Why use an SSR instead of a relay on a PID?

[Tazza]

Is there a good reason to use an SSR instead of a relay on a PID? Sure, an SSR will not wear out with normal use, a relay will wear out over time, but it will take a long time to do.

A relay generates pretty well no heat, an SSR generates heat that you need a heat sink to dissipate depending on the size of the pot it is powering.

Is there a reason i have missed why i should use an SSR over a relay? i have 240v AC coil relays, easily hooked up to the relay outputs on the PID.

[dragon813gt]

Relays will burn out in short order. They aren't designed to take the cycling that a PID puts out and a SSR can handle. A relay will work but will eventually fail. Personally I don't want to hear them opening and closing all the time as well.

[Tazza]

That is what i figured too, the contacts have a life span, an SSR ideally can switch till dooms day.

I'll have to see if my setup will allow for the addition of an SSR, biggest issue is where to put a heat sink..... The case is enclosed, so the heat will have no where to go.

[dikman]

What sort of case? If metal, just bolt the SSR to the case (and drill a few vent holes). If plastic, like mine, I just fitted whatever heat sink bits I had that would fit and drilled some cross-flow ventilation holes in the sides.

[Drew P]

I'm not sure I believe this. A regular old relay can handle many many thousands cycles and are used in PID controlled furnaces all the time. Now, they do make noise.

[hermans]

With the Rex C100's I have purchased cheaply from Ebay, it is also a matter of the relay contacts not being able to deliver the current needed to fire up element of a electric casting furnace like the RCBS Pro-Melt.
So changing over to a SSR is an relatively easy solution to this problem, and yes, it is also quiet when switching on and off.

[Drew P]

Isn't the Rex c100's relay contacts rated for 5 amps? Too bad, thats almost enough to melt lead.

[OS OK]

You have to use a relay 'rated' for the load it controls. Properly sized they will provide service for the life of your pot. If there is a 'short' circuit where many amps rush through they may or may not fail before the circuit breaker trips but they are cheaply replaceable, in that way they provide 'isolated' switching without having a fault current go through your solid state components ruining the entire package. Some relay models have replaceable contacts. Some 20 amp rated 'ice cube' styles are almost completely silent. Don't load your relay more than 80% of its contact rating.
In machine motor control panels I built I used to always use isolation relays between the solid state control and the fractional HP motors having internal thermal protection, heating elements or other heavy current items for the above reasons. We used isolation relays for the actual 'control' circuits on board the machine because of accidental damage of control panels or limit switches etc. resulting in high current shorts. It was my design to keep high fault currents out of the solid state controls because when they melt down it usually involves more than replacing a simple SSR.

[leadman]

There was a member, now banned that was making PIDs with a relay in them. I did hear of several of these going out.
When my pot or oven is up to temperature the PID cycles about once a second or so. I have cast for 8 hours at a time so I don't know how long the relay would last.
The cheap Chinese SSRs do not last forever either. I have a couple not work from the start and had a batch of 25a SSRs that only last several sessions. The 40a SSRs seem to last longer. I have about 3 years on a couple of them.
The Rex unit is labeled for either a relay or SSR so make sure you match them up as one marked for an SSR will not power a relay and one marked for relay will not power an SSR. I did buy a couple that were mis-marked and had me scratching my head for a bit.

[OS OK]

There was a member, now banned that was making PIDs with a relay in them. I did hear of several of these going out.
When my pot or oven is up to temperature the PID cycles about once a second or so. I have cast for 8 hours at a time so I don't know how long the relay would last.
The cheap Chinese SSRs do not last forever either. I have a couple not work from the start and had a batch of 25a SSRs that only last several sessions. The 40a SSRs seem to last longer. I have about 3 years on a couple of them.
The Rex unit is labeled for either a relay or SSR so make sure you match them up as one marked for an SSR will not power a relay and one marked for relay will not power an SSR. I did buy a couple that were mis-marked and had me scratching my head for a bit.

