Issue with Inkbird ITC308S - anyone else had this?

[peebee]

I got my brand new replacement just after Christmas: Thanks @Inkbird!

It seemed quieter, no obvious pulsing on/off of the compressor. The ITC-308 is controlling the keezer at 6C, and is set to switch on the compressor when temperature reaches 7C (differential set at 1C, both CD and HD).

But I'm not in the conservatory (where the keezer is) much at this time of year. Tonight though I was picking tomatoes for dinner (yeap, I've still got tomato plants!) and the keezer came on; five seconds later it went off! I thought I'd better watch it closely for a time, so got ready to pour some beers (I avoid cold beer at this time of year, I just can't get the fascination with drinking cold beers sitting in front of a fire). Didn't wait long and the keezer was back on: This time I could watch the ITC-308 counting down; 6.7, 6.6, 6.5 ... 6.0 and the keezer turned off again (about 3 seconds). A few minutes later it was doing it again! Within 15 minutes it had pulsed the compressor on-off three times.

The "brand new replacement" has done diddly-squat. Thanks @Inkbird for the rapid replacement of the dud unit. But no thanks for putting the problem right, because it didn't!

 

[ssparks]

I received my replacement today, will hopefully be trialling it next weekend so I hope for better results. If not I may consider trying STC1000 as an alternative.

 

[peebee]

FIX! (Perhaps?).

Having had a new replacement that did not fix the issue, I decided to revisit my previous attempt to fix the issue with a bit of logic. My previous idea wouldn't work 'cos my thinking was off (well, I can always claim my bang-on-the-head accounts for that) but I was heading down the right path.

As fixes go, people are not going to like it! It means doing something quite counter to what one might think "right". If I was @Inkbird I could not suggest this "fix" because it would really annoy people with the issue! It came to me because trying to observe the issue I began to realise the defective device was holding the keezer temperature very close to my set temperature; much closer than the set differential (CD) would suggest. Anyway; enough of that.


The error is the temperature will suddenly jump +1.0 to +1.5°C, triggering the compressor to start up only to stop it a few seconds later as the temperature fall back to correct reading which is below or equal to the set temperature. For me this error seems to occur at least every 10 minutes.

So the "fix":

Configure the set temperature ("TS", for "temperature set") where you want it. For me this is 6°C.

Configure the cold differential ("CD", for "cooling differential value") to desired differential plus maximum error skip. For me this is 1C differential, perhaps a bit high so make it 0.5°C, plus 1.5°C (max. error skip witnessed) which is 2.0°C (told you you wouldn't like it!).


Now if the error occurs while the temperature is below 6.5°C the compressor does not turn on (the temperature must indicate 8.0°C to trigger that). If the temperature is above 6.5°C, the error will cause the compressor to start, but as the displayed temperature falls back to reading normal the temperature will still be above 6.5°C and the compressor stays on until reading 6.0°C.

Voila! I'll come back in a few days to post results; or else post that it has all turned into a bag-of-worms. Not bad for a looney (unfortunately, I'm not kidding!).

 

[Inkbird]

Hi,
If the temperature of the use environment is relatively high, when the cooling device is not working, the temperature may rise to 7℃ quickly. And then Inkbird turns on the cooling output immediately when it feels the change, and the reading slowly drops.
And the following is how ITC-308 works：
The Inkbird ITC-308 temperature controller is an on/off model thermostat and will control a temperature range by turn on/off the heating/cooling device according to your setting.
When temperature drops below desired temp-heating difference value it will automatically switch on heaters until desired temp is reached. The same, if the temperature rises above desired temp+cooling difference value it will automatically switch your cooling devices (could be a chiller or fans).
For example, your temperature range is 5°C ~7°C.
TS (temperature set value) =6,
HD (heating differential value) =1,
CD (cooling differential value) =1.
Then it will turn on heater at 5°C (TS-HD), turn off at 6°C. Turn on cooler at 7°C (TS+CD), turn off at 6°C.
It helps keep an ideal temperature range not a fixed temperature point, hope you know it.
Under the temperature of the use environment is relatively high, if your temperature differential values can not be set lager and you won't it often turn on/off the heating/cooling device, sorry the Inkbird may not meet your needs.
Everyone has different needs for temperature control, we can understand your needs and tensions. In order to reduce your losses on it, you can send the device back to us and we will give you a refund directly.
Regards

 

[peebee]

Hi,
If the temperature of the use environment is relatively high, when the cooling device is not working, ...

