from the blue capacitor, on this side of the board. Poking it with a screwdriver, changed the noise...
...following the PCB tracks, reveal it to be in series with R3 (100R), on the other side of the board. The pair of them are connected across MT1/MT2 of the TRIAC.

They are a snubber circuit, are they not?!?

I got out my trusty side cutters, and snipped one of the capacitor's legs and problem solved :p

The SSR no longer appears to be ON with no input and the 'clicking' has ceased.

It just needs re-assembling now ;-)
 

premelec

Senior Member
Interesting - i was thinking MLCC rectangular caps might be microphonic - i guess all these ceramic caps could be - do you know the cap value? BTW - I don't think I'd consider this unit reliable.... ;-0
 

PhilHornby

Senior Member
No, this was one component that was well and truly 'potted' - it ended up in a million pieces, when I tried to remove it.

I assume it had broken down internally - and that's why I had current flowing, when I hadn't asked for it (quite a late development that one).

I think your initial diagnosis of a 'spark' might not have been far off the mark !
 

techElder

Well-known member
Capacitors don't tick; they break down with voltage. The break down will be internal with cheap capacitors.

The snubber circuit helps the triac turn off with high inductive loads. It also helps keep it from triggering with transient voltage events.

This part is overdue for the trash can.
 

PhilHornby

Senior Member
That capacitor 'ticked' - see youtube footage in first post. With capacitor removed, the sound is gone.

I know what a snubber circuit is supposed to do. With that capacitor, it wasn't doing it!

I'm not throwing it away, until it stops working again!
 
Last edited:

Jeremy Harris

Senior Member
The ticking was almost certainly the capacitor breaking down internally. By removing it you may well have created a bigger risk of something going bang. This cheap Chinese junk doesn't include spare components that can be safely removed - they will have already "value engineered" the thing to just barely work OK................

Frankly, I wouldn't give stuff like this house room, and I'd never, ever, have it inside the house unless I was sitting next to it when it was running, with an inert fire extinguisher to hand. I've seen far too much intrinsically unsafe Chinese crap to even think about using it where safety may be an issue.

It definitely is not potted either, it has a weak dribble of some sort of resin dribbled over it. Potting is there to protect the components and reduce the risk of a fault causing a fire. In a decent brand of SSR the whole cavity will be filled with potting compound to encapsulate all the components. You can tell decent ones because they feel heavier, just because of the added weight of the potting compound.
 

Reloadron

Senior Member
They are a snubber circuit, are they not?!?

I got out my trusty side cutters, and snipped one of the capacitor's legs and problem solved

The SSR no longer appears to be ON with no input and the 'clicking' has ceased.

It just needs re-assembling now ;-)
This is a good read on SSRs from our friends at Omron.

1. Leakage Current
A leakage current flows through a snubber circuit in the SSR
even when there is no input. Therefore, always turn OFF the
input or load and check that it is safe before replacing or wiring
the SSR.

Measuring a SSR output with for example a high impedance meter and seeing an output is not unusual and actually normal when there is no load. This is a result of the SSR internal snubber circuit.

Ron
 

hippy

Technical Support
Staff member
This is supposedly a 40A SSR ... BTA16 implies a 16A TRIAC.
This seems to be a quite common observation from those who have dismantled that manufacturer's SSR units; that the fitted Triac is rated half or less than the stated rating on the case.

The 'clicking' noise appeared to be emanating from the blue capacitor, on this side of the board. Poking it with a screwdriver, changed the noise...
...following the PCB tracks, reveal it to be in series with R3 (100R), on the other side of the board. The pair of them are connected across MT1/MT2 of the TRIAC.
Something which is clicking with 100R in series would worry me. I would guess it's either faulty or damaged, an insufficient component or circuit design, or it is just plain not suitable for the way you are using it.

And I wouldn't recommend just taking it out of circuit. If there is some inductive or other effect which is causing a capacitor to click then that effect will likely still be there without the capacitor and it is difficult to predict what adverse consequences that might ultimately have.

I would suggest the clicking is a fortuitous warning that you are heading towards some kind of failure and you don't want that to be something which has far more serious consequences if the Triac or anything else breaks down in a similar manner.
 

PhilHornby

Senior Member
Ah - now there is a brand I have heard of! They seem to have an interesting product line and comprehensive documentation. I'll investigate further.

him as well said:
Measuring a SSR output with for example a high impedance meter and seeing an output is not unusual and actually normal when there is no load.
No, it had reached the stage where my attached 46W fan was operating with the SSR supposedly off. It didn't do this during initial experiments - only later, when that capacitor was on its last legs.

I think I found my TRIAC: https://www.yoycart.com/Product/547985280143/ :(

That really doesn't fill me with confidence ...

I don't for one minute think this SSR is what it says it is (i.e Fotek). It seems the Chinese are now counterfeiting one another, as well as the rest of the world!
 

Jeremy Harris

Senior Member
Worth noting that the snubber is there to reduce the risk of a high dV/dT from triggering the triac, plus a excessively high dV/dT trigger may well cause the triac to fail, too, probably fairly dramatically. I bought a batch of fake high power mosfets that worked in most respects, but violently exploded when exposed to a high, but well within spec, dV/dT. Lucking they were inside an alloy case, but the collateral damage to the rest of the stuff inside the case was serious.
 
Top