Relays vs. SSR's

Guys, (and gals too) I'm looking for some feedback on an electric rig I'm planning on setting up. It is inspired by Kal's system at theelectricbrewery, but differs in one important respect: I plan on using lighting relays instead of SSR's. There are a couple important reasons I'm going in that direction, which it isn't really important to me to discuss at this point. Suffice it to say I plan on using lighting relays to control my 240v heating elements, relays which I am very certain can handle the switching cycle of my brewing system for at least 5 years.

In order to use relays in a system with PIDs I would use a PID with a relay output control. The one I'm looking at is Auber SYL-2342, which is the exact model Kal's system uses with the exception that it has a relay contact output instead of a 12vdc SSR output. The hitch here is that the timing cycle needs to be set at a minimum of 20 seconds since relays are not designed to switch as quickly or as often as SSRs. (that's the whole reason he designed his system with them). This is an important issue to point out: the timing cycle in an SSR PID set up is 2 seconds by default. This means if the PID requires the system to fire the heating element 10% of the time it will turn it on for .2 seconds and off for 1.8 seconds. If the timing cycle were set at 100 seconds then under the same scenario it would turn on for 10 seconds and off for 90 seconds. In my setup the least I could set it for would be 20seconds. This would mean my relay would cycle on/off a maximum of 3 times per minute. And let me repeat: I have chosen relays which can handle this switching cycle.

My dilemma is that I will lose precision of temperature control under my planned scenario. I just don't know how much precision I will lose. 2-3 degrees I could tolerate. 10-12 degrees I could not, for reasons I would think are fairly clear. So I'm wondering if anyone else out there has tried this before and could give any feedback. OR if anyone with an existing PID SSR setup would be willing to set up a simple test: change your default timing cycle from 2 seconds to 20, 40 or 60 seconds and see if you can hold 5 gallons of water at 150 degrees for 60 minutes. If not, how much temperature drift did you get? This will give me a very good idea of the performance I can expect from my planned system.

Please note: I am not asking anyone to mess with their default PID settings unless you are confident you understand what you are doing and know how to make this change and then return it back to your previous setting.

Also, please consider the fact that I am not currently looking for advice on whether or not to use relays in place of SSRs. "Just use SSRs" isn't really the feedback I'm looking for. I have good reasons to be exploring the option of building my system this way, but I'm hoping to get a little help/feedback on the possible temperature drift before I start putting money into it.

Thanks in advance for any help you can give!

If you are set on not using ssr's then you have already decided to trade the simple feedback fast response system that ssr's give you with a slower responding control loop. You might be able to use some feed forwarding controls to keep the temperature closer to your set points, but you have to really know what you're doing and it complicates things a bit. Everyone's setup is a little different, so you'll just have to tune your loops until you find some level of control you are happy with.

I dont think your accuracy is going to suffer greatly. With the fairly large volumes of water and wort we are working with 1-2 degrees in an hour in an insulated cooler style MLT is normal. The largest issue you might encounter would be over shoot and waiting for the temp to settle back to mash temp before proceeding. If you are stirring you pots your overshoot should be minimal if your PID's are even close. Over time any stainless or Alli pot is gonna lose some temp, but if using a insulated Mash tun that will be completely minimized to a much more tolerable level quite easilly. Just some out loud thinkin'
Wheelchair Bob

geoffey said:

Also, please consider the fact that I am not currently looking for advice on whether or not to use relays in place of SSRs. "Just use SSRs" isn't really the feedback I'm looking for.

Click to expand...

Yup. That would have been me. That's the feedback I originally gave Geoff when he emailed me... he wants to stick with the relays so since I don't have any experience with them I suggested he post here for ideas.

Good luck!

Kal

I would really like to know the reason behind this, not to say one is better or you NEED to do this or that, just for my own sanity, when you get time please post your reasoning behind using relays instead of ssr's.

Rbeckett said:

I dont think your accuracy is going to suffer greatly. With the fairly large volumes of water and wort we are working with 1-2 degrees in an hour in an insulated cooler style MLT is normal. The largest issue you might encounter would be over shoot and waiting for the temp to settle back to mash temp before proceeding. If you are stirring you pots your overshoot should be minimal if your PID's are even close. Over time any stainless or Alli pot is gonna lose some temp, but if using a insulated Mash tun that will be completely minimized to a much more tolerable level quite easilly. Just some out loud thinkin'
Wheelchair Bob

Click to expand...

