PID Temps fluctuating a lot

[BredStik]

Hi,

I'm setting up an inkbird itc-106vh with a rtd pt-100 sensor and I'm experiencing weird fluctuations once in a while. I shot a quick video to demonstrate the behavior.

https://youtu.be/EVnTj3hTvIc

Is that normal?

 

[BrunDog]

No. Check your connections. Then try a different probe.

 

[whovous]

The three tiny wires from the probe are flimsy as all get out. I've just bought replacement probes for mine. I got two because I expect it to break again.

 

[BredStik]

No. Check your connections. Then try a different probe.

I tried a different probe with the same results... Connections "seem" ok. Probes are pretty flimsy though. Maybe I should try a better one... Any suggestions?

 

[schematix]

what are you doing when it does that?

 

[BredStik]

what are you doing when it does that?

Getting to targeted mash temp (HERMS). It doesn't seem to do it as much when temp is reached (and maintained).

 

[schematix]

Have you tuned it?

Is your flow rate consistent?

 

[BrunDog]

The output is full on as it should be but the temp should not be varying like that. If it were a RIMs tube it would fluctuate only if overshooting. RTDs are low impedance and relatively impervious to noise but it could be an electrical interference. Do you have any connectors or terminals inline? If so try wiring directly to the PID.

 

[schematix]

I missed the constant on OUT light the first time i watched the video. On second thought I'm leaning towards electrical/noise issue.

Can you post a picture of the inside of your panel, making sure the connections for the RTD and controller are visible.

 

[wobblypops]

Did you do a auto tune. Should been done for setting it up to work properly with probe

 

[schematix]

Did you do a auto tune. Should been done for setting it up to work properly with probe

i thought that too at first too but the output was on steady and the temps were erratic... that pretty much rules out tuning.

 

[BredStik]

It is wired directly to the PID.

I've got the PID, a small 12v supply, a fan, a 110v outlet and the SSR crammed into a 6 x 6 x 4 box, so it is a bit overcrowded in there... Wiring is a mess, to say the least. It could very well be interference, but while filming, the probe wiring was out the cover, not touching or crossing much wires.

 

[BredStik]

It's been auto tuned and the pump was running at a steady rate (through another supply)

 

[augiedoggy]

I tried a different probe with the same results... Connections "seem" ok. Probes are pretty flimsy though. Maybe I should try a better one... Any suggestions?

I use these.. the teflon shielded cable is superior to the cheaper stainless braided stuff many of the cheaper probes come with.
https://www.ebay.com/itm/RTD-PT100-...431862?hash=item28339a24f6:g:59QAAOSwaZdZv08r

 

[BredStik]

I use these.. the teflon shielded cable is superior to the cheaper stainless braided stuff many of the cheaper probes come with.
https://www.ebay.com/itm/RTD-PT100-...431862?hash=item28339a24f6:g:59QAAOSwaZdZv08r

Thanks! I might want to order one of these...

 

[jrcrilly]

Bubbles in the liquid would do that...

 

[BredStik]

It seems it does it when holding probe in my hand as well... It doubt it's the issue, thanks for the input!

 

[JONNYROTTEN]

Couldn't it just be from the element on full power and the water moving around in the BK hitting the temp probe at slightly different temps. He says it balances out once at temp. Which is really the only thing that matters

 

[BrunDog]

I don’t agree... the PID calculations are dependent on correct input to calculate accurate error.

I would swap out the PID and see if that is the source.

 

[BredStik]

Couldn't it just be from the element on full power and the water moving around in the BK hitting the temp probe at slightly different temps. He says it balances out once at temp. Which is really the only thing that matters

The probe is located at the end of a HERMS coil, temp is ramping slowly and doesn't change drastically. It does seem to keep my temps OK, but found it weird it fluctuates like that...

 

[BredStik]

I don’t agree... the PID calculations are dependent on correct input to calculate accurate error.

I would swap out the PID and see if that is the source.

I have a REX C100 lying around, might give a try... All that wiring, ugh.

 

[augiedoggy]

I highly doubt its the pid myself from watching the video.
I would remove the probe and test in different glasses of water to see if its accurate and still jumps around.. I didnt watch the video herms temps could possible be somewhat erratic especially if the liquid in the kettle isnt constantly being stirred around well. This is why I went from herms to rims myself.

 

[BredStik]

I played a little bit with digital filtering and it seems it almost completely resolved the issue, albeit affecting the response time a little. On my particular model, it can be set from 0 (off) to 20 (full on). I set it to 3 and noticed an immediate improvement. It still makes some "jumps" but doesn't go back down like it did before.

It might read the following values for example:

80.1
80.2
80.3
81.2
81.1
81.2
81.3

 

[schematix]

Just my 2 cents, I think you have an underlying issue worth sorting out rather than risking a batch.

I like the idea to test in a glass of water of various temps.

These types of devices are inherently very stable. You shouldn't hardly see any noise if its in a cup of water.

 

[BredStik]

Just my 2 cents, I think you have an underlying issue worth sorting out rather than risking a batch.

I like the idea to test in a glass of water of various temps.

These types of devices are inherently very stable. You shouldn't hardly see any noise if its in a cup of water.

I'll give it a try, thanks!

 

[BrunDog]

I agree with @schematix. Something is still not right. Digital smoothing helps but is appropriate for high impedance sensors, not RTDs. Adding it is a bit of lipstick on a pig.

 

[BredStik]

I did some test by isolating the inkbird pid from the other components in my box and got similar results, although since I put the probe in boiling water I only get the sudden jumps up, not down.

I think my probe just has a slow response time. Plunging the room temp probe in boiling water, it took at least 30 seconds to reach 100c displayed temperature, making sudden jumps along the way, as if it had trouble keeping up...

 

[schematix]

That just doesn't happen. The probe should read smooth and steady. Temperature of water doesn't change erratically and neither should your probe. I'm leaning towards a faulty component or incorrect wiring.

FWIW you don't need to do this with 100C water. Room temp and just some heated water is fine.

Do you have another PID, or another probe you can test with? Isolate and test.

Also how is it wired? Are you using any plugs/sockets? How is your panel grounded? Are all your components grounded? Back panel grounded? Need lots of details here. Picture worth a 1000 words...

 

[rashadx100]

I did some test by isolating the inkbird pid from the other components in my box and got similar results, although since I put the probe in boiling water I only get the sudden jumps up, not down.

I think my probe just has a slow response time. Plunging the room temp probe in boiling water, it took at least 30 seconds to reach 100c displayed temperature, making sudden jumps along the way, as if it had trouble keeping up...

I use k type thermocouples and they tend to get worn out at the connection to the probe itself, then it shorts out and gives sporadic numbers. I always have an extra probe on hand just in case. I suggest buy a backup probe and see if u get the same result. Inevitably your gonna need a new one anyways.

 

[rashadx100]

I use k type thermocouples and they tend to get worn out at the connection to the probe itself, then it shorts out and gives sporadic numbers. I always have an extra probe on hand just in case. I suggest buy a backup probe and see if u get the same result. Inevitably your gonna need a new one anyways.