TETB kegerator build

[Agentgonzo]

As the heaters would be turning on and off very frequently

Why do you need the heater to be turning on/off very frequently? Given it's low power, and there is a high thermal capacity (in your fridge), you could turn it on for minutes at a time etc, rather than fractions of a second.

My rudimentary heating setup is just with a smart plug that I turn on for a fraction of an hour, then off again at the appropriate time. Then turn on at the next hour/half-hour

 

[The-Engineer-That-Brews]

Why do you need the heater to be turning on/off very frequently? Given it's low power, and there is a high thermal capacity (in your fridge), you could turn it on for minutes at a time etc, rather than fractions of a second.

My rudimentary heating setup is just with a smart plug that I turn on for a fraction of an hour, then off again at the appropriate time. Then turn on at the next hour/half-hour

Yes that kind of setup would work fine
In fact I wanted to keep everything in the same box, and those little PCB mounted SSRs are quite a lot cheaper, more compact and a lot easier to control than a mechanical relay would have been

 

[Agentgonzo]

Yes that kind of setup would work fine
In fact I wanted to keep everything in the same box, and those little PCB mounted SSRs are quite a lot cheaper, more compact and a lot easier to control than a mechanical relay would have been

Ok, so the very frequent on/off is a consequence of the design, rather than a driver of the design?

 

[The-Engineer-That-Brews]

A small problem

All plain sailing so far, but when I powered up the box to test it one of the fridges had stopped working. Hmmm.
A bit more investigating revealed that I'd lost a couple of the transistors powering one of the fridge relays - not good.
It took quite a while to properly identify the culprit: a tiny solder hair:

Unfortunately due to Sod's Law the effect was to connect the unregulated 12V supply to a part of the circuit where it definitely wasn't wanted.
Ho hum... that moment's inattention probably cost me a day but it did give me the opportunity to swap the bipolar drive transistors with slightly more suitable MOSFeTs.

New circuitry tested and the hardware side of the controller done.

Next step: software update.

 

[Rodcx500z]

I love your attention to detail, my effort is a tube heater and inkbird which works fine for me as electrickery is not my forte, i can't see electricity but i know it will hurt me

 

[gavp1979]

I always read these builds in awe and wish I understood them better, the software side of things I am fine with but electronics and circuitry is not my thing

 

[Buffers brewery]

Loving the complexity TETB! When I was building my model railway (nothing to do with brewing ) I experienced the same issues with some PCBs I built for point control. Ended up adopting a routine of after "finishing" a board I scraped the gaps between the tracks with a watchmakers screwdriver. I was just switching 12v relays so used Darlington pairs to switch my relays

 

[The-Engineer-That-Brews]

Software Update

Fridge control needn't be complicated...

But where's the fun in that. Anyhow there can be one or two problems with such a straightforward approach.

Firstly the temperature of a full keg or FV takes quite a while to respond to changes in the surrounding air temperature.

So if the beer's more than a degree or two above the set temp the fridge may run continuously for a long time. That means the air temp inside it can get seriously low, causing icing up and possibly even freezing the beer lines. It also means the condenser and compressor can get pretty hot - domestic fridges aren't built for a 100% duty cycle.

Secondly unlike a traditional mechanical thermostat, the temperature measurements from a digital sensor tend to jitter a bit.

So when the temp is very close to the set point the controller sees lots of 'too warm' / 'not too warm' in close succession and tries to turn the fridge on and off very fast. This is a really bad idea because the compressor takes a much higher current when it's starting up and will rapidly wear out if you don't run it for a bit once it's going.

For these reasons when I wrote the V1.0 software I added a couple of timers that limited the maximum and minimum time the controller would run the fridge for without a bit of a rest. I also added a 'min air temp' setting that stopped it taking the fridge too cold.

That worked pretty well but I realised after a bit of experience using it that what I really needed was a way to make the fridge temp track the temp of the beer with a controllable offset.
Sort of like this (timers etc not shown):

By the time I'd realised this however one or other fridge was in pretty much continuous use so I didn't feel much like messing with the controller.

Adding the heaters into the mix gave an opportunity to implement this idea and at the same time add a similar algorithm to 'track' the temperature of the heat belt to a certain offset above the temperature of the beer. Also an additional timer was needed to prevent the controller from switching too quickly from heating to cooling and getting into some kind of death-spiral; and some safety lockouts to behave sensibly if one or more sensors failed.

Implementing this this got quite complicated and I ended up writing it as a state machine.
Anyone who's sufficiently interested can see the relevant bit of the code on GitHub here.

