Stc-1000+

[alphaomega]

I'd thought about this, but was concerned about screwing up the look-up table by not connecting the probe directly to the unit.

How did you figure out that it won't throw off the measurements?

Actually, I haven't... I've so far just tested that the main board is wired correctly. But it is a 10k NTC thermistor, so the resistance over the probe changes pretty dramatically With temperature. I can't imagine that the connectors will interfere.

 

[wilconrad]

I'd thought about this, but was concerned about screwing up the look-up table by not connecting the probe directly to the unit.

How did you figure out that it won't throw off the measurements?

The mini xlr is just a connector, the probe is still connected "directly to the unit". Maybe I'm not understanding the question. Are you worried about the resistance of the connector affecting the probe reading? If so, that should be negligible if installed/connected correctly.

 

[day_trippr]

The sensor is likely a 10K thermistor, while a connector resistance will measure in the milliohm range. You could stick a hundred connectors in series with the sensor and not move the needle...

Cheers!

 

[krazydave]

FWIW I've been connecting my probes to my STCs for years now using a 1/8" mono headphone jack. It may not be XLR, but it's the same concept of having a connector inline.

I was a little concerned at first also, so I did some tests when I first built them of comparing the temps through the connector and directly connected to the STC. The temp values were always reading exactly the same either way.
I also have had a second thermometer alongside my STC probe and it's always been right in the same ballpark.

 

[ogunshay]

But it is a 10k NTC thermistor, so the resistance over the probe changes pretty dramatically With temperature. I can't imagine that the connectors will interfere.

After looking at a table for a 10K NTC thermistor, I can see why this shouldn't really be a concern - R at 20 C (68 F) can be on the order of 12K ohms.

On a related note, could someone change the lookup table that is loaded via Arduino? If using another probe, for instance?

 

[alphaomega]

Changing the lookup table for another probe would absolutely be possible, BUT the hardware is designed with the probe as one part of a voltage divider, so you can pretty much only change it to another 10k thermistor (with another curve), so there is not much point in doing so. You may as well use a 10k NTC Beta 3435 1% sensor and save yourself the trouble.

 

[priest77]

I've also ordered this probe because the one that come with my STC was the rubber one, and I think the metal one will be more sensitive, anyway I will compare them...

 

[gueykewl]

For anyone curious, I've now placed two separate orders on Amazon for the STC-1000 from AGPTek ($21.68 with Prime account). Both orders consisted of two STC's and all received were v1.0 (flashable) hardware.

Not saying that you're guaranteed by any means to get the right hardware from them, but I've had good luck so far!

I just purchased another one of these units on Amazon from AGPTek but it was a v1.1 I'll just use this one for my kegerator.

details:
AGPtek Digital All-purpose Temperature Controller STC-1000 w/Sensor
Sold by BrainyDeal

 

[Spellman]

Is there any possibility to add a feature to disable the alarm? I'm wiring up a control box and getting it all set up. Right now I don't have phone jacks wired to all of my temp sensors yet and they are a bit loud! Or is there a way to soft power them off while the alarm is running? I've tried plugging a sensor into the jack, soft powering one off, then pulling it out to plug into another to do the same, but the first one powers back on when I remove the sensor. I should have wired in a manual power switch to each one, but alas I did not think of that until they all started screaming at me!

 

[balrog]

I have a question of settings, folks--what's everyone using for hd, cd, hy? I ask because I finally have a simple setup using a little dorm fridge with 2" rigid foam board around the 5G ale pale, in the cellar, currently using STC-1000+ in simple "th" mode to keep 63°F with the STC's probe taped (insulated) against the pail. All that being said, with a "hy" of 1°F and 2m heat delay, 5m cooling delay, it's cycling about 3-5x/hr, but ranging 62.9°F--64.1°F. How do I with no knowledge balance cycling fridge per hour with hy value? I am presuming going 0.5°F for hy it will just run more?

 

[wilconrad]

Is there any possibility to add a feature to disable the alarm? I'm wiring up a control box and getting it all set up. Right now I don't have phone jacks wired to all of my temp sensors yet and they are a bit loud! Or is there a way to soft power them off while the alarm is running? I've tried plugging a sensor into the jack, soft powering one off, then pulling it out to plug into another to do the same, but the first one powers back on when I remove the sensor. I should have wired in a manual power switch to each one, but alas I did not think of that until they all started screaming at me!