I'd tend to agree. Really there are two major differences between SSRs and Relays that are relevant.

1) SSRs can cycle more often than Relays
2) SSRs don't actually shut off, where Relays do.

So... the way I see it if your ok with a little loss of temperature control a Relay would work fine and may actually provide some theoretical safety addition. As far as the accuracy/precision factor goes... given the volumes I also would tend to agree that the relay would be just fine. When I mash in my keggle I only turn my gas burner on once ever 30 minutes or so. A relay could easily handle that. Plus the worst thing that happens is you decide later you want the SSR you could always swap parts later. Or you could get a SYL-2362 and then use a separate relay or SSR (the SYL-2362 has both outputs).

Efaden, thanks for the heads up on that Auber model. I didn't know they made one with both output types. That is a great solution for me. One of my fears was dropping $100 on the PIDs for relay control and having the system not work to my satisfaction and having to repurchase PIDs for SSRs.

geoffey said:

Efaden, thanks for the heads up on that Auber model. I didn't know they made one with both output types. That is a great solution for me. One of my fears was dropping $100 on the PIDs for relay control and having the system not work to my satisfaction and having to repurchase PIDs for SSRs.

Click to expand...

No problem. Seems like a good way to go...

I use a relay to switch the elements in my HLT via PID. It has worked fine for a few years now. An auto tune and a little tweaking of the delay time and I don't get any overshoot. The only real difference is the noise from the contactor.

Thanks for the great feedback so far everyone. I'm fairly encouraged I'm going to be satisfied with my results. May take a few months to get there, but I'm excited to get underway soon.

To satisfy curiosity, there are a few reasons I'm interested in using relays in place of SSRs. First of all, I work in the commercial lighting control industry and work with these types of relays on a daily basis. I'm extremely familiar with them and very comfortable with wiring them. I have put together a plan using parts and pieces from my industry which I can source at little to no cost, including the touchscreen front end which I can provide the custom programming for myself. Incorporating SSRs into this system wouldn't necessarily be difficult. But it's out of my comfort zone and a bit of another unknown to me.

My control panel will be simply the touchscreen and two PIDs. I'm planning on having everything else installed above the ceiling, etc. I'm not that interested in portability.

More on this later if anyone is interested. Thanks for the help so far. I'd love to hear more from others!

geoffey said:

Suffice it to say I plan on using lighting relays to control my 240v heating elements, relays which I am very certain can handle the switching cycle of my brewing system for at least 5 years.

Click to expand...

Geoffey,

What specifications / brand / model number could I search so I could price a relay like that?

lschiavo said:

I use a relay to switch the elements in my HLT via PID. It has worked fine for a few years now. An auto tune and a little tweaking of the delay time and I don't get any overshoot. The only real difference is the noise from the contactor.

Click to expand...

Ischiavo,

I am also very interested to know what relays you're using. It sounds like you have something that works, and might work for my situation...

I kinda made a mistake. I've been gathering parts for a three keggle, single tier HERMS system, and want to use a 3500W 240V element to control the temp in my HLT. Wanting to keep things as simple as possible, I bought a couple of Johnson A419ABC-1C controllers and planned to use one to control the heating element in the HLT, and keep one for future projects. I read on Johnson's website that the controller has a built in relay capable of switching a 16 amp load.

Today, I was looking back at the instructions for the controller and realized that it's capable of a 16 amp load at 120V but only 8 amps at 240V. A 3500W element at 240V will draw almost 15 amps. So I am considering one of two options: 1) Attempt to return the Johnson controllers and buy a PID controller and use a SSR. I don't really want to do this - I got a good price for the A419s ($119.07, delivered, but I had to buy two of them). Plus, I want to use one for temperature control of a fermentation chamber at some point. Option # 2) Use the Johnson controller to control a contactor or lighting relay, as is being discussed in this thread.

I'm planning on constantly recirculating the wort through a HEX coil in the HLT using one pump, and constantly recirculating water from the bottom of the HLT back to the top of the HLT, to keep it stirred up and maintain a uniform temperature in my HLT. Would a lighting relay work for me and give me a relatively stable mash temperature? Do I need to look for a specific type / duty cycle / IEC Utilization Category?

