PID Controller - Auber SYL-2352

[NineInchNails]

Auber Instruments SYL-2352 PID Controller
Auber Instruments:
Auber 1/16 DIN PID Controller (SSR control output) SYL-2352 (1.89" x 1.89" x 4") (http://www.auberins.com/index.php?main_page=product_info&cPath=1&products_id=3) $45.00
10" K type high temperature thermocouple for heat treatment (http://www.auberins.com/index.php?main_page=product_info&cPath=20_3&products_id=22) $22.62
K Type Thermocouple Mini Connector, Plug (http://www.auberins.com/index.php?main_page=product_info&cPath=7_34&products_id=162) $3.30
120V 15A US Socket, Panel Mount, NEMA 5-15R (http://www.auberins.com/index.php?main_page=product_info&cPath=7_34&products_id=274) $1.95
Panel mount connector for K thermocouple (http://www.auberins.com/index.php?main_page=product_info&cPath=20_3&products_id=119) $5.89
K type thermocouple extension wire, multistranded (12") (http://www.auberins.com/index.php?main_page=product_info&cPath=20_3&products_id=179) $1.25
Thermal Grease, Silicone Compound (http://www.auberins.com/index.php?main_page=product_info&cPath=2_30&products_id=123) $2.17
Shipping $7.02
Total = $89.70

eBay:
Solid State Relay SSR-40DA 40A 3-32VDC Output 24-380V (http://www.ebay.com/itm/Solid-State-Relay-SSR-40DA-40-AMP-3-32VDC-Output-24-380V-AC-SSR-40A-Module-/321123577566?pt=LH_DefaultDomain_0&hash=item4ac474eede) $10.99 Free Shipping
Black Aluminum Project Box Enclosure 8" x 5.7" x 2.7" (http://www.ebay.com/itm/Aluminum-Project-Box-Enclousure-Case-Electronic-DIY-Big-/250826353904?pt=LH_DefaultDomain_0&hash=item3a666a6cf0) $18.49 + $4.99 Shipping = $23.48

Heatsink USA:
Heatsink 3.5" x 2.75" ( http://www.heatsinkusa.com/3-500-wide-extruded-aluminum-heatsink/) $13.87 (including shipping)

Radio Shack:
RadioShack 6-Position Dual-Row Barrier Strips Model: 274-659 ( http://www.radioshack.com/product/index.jsp?productId=2103983) $2.99
Toggle Switch with Center Off 10A Model: 275-1533 ( http://www.radioshack.com/product/index.jsp?productId=2062537) $4.49

Lowes:
1/2" BX-MC-Flex Connector Item #: 108659 Model #: 49212 (http://www.lowes.com/pd_108659-15527-49212_0__?Ntt=49212&UserSearch=49212&productId=3150213&rpp=32) $3.04

Grand Total So Far = $148.06

Here’s the wiring diagram of what I have done so far. At this point I DO NOT have any fuses installed though I plan to. I originally found this wiring diagram in this post by Fredx10sen (http://castboolits.gunloads.com/showthread.php?34547-How-I-wired-a-PID-to-control-temperature&p=1240893&viewfull=1#post1240893). I modified the drawing a bit, made it accurate for the Auber SYL-2352, added the toggle switch, changed some pics and more clearly labeled the wires (live, neutral, ground).
http://i1303.photobucket.com/albums/ag142/boolits/PID%20Controller/PIDWIRINGDIAGRAM3_zps424af218.jpg (http://s1303.photobucket.com/user/boolits/media/PID%20Controller/PIDWIRINGDIAGRAM3_zps424af218.jpg.html)

I used 12 AWG wire for most everything. I used much smaller wire to supply power to the PID (110V) and from the PID to the SSR (which is 12V out).

