SSR vs Mech Relay/Contactor for boil?

[kenc_zymurgy]

Here's my 3rd, and last question on the subject (until I do more research!) on my Electric Brewing plans:

I assumed I'd be using SSRS to control a 1650 ~2250W element, like most brewers. I have some experience with SSRs on other projects, and I know that heat-sinking is critical when you get up to tens of amps. And that many of the cheap ebay/ali-express units are counterfeit (seems mostly authentic units, but a 10 Amp unit is relabeled as a 40 Amp?). So for > 20 Amp SSR, I'd want to buy from a major US distributor, and those aren't cheap. Plus heat sinks.

Then it occurred to me that the relay/contactor for a home A/C unit ( 20 ~ 40 A @ 240V) cycles a few times an hour for several months for years. I did some research, and for < $15 these are rated for 100,000 cycles. And they are dual pole, so could be used on 240 V as well, and I could parallel the contacts for 120V, which should further prolong life.

https://www.arrow.com/en/products/t92p7d22-12/te-connectivity ( 3 minimum, < $40 with shipping to me)

Some math - per brewing session, 1 cycle to heat the strike water, 1 cycle to heat the sparge water, and 1 cycle to get to a boil. Not sure how many cycles to control a 60 minute boil, but I'd guess that a 10 second period ( X seconds ON, 10-X seconds OFF) would be more than fast enough to maintain a consistent boil. So that is 6 cycles per minute, and 360 + 3 cycles per brew session. round up to 400, is100,000/400 = 250 brews. That's beyond my lifetime (I'm a geezer).

So I'm thinking the mech relay has advantages for this sort of set-up (not for a RIMS/HERMs though, I think those need faster switching and/or true PWM?). Other advantages are fully OFF (no leakage current from SSR). And typical failure is OFF rather than ON for an SSR?

I'm competent enough with Arduino programming, that I could easily set up a $4 Pro-Micro for 0~100% duty cycle on one pot, and the time period on another. I'd include some safety start ups, and a safety time out (I have his nightmare that I get locked out of the house with my pots on!).

About the only con I can think of is hearing the relays clunk on/off every 10 seconds for an hour, but I usually wander in/out during the boil, so I don't think that will bug me too much.

Anything else I'm missing?

 

[Bobby_M]

Mechanical contactors may be rated for 100k cycles, but not at 20 per minute. You're solving a problem that doesn't exist. Here's a legit 40amp SSR for $17.

https://www.auberins.com/index.php?main_page=product_info&cPath=2_30&products_id=30

 

[kenc_zymurgy]

Mechanical contactors may be rated for 100k cycles, but not at 20 per minute. You're solving a problem that doesn't exist. Here's a legit 40amp SSR for $17.

https://www.auberins.com/index.php?main_page=product_info&cPath=2_30&products_id=30

OK, that's a much better price for a legit SSR than I was finding at the big distributors. But I'd still need to add a $10 heat sink, and assure I've got a good thermal transfer with the compound. That's probably larger than just the relay.

I wouldn't think the rate of cycling would affect the cycle life of a relay. Isn't 100,000 still 100,000, if it's 20/minute (though I also assumed a "cycle" meant an ON/OFF cycle?) or 2 per hour?

I've actually bought some of the FOTEK SSR on ali-express for another project. Super cheap, $5 for a (labelled) 40 Amp unit. I doubt it is actually 40 A capable, but I was using it in a 1 Amp circuit (none of the SSRs above a rated 10A need any heat sink at all at below ~2 Amps). Though I didn't feel good about supporting a likely fraud.

BTW, I did some research on these cheap SSRs a while back. I looked up specs on the Triacs that are used, and it seems that regardless of Amp rating, they have approximately the same Voltage drop at X amps, so it's hard to tell if a 40 A label is using a 10 A rated Triac, without actually doing a destructive test.

 

[doug293cz]

A 10 second cycle time at boil will give you noticeable pulsing in your boil. At that cycle time it will stop and restart. Brewers using SSR's at a two second cycle time see noticeable pulsing.

Brew on

 

[sibelman]

A 10 second cycle time at boil will give you noticeable pulsing in your boil. At that cycle time it will stop and restart. Brewers using SSR's at a two second cycle time see noticeable pulsing.