Or am I way off base with the Johnson controller and just need to back up and punt (go with the PID idea)?

The other thing to consider is that moving from solid state relays to mechanical relays you dont have the heat generation problem due to the losses within the SSR. I think its a good idea, and dont think you will lose much accuracy at all.

AggieBrewer01 said:

Geoffey,

What specifications / brand / model number could I search so I could price a relay like that?



Ischiavo,

I am also very interested to know what relays you're using. It sounds like you have something that works, and might work for my situation...

I kinda made a mistake. I've been gathering parts for a three keggle, single tier HERMS system, and want to use a 3500W 240V element to control the temp in my HLT. Wanting to keep things as simple as possible, I bought a couple of Johnson A419ABC-1C controllers and planned to use one to control the heating element in the HLT, and keep one for future projects. I read on Johnson's website that the controller has a built in relay capable of switching a 16 amp load.

Today, I was looking back at the instructions for the controller and realized that it's capable of a 16 amp load at 120V but only 8 amps at 240V. A 3500W element at 240V will draw almost 15 amps. So I am considering one of two options: 1) Attempt to return the Johnson controllers and buy a PID controller and use a SSR. I don't really want to do this - I got a good price for the A419s ($119.07, delivered, but I had to buy two of them). Plus, I want to use one for temperature control of a fermentation chamber at some point. Option # 2) Use the Johnson controller to control a contactor or lighting relay, as is being discussed in this thread.

I'm planning on constantly recirculating the wort through a HEX coil in the HLT using one pump, and constantly recirculating water from the bottom of the HLT back to the top of the HLT, to keep it stirred up and maintain a uniform temperature in my HLT. Would a lighting relay work for me and give me a relatively stable mash temperature? Do I need to look for a specific type / duty cycle / IEC Utilization Category?

Or am I way off base with the Johnson controller and just need to back up and punt (go with the PID idea)?

Click to expand...

I am using relays from Wattstopper, part number BZ-150. You could probably still use your Johnson controller to run these, you'd need to use two of them (one for each 120v circuit of your 240v heating element). The BZ-150's are controlled via low voltage input. You would take the 24vdc lead from the BZ-150 and run that through your Johnson Controller contact, and run a return wire back to the BZ-150 and connect it to the "control input" wire. To wire two BZ-150's you'd just wire them in parallel on the low voltage side. (I am sure I could come up with a wiring diagram for you, it's really quite simple.)

I chose this model because it has a "hold off" input which can be used in our application as a safety measure. I created a virtual button on my touchscreen which has to be pressed before the system will allow the elements to fire.

My build is almost complete and I will be posting pictures of my progress here in the next couple weeks.

geoffey said:

I am using relays from Wattstopper, part number BZ-150. You could probably still use your Johnson controller to run these, you'd need to use two of them (one for each 120v circuit of your 240v heating element).

Click to expand...

Be careful here - These Wattstopper relays are not necessarily a bad idea, but there's a couple big asterisks to keep in mind...
They are rated for 20 amp loads on 14ga wire. Note 14ga wiring in the home is rated for 15 amps. Granted this is stranded, and as long as your lead length is short it should be OK. BUT, if you have 10' of 14ga cable running 240v to the element, it'll probably get hot. IF you use larger wire to the element, the 14ga at the relay effectively becomes a "fuse" as the weakest link in the circuit. The 14ga wire can heat up. As heat increases, so does resistance, causing more heat. It's a vicious circle.
Also, there are NOT (2) 120v circuits in a 240v element. There's ONE 240v circuit. You can switch only one leg of the 240v circuit, and the element will be 100% off. People switch both legs for safety (Touching the unswitched leg to ground will obviously be a very bad idea, even though the element is not drawing current and creating heat)
I mention this to emphasize one point: If you use 20 amp rated relays, and utilize 2 of them thinking you're going to have a 40 amp switching capacity, you're going to risk a fire if you drive 5500 watt elements. I'm sure there's some fudge factor with the 20 amp rating, and a 5500 watt element will pull 23 amps. But that's still 3 amps OVER the rated load of the relay (Or SSR).
Now, if your household electrical system is only delivering 215 volts under load, you're going to be pulling 25.5 amps through your 20 amp rated device. This is where it gets risky.