Generally temp. controls have a 'cut in' and 'cut out' setting providing a certain degree of spread there. Say for instance I want it to cut in and heat up to 650 degrees and then cut out. Then it will not cut back in until the temp drops down to say 640 degrees. That way it will not sit there and burn itself out switching in and out trying to stay on an exact temp. I think that's the problem you are experiencing there.

Air compressors use different controls but the idea is the same. My compressor cuts out at 125 lbs/sq/in, PSI. It will not cut in, restart until it drops down to about 112 PSI or so, that is called a 'differential'. It's adjustable. I think your PID should be the same…of course I might be talking out of my butt as I don't have a fancy control on my pot…yet!

[dikman]

It's been mentioned here in other threads - just because the label rates the (PID) internal relay at 5 amps (for example) doesn't mean it can actually handle that load! The common consensus seems to be that those internal relays are probably over-rated somewhat (advertising zeal!). They should only be used to drive an external, heavy duty relay if you want to switch a pot's heating element.

SSR's are generally used because, for most people, they provide simpler wiring requirements. Plus they're cheap enough that it pays to keep a couple of spares around.

[Vann]

About the worst thing that can happen by using a relay is the contacts can weld closed and not break. This would cause it to heat up until the unit either burned out, tripped a breaker, or gets unplugged. Not very likely but it could happen.

[z28z34man]

Whenever you make ore brake a electrical load with an air gap there is a small arc that arc erodes the contact some that is why relays are rated for booth amps and cycles. A pid will cycle a relay on and off quite a bit and kill it in short order.

[dragon813gt]

Generally temp. controls have a 'cut in' and 'cut out' setting providing a certain degree of spread there. Say for instance I want it to cut in and heat up to 650 degrees and then cut out. Then it will not cut back in until the temp drops down to say 640 degrees. That way it will not sit there and burn itself out switching in and out trying to stay on an exact temp. I think that's the problem you are experiencing there.

Air compressors use different controls but the idea is the same. My compressor cuts out at 125 lbs/sq/in, PSI. It will not cut in, restart until it drops down to about 112 PSI or so, that is called a 'differential'. It's adjustable. I think your PID should be the same…of course I might be talking out of my butt as I don't have a fancy control on my pot…yet!

PIDs are attempting to maintain to a specific setpoint. There is no cut in and cut out. We aren't talking a well pump here.

You can absolutely use a conventional relay. But it's not a matter of if but when it will fail. They have a lifespan that is a lot shorter than a SSR. I'm still on the original SSR and going on five years. W/ how quickly the PID cycles I would have had to replace a conventional relay by now.

[6bg6ga]

I have to pull the BS cord. Relays will work fine. Anyone here ever run machines? I ran machines for years in a factory and guess what? They had a bank of relays in them and this was before the age of SS relays and logic. Want to talk cycles? I would hate to try to figure out the cycles because the number of cycles times a single years would scare you. The simple plane fact is either will work. Ever wonder what is inside that SS relay? Its solid state which means its internals which I will not bore you with also have a life cycle. Each either a SS relay or a mechanical relay HAVE a known life in cycles. One makes a click and one is silent. The trick is to pick a relay that is capable of the load it controls. And NO a PID will not kill a relay in short order. Stupidity however in design will. I puchased a bullet caster a few years back and its name and model isn't important. The problem I encountered with this unit were staggering. It ran for a whole 4 hours before the PID went kaput. I ended up re-doing the complete control system a complete re-wire on it before it was 100% usable. I also had to install a fan to cool the SS relay in order to obtain workability. I finally set the control unit aside and designed an old fashoned relay controled PID setup and it hasn't missed a beat. This is because I understand electronics and control. Now, I still have the original control box and its workable but I have also proven that the old fashoned relay control system works well also. I have about 20K cycles on the relay and it still functions correctly and no burned contacts to date.

[6bg6ga]

Just to figure a years worth of cycles on one machine It used to run 1,440,000 cycles a year based on production for three shifts and a years production. Since I also kept track of maintenance on the machines in my area I can also tell you that some ran over 5 years at this rate before any relays were changed.