I used AutoSketch to draw a template to plot the holes for drilling into the heatsink and the top of the project box. I also used AutoSketch to draw a template for the front & real panels to drill & cut holes. If I can figure out a way to post the template so that it can be printed I will.

http://i1303.photobucket.com/albums/ag142/boolits/PID%20Controller/PID01a_zps15ec49f7.jpg (http://s1303.photobucket.com/user/boolits/media/PID%20Controller/PID01a_zps15ec49f7.jpg.html)

http://i1303.photobucket.com/albums/ag142/boolits/PID%20Controller/PID02_zps6b913f3a.jpg (http://s1303.photobucket.com/user/boolits/media/PID%20Controller/PID02_zps6b913f3a.jpg.html)

http://i1303.photobucket.com/albums/ag142/boolits/PID%20Controller/PID09_zps63e82287.jpg (http://s1303.photobucket.com/user/boolits/media/PID%20Controller/PID09_zps63e82287.jpg.html)

http://i1303.photobucket.com/albums/ag142/boolits/PID%20Controller/PID10_zpsd40f8232.jpg (http://s1303.photobucket.com/user/boolits/media/PID%20Controller/PID10_zpsd40f8232.jpg.html)

Heatsink attached to the exterior of the project box:
http://i1303.photobucket.com/albums/ag142/boolits/PID%20Controller/PIDController06_zps5f54de0b.jpg (http://s1303.photobucket.com/user/boolits/media/PID%20Controller/PIDController06_zps5f54de0b.jpg.html)

http://i1303.photobucket.com/albums/ag142/boolits/PID%20Controller/PIDController07_zps3d093a0f.jpg (http://s1303.photobucket.com/user/boolits/media/PID%20Controller/PIDController07_zps3d093a0f.jpg.html)

Solid State Relay attached to the underside of the heatsink (using thermal grease) inside the project box:
http://i1303.photobucket.com/albums/ag142/boolits/PID%20Controller/PIDController08_zpsd7efab57.jpg (http://s1303.photobucket.com/user/boolits/media/PID%20Controller/PIDController08_zpsd7efab57.jpg.html)

http://i1303.photobucket.com/albums/ag142/boolits/PID%20Controller/PIDController09_zpsa08bce16.jpg (http://s1303.photobucket.com/user/boolits/media/PID%20Controller/PIDController09_zpsa08bce16.jpg.html)

http://i1303.photobucket.com/albums/ag142/boolits/PID%20Controller/PIDController11_zps7b57ac64.jpg (http://s1303.photobucket.com/user/boolits/media/PID%20Controller/PIDController11_zps7b57ac64.jpg.html)


How The Toggle Switch Works:

With the switch flipped UP the PID is in control of the pot.
With the switch in the CENTER position ALL power is off.
With the switch flipped DOWN you get power to the PID and the pot. Basically the solid state relay is bypassed providing power to the pot and the PID just functions as a digital thermometer.


Once it was all assembled I tested the accuracy in a large glass of crushed ice & water and with a pot of water at a rolling boil. Both temps read accurate so no calibration was required.

The extruded aluminum project box is real nice looking and heavy duty. Plenty of room inside for all the components with room to spare for potential upgrades, modifications or additions. I plan to get 4 rubber feet to attach to the underside of the box so it doesn't slide on smooth surfaces and so debris won't scuff the underside of the box.

I have not determined how I will clamp my thermocouple to the pot yet. I'm rolling around some ideas and I've seen a few examples in this forum.

[M Hicks]

I like the schematic NIN. That is pretty helpful to someone trying to make one of these. I like the idea of being able to use the PID as just a thermometer even if it may negate the reason for making the thing in the first place.

[NineInchNails]

Thanks M Hicks! Also you can switch the pot on, manually monitor the temp during warmup, when when the pot almost reaches the desired temp you can flip the toggle switch to 'PID' then set the PID to 'auto-tune'. From what I understand, heating up the pot (near the desired set value) prior to setting the 'auto-tune' makes it more accurate or something like that.

I liked the thought of being able to use the the PID as just a thermometer for smelting. It doesn't require a switch to be able to do that though ... ya just don't have to have anything plugged into the PID and it would work as a thermometer.