If you're comfortable with a soldering iron, you could remove the speaker/buzzer from the board on each STC. Or you could just finish wiring up the sensors...

 

[mattd2]

If you're comfortable with a soldering iron, you could remove the speaker/buzzer from the board on each STC. Or you could just finish wiring up the sensors...

Or just jump the sensor pins with a 10k resistor in the meantime

 

[alphaomega]

Is there any possibility to add a feature to disable the alarm? I'm wiring up a control box and getting it all set up. Right now I don't have phone jacks wired to all of my temp sensors yet and they are a bit loud! Or is there a way to soft power them off while the alarm is running? I've tried plugging a sensor into the jack, soft powering one off, then pulling it out to plug into another to do the same, but the first one powers back on when I remove the sensor. I should have wired in a manual power switch to each one, but alas I did not think of that until they all started screaming at me!

That is a bug. It makes no sense for alarm to sound when unit is soft off. I'll try to fix that as soon as I can.

 

[Spellman]

If you're comfortable with a soldering iron, you could remove the speaker/buzzer from the board on each STC. Or you could just finish wiring up the sensors...

The thought definitely crossed my mind!

 

[Will_D]

Have run a few tests and I am delighted with the performance.

My one question (prompted by a few posts back re. settings for hy, dh and ch):

I notice that my fridge undershoots a bit so much that the heater then comes on.

One fix maybe would be to allow a much longer 'hd' say from 4 mins to 60

Would this be possible in next version?

We could then tweak the delays to run both the heater/cooler a bit less!

Maybe also a bit more time delay in the debounce logic as sometimes it skips from say pr0 to pr2

But all inj all everyone in Ireland thinks its great. We have a maker fair comming up in July so I will be showing them YOUR brilliant work!

Chhers for now Will

 

[alphaomega]

Have run a few tests and I am delighted with the performance.

My one question (prompted by a few posts back re. settings for hy, dh and ch):

I notice that my fridge undershoots a bit so much that the heater then comes on.

One fix maybe would be to allow a much longer 'hd' say from 4 mins to 60

Would this be possible in next version?

We could then tweak the delays to run both the heater/cooler a bit less!

Maybe also a bit more time delay in the debounce logic as sometimes it skips from say pr0 to pr2

But all inj all everyone in Ireland thinks its great. We have a maker fair comming up in July so I will be showing them YOUR brilliant work!

Chhers for now Will

Hi Will!

Thank you so much for the positive feedback and excellent suggestions for improvement!
First off, hy, cd, and hd settings. I wish I could help you more, but I have not had a chance to actually use STC-1000+ myself. I'm just finishing my fermentation fridge as you might have seen previously in this thread. Intuitively, I would say choose cd to be the minimum 'rest' time you think is acceptable, then choose hy to be the tradeoff of how much swing you can tolerate vs how short cycles you can tolerate. This depends on your setup, so I think you need to get a 'feel' for this setting.

Honestly, I hadn't really though about the hd setting as a means to prevent heater coming on when temperature undershoots (frankly I thought it would be kind of unimportant). If it is to be used like this, then it would need last longer that the entire cooling cycle as well (so yes, probably a lot more that 4 minutes needed). Looking at the code, I can see that it probably would be a pretty easy and cheap fix, so I will definitely add this to the TODO list (I'll create an issue for it).
Related to this, I wonder if BOTH delays maybe should be reset when either relay is turned off. This might be a bit confusing, but kind of makes sense when you think about it. The same situation could occur if your fridge is in a place colder than desired ferm temp, then you might not want to turn on fridge directly on an overshoot.
What do you guys think?

Since keypress/menu logic is moved to its own timer, increasing debounce time a bit should be a relatively easy fix. I'll look into that as well.

Again thanks for excellent feedback/suggestions! This was very helpful info!
And I can't thank you enough for bringing STC-1000+ to the maker fair!!! That is excellent marketing! Thanks so much for spreading the word! If at all possible, I would absolutely love to see a few pics.

Cheers!
//mats

 

[balrog]

Or just jump the sensor pins with a 10k resistor in the meantime

I like this idea. But does anyone need/see need for/use the buzzer on these?

I would say choose cd to be the minimum 'rest' time you think is acceptable, then choose hy to be the tradeoff of how much swing you can tolerate vs how short cycles you can tolerate. This depends on your setup, so I think you need to get a 'feel' for this setting.