[dragon813gt]

The life cycles are listed w/ the relays and SSRs. Mechanical relay lifespans are small in comparison. There is no doubt they will work. No one is saying otherwise.

[hatti]

i have let out the magic smoke - lots of smoke :razz: - from one SSR - the pot was dripping and lead eventually shorted the heating element.
had no circuit breaker either,but PID did not get damaged in any way.

IIRC i read somewhere that SSR's are optically coupled,therefore they cannot damage the controller when they fail.

[Vann]

I agree a relay will work. I work in a pvc compounding plant, the largest in North America as a matter of fact. We don't use relays in our TCU's we use contactors, which is just a big relay with 480 vac running through it. Most or the larger ones are either size 4 Allen Bradley or size 3 Benshaw. If you know much about electrical you'll know what I'm talking about, the smallest ones use Allen Bradley C43's and C60's.

So far in the last 4 years I've had the pleasure of rewiring 3 TCU's due to a stuck contactor. My best advice is if you use a relay go oversize and install some kind of failsafe.

[Half Dog]

Mechanical relays and SSR's work. Each are different and with the differences comes pro's and con's. SSR's turn on and off when the voltage on the switch leg crosses 0 volts. The relays just cycle but because they do not turn off at the 0 volt level, arcing occurs. The arcing causes contact failure. The SSR is sensitive to spikes and it can be damaged within a fraction of a voltage cycle (60 hertz). With a resistive load (heaters) there is no need to choose a relay that has a contact rating that is greater than the current draw.

If you choose the SSR, it can be mounted to a metal surface but it is advised to use a thermal compound between the back and the metal.

[Mal Paso]

Modern SSRs switch at zero volts in the alternating current cycle generating almost no electrical noise.

SSRs handle high speed switching much better than mechanical relays.

With a LEE 4-20 a SSR will generate 6 watts of heat. Bolt it to a metal case.

Amazon sold me a 25A SSR with a fancy heat sink a couple years ago for $12. You'd be hard pressed to find a mechanical relay that would do as well at that price.

[edctexas]

Depending on how the PID is designed, good designS should be capable of applying as little as one half cycle of AC. Most of the PIDS we buy are digitally implemented. This means that they may output longer bursts of AC power. If the bursts are short (like 1/60 of a sec) a relay will die in short order. Digital PIDS output longer power bursts so relays may give useful life.
Ed C

[scarry scarney]

My vote is for the SSR as well. I have had relays contacts corrode away, just like the points of an old car. (If your old enough to remember points, that is). How many of you recall using a file, to reshape the contact pads on your points. I know that while casting lead, it's not going to cycle as much as a set of points on an old chevy 350.

[ipijohn]

I am SSR all the way (I have 5 PID's that I use for different purposes around my house) mechanical relays are for turning on/off less frequent duty cycle devices such as refrigerators, furnaces etc. because for OEM use they are cheaper.

[jmorris]

Life span for an SSR is greater than an EMR as it has no moving parts.

http://m.electronicdesign.com/components/electromechanical-relays-versus-solid-state-each-has-its-place

[Jim_P]

Well, here's one that often gets missed. Relays generate something called Back EMF. When the magnetic field collapses after the current is turned off, that residual magnetic field induces a current in the coil of the relay/solenoid. The only place for it to go is back to the PID and whatever electronics are used to control the output of the PID. This Back EMF is large enough sometimes to destroy electronic components. The way around this is two diodes back to back across the relay solenoid to dissipate this induced current.

So if you're going to use a relay, don't forget the diodes.

[Tazza]

All very good points, a relay is simple as no heat sink is required. The case i have is plastic and pretty full of control gear for my master caster. Yet if i can find space, i could always cut a hole through the case and install an external heat sink.

[bangerjim]

If I ever build one of my high-end 1/2 DIN dual channel industrial controllers up for controlling my little lead pot, I will probably use one of the 60A rated relays I have in the shop! Industrial relays of that kind are rated in decades of cycle time, have NiAg "No-Korode" contacts, and are not these cheap little things you find on evilbay. And it is nice to hear the audible clicking for feedback!