[Lights]

Nice build. I just bought this off of fleabay for my build hope it is the right stuff. http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=280879156946

[NineInchNails]

Nice build. I just bought this off of fleabay for my build hope it is the right stuff. http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=280879156946

I saw that one when I was shopping around. It should work fine. It seems to be a pretty darn good deal for a complete package like that. I knew I wanted a longer thermocouple so I could use it for smelting in my cast iron dutch oven. I'm not sure how long your thermocouple is, I didn't see a specification in that eBay link. Hopefully it's a usable length. Thermocouples aren't that expensive so you can pretty much get what you need or even extras to do what you need to do.

[Frozone]

.... I like the idea of being able to use the PID as just a thermometer even if it may negate the reason for making the thing in the first place.

It's not Just that - if the PID 'thinks' it is in control, but nothing it does has any effect, then it will mess up it's internal parameters.
This means you will need to retrain it before it can work properly as a PID.
Not too bad, it only takes about 15 minutes to train the unit, once you flip that switch.


Nice build. I just bought this off of fleabay for my build hope it is the right stuff. http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=280879156946

That's a RTD sensor. The Absolute maximum of that type is only 788º F, and that requires a Very spendy version
Most will only do ~ 400ºF, many only 260ºF

Oh and BTW My Kit, no errors, $142 shipped - complete (includes a wiring harness) and ready to build.

[NineInchNails]

It's not Just that - if the PID 'thinks' it is in control, but nothing it does has any effect, then it will mess up it's internal parameters.
This means you will need to retrain it before it can work properly as a PID.
Not too bad, it only takes about 15 minutes to train the unit, once you flip that switch.

Interesting. I had assumed that the PID 'learns' the pot after the auto-tune process and when in normal operation it followed those learned settings so to speak. I didn't realize it learns as it runs in normal operation as well. Poor assumption on my part, but I'm as new as can be with PIDs.

Just for fun I hooked an electric double boiler up to it, set the temp, turned on 'auto-tune' and set a stop watch. 8 minutes later auto-tune was complete! It has been holding within 1 degree F for the last 30 min. Even during the auto-tune the temp stayed very very close to the set value. I'm going to auto-tune again and double check how long it takes.

It is super simple to change the wiring on the toggle switch if need be, but so far it appears to be more beneficial to leave it as is. I imagine that this double boiler may be a poor comparison to a lead pot, but IF the PID 'learns' the pot anywhere near as easy, fast and acurrate ... I'm thinking I'll keep the switch wired up just as is.

[NineInchNails]

Bummer Lights ... Frozone is right about the temp probe. Check out the pic of the temp probe in your eBay link. It has 3 wires (thermocouples only have 2) and in the description they call it "Element: Pt100". If you look up a PT100 you'll find more info on it. No worries though. Thermocouples are relatively inexpensive and for $42.99 you can afford to buy a proper thermocouple.

[Frozone]

Interesting. I had assumed that the PID 'learns' the pot after the auto-tune process and when in normal operation it followed those learned settings so to speak. I didn't realize it learns as it runs in normal operation as well. Poor assumption on my part, but I'm as new as can be with PIDs.

Just for fun I hooked an electric double boiler up to it, set the temp, turned on 'auto-tune' and set a stop watch. 8 minutes later auto-tune was complete! It has been holding within 1 degree F for the last 30 min. Even during the auto-tune the temp stayed very very close to the set value. I'm going to auto-tune again and double check how long it takes.

It is super simple to change the wiring on the toggle switch if need be, but so far it appears to be more beneficial to leave it as is. I imagine that this double boiler may be a poor comparison to a lead pot, but IF the PID 'learns' the pot anywhere near as easy, fast and acurrate ... I'm thinking I'll keep the switch wired up just as is.

The PID learns as it runs. That's what the acronym stand for - the antilogarithm used to make it smart.

The 'training' time will vary with the thermal mass and the heating element wattage.

The only Other problem with your switch idea is: What happens if you don't get back to shut it off? 1200º+ lead temperature.

The SSR/PID is More than capable of heating the pot from cold, IF it is capable of the job to begin with.

[NineInchNails]

The SSR/PID is More than capable of heating the pot from cold, IF it is capable of the job to begin with.

