How hot should an SSR run, (and is this a sign of a DOA one?)

I'm working on getting my new brewery up and running. Did my first autotune try today.

I gave two AT tries, one from 50F to a SV of 165F, and one from 145F to a SV of 165F. The first time, the temp overshot to 208, at which point I shut it down. The second time, it overshot to 195, at which point I shut it down.

40A SSR w/ standard heatsink, 5500W element pulling ~23A IIRC, SYL-2352 running the SSR.

Wiring diagram here: https://www.homebrewtalk.com/f170/diagram-critique-my-brutus-20-10-a-279903/

Once the PID passed the SV during the autotunes, the "on" light on it shut off. So did the LED on the SSR. But the element was pumping out heat. Not as much as before, when the PID was at 100% for the AT, but it was still boiling water. When I power cycled the Line/Load side of the SSR, that fixed the problem and allowed the system to start cooling.

So it looks like my SSR is sticking "on". Power cycling it after it's stopped getting a control signal unlatches it, at least that's the way I read it. OH, and the SSR heatsink is frigging HOT during all this, (even once the SSR has supposedly shut "off"). Like we're talking burn your skin hot. It's in an open box with plenty of ventilation.

So, blown SSR?

I'd say either the SSR stuck on or the PID control signal stuck despite the LED going off. You'd need a voltmeter to confirm which. I've not really looked at the schematic as I just woke up from a nap and honestly don't feel like thinking. It's always good to double check wiring though. When you built it, did you wire everything at once? Or wire one circuit, test, wire one circuit, test etc.?

Thanks Irregular Pulse, I suppose I should have posted the pic...you actually chose an earlier iteration from that thread. Here's what it is now:


Coderage, first time I turned it on was when the wiring was all done. I suppose I'm going to to have to man up and give it full power while it's open so I can test with a multimeter....carefully. That bit scares me .

(click pic above for bigger)

Sounds like a wiring problem or defective SSR. I don't know how common that is. I just got mine installed. My SSR heat sink was running hot as hell. See sig for fix.

My 40 amp SSR's get to about 120F when things are running 100%, they are outside the box. Can you swap the SSR's to test?

random thought is there anyway to turn on the system but not turn on the heating element? see if it is heating at the start, see if it is a wiring problem and not the ssr?

almost sounds like the element is still running on 120

samc said:

My 40 amp SSR's get to about 120F when things are running 100%, they are outside the box. Can you swap the SSR's to test?

Click to expand...

Yeah, mine are definitely way above 120F. Based on sticking my hand in 146F water, I'd say they're pushing 150-160F. (this was pushing 100% too, since in AT it runs 100% or 0%). I can't swap SSRs with any degree of happiness....the other is a 25A SSR....I don't like running it that close to redline...

bf514921 said:

random thought is there anyway to turn on the system but not turn on the heating element? see if it is heating at the start, see if it is a wiring problem and not the ssr?

Click to expand...

Not sure I follow. I can hit the disconnect on the latching relay that provides power to the Line/Load side of the SSR, and when I do that the element definitely goes dead. I can shut off the PID providing control voltage to the SSR, and when I do that I'm PRETTY sure the element stays live, (I think, I need to check this tomorrow).

bf514921 said:

almost sounds like the element is still running on 120

Click to expand...

My thoughts too...seeing as it was only pumping out partial power and not full power....but there's no way that element can see neutral....

Stevo2569 said:

Sounds like a wiring problem or defective SSR. I don't know how common that is. I just got mine installed. My SSR heat sink was running hot as hell. See sig for fix.

Click to expand...

Yeah, definitely gonna be mounting a fan in there, regardless of my current problem.



Thanks for the responses folks. Tomorrow, if I get a chance, I'm gonna fire it up live and get in there with a multimeter....hopefully get more info.

I recently had a similar problem, make sure your PID is not in the manual mode (where you select a %value). Mine was somewhat confusing initially, however when i eventually got it set in its PID mode the auto tune worked fine.