For the average person on here, a 25+A SSR is the best choice.

banger

[Tazza]

I bet one of those relays is worth quite a bit, but designed to last.

The relays i have were from ebay, no doubt made from the cheapest metals they could find.

The SSRs i have, i believe are 25a. I bought them a while ago for a project that never happened.

[bangerjim]

My controllers sell for ~$850 each and those relays are around $200 each. I design and sell industrial control systems!

I do not use SSR's due to the normal leakage current thru them. To a little lead pot, that is not a factor.

Of course, being "down under" your current circulates in a different direction!?!??!?!? HA......HA!

banger

[Tazza]

Wow, now that's a relay! The sad thing is, that over the years of me pulling gear apart, i may have come across many expensive quality relays that were scrapped or thrown out. You know how it goes, what do you keep and what do you throw out?

hehe, now that's why things run backwards here :) well that and the government....

[Vann]

I've been scraping for a while I've managed to save 3 Eurotherm 3612 controllers and a Fuzzy PID. I really like the Fuzzy's, they are cheap, easy to program, and seem to last forever compared to some of the PIDs I work with. We replaced 3 Sterico units last year so I was able to go through and salvage what I wanted. I ended up with 3 5hp and 1 10hp Allen Bradley contactors. With the Eurotherm and Fuzzy PIDs we have been lucky and haven't had to use snubbers, at least that's the way the units are built.

[jmorris]

Mine is just mounted to a flat plate for the heat sink in my "control" box. It has been turning on and off a 3500 watt element for 11-12 years now.

http://castboolits.gunloads.com/attachment.php?attachmentid=161726&d=1456274429

[Tazza]

Sadly i don't get many control boxes to pull to bits. The last one my, dad sold to a friend of his before i got to it. It had all sorts of fancy stuff, PLCs, power supplies, motor speed controllers. No temperature controllers though.

I'll have a look to see what space i have left for an SSR.

[Mal Paso]

Mount the heat sink through the side or top leaving the wiring inside.

Here's a recycled Intel CPU cooler out the top of my Auber Box.

[Tazza]

I like it! Simple, yet effective

Would easily move enough heat away from the SSR.

[dikman]

Tazza, if you use a 40 amp SSR it will barely get warm when used with a typical lead pot. I used a 25 amp and have had no problems with minimal heat sinking.
On the subject of relays, prior to digital circuitry telephone exchanges consisted of relay switching - untold thousands and thousands of them, switching almost continuously. Failure rates were extremely low.

[Tazza]

Good point about the phone network using relays, but i'd bet that the materials they used were just a little better than what we get out of china these days :)

I examined my control box this afternoon, i think i have found a spot to mount it. I just need to modify my PID to make it work with an SSR, but that looks easy enough with the youtube instructions.

Hopefully i'd have it all up and running on the weekend, then i can get it to auto tune and see how well it holds temperature.

[bangerjim]

Good point about the phone network using relays, but i'd bet that the materials they used were just a little better than what we get out of china these days :)

I examined my control box this afternoon, i think i have found a spot to mount it. I just need to modify my PID to make it work with an SSR, but that looks easy enough with the youtube instructions.

Hopefully i'd have it all up and running on the weekend, then i can get it to auto tune and see how well it holds temperature.

Those old telephone relays were for VERY low current and ~70vdc.....and would not work on a 10-15A lead pot. Contacts did not usually corrode, but could not handle any current either. And the coils would not work with the outputs of those cheap controllers everybody uses.

banger

[Tazza]

Where i can, i try and use a DPDT relay and parallel the poles. So a 6a relay can technically handle 12a.

[Mike W1]

Those old telephone relays were for VERY low current and ~70vdc.....and would not work on a 10-15A lead pot. Contacts did not usually corrode, but could not handle any current either. And the coils would not work with the outputs of those cheap controllers everybody uses.

banger

Actually when I worked in the switchroom the voltage was 48 VDC and the contacts were valuable enough that we clipped the ends off the old relays and they were recycled. Think my spare SSR was about $3 shipped so that's a no-brainer as to what to use.

