Kiln Performance / SSR vs Direct Wired

I have recently completed my osPID project and was testing it out on my kiln. The controller works as expected, so it is not in question. I noticed that when the kiln is wired to the ssr, even though the PID output is 100%, it does not reach desired temperatures as fast as it does when it is wired directly to mains. Is it typical for an SSR to have high internal resistance?

Specs: Kiln runs on 240V mains. Each leg of 120V is connected to neutral. I have two SSRs for each leg connected to my PID controller (based on osPID). SSR1 - RS3-1D40-21 SSR2 - Crouzet GN 84131120 SSRs are connected to aluminum head sink. Fan is blowing on heat sink/ssr.

When the kiln is wired directly to mains, it can reach over 800C in under an hour. When it is wired through the SSRs, it is only around 600C after an hour, even though the PID output is 100%.

Are the SSR the incorrect type? From what I can tell on the datasheets (http://www.cpu.com.tw/kh/m/ssr/gn-series2.pdf and http://www.nteinc.com/relay_web/pdf/RS3.pdf), they are both zero-cross type, which is what I think I should be using.

Please let me know of any other details I can provide if this is unclear.

Add SSR in parallel and control with the same signal. This will double the current capability and should get the oven to 800 deg in an hour.

Is it typical for an SSR to have high internal resistance?

No. Typically less than 2vac across the AC SSR terminals when on.

Measure it with a voltmeter when ON.
If the voltage is high DC, one of the internal SCRs is open.
If the voltage is always 120vac, then that SSR is open.

Replace the bad SSR.

Have you measured the AC voltage across the kiln heating elements? What is the difference when connected directly, compared to connected through the SSRs?

The data sheet for the RS3-1D40-21 says on-state voltage drop should be 1.6 volts. Not sure about the other unit you have.

Is the kiln wired for 240V? You COULD switch only one side of the 240V as a test.

How much current does the kiln draw?

What are you doing with the kiln? Are you planning to control rampup, soak, rampdown etc for some reason?

kellysumrall:
SSR1 - RS3-1D40-21
SSR2 - Crouzet GN 84131120

both those have a voltage drop of about 1.6V

When the kiln is wired directly to mains, it can reach over 800C in under an hour. When it is wired through the SSRs, it is only around 600C after an hour, even though the PID output is 100%.
Well it cannot be driving the SSR’s at 100% since the power loss in the SSRs totals a
few percent, not enough to account for the loss in performance.

Hi, what is the power rating of the heating element? Have you got heavy enough wiring in your project to carry the current required. Can you please post a copy of your circuit, in CAD or picture of hand drawn circuit in jpg, png or pdf format and a picture of your project?

Tom.... :)

I have attached a basic schematic showing the resistance for the two scenarios, without the ssr and with the ssr. All readings taken with the kiln unplugged and the readings taken from the plug prongs itself. The ssr was powered on by setting the temp controller to a set point higher than ambient temperature and the output duty cycle was at 100%.

raschemmel:
Both SSRs are being powered in parallel using one control wire.

terryking228:
I’m not sure how much current the kiln draws. The manufactures plate on the kiln indicates 30 amps, but, if I’m calculating correctly, a 5 ? element will use 22 amps. Since there are two elements, this would indicate 44 amps. I’m not sure how to take a direct reading. My multimeter only reads up to 20 amps.

I am planning on using ramp/soak profiles for controlled firing of ceramics and glass. Since the PID temp controllers I have purchased on ebay didn’t have the ability to wait for a temperature threshold to be crossed before moving to the next step, I decided to take this adventure with Arduino and osPID.

Thoughts:

The Kiln has a very long time constant. It may take 4 days to totally cool to room temp, right? Are you sure you are starting at the same 'cold' temperature in your measurements?

Can you measure the AC voltage across the SSR's when they are turned on? What is that value?

Then temporarily apply 5V directly to the SSR control inputs. Is the voltage drop across them the same?

Later, keep us up to date on your temperature profile work. I plan to automate a kiln this winter (November in USA) and I had done some work a few years ago with people who were using PC's to control kilns, and played with making a Kiln Control Language (KCL) to be able to create and save profiles for Warm Glass etc.. I'd like to get back to it. It would be nice to be able to edit a text file with the profile definition and then download that to an Arduino/Mega etc.

terryking228:

terryking228: The Kiln has a very long time constant. It may take 4 days to totally cool to room temp, right? Are you sure you are starting at the same 'cold' temperature in your measurements?

This is a 1970s Paragon top load kiln. It looses quite a bit of heat at the moment (will be reworking it). It will drop from 1050c to about 100c in about 8 hours.

terryking228: Can you measure the AC voltage across the SSR's when they are turned on? What is that value?

I'm not sure how to go about doing this without dismantling some of the kiln control paneling. I'm willing to if it helps resolve this ssr delimma.

terryking228: Then temporarily apply 5V directly to the SSR control inputs. Is the voltage drop across them the same?

I took a reading across the ssr while uncoupled from the kiln and came up with the same readings (4.9 K? and 73K?).

As for your last thought about your planned future work, check out http://www.ospid.com/blog/. This may give you a head start on your project. It does load a text based profile from a separate interface. I've been very happy with the code.