On the topic of relays vs SSR's, the only thing I like to mention is MTBF on the relays. Givin the low switching rate, everything should be fine. But, if you run into the higher switching rates you can get close to the rated lifetime cycle count on the relays in a hurry (A year or 2 maybe?)
If a relay is rated for 1 million cycles, and you have a 10Hz cycle rate, that's 27.7 hours of use before you actually hit the MTBF. Just something to keep in mind - The math adds up quick!

Sweetsounds- appreciate the input here. I'm not 100% following you though. You say it's not 2 120v circuits, and that's where you lose me. I am feeding the elements with a double pole breaker, each pole is a 120v circuit. Each relay is controlling only ONE of the two poles. Are you saying that only one of the two relays is a power handling device in this setup? I am of the understanding that each relay is handling about 13amps each. (I'm also using 10 gauge wire).

I've already done some testing of this system, albeit not of a very thorough nature, (more testing to come) but if what you are saying is true then I'd have easily fried one of those relays already, and I haven't.

geoffey said:

Sweetsounds- appreciate the input here. I'm not 100% following you though. You say it's not 2 120v circuits, and that's where you lose me. I am feeding the elements with a double pole breaker, each pole is a 120v circuit. Each relay is controlling only ONE of the two poles. Are you saying that only one of the two relays is a power handling device in this setup? I am of the understanding that each relay is handling about 13amps each. (I'm also using 10 gauge wire).

I've already done some testing of this system, albeit not of a very thorough nature, (more testing to come) but if what you are saying is true then I'd have easily fried one of those relays already, and I haven't.

Click to expand...

No! (Sorry, but this is important)
240 volts is NOT 2 120v lines. You have 4 wires in your electrical panel. A ground, 2 Hot's and a Neutral. What we are talking about here is electrical potential measured between any 2 lines.
Hot to Neutral = 120 volts
Hot to Hot = 240 volts
Hot to Ground = Sparks, fire, death

Ground is NEVER part of an electrical circuit - It's there to give the current somewhere to go besides through your body, and cause the breaker to trip
Neutral IS part of an electrical circuit. But ONLY 120 volt circuits, because the electrical potential (Voltage) between hot and neutral is 120 volts.

Your 240 volt element connects to both hots and is 240 volts. 240 volts alone, and not 120 volts (Because the neutral is not involved). It's not (2) 120 volt lines, just as your vacuum cleaner, when plugged into your 120 volt receptacle is NOT 2 60 volt lines. Follow?

A 120v element would use one hot and the neutral - Because the electrical potential between hot and neutral is 120 volts! (This is how every standard outlet in your house is wired)

This is voltage, and it's critical you understand how that works.

Now, for load: (Even more critical, because voltage doesn't start fires, Current starts fires)
Current creates heat - This is, after-all, exactly what we're all using it for.

Back to the vacuum cleaner analogy:
If your vacuum cleaner draws 5 amps at 120 volts, it's not drawing2.5 amps per wire. It's drawing 5 amps in one wire, and out the other (Over-simplification, but accurate)
Same for the element in your rig - it's drawing 23 amps in one wire and out the other. Kinda like a battery, except instead of electricity flowing from positive to negative (Direct current), it flows back and forth, switching directions 60 times per second (Alternating Current) (This is where 60 Hertz comes from)

So - At the end of the day, you are going to pull 23 amps through your 20 amp relays. (Assuming 1 5500 watt element) It doesn't matter if you use one or two or ten relays, It's still 23 amps at 240 volts.

One saving grace about AC is duty cycle. Since the current switches directions 60 times per second, it's actually 0 volts 60 times per second, too. Passing through 0 essentially gives the circuit a "break" and it can cool off. This is why a switch may be rated for 10 amps AC, but only 4 amps DC - direct current never goes to 0 volts, and things heat up faster because the contacts never get time to "rest".
It will heat up. When it gets hot, resistance increases. This creates more heat, causing higher resistance. Once you reach a critical mass, you get Chernobyl.

Again, I'm not saying they are bad relays. I'm just saying you need to be aware of the limitations of your device, and how close you may be running to those limits.