[dikman]

True, most of those relays weren't designed to handle more than probably 3 amps, but as Mike said many had precious metal contacts and also used "wetting" ccts. to reduce arcing. Point is that if a relay is designed properly they are very reliable. But I still wouldn't bother with one.

Tazza, it's easy enough to change the PID over (at least, my Rex C100 was). I opened it up, worked out which relay was the controlling one, desoldered it and installed two links to bypass it. Simple).

[Tazza]

I think i will have a stab at it on the weekend, see if i can avoid letting out the magic smoke inside the PID :)

[6bg6ga]

If I understand you correctly you wish to parallel two 6A contacts. I wouldn't count on that working. Instead source a relay with with higher amp rating. Most PID's have relay contact that either work all the time or can be programed to work. Use the 3-4 A PID relay contact and run control voltage thru these contacts to activate an external relay that is rated 15A. To recap for you... You could use a 24Volt supply and run this thru the PID relay contacts and from there to your external relay that is rated for the load. The relay points of the external relay will switch the 120 or 220 or what you have in the land down under. Across the external relay coil you will use a 1N4007 diode to take care of the back EMF cause by the coil deactivation. It doesn't have to be a zenier in conjunction with a diode to work just a simple 1N4007 diode and this diode is rated 1A@1000v which is enough to take care of the back EMF pulse.Stay away from the ebay relays and use a control voltage that will work well. A lot of control circuitry is done with 24VDC and it should be simple for you to either build or purchase a 24VDC supply. If you need help PM me and I will send you diagrams.

[6bg6ga]

If you wish a visual indication you can easily run a LED or multiple LED's with a properly sized voltage dropping resistors to give yourself a display that shows cycle on or cycle on and cycle off. As long as you use a good quality relay you will get years of service from a single relay. As mentioned in posts above relays will work. Every control design can have its drawbacks. Some feel SSR's are industructable and granted they can have a long life but if not sinked correctly their life can be short. I used the cut a hole in the box method on a new piece of equipment and found that dispite the fact that I had air moving in the box it wasn't enough to properly cool the SSR. It ended up taking a heat sink 3 time normal size that was fan cooled to work in the enclosed box so as I have said design is everything. I had tried to make an original control box work when I should have scrapped the box and purchased a new one that was bigger that would have allowed one to either heat sink and fan cool an internally mounted sink or externally mount the sink with a fan. I finally did get the control box working without incident but opted to build yet another control box with a relay and use it.

[Tazza]

The paralleled contacts is just an example. The relay i used in my other control box was rated at a suitable current, i just joined them up as i only needed to switch one pole and the other was un-used, i figured it would make the contacts last longer. I didn't do it so i could use a 6a relay instead of a 12a.

[jsizemore]

I think i will have a stab at it on the weekend, see if i can avoid letting out the magic smoke inside the PID :)

I found a relay with 120 vac control.

[Tazza]

The SSR has been mounted and hooked up. The heat issue is not an issue, when running, it only gets a little hotter than ambient temperature.

After the auto tune, and when not adding ingots, it holds the temperature very steady. I set it to 340c and it stayed there.

I'll get pictures off my phone when i get a few minutes spare.

[Tazza]

http://i903.photobucket.com/albums/ac233/tazza_/master%20caster/20160228_103809.jpg

It fitted pretty well up in the right corner

[dikman]

Just curious, what else have you got running in that box?

[Mike W1]

Just curious, what else have you got running in that box?

Was wondering that also!

[Tazza]

Sorry, i didn't mention that it was for an automated master caster.

[dudel]

Tazza, if you use a 40 amp SSR it will barely get warm when used with a typical lead pot. I used a 25 amp and have had no problems with minimal heat sinking.
On the subject of relays, prior to digital circuitry telephone exchanges consisted of relay switching - untold thousands and thousands of them, switching almost continuously. Failure rates were extremely low.