Brew on

Agreed. I'm wondering whether brief pulsing is a problem?

 

[Bobby_M]

You understand that boil control does mean turning the element on and off up to once per second right? A good boil control at 50% power is 500ms on and 500ms off. If you're talking about doing 5 gallon batches with a 2250w element, you don't need to worry about that because the boil will be fine at 100% output (no cycling).

 

[sibelman]

I'm wondering whether I ought to be changing out the timing capacitor in my SSR controller to reduce the cycle duration, which is noticeable. btw, my element is 5500W.

But mostly I'm still wondering: what are the ill effects of a bit of "pulsing" on the process or the product?

 

[kenc_zymurgy]

OK, I wasn't aware that the switching needed to be that fast. I was thinking that volume of liquid would have enough thermal mass to even out over a 10 second period, but I guess not. Switching a big honking relay once a second would probably drive me nuts (I do split boils, so there'd be two of them), and the lifetime might be short enough to just not make sense (but still ~ 55 brew sessions, about 4~5 years for me, so not a crazy bad life expectancy).

So I'll do some more research on the SSRs and heat sinks. I'm thinking I'll add one big toggle switch to select SSR or straight line - no need to run that SSR hot when I'm heating strike/sparge water, or bringing the wort to a boil. Switch the SSR in only during the boil. Plus a main ON/OFF switch that I'd put a shield over to make it easy to turn OFF, but would take a little effort to turn ON for a bit of added safety.

 

[kenc_zymurgy]

I'm wondering whether I ought to be changing out the timing capacitor in my SSR controller to reduce the cycle duration, which is noticeable. btw, my element is 5500W.

But mostly I'm still wondering: what are the ill effects of a bit of "pulsing" on the process or the product?

I was just catching up on some of the threads about slow boils and low evaporation rates (5%), so I wonder. As long as you hit a target boil-off, would a pulsing boil even matter? Those threads seem to be saying that even a simmer gets the job done.

 

[gypsylore]

Industrial automation guy here.

Mechanical relays and contactors are not designed for rapid cycles. During operation there is always a small voltage jump as the internal contacts near each other while closing. This builds up heat and carbon over time (how fast depends on voltage, amperage and cycles/minute). 2-3 operations per minute is pretty rapid for switching relatively high currents. More then that will burn out your relay's contacts over time.

I'm a very new brewer (just 3 batches under my belt) so I can't say anything about a pulsing boil effecting a brew 'cause I haven't a clue if that'll effect anything or not. I can say, however, that I would advise using a SSR over a mechanical solution. They are specifically designed to do rapid switching after all.

 

[kenc_zymurgy]

Industrial automation guy here.

Mechanical relays and contactors are not designed for rapid cycles. During operation there is always a small voltage jump as the internal contacts near each other while closing. This builds up heat and carbon over time (how fast depends on voltage, amperage and cycles/minute). 2-3 operations per minute is pretty rapid for switching relatively high currents. More then that will burn out your relay's contacts over time.

I'm a very new brewer (just 3 batches under my belt) so I can't say anything about a pulsing boil effecting a brew 'cause I haven't a clue if that'll effect anything or not. I can say, however, that I would advise using a SSR over a mechanical solution. They are specifically designed to do rapid switching after all.

I think I'm convinced. I guess I was thinking the SSR requirement was from the RIMS/HERMS cases where fast control is probably needed, and that *maybe* a relay could work for the boil.

I didn't think about the heat build up on the contacts from each high current switch operation, I'll take a closer look at the datasheets, I don't think that's mentioned, but then again, they probably figure with a 100,000 life, relays would not be used in that way. But since my use would be so intermittent (one or two hours a dozen times a year), I thought I might get away with it. The simplicity and total disconnect are attractive features.

I guess I was also thinking about those electric cooktop "Infinity" controls (mechanical/thermal switch - like the old car turn signals). I haven't cooked on an electric cooktop in decades, but I seem to recall being able to maintain a steady light boil/simmer. I guess the difference is those big round elements (plus the directly heated pot) had a lot of thermal mass?