The SSR's shown in the schematic may be in parallel if you want to consider them as being on the same circuit as opposed to on different circuit breaker lines (do they both go off with the same breaker) , but that is not what I meant. What I meant was DOUBLE the number of SSR's that you have. There can only be two reasons for the difference in the time it takes to heat the oven. A- The oven does not draw enough current due to voltage drop across SSR. B- The voltage across the one or both of the heating elements is lower than without the SSRs

Since I (A) = V (V) / R, Increasing R by adding the SSR both decreases the current and reduces the voltage across the load (by 1.6 V according to Mark) because you are adding an equivilent series resistor between the source and the load. This added resistance drops voltage and decreases the current. What I meant by add SSRs in parallel is add two more SSRs , one in parallel with each of the existing ones. Two series resistors in parallel have half the resistance of one, consequently you will increase the current and since the voltage drop is a function of the resistance they will drop 0.8V instead of 1.6V.

raschemmel: What I meant by add SSRs in parallel is add two more SSRs , one in parallel with each of the existing ones. Two series resistors in parallel have half the resistance of one, consequently you will increase the current and since the voltage drop is a function of the resistance they will drop 0.8V instead of 1.6V.

OK, I now understand what you are saying. The small 1.6 voltage drop shouldn't be causing the temperature discrepancy (1064 c after 3 hours without SSR vs 650 c after 3 hours with SSR). I'll measure the voltage drop across the terminals to be sure it is 1.6...

...OK I took a reading across the SSR output terminals with the kiln plugged in. I set the temp controller to turn on the SSRs and LED indicators on both SSRs are on. The RS3-1D40-21 measured 1.09 V across output terminals while the Crouzet GN 84131120 measured 122.5 V. I'm starting to think the Crouzet GN 84131120 is the cause of the problem. Since I don't know enough about the internals, I can't rule out the possibility that I'm not using it correctly. Research tells me that a failing SSR will likely fail in a constant ON state. This is not the case. I think I'll run down to the local electronics outlet and get a replacement for the Crouset. I'll let you know the results.

It appears you have answered your own question. Clearly an SSR measuring 122V across it is in the OFF state or just plain OPEN internallly. I would suggest jumpering across the bad SSR and rerun the test. You still need to replace it but I think you know what to do now.

I actually had a spare SSR that I had forgotten about. I replaced the Crouzet GN 84131120 with the spare and fired up the kiln. The new SSR (doesn't have a part number other than SSR-50DA) measured 1 V drop across output terminals. I test fired the kiln for about 30 minutes and it was performing as well as it did when not using the SSRs (reached almost 700 C).

In summary: I measured the voltage drop across the SSR output terminals while the kiln was firing. One SSR measured 1.09 V and the other measured 122.5 V. I replaced the SSR that measured 122.5 V and took another reading with the new SSR. This one measured 1 V drop. The kiln is doing much better now.

Thank you all for your help.

You get all the credit . ( or most of it …)
Everyone did their job.

Hi, that SSR that you replaced is actually two SCR in parallel but back to back to switch full AC. I'd say it has one failed SCR or failed driving circuit, so it was only acting as rectrifier when turned on, not a switch.

Tom.... :)

TomGeorge: Hi, that SSR that you replaced is actually two SCR in parallel but back to back to switch full AC. I'd say it has one failed SCR or failed driving circuit, so it was only acting as rectrifier when turned on, not a switch.

Based on the comment, the bad SSR is only using half the ac wave. That would explain why I would see the element glowing orange, yet, not as bright orange as the other element.

Thanks for the comment Tom.

Half a wave is better than no wave... :D

And significantly better than some gestures....

Actually I was thinking about this problem, you could back up the SSR's with relays, arranging to switch SSR on just before relay, and relay off just before SSR. That way for most of the time the voltage drop of the SSR's is avoided since the relay has much less drop (I hope), and the sparking on the relay contacts is reduced as they only deal with switching 1.6V, thus prolonging the contact life by a substantial factor.

Then you would not need to fan-cool the SSRs (in fact they probably wouldn't need heatsink).

Really glad you got this working!

How hot do the SSR's get?

The diagram you show has the two 5 ohm heater sections in series. So 10 ohms across 240 volts ~ 24 amps.

24A * 1.6V = 38 watts of heat dissipated in the SSRs. So a decent heatsink.

The idea of using a relay across the SSR contacts is possible. Maybe relays like THIS: (DISCLAIMER: Mentioned stuff from my own shop... :-) )


As for your last thought about your planned future work, check out http://www.ospid.com/blog/. This may give you a head start on your project. It does load a text based profile from a separate interface.

I'm looking forward to that. I want to do a low-cost Arduino example. And I'm interested in getting back to KCL (Kiln Control Language) especially for doing Warm Glass (Fused glass and slumped glass) control profiles. I designed a couple of specialized computer languages at IBM and those were my most favorite projects. It was weird months later to walk into a lab and look over someones shoulder and realize they were writing code in "my" language. One of the concepts in my "language" was to separate the data-driven part into Table Editors, and keep the Procedural Part as a free-form language. I am working on applying that concept to Home Automation. But, as my wife, children and grandchilden can attest: Terry Digresses. :cold_sweat:

terryking228: How hot do the SSR's get?

I'm not sure. I'll need to check next time. I also need to use some pyrometric cones to calibrate the thermocouple. One thing I know for sure, the kiln leaks heat and the garage gets very hot.

In case you or anyone else is interested, here is a link to the wiring of my kiln (A82B). http://www.paragonweb.com/files/wiringdiagrams/A82WD.pdf I am replacing the kiln sitter with the PID device.