Hm. I'm not sure I know this "feel" thing of which you speak. I mean, how much cycling is "too much" for a little cube dorm fridge? I hate to find out by destroying it through over-cycling, but will have to data log it to even know cycle rate and duration.

As it stands now, with my simple setup, I'm in the fairly stable cellar, room temp is just above my setpoint, and my little cube fridge is small enough that it doesn't overshoot by more than 0.1°F to 0.2°F so my SP=63, hy=1 looks like it's swinging between 62.9 and 64, about 1 hr regular cycle.

Related to this, I wonder if BOTH delays maybe should be reset when either relay is turned off. What do you guys think?

Yes definitely on this--I hadn't thought about it but it makes sense to me that any delay is reset when any relay has turned off. Machine state finishes a cycle (heat/cold), dealys are reset to 0, nothing happens for hd/cd.

 

[helibrewer]

In my system, I put the probe in about 1 cup of water, that amount seems to be good to act as a buffer and help prevent the overshoot problems. Cycling is just a factor of how well your fridge is insulated...if it loses heat rapidly, it will cycle a lot. You can try a longer delay setting.


Sent from my iPad using Home Brew

 

[alphaomega]

I like this idea. But does anyone need/see need for/use the buzzer on these?

Hm. I'm not sure I know this "feel" thing of which you speak. I mean, how much cycling is "too much" for a little cube dorm fridge? I hate to find out by destroying it through over-cycling, but will have to data log it to even know cycle rate and duration.

As it stands now, with my simple setup, I'm in the fairly stable cellar, room temp is just above my setpoint, and my little cube fridge is small enough that it doesn't overshoot by more than 0.1°F to 0.2°F so my SP=63, hy=1 looks like it's swinging between 62.9 and 64, about 1 hr regular cycle.

Yes definitely on this--I hadn't thought about it but it makes sense to me that any delay is reset when any relay has turned off. Machine state finishes a cycle (heat/cold), dealys are reset to 0, nothing happens for hd/cd.

Personally, I'm no fan of the buzzer. I don't really see the point of it, but hey, it's there.

The 'feel' thing I'm referring to, is that the amount of cycling will be dependent on what fridge you have and your batch size etc. I would say keeping an eye on it during active fermentation and just make a guesstimate if the cycling is too frequent could be good enough.

I'm not sure what you mean by 'nothing happens for hd/cd'. But I guess you agree on that the proposed change would be a good idea?

In my system, I put the probe in about 1 cup of water, that amount seems to be good to act as a buffer and help prevent the overshoot problems. Cycling is just a factor of how well your fridge is insulated...if it loses heat rapidly, it will cycle a lot. You can try a longer delay setting.

Sent from my iPad using Home Brew

The consensus is that attaching the probe to the side of the FV and covering with insulation is best practice (short of using a thermowell). You want to measure beer temperature, not fridge temperature. Also, submersing the probe directly is probably not a good idea.

 

[helibrewer]

For fermentation I use a thermowell, for my serving fridge I use the small water sample, I have found too much of a difference using the insulation method.


Sent from my iPad using Home Brew

 

[uatuba]

Keep in mind that these were originally intended to be sold as aquarium temperature controllers. As such, there are two things we should realize: the alarm makes sense and the probe is submersible.


Sent from my iPad using Home Brew

 

[alphaomega]

For fermentation I use a thermowell, for my serving fridge I use the small water sample, I have found too much of a difference using the insulation method.

Sent from my iPad using Home Brew

That makes a lot of sense.

Keep in mind that these were originally intended to be sold as aquarium temperature controllers. As such, there are two things we should realize: the alarm makes sense and the probe is submersible.

Sent from my iPad using Home Brew

That also makes a whole lotta sense. However, I still wouldn't trust submersing the probe for extended periods. I've had water creep into probes before and it can be a sneaky error. Even if they're supposed to handle it, I'd avoid it just to be on the safe side.

 

[alphaomega]

Ok, I have been working a bit on the stuff mentioned earlier and have pushed a new version to the work branch.

There are some changes (that also concern thermostat function) that I want to 'mature' before make an official release.

So, what's new?