I also did some more digging into specs on the SSR - a 45A rated one can run 10A in free air w/o a heat sink (not that I'd recc it, that's right at the limit), but a 1650W element on 120V is only 13.75A, and if I only switch in the SSR for the boil, I'll be below 100% duty cycle, so I guess the heatsink performance isn't as hard to achieve as I was thinking.

https://documents.opto22.com/0859_Solid_State_Relays_data_sheet.pdf
Thanks to all for the feedback.

 

[TLaffey]

I think I'm convinced. I guess I was thinking the SSR requirement was from the RIMS/HERMS cases where fast control is probably needed, and that *maybe* a relay could work for the boil.

I didn't think about the heat build up on the contacts from each high current switch operation, I'll take a closer look at the datasheets, I don't think that's mentioned, but then again, they probably figure with a 100,000 life, relays would not be used in that way. But since my use would be so intermittent (one or two hours a dozen times a year), I thought I might get away with it. The simplicity and total disconnect are attractive features.

I guess I was also thinking about those electric cooktop "Infinity" controls (mechanical/thermal switch - like the old car turn signals). I haven't cooked on an electric cooktop in decades, but I seem to recall being able to maintain a steady light boil/simmer. I guess the difference is those big round elements (plus the directly heated pot) had a lot of thermal mass?

I also did some more digging into specs on the SSR - a 45A rated one can run 10A in free air w/o a heat sink (not that I'd recc it, that's right at the limit), but a 1650W element on 120V is only 13.75A, and if I only switch in the SSR for the boil, I'll be below 100% duty cycle, so I guess the heatsink performance isn't as hard to achieve as I was thinking.

https://documents.opto22.com/0859_Solid_State_Relays_data_sheet.pdf
Thanks to all for the feedback.

One more thing in favor of an SSR: A relay coil is inductive and itself requires a certain pull-in current. Assuming it works at all, the relay will likely kill your Arduino in short order if you don't put a relay driver between your Arduino and the relay. This would add to your cost and reduces reliability.

 

[kenc_zymurgy]

One more thing in favor of an SSR: A relay coil is inductive and itself requires a certain pull-in current. Assuming it works at all, the relay will likely kill your Arduino in short order if you don't put a relay driver between your Arduino and the relay. This would add to your cost and reduces reliability.

They are 12V coils, so an Arduino would definitely need a transistor and you always need a clamp diode across a relay coil. That's pretty minor, and highly reliable, and I've got plenty of those parts in the bin, but it is one more thing. Oh, and then I'd also need a 7805 to drop the 12V to 5V for the Ardunio , or a separate wall-wart. It all adds up a little bit.

 

[Bobby_M]

Do we need to scrape together a collection basket to buy you a $10 heat sink? I've never seen so much agony over saving money for what is such an industry standard solution that we know works.

 

[moorejl57]

I bought two used Crydom 50 amp SSR pulled from industrial equipment on ebay for $18 that run at room temp for a 10 amp load on my DIY air popper coffee roaster. They are designed to be mounted to a panel, but even at a 20 amp load I doubt they would need a heat sink. The cheap fake ones are a fire hazzard and I wouldn't use them with or without a heat sink.

 

[kenc_zymurgy]

Do we need to scrape together a collection basket to buy you a $10 heat sink? I've never seen so much agony over saving money for what is such an industry standard solution that we know works.

No, I'm just doing my usual over-thinking. And I do like to (in my mind) challenge the status quo, not so much that I think the established conventions are wrong, but it helps to fix in my mind *why* those are the conventions. And I probably got a bit scared from the reports of these fake SSRs.

And heck, if a hot plate would work (sure seems marginal to me), that means no drilling into a pot and extra fittings. The simplicity is attractive, the downsides almost certainly rule it out. And now I see the hot plate being used is commercial grade, and big $.

I'm convinced. So with that in mind - where do I get known-good SSRs at a good price? I think someone linked earlier to some pretty cheap heat sinks on ebay like these ($4?)

https://www.ebay.com/itm/3-Pcs-Alum...632752&hash=item5b736d6e15:g:K~8AAOSwLB5epwes
And this place has been mentioned for SSR:

https://www.auberins.com/index.php?main_page=product_info&products_id=30