TRUE. Absolutely true. I highly value your input, knowledge and experience. I do appreciate it!

If the toggle switch were simply on & off and nothing were plugged into the PID then it could still be used as a digital thermometer. Yes, as you mentioned previously, an auto-tune may be necessary prior to running it on a pot (or anything else) to be accurate, but that is not all too inconvenient. In effect the 3 position toggle is ... pretty much unnecessary.

I think I may have to simplify my build and the wiring diagram to reflect a simple on & off switch. I still plan to add fuzes too.

Thanks Frozone!

[el34]

Nine, you did an excellent job! It looks very professionally designed and assembled.

And well documented, hopefully will help de-mystify the PID thing for some more folks.

It was my understanding that the autotune function first is put in 'learn' mode by pressing its button for 3 sec until the AT light came on. Then it would calculate the P, I, and D variables based on what it discovers about the time relationships between applying heat and feeling the resulting temp, including the rate of change while heating, and store those values in non-volatile memory. The AT light goes out indicating that learning is finished.

So if that is correct, the stored values won't be altered unless it is manually put back into the learn mode. And if that is correct, it'll always know how to handle your pot. If you want it to control something else it'll have to again be put in learn mode to discover the new situation characteristics.

Lot of ifs there, Frozone knows probably more than many others including me, and might be able to straighten us out.

[el34]

I just bought this off of fleabay for my build hope it is the right stuff...

It's already been pointed out that your kit comes with something other than a thermocouple for measuring the pot temperature. It'll work but has limits as Frozone pointed to. If you go shopping for a TC pay attention to their length (4" fits a Lee 4-20 pot very well) and that it has a stainless steel stem and steel braid over the wires. There are a few different 'types', type K being the most common and lowest cost. If you plan to have a connector between the TC and the controller, be sure to get one specifically for a type K thermocouple or you'll introduce measurement errors. Auber has it for a few $, Frozone probably does too.

[6bg6ga]

I like your thread..

A couple of points ... The probe and having the unit wired normally to provide heat when needed. Forget the switch.

Nice detailed schematic and nice build sheet.

Too bad someone had to come in and advertise his kits for sale on your thread as he has numerous theads to the mention his kits.

Also... You used a very nice PID control unit and it beats the Mypin unit....I have both and the cheaper Mypin doesn't stand up like the

Auberins unit. WWW.Auberins.com

I would advise people to consider the Auber unit over the Mypin unit.

For those not wanting to work with the SS relay it is very simply to drive a relay using the output of the PID control unit and a few

simply parts and a small power supply to power the relay.

[500MAG]

:popcorn:

[6bg6ga]

:bigsmyl2:
:popcorn:

[Frozone]

........
It was my understanding that the autotune function first is put in 'learn' mode by pressing its button for 3 sec until the AT light came on. Then it would calculate the P, I, and D variables based on what it discovers about the time relationships between applying heat and feeling the resulting temp, including the rate of change while heating, and store those values in non-volatile memory. The AT light goes out indicating that learning is finished.

So if that is correct, the stored values won't be altered unless it is manually put back into the learn mode. And if that is correct, it'll always know how to handle your pot. If you want it to control something else it'll have to again be put in learn mode to discover the new situation characteristics.

No, that is true Only if the 'I' and 'D' functions are turned off.
<with a Mypin can you turn 'I' and 'D' OFF, I don't remember if Auber can>

The PID initially sets the values based on time to raise the temp a certain amount, how much the temp continues to rise after heat is turned off, how fast it cools over a certain time period and lastly how long it takes to stop the drop in temp.

That works fine if the thermal mass never changes, But if you actually cast with the pot you would soon develop errors as the lead drained off.
So that is what the 'Derivative' is for, it is the prediction of future errors.

Each time the PID makes a correction (usually 3 times a second) it checks that the 'D' was correct (it also updates 'I').
If an unpredicted rate of heating or cooling is found the PID updates the 'I' and 'D' values, It then changes 'P' to reflect those differences.

This is why it runs within 1º while 'idling' but will start to drift as you cast. It takes a little while for the algorithm to catch up.