• The 'power off' function is reworked. Alarm should not be a problem any more when in power off mode. This is a bigger change than it might seem, so I'd be glad to have some help testing. When powering on again, a reset is forced.
• Allow longer heating delays.
• Reset both cooling delay and heating delay counters when either heating or cooling cycle ends. Hopefully this could help avoid unnecessary cycling on over/undershoots by setting appropriate cd/hd values.
• Increased button debounce time. Took it as far as I felt comfortable (i.e. without buttons feeling 'sluggish'), unfortunately button acceleration is not quite as smooth anymore, but selecting profile to run should be a bit easier.
• Added a check when writing EEPROM to avoid unnecessary writes.

Please help out testing if you can.

Cheers!
//mats

 

[balrog]

Ok, I have been working a bit on the stuff mentioned earlier and have pushed a new version to the work branch.

There are some changes (that also concern thermostat function) that I want to 'mature' before make an official release.

So, what's new?

• The 'power off' function is reworked. Alarm should not be a problem any more when in power off mode. This is a bigger change than it might seem, so I'd be glad to have some help testing. When powering on again, a reset is forced.
• Allow longer heating delays.
• Reset both cooling delay and heating delay counters when either heating or cooling cycle ends. Hopefully this could help avoid unnecessary cycling on over/undershoots by setting appropriate cd/hd values.
• Increased button debounce time. Took it as far as I felt comfortable (i.e. without buttons feeling 'sluggish'), unfortunately button acceleration is not quite as smooth anymore, but selecting profile to run should be a bit easier.
• Added a check when writing EEPROM to avoid unnecessary writes.

Please help out testing if you can.

Cheers!
//mats

Will grab tomorrow and take a look. What I meant in prior post about HD/CD reset is exactly what you implemented I believe. Either delay does not start counting until any relay turns off. I had thought the factory default STC method was to start the delay counter when cycle was requested--if cooling was called for, start cooling delay, but maybe the delay really started at the end of cooling cycle.

 

[rjmaster19]

Extremely cool project. And thanks a bunch for sharing.

 

[Spellman]

I'll see if I can look these changes over tomorrow as well. I got all of my sensors wired up today and noticed one is about 5-6 degrees lower than the others. I haven't calibrated them yet and I suspect they may be a reading a degree or two higher than they should, but I noticed that the compensation limit is 5*, is it possible to change that to a higher value? I know most people won't need that much, and after proper calibration I might not either, but it could be useful to some people. Thanks again for all of your hard work!

 

[Spellman]

One minor typo in the manual on page 13:
SP Set setpoint 40 to 140°C or 40 to 25°0F

 

[wilconrad]

Ok, I have been working a bit on the stuff mentioned earlier and have pushed a new version to the work branch.

There are some changes (that also concern thermostat function) that I want to 'mature' before make an official release.

So, what's new?

• The 'power off' function is reworked. Alarm should not be a problem any more when in power off mode. This is a bigger change than it might seem, so I'd be glad to have some help testing. When powering on again, a reset is forced.
• Allow longer heating delays.
• Reset both cooling delay and heating delay counters when either heating or cooling cycle ends. Hopefully this could help avoid unnecessary cycling on over/undershoots by setting appropriate cd/hd values.
• Increased button debounce time. Took it as far as I felt comfortable (i.e. without buttons feeling 'sluggish'), unfortunately button acceleration is not quite as smooth anymore, but selecting profile to run should be a bit easier.
• Added a check when writing EEPROM to avoid unnecessary writes.

Please help out testing if you can.

Cheers!
//mats

I flashed with 1.05 this morning and like the new debounce time, nice improvement! Button acceleration still works well enough, so the tradeoff you made here was a good one, in my opinion. Will let it run for a while and see if any issues pop up, but so far, so good. Thanks Mats!

 

[Spellman]

I flashed mine as well and love the ability to turn off without having to have a sensor plugged in, another great feature! I haven't had much time to test much else yet. Been hard at work making a fridge to house my coolant reservoir.

 

[disney7]

Alpha, I'm still on version 1.04, so I don't know if the following behavior has been changed in 1.05.

I finished a fermentation a few weeks ago and turned the STC+ off using its power button. I then unplugged it and set it aside.

When I went to use it again, I plugged it in and the display flashed, "-8", and went back off. I unplugged and plugged it back up a few times and got the same result. At that point I thought the STC had gotten fried by a power surge, cosmic rays, etc.

Then I remember that I had turned it off earlier. Plugging it back up and holding down the power button to turn it back on took care of things.

Is there any way to indicate that it is in the off state when it is plugged up? Maybe flash "OFF" for a couple of seconds?