I should add that (as usual) the most complicated part of the whole exercise was making a sensible user interface that worked on a 20 col 4 row LCD display with a single click-turny knob

Next... hacking the fridges (again)

 

[Agentgonzo]

I know you're talking about cooling, but given you have a heat-belt as your heat source, and (I think) you're measuring the temperature of the beer with an iSpindel or similar, rather than the air temperature in the fridge, why don't you wrap the heat belt around the fermenter and heat the fermenter directly rather than heating the air inside the fridge? (unless I've mistaken somehing)

 

[The-Engineer-That-Brews]

I know you're talking about cooling, but given you have a heat-belt as your heat source, and (I think) you're measuring the temperature of the beer with an iSpindel or similar, rather than the air temperature in the fridge, why don't you wrap the heat belt around the fermenter and heat the fermenter directly rather than heating the air inside the fridge? (unless I've mistaken somehing)

Nope - using separate temperature sensors at all points: stay tuned and hopefully all will become clear ;-)

 

[The-Engineer-That-Brews]

Hacking the fridges (again)

With the controller done the next job was to install the heaters in the fridges themselves. Although one fridge is normally just for dispense I decided to put a heater in it anyway so I could press it into service as a secondary fermentor if needed ... or ... mulled wine, anyone?

It was a pretty straight forward choice to put the heater power cables through the top of the fridge and out at the back like the other connections. In this case though I decided not to add connectors because of the potential for water spills etc in that area. Fortunately the heaters had decent enough length of cable to reach the controller without having to be extended. I replaced the moulded plugs with 3pin IEC connectors to match those on the controller.

That was just a simple job of cutting the moulded 13A plug off and feeding the cable through from the inside. Once I'd done that though I realised that the belts didn't really fit well on the FVs.
The problem was that the belts are made like a loop, with the power cable threaded through a hole at the opposite end:

The power cable is rubberised so it grips OK - but in practice it was quite hard to get the belt properly tightened up and it was obvious that with a taper-sided bucket it was quickly going to work loose and drop down like a pair of badly elasticated doodahs.

Fortunately a while ago I was snagged by one of those 'other items you might be interested in' while doing a bit of online component shopping, and bought an assortment of various different sizes of springs. When they turned up it was a case of Hmmm... they're very shiny, but what on earth will I ever use them for? But now I dug them out and added a sort of spring and hook arrangement to keep the belt taut. This did of course however entail removing the new plugs, taking the lids off the fridges and re-threading the cables *sigh*.

The heating belt does also provide a convenient way to mount the temperature sensors (these are DS18B20 one-wire digital sensors):

As you can see, I also recently decided that I absolutely must have a labelling machine.

 

[The-Engineer-That-Brews]

@Agentgonzo the current version of the controller uses three sensors:

1. air temp (left, taped to fridge wall behind the FV)
2. heat belt surface temp (middle, zip-tied to the surface of the belt under a foam pad)
3. beer (right, held against the side of the FV under another foam pad).

I suppose there's nothing stopping me from dangling a sensor in the beer itself, but I'd have to get one with a food-safe silicone cable and those are a bit hard to get hold of :-)

 

[Agentgonzo]

Things look so much better if they have labels and are colour coded

 

[Agentgonzo]

I suppose there's nothing stopping me from dangling a sensor in the beer itself, but I'd have to get one with a food-safe silicone cable and those are a bit hard to get hold of :-)

From what other members on the forum are saying, sticking your temp probe to the outside of the fermenter is plenty good enough.

 

[The-Engineer-That-Brews]

So now it's all in operation, and I'm currently fermenting in both fridges.
There was a slight iffy moment when I realised I'd got the heaters the wrong way round... but fortunately I was able to spot that and swap the plugs over straight away (phew).

In case anyone's wondering, the ^ symbol on Fridge 1 means 'heating' and the * on Fridge 2 means cooling - there are various other symbols too, but I can't remember what they mean!

 

[Buffers brewery]

There was a slight iffy moment when I realised I'd got the heaters the wrong way round

Looks like you need to press your labelling machine into service again

 

[Buffers brewery]

Waiting for you to get a 3-D printer

 

[The-Engineer-That-Brews]

Waiting for you to get a 3-D printer

Do you know what, I’m not that tempted. I’ve got a friend who’s very into making animated model StarWars robots and he‘s had about four different 3D printers before he got one that actually prints reliably…

 

[Buffers brewery]

Do you know what, I’m not that tempted. I’ve got a friend who’s very into making animated model StarWars robots and he‘s had about four different 3D printers before he got one that actually prints reliably…

Watch this space

 

[The-Engineer-That-Brews]

I love your attention to detail, my effort is a tube heater and inkbird which works fine for me as electrickery is not my forte, i can't see electricity but i know it will hurt me

To be fair it probably comes across better planned and executed condensed like this into a blog… In reality there was quite a lot of head-scratching and it was all spread out over a few weeks :-)