Edit: Sorry - You're using mechanical relays, not SSR's. Edited for accuracy
Edit-Edit: I'm going to change my "Relays! Nifty!" view of your build - Just buy the SSR's man - They're cheaper than melting your rig and starting over.
Edit-Edit-Edit: I'm back to "Nifty! 30 amp Relays!" Carry on

You haven't mentioned your elements yet - How many watts are you asking the relays to switch?

I get what you are saying. I'm not an electrician or an electrical engineer by any means. But i DO have a basic understanding of electricity - understand AC vs DC and understand 120v is a hot and a neutral and that 240v isn't (unless you're in Europe). But you will have to explain how I am passing 240v through any of my relays when none of them are connected to 2 hots. And how this would be any different using SSRs.

I am not an electrician, and this might be an over-simplification, but this is how I understand it. Here is a highly professional-looking (insert sarcasm here) drawing that might help. In the pictures, the yellow boxes represent relays and the blue loops represent heating elements. The entire circuit through the heating element is considered a 240V circuit. I assume you agree that you are passing 240V through the heating element. It is one part of the circuit. It is also important to realize that all other parts of that circuit "see" the 240 volts and whatever amp load the element is drawing as well.

Look at the bottom drawing - I purposefully drew the relay on the 'end' of the circuit, where the heating element was drawn in the others. (This would be wired exactly the same way as the top left picture.) Maybe if you visualize the relay on the 'end,' you will be able to understand how it is all part of the same circuit, therefore subject to the same voltage and amp load. Does this make sense? SweetSounds, please correct me if this is not the information you are intending to convey.

geoffey said:

I get what you are saying. I'm not an electrician or an electrical engineer by any means. But i DO have a basic understanding of electricity - understand AC vs DC and understand 120v is a hot and a neutral and that 240v isn't (unless you're in Europe). But you will have to explain how I am passing 240v through any of my relays when none of them are connected to 2 hots. And how this would be any different using SSRs.

Click to expand...

I'm sorry, but you don't understand.


You are missing the key point:
Each relay IS connecting to two "Hot's"!
The one going into it and the one coming out of it.

Because, 240 volts is the voltage of the CIRCUIT and the voltage of every single wire in it.
Your element is 240 volts. It doesn't matter if you're measuring at terminal 1, terminal 2, or if you could measure somewhere in between. It's still 240 volts. I'm having trouble coming up with another way of explaining it...

Let me try it this way.
Lets pretend you have a 9 volt battery, and you connect it to your heating element.
Do you have 4.5 volts per wire? No, you have 9 Volts! 9 Volts of DC electricity passing from the positive terminal on the battery, to the negative terminal on the battery. The heating element just happens to be in the electricity's way. But believe me, it's still 9 volts. On every wire in the circuit.

That's all there is to it. Forget for a minute about hot and neutral and heating elements. The voltage running through any single wire in any circuit is what it's fed by.

Or maybe this is an easier way to understand:
If you look at Aggie's drawing - The top left one. You know that if you put a volt meter across the terminals of the heating element, you'll get 240 volts right?
Now, if you put your volt meter across the terminals of the relay, What will you measure?
240 volts. Have we suddenly created a new hot? No, it's just a wire. It's got 240 volts on it. Just like every other wire in that circuit.

All 3 wires in that simple circuit are carrying 240 volts. And if it's a 5500 watt element, all 3 wires in that circuit (And the relay itself, the 14 gauge wires connecting it, and the 20 amp rated contacts inside it) are carrying 23 amps!

SSR's are designed for quickly switching loads. They don't have a "Break current" limitation. This is my biggest concern with using little mechanical relays over their rated load. You can literally weld the terminals together when they touch.
I say use SSR's because they are "normal" It's what they do. They can handle the load. They can handle the switching rate. You only need 1 or 2 of them, and they are not that expensive.
Again, your relays might be fine! And kids playing with matches in the basement don't always start the house on fire. You need to fully understand what you've built and what it means to each component you pass 115% of its rated current through.

This is why most people here use 40 amp SSR's for their elements. The elements draw 23. It's a safety factor, and the SSR runs cooler because it's at 1/2 of its rated load.