A perfect example of this is when you put the sprue back in the pot.
When you start casting, the temp drops a few degrees ( ~ 10).
Then, if you maintain a rhythm, the pot will start to maintain that 1º margin again. When you quit casting ( or significantly slow down),
You can watch the temp rise that same amount (~10º) for the same amount of time before it settles to within that 1º again.

< i decided to expound on this>

'P' = Proportional. The amount of time it takes to raise the controlled device a given amount of degrees. Because this is dependent on the Size of the thing (the thermal mass) to be heated and the wattage of the heater ( and other factors), it is variable IF the mass is variable.

'I' = Integral. Just as it sounds, this is the integration of past errors.

'D' = Derivative. This is the prediction of the next error

The unit makes a correction, based on 'P', it then checks if the correction matches 'D', It then uses 'I' to make a change to 'P' and starts the correction process over.
</>

[NineInchNails]

That makes sense. The PID has to be able to compensate for changes on the fly.

I'll be sure to post another diagram showing a simple on/off switch.

[el34]

Frozone, I truly appreciate the detailed explanation. I am generally familiar with the PID algorithm but never needed a deeper understanding. I guess I don't need it now either, just hook it up and be happy, but I really like knowing what's happening beneath the surface.

From what I thought I knew, the D principally deals with rate of change, as in calculus, as part of predicting a stabilized temp.

So if it is always automatically updating itself, what special thing happens as a result of manually invoking the Autotune function? Its description led me to believe that when it finished, it had discovered and saved control parameters, and would not change them until Autotune was fired off again.

I understand the implications of altering the thermal mass, and that the rules for predicting tight temp control must follow that change.

[NineInchNails]

Not much of an update, but I got around to installing self-adhesive rubber feet on the underside of the enclosure. Now it doesn't slip or slide around at all. It grips to any surface very well. A much needed improvement.

http://i1303.photobucket.com/albums/ag142/boolits/PID%20Controller/PIDControllerRubberFeet_zpsf1513dec.jpg (http://s1303.photobucket.com/user/boolits/media/PID%20Controller/PIDControllerRubberFeet_zpsf1513dec.jpg.html)

[Lights]

It's already been pointed out that your kit comes with something other than a thermocouple for measuring the pot temperature. It'll work but has limits as Frozone pointed to. If you go shopping for a TC pay attention to their length (4" fits a Lee 4-20 pot very well) and that it has a stainless steel stem and steel braid over the wires. There are a few different 'types', type K being the most common and lowest cost. If you plan to have a connector between the TC and the controller, be sure to get one specifically for a type K thermocouple or you'll introduce measurement errors. Auber has it for a few $, Frozone probably does too.

Thanks for the info el34. I am looking for a TC right now for my build. I got some of the K type connectors already.

[Frozone]

I have what you need.

$13.50 shipped to your door.

[NineInchNails]

My SSR just crapped out on me. Very strange.

The PID is sending a 12V signal to the SSR. The indicator light on the SSR is illuminating, but the SSR is just not allowing current to pass through. It has been working flawlessly. Just this past weekend the SSR wouldn't function, but then began functioning. It's like it doesn't want to go and once it starts to work it works fine. It seems to have a hard time getting started.

[el34]

I've killed a couple but only because of over-current. Way over. I doubt you're doing that. Maybe just replace it?

[NineInchNails]

Yeah the eBay seller has a 30 day return policy. I'll send it back to have him replace it and see if this one was just a fluke. I've read that there are a lot of eBay sellers that sell clone copy SSRs and other clone electric components. Supposedly they stop working because they are counterfeit.

If the 2nd one takes a dump I'll probably just order one from Auber. I assume theirs are good quality. It better be for over twice the price + shipping.

[Mal Paso]

Yeah the eBay seller has a 30 day return policy. I'll send it back to have him replace it and see if this one was just a fluke. I've read that there are a lot of eBay sellers that sell clone copy SSRs and other clone electric components. Supposedly they stop working because they are counterfeit.