Edit: I re-read your initial post, and it seems doubtful my above suggested situation is what you have, especially if the elements output decreased after reaching the set value.

Where did you get your SSRs from. I'm running into the same problem on my build. They both get cooking to the point that it smells bad in the box. My heat sinks are inside, with no exhaust fans, but I also have a box big enough to stuff a small kid in. I think they are just bad SSRs, but waiting 2 weeks for new ones from China is really a crap option.

I'm running mine on a BCS setup, but the high voltage side is the same setup. I can shut off the elent by simply shutting off the contractor and turning it back on.

hatrickwah said:

Where did you get your SSRs from. I'm running into the same problem on my build. They both get cooking to the point that it smells bad in the box. My heat sinks are inside, with no exhaust fans, but I also have a box big enough to stuff a small kid in. I think they are just bad SSRs, but waiting 2 weeks for new ones from China is really a crap option.

I'm running mine on a BCS setup, but the high voltage side is the same setup. I can shut off the elent by simply shutting off the contractor and turning it back on.

Click to expand...

Cheapos from china, whoever the lowest price was that day... If I get a replacement it'll be from the US to get it faster, but of course the ones shipped from the US are still cheapo Chinese imports.

Alright guys, got to dig into it a bit more.

Rechecked all my wiring, everything's good...(actually, realized I had switched 120V going to where I wanted live 120V for one of my latching contactor relays, and fixed that).

Found out a bit more about my problem. When I first switch the system on, it functions as expected, (power to the BK element is cut when the PID stops sending a signal). After it runs at 100% for a few minutes, the SSR ceases to function correctly, and once again does not cut power to the BK element until I power cycle the Line/Load side of the SSR. I confirmed via multimeter that the SSR is giving full 240V power to the element whether it's getting a control signal or not, once it's been running for a few minutes.

My thought is that the SSR is half bad, and that as it heats up, it develops a short that does not switch off Line power when the control signal stops. It didn't have to heat up much....the heatsink wasn't burning hot after only a few minutes at 100%, but apparently that's one Key.

Another key is that it seems to need current coursing though it to stick "closed". If I disconnect the element, the power to the outlet cycles as it should with the control signal.

It's weird, but as best as I can figure, the SSR is heat sensitive, and either heating up too much under 100% power and sticking, or just sticking when it should open once it's heated up, (still not sure how hot it's supposed to run "normally").

Again, no fan yet on the heatsink, but the heatsink IS open to room air. I'm going to get another 40A SSR and hope the new one works.

nfirmed via multimeter that the SSR is giving full 240V power to the element whether it's getting a control signal or not

Click to expand...

this is actually normal, and it should happen regardless of temperature. there will be a very low current, but full voltage, signal across the switched terminal even when the SSR is supposedly off. if you measure the current though, it should only be a few mA while the relay is off.

how much current is this 40A SSR being made to handle? using a 40A SSR to switch 40 full amps will produce much more than twice the amount of heat than switching only 20 amps would.

audger said:

this is actually normal, and it should happen regardless of temperature. there will be a very low current, but full voltage, signal across the switched terminal even when the SSR is supposedly off. if you measure the current though, it should only be a few mA while the relay is off.

how much current is this 40A SSR being made to handle? using a 40A SSR to switch 40 full amps will produce much more than twice the amount of heat than switching only 20 amps would.

Click to expand...

How could there possibly be low current but full voltage? There should be a large voltage drop across the SSR when it's "off", right? Otherwise what's stopping full current from flowing?

I do read 3-4 volts when the SSR is off before it heats up, and the SSR cycles between 4 volts and 240 volts during the first few minutes of operation. Once it heats up though, it sticks at 240 volts, regardless of whether it's "on" or "off".

It's a 5500W element being controlled by the SSR, at 240V, so 23 Amps.

Do you have a very good thermal / mechanical connection between your SSR and 'sink? If so, you might try a much smaller load to see if it continues to switch on /off if there's is less heat buildup. By smaller load, I mean say a couple of 120V, 60W lamps in series. This might help to determine if it's a bad SSR due to internal heating....