Thanks, but not really helpful.

... The error is the temperature will suddenly jump +1.0 to +1.5°C, triggering the compressor to start up only to stop it a few seconds later as the temperature fall back to correct reading which is below or equal to the set temperature. For me this error seems to occur at least every 10 minutes. ...

That is the error I'm fixing! It IS an error in the ITC-308, NOT me, or anyone else here, having too high an expectation of the device. When I said "suddenly jump" I meant instantly not "quickly". And the "fall back" I said happens in a "few seconds", perhaps I should have said 2-5 seconds.

Interestingly, if someone has a ITC-308 but never changes the differential, or "CD" value (set to 2.0°C by default), they may never notice the issue.

The ITC-308 has a design defect. Your message gives the impression that Inkbird has nothing to do with it. That is not the impression that we have all appreciated previously.

 

[peebee]

Interesting development today, but not one that will make significant changes to what I've already said.

My ITC-308 was displaying 6.9°C. Next time the temperature skips it should go above 8.0°C. I elected to wait and see what happens. Not long; it skipped from 6.9°C up to 8.0°C (coincidentally!) in a blink. Nothing unusual, but what happened next was. The compressor started up as expected and I prepared to watch it rapidly "count-down" to 6.9°C. It didn't. The display snapped back to 6.9°C in one skip. It took less than one or two seconds to skip to 8.0°C and return to 6.9°C again. If this is normal, it would explain why I've rarely caught it "counting down".

Of course, this time the compressor wasn't pulsed on and off. It kept running until 6.0°C (the temperature set value) was reached. So the "fix" seems to be working (7.0°C is the highest I've seen it reach).

I shall try to get some logging on the job.

 

[LED_ZEP]

If it's been suggested before forgive me as I've not read the entire thread. Couldn't you use a micro controler (Arduino or the like) between the temp prob and the Inkbird to filter the temp fluctuation out. i.e. it's programed to ignore the change if it's not maintained for at least 5 seconds before relaying the signal to the Inkbird?
I suppose if you're tech savvy enough to know how to do this you would sack the Inkbird and make your own programable relay switch, with built in WIFI so you can monitor the temp from your phone (I have an idea for this lockdown).

 

[chrisb8]

If it's been suggested before forgive me as I've not read the entire thread. Couldn't you use a micro controler (Arduino or the like) between the temp prob and the Inkbird to filter the temp fluctuation out. i.e. it's programed to ignore the change if it's not maintained for at least 5 seconds before relaying the signal to the Inkbird?
I suppose if you're tech savvy enough to know how to do this you would sack the Inkbird and make your own programable relay switch, with built in WIFI so you can monitor the temp from your phone (I have an idea for this lockdown).

This is something that I wondered about as a possible option to be added to the inkbird settings. So you could set it up so that the read temperature has to remain above the trigger point for a number of seconds before the switch is turned on. It should certainly solve the issues a number of us have had.

 

[chrisb8]

Thanks, but not really helpful.


That is the error I'm fixing! It IS an error in the ITC-308, NOT me, or anyone else here, having too high an expectation of the device. When I said "suddenly jump" I meant instantly not "quickly". And the "fall back" I said happens in a "few seconds", perhaps I should have said 2-5 seconds.

Interestingly, if someone has a ITC-308 but never changes the differential, or "CD" value (set to 2.0°C by default), they may never notice the issue.

The ITC-308 has a design defect. Your message gives the impression that Inkbird has nothing to do with it. That is not the impression that we have all appreciated previously.

I have to say that, as helpful as inkbird have been in replacing the faulty units, they really don't seem to have grasped the issue a number of us are having. I even sent a video of the error occurring showing the temperature jumping up in an instant and then counting back down again within a few seconds. They replied with a very similar explanation given above which suggests this is normal functionality and caused by a high external temperature. Whatever they say, 20L of liquid in my brew fridge cannot be heated by 1 degree plus within a fraction of a second and cooled back down again in a few more!

 

[chrisb8]

On the plus side they did kindly replace my unit even though it was out of warranty and they claimed there was nothing wrong with the way it was functioning. For that I cannot fault them. So far so good after a week but I will report back once it has had a good time controlling the brew fridge, especially when I cold crash as that's when it usually goes totally haywire.