Guys, thanks for helping clarify this issue for me. I think I understand much better and for safety sake I will make a change. Who knows, maybe you helped me avoid burning my house down! One easy change would be to swap the relays out for a different model which is rated for 30amps at 347v for resistive loads. OR I could just go ahead and buy SSRs but I'd need to put some further thought into how I'd install those in my system.

Back to the original question: Mechanical Relays versus SSRs

Note that SSR stands for Solid State Relay. If you switch the mechanical relays just as fast as the SSR you should get the same level of control (holding of desired temperature). The question is: how long will a mechanical relay last switching as often as every two seconds? As I remember that is the fastest cycle time of most of these small PID controllers.

If you are happy with the life of that relay the only other issue might be the small radio pulses - like tiny lightning static every two seconds.... which may. or may not, be noticeable.

I would be very curious how long they last in that service - a quality relay will do many thousands of cycles - so it may be a few years.

I think Geoffey will be fine if he uses relays of the appropriate amp rating and adjusts his cycle times so they are not switching frequently. In fact, that is what I am going to go with, since I can adjust the Johnson controller so it doesn't 'short cycle'. (Disclaimer: I have no experience with brewing PID controllers and have no advice as to how to switch the frequency, but it sounds like he has that part handled.)

I explained to an electrician friend of mine what I wanted to do, and he happened to have a uesd relay laying around that he gave me. In light of the 'MTBF' and life expectancy issue, I hope it's not too old, but I guess I'll see how long it lasts. I think he said it came out of an industrial water heater. It is good for 40 amps, and is actually meant for 3 phase, so it has 3 poles (or sets of contacts; I'm unsure of the terminology). Anyway, I'll use two of the poles to switch the hot legs going to the heating element. I might use the third pole to run an LED or something, depending on how crazy of a control panel I want to build. Geoffy, let us know what you decide to go with and how it performs. There is a possibility that my used relay will burn out sooner than later and I might need a replacement

I agree - With a 30 amp rated relay, the only thing you have to worry about is mean time between failure and stratification (In my estimation)

I could see an issue, unless you are recirculating, where if the element is on for 30 seconds and off for 90, that the heat may not evenly distribute.

But I have nothing to back that up.

Carry on and let us know how it works out!

@SweetSounds Thanks for your input on this thread. You did a great job of explaining things.

I've just purchases some relays and SSRs this week for my rig. The SSRs will be for control and the mechanical relays will be for main power shutoff. I actually found the 40 amp SSRs to be cheaper than the 30 amp DPDT mechanical relays. My system will be a little different because my relays use a 12 volt DC coil instead of 120 volt AC, but those are even cheaper than the 120 volt coil relays.

Mechanical Relay model # JQX-30F 2Z about $7.50
Solid State Relay model # SSR-40DA 40A /250V 3-32VDC about $5.50

These prices are on eBay with free shipping. You just have to wait a few weeks to clear customs.

bendiy said:

@SweetSounds Thanks for your input on this thread. You did a great job of explaining things.

I've just purchases some relays and SSRs this week for my rig. The SSRs will be for control and the mechanical relays will be for main power shutoff. I actually found the 40 amp SSRs to be cheaper than the 30 amp DPDT mechanical relays. My system will be a little different because my relays use a 12 volt DC coil instead of 120 volt AC, but those are even cheaper than the 120 volt coil relays.

Mechanical Relay model # JQX-30F 2Z about $7.50
Solid State Relay model # SSR-40DA 40A /250V 3-32VDC about $5.50

These prices are on eBay with free shipping. You just have to wait a few weeks to clear customs.

Click to expand...

Glad to help!
And sorry if it came off a little harsh (After re-reading some of the posts)
But I think you're on the right track!

Keep the progress updates coming

Just a quick update on this old thread I started:

I completed the build in September and have brewed 5 batches so far. I ended up changing my relays out to a 30 amp latching relay which was rated to handle the requires load (ended up installing a Wattstopper relay panel, for those interested). A quick thanks once again to Sweetsounds for helping me to fully understand the implications of the system I was proposing

I ended up having more trouble with the PID controller than anything else. Auber model SYL-2362 is what I purchased because it has both a relay output AND SSR output. I expected this would give me some future flexibility if I ever decided to go SSR control. The problem I didn't realize is that this model doesn't have fuzzy logic. I'm not 100% on what fuzzy logic is, but I'm now convinced its important on a system like this.