Junkster said:

Do you have a very good thermal / mechanical connection between your SSR and 'sink? If so, you might try a much smaller load to see if it continues to switch on /off if there's is less heat buildup. By smaller load, I mean say a couple of 120V, 60W lamps in series. This might help to determine if it's a bad SSR due to internal heating....

Click to expand...

Hrm...well, it's screwed on tight...and the heatsink gets extremely hot, so I'd assume it's a good contact.

I like your idea of testing it with light bulbs or something, but unfortunately i'd have to buy stuff to cobble that together, (or canabalize my house). I just bought a new 40A SSR on amazon for $10, should be here Thursday. Will give it a go and if it TOO fails, then dig in deeper .

If you have access to a clamp-on ammeter, you could easily check how much current your element is pulling - just for some additional info.....

That super wide pic makes this thread nearly unreadable for me. I probably would not have been much help anyway.

lschiavo said:

That super wide pic makes this thread nearly unreadable for me. I probably would not have been much help anyway.

Click to expand...

sorry lschaivo, I shrunk mine, but IPs is still big. Looks okay on my screen, but you probably are running a different resolution.

hmm 4 volts when off, then 240 when on? i think when i tested mine volts was 240 either way, its the amps that changed, but then again that was awhile ago, either way woulb be glad to hear when you install a new ssr to see if that was it.

shortyjacobs said:

How could there possibly be low current but full voltage? There should be a large voltage drop across the SSR when it's "off", right? Otherwise what's stopping full current from flowing?

Click to expand...

With no load, you might read anything. Tie a load to it and the voltage should drop to zero when off (nearly, minus leakage current time load resistance), just like you say.

I use Crydom SSRs. Yes, a lot more money than the chinese ones, but they work great. I've used 2 dual 25 amp Crydoms on my 5500W elements for 75 batches without a single problem. About 20 of those batches were without a heat sink (used metal box as heat sink - bad idea). If you don't mind spending a few extra bucks, search for Crydom and know you're going with an industry standard.

Further, take a good look at what heat sink you're using. If it doesn't spec °C / W, then it probably isn't a good choice. Mine is 0.7 °C / W, which means for every watt dissipated by the heat sink the temp will rise 0.7°C. How many watts will be disappated you ask? Guess about 1.5V * Amps. So if you are drawing 23A like I, it'll be 23*1.5 = 34W. So, my heat sink / SSR will rise about 33*0.7 = 24°C (above ambient). Sorry if that is too much info, just blabbing too much as usual. Cheers!

shortyjacobs said:

sorry lschaivo, I shrunk mine, but IPs is still big. Looks okay on my screen, but you probably are running a different resolution.

Click to expand...

No Problem. I guess I'm in a bad mood?

I dont see a problem with your diagram either. The latches make it fun though. As stated do not rely on AC voltage to test your SSR. Verify DC input to the SSR and check if it is reacting to the DC input. I think your problem is in the DC wiring. Do you have anything tied to the DC output of the PID? I tried that and had a polarity issue that messed with me for a while.

passedpawn said:

With no load, you might read anything. Tie a load to it and the voltage should drop to zero when off (nearly, minus leakage current time load resistance), just like you say.

I use Crydom SSRs. Yes, a lot more money than the chinese ones, but they work great. I've used 2 dual 25 amp Crydoms on my 5500W elements for 75 batches without a single problem. About 20 of those batches were without a heat sink (used metal box as heat sink - bad idea). If you don't mind spending a few extra bucks, search for Crydom and know you're going with an industry standard.

Further, take a good look at what heat sink you're using. If it doesn't spec °C / W, then it probably isn't a good choice. Mine is 0.7 °C / W, which means for every watt dissipated by the heat sink the temp will rise 0.7°C. How many watts will be disappated you ask? Guess about 1.5V * Amps. So if you are drawing 23A like I, it'll be 23*1.5 = 34W. So, my heat sink / SSR will rise about 33*0.7 = 24°C (above ambient). Sorry if that is too much info, just blabbing too much as usual. Cheers!