 

[LED_ZEP]

I have to say that, as helpful as inkbird have been in replacing the faulty units, they really don't seem to have grasped the issue a number of us are having. I even sent a video of the error occurring showing the temperature jumping up in an instant and then counting back down again within a few seconds. They replied with the exact same explanation given above which suggests this is normal functionality and caused by a high external temperature. Whatever they say, 20L of liquid in my brew fridge cannot be heated by 1 degree plus within a fraction of a second and cooled back down again in a few more!

I've used a cheap one like This for years and never had a "temp spike issue". They're only good for heating as their operating temps are between 16 - 40c but the point is they're dirt cheap and no temp spike. It'll bounce a little between each 10th of a degree but differences greater that are fine. I set mine to 1 degree on/off (21c on 22c off) but it would work fine down to 2/10 of a degree.

 

[Inkbird]

Hi,
Maybe our previous explanation is not clear? An explanation if we can here again.
The working principle of our controller is like this. It is on/off model, so it will help control a temperature range by turning on/off your heating/cooling equipment.
Under the temperature of the use environment is relatively high and HD/CD differential values are only 1C, When the cooling equipment stops, 1C will arrive soon, and then Inkbird will turn on the cooler again to maintain the temperature range you want. It is not the unit error, it may doesn't suit your requirements. If you want to maintain the temperature at a stable point, this is contrary to the working principle of ITC-308. Like the above reply, if the product is not suitable for you, you can send it back to us, and we are happy to give you a refund.
In addition, the thermostat is used to keep the temperature within a stable temperature range, so it uses a stable reading type probe, which will take a little time for the final reading. For example, when the cooler is turned off, the probe has already sensed the change in the surrounding temperature, but because the probe has a delay, sometimes it will not immediately show a slight change. After the reading is updated, the final stable reading will be displayed after about 3-5 seconds. This is why sometimes you will not see a gradual change in the reading.
Hope these can explain your confusion. If you have questions about the Inkbird units, could you please contact us? We will definitely give you a satisfactory solution.
Regards

 

[peebee]

... Hope these can explain your confusion. If you have questions about the Inkbird units, could you please contact us? We will definitely give you a satisfactory solution. ...

There is no confusion. I think what you mean is you hope we are confused; then perhaps we might shut up?

But I do appreciate that you can't admit the ITC-308 has a design fault without withdrawing the unit from sale because you obviously do not have a fix.

The fix I've recommended to readers here is performing okay. Even though I have the cooling differential set to 2.0°C, the fault ensures cooling is switched on when the temperature gets above between +0.5°C and +1.0°C of set temperature. So I have the set temperature as 6.0°C and the fridge it is controlling never gets above 7°C (if the ITC-308 worked correctly the temperature would reach 8°C regularly). Because my "fix" relies on the fault, there is a little uncertainty as to when cooling is triggered on, but an uncertainty of 0.5°C is no problem for me.

I have three defective ITC-308s now, one being a replacement received a few days ago. The defect is fairly predictable so I'm fairly happy with this. Thank you.

 

[Drunkula]

I never liked that when you use the app to change the settings it ignores the compressor delay. And I ran extensive tests on using scenes from the app and if you changed multiple parameters sometimes it skipped some of them. The pro version of the app would just cane the settings back to defaults which could be absolutely disasterous. No idea if that's changed.

@peebee I'm going to run some tests, too. I used to have a timer plug on the extension just to see how long it had run over a few days but I need to come up with something a bit more wily than that. I'm using a freezer as a keezer / fridge with a standard 308 probe in a huge block of blue-tak strapped to a bottle of water, so it's got good thermal ballast. Any huge skips that realistically couldn't happen will trigger the alarm and make a noise and notify on the app.

I'm on it.

 

[Inkbird]

Hi,
Regarding the temperature jump of 1C when the refrigerator is not working, we also gave an explanation in the previous comment.
Our products are used to maintain a temperature range rather than a temperature point, it cannot be as accurate as the kind of instrument in a professional laboratory. It is an on/off type product. If the actual temperature is relatively high and the CD value is very small, it is difficult to avoid it often turning on/off the device. Hope you can understand. And you can try to set the CD value larger to avoid this, hope it can help you a bit.
We also want to make a product that can meet your needs. As we mentioned before, we have been constantly improving our products according to customer needs, but product development takes a long time and cannot be completed within a period of time. We will try our best to do it and hope our products can meet your needs more in the future.
Regards

 

[Drunkula]

@Inkbird I think everyone knows how the product works, but while the standard 308 standard functions exactly as you'd expect, the same as other on/off with a hysteresis range products like an ITC-1000, it's the wifi version that's experiencing temperature spikes momentarily that are causing these problems. It's clearly not expected behaviour and a design feature if other units using the exact same temperature model aren't having problems.

