PID control of gas forge

[KenH]

Hello all, There have been some good threads on PID control, and from reading it seems like a gas forge really needs to be a blown burner rather than atmospherically aspirated burner.   With the blown forge, it's the blower speed that gets the PID control - is that correct?  Then a solenoid valve that's also controlled to turn OFF/ON as required?

 

I've got the Auber SY-2362 PID controller to play with - neat piece of equipment.  I tested it today with a small single burner forge using a side arm burner I built from drawings on http://zoellerforge.com/sidearm.html site.  I've been using the gas forge a while, seems to burn good.  Today controlling at 1525, with controller set for OFF/ON control rather than PID, it seems to do "ok" holding ±5ºF range at best - with some larger excursions from setpoint.  These I think were due to a good bit of cool wind blowing into forge - and the thermocouple is exposed directly inside the forge. I've got the ceramic sheath from Auber to protect the thermocouple - I think that might help control temps better.

 

In reading the SY-2362 manual it says PID control shouldn't be used with inductive loads like solenoids, but rather the OFF/ON control should be used.  My question, does anyone use PID on solenoids?  OR is only OFF/ON control used?  I know PID control will cause the solenoid to open/close often and pretty quick.

 

I'm thinking I might wish to add a blower to burner, but do need to do some reading and learning before.

 

Thanks to all for any suggestions and ideas.

 

Ken H>

 

 

[timgunn1962]

Ken,

 

I've not tried PID control on a forge myself. To be honest, I don't think I can see enough benefit to justify the effort for anything except Heat treatment.

 

I've seen PID control used on a Don Fogg-style 55 gallon Heat Treat forge and it was excellent.

 

To be honest, it wasn't very much better at holding a given temperature than a manually-controlled Don Fogg-style 55 gallon HT forge though. 

 

If I was building a gas forge for myself for HT, I'd probably go with the PID-controlled Don Fogg design, simply because I am too lazy to spend the time tweaking the temperature manually and I do enough process control in my day job that the PID stuff is almost second nature.

 

For anyone who makes a living from making and HT-ing stuff, the PID would pay for itself very quickly by eliminating the need to spend the time making manual adjustments each time.

 

It is specifically for heat-treatment though.

 

In your shoes, I would not add the ceramic sheath. If you do, it will slow the response of the thermocouple to temperature changes in the forge; the display will tell you the thermocouple junction temperature is more stable, but the actual workpiece temperature will almost certainly see bigger swings. Been there. Done that.

 

If/when you try it, I'd strongly recommend using a fast-responding thermocouple and a hand-held display to check what is going on before and after you fit the sheath (I wouldn't believe someone I've never heard of telling me stuff that seems counter-intuitive on the internet either).

 

The cycle time on your On/Off control could actually be pretty short, depending on your Hysteresis setting and the thermal mass of your forge. It's worth timing it, as whatever it turns out to be should give a "safe" cycle time for PID control ("Safe" meaning no worse than you have at present; I'd make sure I had, or could quickly get, a spare coil for the solenoid valve and maybe a complete valve).

[KenH]

Tim, Thanks for the input - your comments agree pretty much with what I've found. As in writing sometimes, I didn't directly mention this was for heat treating - note the 1525ºF temperature mentioned.  For forging I don't see any need for PID control, I just crank the forge as high as it'll go with 5 to 7 PSIG. Monitoring with the K type TC it seems about 2100ºF to 2200ºF is about max for my little forge, depending on if I've got both burners going, or just a single burner.

 

I am wondering if the blade that's inside will see the fast swings in temp as would the bare open end of TC - I'd planned on blade lying on a nice heavy steel plate as a heatsink.  I should be able to watch any slight color change of blade to see if it's holding fairly steady or swinging much as the burner is cycling off/on.

 

I see what you mean by Dan Fogg's forge - as a blown forge it should be easier to use good PID control for.  I've got most of the stuff required to add the blower to my forge, but I think I'll wait a bit.  Today I used the PID control on a toaster oven I plan to use for tempering - held about ±2ºF - maybe 3 with the empty oven. Again, I think with some mass inside will help stabilize temps some.

 

While I'm not a PID control expert, I've done a bit in chemical plants of yrs past using stand-alone controllers.  For the last 15 yrs, this control has been with large DCS systems so I never got involved in that.  Just amazes me that I've got a PID controller at home with the same stuff I tuned in chemical plants.

 

Thanks again for the comments and ideas - this sure is a GREAT site.

 

Ken H>