Click to expand...

Haha, my heat sink is just a chunk of aluminum from China...no documentation. Great info though! 24C above ambient would only have put my SSR at 44C, or 111F. It was definitely hotter than that...so either my heatsink has a much higher spec than yours, or the SSR is burning up more power than it should...(my bet's still on the latter). I didn't know that heatsinks were specced like that, nor had I heard of Crydom, so thanks for the info!

lschiavo said:

No Problem. I guess I'm in a bad mood?

I dont see a problem with your diagram either. The latches make it fun though. As stated do not rely on AC voltage to test your SSR. Verify DC input to the SSR and check if it is reacting to the DC input. I think your problem is in the DC wiring. Do you have anything tied to the DC output of the PID? I tried that and had a polarity issue that messed with me for a while.

Click to expand...

Yeah, that's one thing I didn't look at, (DC power to the SSR). I get spooked by fiddling around much inside the panel while it's live....but I'll give it a shot, (probably if the new SSR I ordered doesn't fix the problem, since I won't have a chance to fiddle with it until after the new one gets here...). The DC side of the panel is VERY straightfoward though. I have some 22 ga wire going directly from PID to SSR and back...no breaks/junctions/etc.

shortyjacobs said:

Haha, my heat sink is just a chunk of aluminum from China...no documentation. Great info though! 24C above ambient would only have put my SSR at 44C, or 111F. It was definitely hotter than that...so either my heatsink has a much higher spec than yours, or the SSR is burning up more power than it should...(my bet's still on the latter). I didn't know that heatsinks were specced like that, nor had I heard of Crydom, so thanks for the info!

Yeah, that's one thing I didn't look at, (DC power to the SSR). I get spooked by fiddling around much inside the panel while it's live....but I'll give it a shot, (probably if the new SSR I ordered doesn't fix the problem, since I won't have a chance to fiddle with it until after the new one gets here...). The DC side of the panel is VERY straightfoward though. I have some 22 ga wire going directly from PID to SSR and back...no breaks/junctions/etc.

Click to expand...

I didn't mean to imply that you messed up the DC. But if the signals are right, testing it may point to the bad component.

lschiavo said:

I didn't mean to imply that you messed up the DC. But if the signals are right, testing it may point to the bad component.

Click to expand...

Gotcha

shortyjacobs said:

How could there possibly be low current but full voltage? There should be a large voltage drop across the SSR when it's "off", right? Otherwise what's stopping full current from flowing?

Click to expand...

Solid state relay - Wikipedia, the free encyclopedia

specifically:
"For AC (bi-directional) operation two MOSFETs are arranged back to back with their source pins tied together. Their drain pins are connected to either side of the output. The substrate diodes are alternately reverse biased in order to block current when the relay is off."

MOSFET - Wikipedia, the free encyclopedia

SSR's block current, not voltage

I do read 3-4 volts when the SSR is off before it heats up, and the SSR cycles between 4 volts and 240 volts during the first few minutes of operation. Once it heats up though, it sticks at 240 volts, regardless of whether it's "on" or "off".

Click to expand...

that makes me believe that the SSR is faulty. semiconductors in general conduct more electricity the hotter they get (yes, the hotter they get the more they conduct, and the more they conduct the hotter they get). however, a mosfet should work at room temperature. if it doesnt allow electricity to flow untill it heats up considerably above room temperature, that means it has been damaged and is not operating correctly. i would just replace it with another one.

audger said:

Solid state relay - Wikipedia, the free encyclopedia

specifically:
"For AC (bi-directional) operation two MOSFETs are arranged back to back with their source pins tied together. Their drain pins are connected to either side of the output. The substrate diodes are alternately reverse biased in order to block current when the relay is off."

MOSFET - Wikipedia, the free encyclopedia

SSR's block current, not voltage




that makes me believe that the SSR is faulty. semiconductors in general conduct more electricity the hotter they get (yes, the hotter they get the more they conduct, and the more they conduct the hotter they get). however, a mosfet should work at room temperature. if it doesnt allow electricity to flow untill it heats up considerably above room temperature, that means it has been damaged and is not operating correctly. i would just replace it with another one.