Anyone else having problems thinks that about sums it up?

 

[peebee]

...
Anyone else having problems thinks that about sums it up?

Yes, about sums it up. But! Mine is not the WiFi version. It seems any ITC-308 can be faulty.

And for @Inkbird: Mine does not have a very small CD value. It is set to 2.0°C which was to try and avoid the erroneous spikes, but I still get occasional erroneous triggers (jumping from 6 to 8°C in an instant). This is a fault! It is not me expecting it to be "accurate as the kind of instrument in a professional laboratory". Suggesting the problem is me having too high an expectation is insulting!

 

[chrisb8]

@Inkbird I think everyone knows how the product works, but while the standard 308 standard functions exactly as you'd expect, the same as other on/off with a hysteresis range products like an ITC-1000, it's the wifi version that's experiencing temperature spikes momentarily that are causing these problems. It's clearly not expected behaviour and a design feature if other units using the exact same temperature model aren't having problems.

Anyone else having problems thinks that about sums it up?

Yes this is true but as @peebee says mine was with the non wifi version. I currently have a wifi version on test and will report back when it's had a full brew cycle.

The point they are missing is not that we don't understand how the thing works or have unreasonable expectations - it's that the probe, sensor or whatever cannot keep a steady reading. Unless it is unreasonable to expect a digital thermometer to read within a +2 degrees error? I think not.

 

[peebee]

I ran into some "issues" with my "fix" that means I'm not going to be able to suggest a universal fix that will apply to anyone.

The fault in the ITC-308 (and a fault it is, I can't see how anyone can recommend one of these defective controllers for controlling fermentation and dispensing fridges) isn't a fault that can be tied down, it seems to change overtime. I was using a "CD" setting (cooling differential) of 2C and it was working fine (spikes were about 1C, so the temperature never read more than 1C over the set temperature). But the spikes now seem to be over 2C, so I never see temperatures more than 0.1C over set temperature. So I've had to set "CD" to 3C, but even at this I've never seen temperatures more than 0.4C over set temperature.

Remember, this ITC-308 is brand new, sent as a replacement for an earlier "faulty" ITC-308.

So the faulty spikes can change in magnitude over time. I've not yet established if the spikes will also reduce in magnitude? Worse (good grief, it can get worse?) I had a set temperature at 6C, "CD" at 2C, so cooling shouldn't come on until 8C: I was seeing a "spike" creating a jump in temperature reading to 7C and triggering the compressor (cooling) to come on, i.e. THE ERRORONEOUS TEMPERATURE SPIKE IS SO FAST THE DISPLAYED TEMPERATURE CAN'T KEEP UP!!! That puts a rather big hole in @Inkbird's arguments that we should expect rapid changes in temperature and have the cooling differential too small.

Conclusion: The ITC-308 design is seriously defective. Avoid using them (or any of the very similar looking rebadged controllers). Inkbird will not help, they have concluded the problem is your expectations being too high!

A glimmer of hope? I'm not seeing reports of broken fridges and freezers that could be attributed to the controllers. Perhaps this idea that quickly switching on and off the cooling compressor will quickly break it is a myth built on historical fridge builds? Also I'm not experiencing problems controlling an external cooling unit (a shelf cooler with the ITC-308 probe in the fermenter, over a meter from the shelf cooler); perhaps the fault is some sort of remote (radio) interference between fridge compressor and controller? Or maybe I've got a rare working ITC-308 attached to the fermenter.

Hopefully, when I eventually wire in the (Inkbird) ITC-1000 units, the problem will go away?

 

[peebee]

... So I've had to set "CD" to 3C, but even at this I've never seen temperatures more than 0.4C over set temperature. ...

Slightly less pessimistic today! I thought I was witnessing increasing magnitude of erroneous spikes, but I'm now thinking maybe my original assessment was a bit short of the reality: My original assessment was the maximum spikes were about 1.5C, but this value was derived from rather casual observations and I now think that value (max. spike magnitude) will be closer to 2.2 - 2.5C.

That doesn't make the issue less serious! But it does mean my "fix" might not be a lost cause? I'll keep working at it.