Been in one of those. The noise was amazing! I remember the conversion to digital. Once the wire frames were removed, the COs were almost empty.

[el34]

Just curious- has anyone ever felt the SSR to get an idea of heat generated? Theoretically the worst case would be just before it's starting to cycle on-off as it nears the set point.

I have an SSR rated for 25A. Its datasheet says at that current it would drop 1V across it. The pot only pulls about 7A so the SSR shouldn't ever dissipate more than 7 watts (1V x 7A), about the same as an old-fashioned night light. Being nitpicky I put a small fan in my controller box between SSR and the outside but the barely-detectable air flow feels the same temp as the surrounding air.

Once the pot is up to temp the on-off cycling dissipates even less heat since it's off a good part of the time.

[JonB_in_Glencoe]

Just curious- has anyone ever felt the SSR to get an idea of heat generated? Theoretically the worst case would be just before it's starting to cycle on-off as it nears the set point.

I have an SSR rated for 25A. Its datasheet says at that current it would drop 1V across it. The pot only pulls about 7A so the SSR shouldn't ever dissipate more than 7 watts (1V x 7A), about the same as an old-fashioned night light. Being nitpicky I put a small fan in my controller box between SSR and the outside but the barely-detectable air flow feels the same temp as the surrounding air.

Once the pot is up to temp the on-off cycling dissipates even less heat since it's off a good part of the time.
The aluminum enclosure I used, is a two piece. One piece is perforated and makes up 3 sides. The other piece is solid and makes up both ends and the top. I mounted the SSR (25A) to the top (inside). So I can feel the heat generated, by feeling the top (outside). It barely gets warm to the touch during and at the end of the initial 20 minute ramp up to temp...but I've noticed it "IS" warm at the end of a hour long casting. session, so the heat does build.

[Tazza]

I think mine is a 40a one, during initial heat up it just gets warm. I installed a very small heat sink, not sure it was really even needed. When the pot is hot, the heat sink is pretty well room temperature, even with it's cycling on and off to maintain temperature.

From what i read, they dissipate about 1 watt per amp, so really bugger all for the melting pot we are running them on. I had expected quit a bit more loss to heat.

[Mike W1]

I install heat sinks in my units so I don't receive a "Gottcha!". :) I worked the graveyard shift at the Palm Desert Central Office in 1979, and you could always tell when some radio station was having a dial in contest. Sounded like a company of Tigers was rolling through the Ardennes. :)

OB

Line Load Control ? used to have to test that once a month but far as I know nobody ever had to activate it.

[dikman]

I only have a 25amp SSR and it barely gets warm.

When we installed a "relay" exchange (Swedish Crossbar system) we had to load test them prior to putting into service and we would listen to the rhythm of the relays as it would often give an indication where to look if there was a problem. The earlier Step by Step switching systems were even noisier!

And yes, dudel, when these were replaced with digital stuff the buildings looked almost empty!!!!

El, always good to hear from you.

[Tazza]

I wonder what the power savings were when they changed over too. The old stuff that took up masses of space with mechanical switches vs digital switching. Also, the cost to repair the newer stuff, i can imagine that the new stuff isn't cheap and you can't just replace one component, you replace a card. Old school, a bad relay, you replaced just the relay.

Gotta love the throw away society we live in now.

[Vann]

I wonder what the power savings were when they changed over too. The old stuff that took up masses of space with mechanical switches vs digital switching. Also, the cost to repair the newer stuff, i can imagine that the new stuff isn't cheap and you can't just replace one component, you replace a card. Old school, a bad relay, you replaced just the relay.

Gotta love the throw away society we live in now.

I can't tell you how much replacing a whole board cost vs just replacing a single relay, I know that in the maintenance field a PLC board cost several hundreds of dollars. The flip side is down time, it can take several hours trying to find a bad relay verses jumping on a computer with Modicon, Allen Bradley, or one of the other PLC programs and being able to find the problem in a matter of minutes. I'd say down time saved by not having to trouble shoot a switch board verses the price of a PLC board that can be sent off and repaired, you'd come out a head with the PLC.