Click to expand...

I don't think the body diodes are relevant. Those body diodes are an inherent alternate path that have nothing to do with the operation of the FETs (they do protect the fets from reverse voltage breakdown, but then make AC operation impossible unless they are arranged back-to-back, like described). Maybe I don't understand what you're getting at, tho, that happens sometimes (mostly later at night ). But yea, SSRs block current. A FET is like a big variable resistor that gets pinched closed as the gate voltage drops.

Hrm, definitely was a DOA SSR. I bought a new one from Lightobject on Amazon. (same Fotek 40A SSR that I had bought from China), and installed it today. Works like a charm! Runs cooler, actually controls things...I'm autotuning right now.

Thanks all!

I suspect that most of you complaining about cheap Chinese SSRs failing don't have a large enough heatsink or are not using thermal compound between the SSR and heatsink.

A SSR will dissipate about 1 watt for every amp its controlling. So, 23 amps for a 5500 watt element = about 23 watts of power and every bit of it is heat. The heat is generated because the heart of a SSR is a semiconductor and semiconductors never turn all the way on.

And as far as heat sinking goes, if you use a good heat sink with 1" fins you will need a heat sink that's about 2" X 3" for 10 amps. This is assuming the heatsink is mounted on the outside of your box with the fins oriented so that air moves up between them. Mont the SSR inside a box and there is no chance for airflow without a strong fan. Pull 23 amps and your heatsink needs to be at least 2.5X that size or 5" X 3" to cool your SSR.

When the commercial guys push power through a SSR they usually go with one of these heatsinks. It will be mounted so that air drafts through like a chimney or they will blow ait through the heatsink.
Heat Sink, Aluminum - Relay Accessories - Relays - 6CXA8 : Grainger Industrial Supply

The thermal compound is the same type of thermal "grease" used to prevent PC CPUs from overheating. It fills in the microscopic spaces between the SSR & heat sink and without it your SSR will run hot, even on a proper heatsink. Here is some thermal compound at a good price.
BestByte.net - Discount Computer Hardware & Reliable Service: Thermal Compounds & Adhesives

thargrav said:

I suspect that most of you complaining about cheap Chinese SSRs failing don't have a large enough heatsink or are not using thermal compound between the SSR and heatsink.

A SSR will dissipate about 1 watt for every amp its controlling. So, 23 amps for a 5500 watt element = about 23 watts of power and every bit of it is heat. The heat is generated because the heart of a SSR is a semiconductor and semiconductors never turn all the way on.

And as far as heat sinking goes, if you use a good heat sink with 1" fins you will need a heat sink that's about 2" X 3" for 10 amps. This is assuming the heatsink is mounted on the outside of your box with the fins oriented so that air moves up between them. Mont the SSR inside a box and there is no chance for airflow without a strong fan. Pull 23 amps and your heatsink needs to be at least 2.5X that size or 5" X 3" to cool your SSR.

When the commercial guys push power through a SSR they usually go with one of these heatsinks. It will be mounted so that air drafts through like a chimney or they will blow ait through the heatsink.
Heat Sink, Aluminum - Relay Accessories - Relays - 6CXA8 : Grainger Industrial Supply

The thermal compound is the same type of thermal "grease" used to prevent PC CPUs from overheating. It fills in the microscopic spaces between the SSR & heat sink and without it your SSR will run hot, even on a proper heatsink. Here is some thermal compound at a good price.
BestByte.net - Discount Computer Hardware & Reliable Service: Thermal Compounds & Adhesives

Click to expand...

I don't know if mine failed BECAUSE it got hot, or it got hot BECAUSE it failed...

I know for sure that it had semi-failed, (worked fine when cool, but it got REALLY hot, and then latched open). I know I was getting good conductivity to my heatsink, because the heatsink was VERY hot.