When I auto-tuned the PIDs the system became extremely sluggish. Trying to change the temp by just 10 degrees would take way too long as the relays continually cycled, even when the target temp was not close yet. I eventually manually tuned the P, I, and D variables until the system was faster to respond. But I consistently overshot my temps by 3-5 degrees.

I decided that was unacceptable for me in the long run and just pulled the trigger on model SYL-2342, which has a relay output and fuzzy logic. After one auto tune I am able to hit my temps on the dot and hold them.

So my suggestion to anyone looking to build a relay based control rig would be to make sure you use the Auber model with Fuzzy logic.

Let's do the math!

Heat capacity of water is ~4 Joules per mL, per degree C.

So, for 20L of beer, that's 20,000mL of beer. So it will take 80,000J to heat it up 1 degree.

5500W is 5500 Joules / second. That means it will take 80,000/5500 = 14 seconds for a 5500W element to heat up 20L of beer 1 degree C.

OK, so a 10 second cycle would probably be OK on a HLT especially if it was really full. Maybe you could get away with a mechanical relay for that.

But you'd want finer control over mash temperature.

geoffey said:

I plan on using lighting relays instead of SSR's. There are a couple important reasons I'm going in that direction, which it isn't really important to me to discuss at this point.

Click to expand...

1) You're a tightwad and you have access to free relays.
or
2) You really like the sound of clacking mechanical relays.

It's okay to fess up to either of those. Plenty of tightwads on this board. You're among friends.

I am using relays for controlling my 240v heating elements. I decided to go this route because 1) they were free, 2) it switches both legs of power with one relay, 3) no bleed through like you get with SSR's. I kind of like the clicking sound because I can hear that it is working.

Getting back to the mtbf of a mechanical relay, a little quick math tells me that assuming you are controlling it with a PID, and you had a cycle rate of 4 seconds, that works out to be about 900 cycles per hour,

Of course, in an actual application, your controller is not going to cycle at that rate continuously.

The point is, that mechanical relay is rack up a huge number of cycles over the life of your system and will most likely be the first thing to fail from simply wearing itself out.

IMHO just another reason to consider an SSR for your current control device.

processhead said:

Getting back to the mtbf of a mechanical relay, a little quick math tells me that assuming you are controlling it with a PID, and you had a cycle rate of 4 seconds, that works out to be about 900 cycles per hour,

Of course, in an actual application, your controller is not going to cycle at that rate continuously.

The point is, that mechanical relay is rack up a huge number of cycles over the life of your system and will most likely be the first thing to fail from simply wearing itself out.

IMHO just another reason to consider an SSR for your current control device.

Click to expand...

I have 3 SSR's and 2 mechanical relays that have been in my brewery for over 5 years and the only one that I've had to replace was a SSR. All relays are subject to failure so it is tough to say which will go out first.

LandoLincoln said:

1) You're a tightwad and you have access to free relays.
or
2) You really like the sound of clacking mechanical relays.

It's okay to fess up to either of those. Plenty of tightwads on this board. You're among friends.

Click to expand...

Nope, neither of those were the reasons I chose to use relays. I believe I went on to explain myself later in this thread. But when building a system with an overall cost of over $1,000 I didn't consider $40 for an SSR and heat sink a breaking point.

processhead said:

Getting back to the mtbf of a mechanical relay, a little quick math tells me that assuming you are controlling it with a PID, and you had a cycle rate of 4 seconds, that works out to be about 900 cycles per hour,

Of course, in an actual application, your controller is not going to cycle at that rate continuously.

The point is, that mechanical relay is rack up a huge number of cycles over the life of your system and will most likely be the first thing to fail from simply wearing itself out.

IMHO just another reason to consider an SSR for your current control device.

Click to expand...

My PIDs are set for a 20 or 40 second cycle, I wouldn't set them to any less. If they wear out its a quick replacement fix for me at no cost since I can source them for free. But the main thing is that at that cycle rate I can consistently hold temps within one degree, which is what's important to me.

For what it's worth, I think the best arrangement is to have a power on switch governing a DPDT mechanical relay and the PID controlling the element via an SSR. That way you get the best of both worlds - the fast switch rate of the SSR and the fail safe mechanism of the mechanical relay.

Not either relay or SSR but both.