If the 2nd one takes a dump I'll probably just order one from Auber. I assume theirs are good quality. It better be for over twice the price + shipping.

I don't think clone or counterfeit apply. There are poorly made parts as well as rejected parts that find their way to secondary markets. Auber and Frozone sort this out for you at a nominal cost and you can be fairly certain the parts were Not found in a dumpster.

No slight to those who recycle but it should say so and most often does not.

[Springfield0612]

NIN, just last week I was on eBay and there was an eBay member there outing a seller for selling knock of SSR's. The buyer went to the extent of opening up the unit and posting pictures of the knock off components that were slapped together inside and posted results of the units failing. I'd be very skeptical of most electronic components on eBay sourced from over seas sellers. I've been taken once on a memory chip for my digital cammera a couple years back. Typically the member won't respond to your request, default on the return payment, PAYPAL will automatically refund your money and the seller will disappear to open another account. $5-$10 US goes along way in a lot of those over seas countries.

[NineInchNails]

NIN, just last week I was on eBay and there was an eBay member there outing a seller for selling knock of SSR's. The buyer went to the extent of opening up the unit and posting pictures of the knock off components that were slapped together inside and posted results of the units failing. I'd be very skeptical of most electronic components on eBay sourced from over seas sellers. I've been taken once on a memory chip for my digital cammera a couple years back. Typically the member won't respond to your request, default on the return payment, PAYPAL will automatically refund your money and the seller will disappear to open another account. $5-$10 US goes along way in a lot of those over seas countries.

I know exactly what you mean. I've seen similar accounts myself. That's the only reason I brought up "counterfeit" to begin with. I have no reason to suspect that I received a knockoff, at least not yet. A lot of the knockoff stuff I saw online, you don't have to look too close to spot the poor quality. My replacement should arrive very soon and hopefully it works for a VERY long time. If that one takes a dump then I'll be quite upset.

A coworker of mine feels that the way I wired my SSR could have caused it to stop working. The switch in the down position bypasses the SSR by supplying power to one of the terminals while at the same time the PID can activate the SSR which has power in the other terminal. He feels as though this could be what caused the SSR to fail. I couldn't say one way or the other.

[el34]

A coworker of mine feels that the way I wired my SSR could have caused it to stop working. The switch in the down position bypasses the SSR by supplying power to one of the terminals while at the same time the PID can activate the SSR which has power in the other terminal. He feels as though this could be what caused the SSR to fail. I couldn't say one way or the other.

I need to respectfully disagree. The control input circuit is totally isolated from the power switching circuit by way of an opto-isolator. The control signal lights up an LED whose light turns on a photo-sensitive transistor which in turn turns on the AC power switch, usually back-to-back SCRs. Neither side is dependent on the other.

[NineInchNails]

I need to respectfully disagree. The control input circuit is totally isolated from the power switching circuit by way of an opto-isolator. The control signal lights up an LED whose light turns on a photo-sensitive transistor which in turn turns on the AC power switch, usually back-to-back SCRs. Neither side is dependent on the other.

I'm not sure if were talking about the same thing or not. Not the 12v (the control side) crossing over to the 110v (circuit side). He, my coworker, was suspicious about having 110v current on both terminals of the circuit side meanwhile the SSR often activating the SSR.

Having current on connections #1 and #2 at the same time is what he was concerned about. Connections #3 and #4 is the control side.

http://i1303.photobucket.com/albums/ag142/boolits/PID%20Controller/FotekSSR-40DA01_zpse537b35b.jpg (http://s1303.photobucket.com/user/boolits/media/PID%20Controller/FotekSSR-40DA01_zpse537b35b.jpg.html)

I just received my replacement SSR today! Will hook it up and give it a go ASAP. I don't see how the SSR could be harmed by how I wired it, but I do not claim to have an understanding of how SSRs work exactly.

[Ranger 7]

I used NineInchNails Parts list and wiring instructions and added a In Power and an OUT power led light. Also fused both the in power and out power circuits.
Grounded the case to incoming ground via the terminal board.

Plugged in and works Great!

Thank you to NineInchNails!