In any case, I have the new SSR, and I mounted a fan to blow across my (much smaller than you show), heatsink. The heatsink barely gets warmer than room temp now, and the SSR functions as expected.

I think it's a definite possibility that I fried the SSR by not using a fan on the wimpy heatsink, and it's also a definite possibility that the SSR was DOA, and it's failure mode was to heat up a helluva lot.

Either way, I'm happy now!

thargrav said:

I suspect that most of you complaining about cheap Chinese SSRs failing don't have a large enough heatsink or are not using thermal compound between the SSR and heatsink.

A SSR will dissipate about 1 watt for every amp its controlling. So, 23 amps for a 5500 watt element = about 23 watts of power and every bit of it is heat. The heat is generated because the heart of a SSR is a semiconductor and semiconductors never turn all the way on.

And as far as heat sinking goes, if you use a good heat sink with 1" fins you will need a heat sink that's about 2" X 3" for 10 amps. This is assuming the heatsink is mounted on the outside of your box with the fins oriented so that air moves up between them. Mont the SSR inside a box and there is no chance for airflow without a strong fan. Pull 23 amps and your heatsink needs to be at least 2.5X that size or 5" X 3" to cool your SSR.

When the commercial guys push power through a SSR they usually go with one of these heatsinks. It will be mounted so that air drafts through like a chimney or they will blow ait through the heatsink.
Heat Sink, Aluminum - Relay Accessories - Relays - 6CXA8 : Grainger Industrial Supply

The thermal compound is the same type of thermal "grease" used to prevent PC CPUs from overheating. It fills in the microscopic spaces between the SSR & heat sink and without it your SSR will run hot, even on a proper heatsink. Here is some thermal compound at a good price.
BestByte.net - Discount Computer Hardware & Reliable Service: Thermal Compounds & Adhesives

Click to expand...

Good advice all around. I used Arctic Silver 5, which is a premium (IOW costs more) paste for mating processors and heat sinks. The overclockers know what I'm talking about But you could get comparable (probably for free) from any computer store - every processor ships with a bit of it I think.

shortyjacobs said:

I don't know if mine failed BECAUSE it got hot, or it got hot BECAUSE it failed...

I know for sure that it had semi-failed, (worked fine when cool, but it got REALLY hot, and then latched open). I know I was getting good conductivity to my heatsink, because the heatsink was VERY hot.

In any case, I have the new SSR, and I mounted a fan to blow across my (much smaller than you show), heatsink. The heatsink barely gets warmer than room temp now, and the SSR functions as expected.

I think it's a definite possibility that I fried the SSR by not using a fan on the wimpy heatsink, and it's also a definite possibility that the SSR was DOA, and it's failure mode was to heat up a helluva lot.

Either way, I'm happy now!

Click to expand...

Most likely you fried the SSR. And it would run hot because the part was damaged and dissipating more wattage.

And BTW, the heat sink specs I was quoting is for free air & no fan. As you already know you can go smaller when using a fan to blow air across the heat sink. How small, I don't know. CPU coolers do this all the time.

OK, a few things.... likely, your PID is not providing a sinking current to shut off your ssr. You need to jump a 10k resistor across the positive and negative output from your pid. This will pull your input to the ssr LOW when it is off. That being said, the ssrs still can stick when they get too hot, so you need to up your cooling capacity. put a little computer fan on the bottom of your box

Lol, thanks champ. This post was 9 years ago, it was just a failed SSR. Put in a new one, put some cooling on it to stop it from cooking itself, and it’s been fine for 9 years now.

kazikame said:

OK, a few things.... likely, your PID is not providing a sinking current to shut off your ssr. You need to jump a 10k resistor across the positive and negative output from your pid. This will pull your input to the ssr LOW when it is off. That being said, the ssrs still can stick when they get too hot, so you need to up your cooling capacity. put a little computer fan on the bottom of your box

Click to expand...

PID's designed to drive SSR's do NOT require any additional components (like resisters) between the PID and SSR (assuming you have the correct SSR type.)

Brew on