Coolmaster

[vlbaby]

recharge,
I think your problem is not hystersis. The circuit has clearly had a set level of hysteresis designed into it already ie 33k resistor.
I think you'll find that the fridge itself, being an inductive load, is causing a power supply spike that is reflecting into the mains and into your little circuit. One easy way of confirming this would be to connect the circuit to a simple incandescent light bulb. The light bulb is a resistive load and wont induce any spikes at all.
Try heating up the sensor slowing and see if your light bulb flutters as it switches on, my bet is it wont.
If i'm right, then the only way to stop this is either fit a supressor to the fridge, or better line filtration to your circuit.
My own temperature controller i built suffers a similar problem of which i am to fix one day. My march pump starting is enough to set mine off crazy.

hope this doesnt confuse things more <_<

 

[Uncle Fester]

ambient temp in brew room currently 15 freezer set at 12)
[\quote]

Ok you got me curious...

Where the hell in Western Australia do you find a room with an ambient temperature of 15 degrees at this time of year?

Do you reside in a morgue??? :blink:

M

 

[recharge]

ambient temp in brew room currently 15 freezer set at 12)
[\quote]

Ok you got me curious...

Where the hell in Western Australia do you find a room with an ambient temperature of 15 degrees at this time of year?

Do you reside in a morgue??? :blink:

M

[post="102911"][/post]​

OK was a cold day in the hills and my brew room is made of 50mm cool room panel i have huge trees blocking all morning sun plus there is an airspace of 200 to 300mm between brew rooom roof and leanto roof above it it gets midday sun (but not on this day as there wasnt any) but very little afternoon sun as it is attached to eastside of a much larger insulated garage. Hence it is usually always much cooler inside than out.(unless of course im boiling large amounts inside.

 

[recharge]

recharge,
I think your problem is not hystersis. The circuit has clearly had a set level of hysteresis designed into it already ie 33k resistor.
I think you'll find that the fridge itself, being an inductive load, is causing a power supply spike that is reflecting into the mains and into your little circuit. One easy way of confirming this would be to connect the circuit to a simple incandescent light bulb. The light bulb is a resistive load and wont induce any spikes at all.
Try heating up the sensor slowing and see if your light bulb flutters as it switches on, my bet is it wont.
If i'm right, then the only way to stop this is either fit a supressor to the fridge, or better line filtration to your circuit.
My own temperature controller i built suffers a similar problem of which i am to fix one day. My march pump starting is enough to set mine off crazy.

hope this doesnt confuse things more <_<

[post="102910"][/post]​

Thanks for reply vlbaby
I will still try with a fan or light globe to see.
Any more info on a supressor or line filter or if you could point me in right direction would be appreciated

Cheers
:beer:
Richard

 

[recharge]

I've been fiddling now for a week.Blown 4 triacs and solved the problem with a exhaust fan.Must be working as a spike control.Now have the problem of cutting in and out over 1/2 degree.Have had 3 new lm311s still the same .Pin5 voltage 8.79v pin6 8.91v .Will try tomorrow 150 ohm resistor.Jaycar only sells the produce but has no service after.

[post="102883"][/post]​

Jaycars forum aint much chop either
:beer:
Richard

 

[spambait]

I have recently built the silicon chip coolmaster which is currently running a chest freezer @12 deg.
I check it every day and the temp is always spot on. But this evening i could hear a clicking in the brew room and noticed it was the freezer trying to click on.
I had a look at the coolmaster controller and the led was flashing dimmly and then a little brighter when the freezer was clicking but wouldnt cycle on.
As soon as i touched the sensor to warm it up a little the freezer started up straight away.

I'm no Electronics expert... Just a guy who can read a circuit diagram well enough to be dangerous with a soldering iron

I built this kit about 6 weeks ago and haven't really had any problems with it apart from working out how to set the thing to run the fridge at the temperature that I want. One thing I did notice was that I would get some funny results when I started mucking about with the the multiturn trimpot while it was plugged in to the fridge. I also tried testing the thing out with a light turned on, and found that the cooler master would act a little like a dimmer switch when I turned the trimpot up and down just past the point where the hysteresis comes into effect.

Another thing worth mentioning is that I have a a friend who is an electronics engineer who built his own controller based on this circuit, but replaced the optocoupler and the mains part of the circuit with a solid state relay. The interesting thing is that this device caused my fridge to make a funny chattering sound, but failed to allow the compressor to fire up. His explanation was that the while the fridge's running current is quite low, the draw-down current is quite high (blew a 3Amp fuse in his circuit), and his relay must have had some sort of protection circuit built-in.

Even more interesting though was that the Coolermaster worked fine. Two things that I noticed though... first was that I needed to give it a slightly wider setting for the hysteresis (can't remember which way to turn VR2) because the fridge would sometimes behave like it did with my friend's controller. The second thing was that I had to ensure I had given the curcuit enough power to do the job, but not so much that it would cause any of the capacitors to start smoking. I'm running the Coolermaster off a 12v regulated plugpack rated at 1Amp. I found that my other unregulated plugpack pushed too much power into the circuit (had to replace that monster 2200f cap due to burning it out), and a 9v supply was too weak to drive the 7809. The other thing I did was to make sure that the temperature sensor was fully sealed against moisture (does funny things if you touch the LM335Z's pins with damp hands!!) and to place the sensor against the fermentor so that the brew temperature would be measured rather than just the fridge - I got a mate to make up a nice little copper plate with the sensor attached that I can mould to the shape of the fermenter, which replaces that aluminium block that came with the kit). Brew temp will change more slowly than the air in the fridge, meaning that the controller won't try and turn the fridge on more often than it needs to.

Like I said, I'm no expert. Just thought it might be nice to share the experiences as it were. I've got the controller running with about 50 litres of pilsner, maintaining a perfect 11C... just try doing that without the fridge in a Melbourne summer !!!

Working with this controller has given me a second obsession... now it's Beer AND Electronics (My missus is going to kill me!). I'm now working on a microcontroller based circuit that will allow me to heat as well as chill, and with a selectable temperature and hysteresis... Got most of the parts and I've written the software, and I'll post the design and code once it's finished.

Salut.

S.

 

[Simon W]

Recharge, DO NOT adjust VR2. It DOES NOT adjust the hysteresis. All it will do is change the temperature(very slightly) that you spent lots of time getting VR1 spot-on 12 degree's!

I think your problem is not hystersis. The circuit has clearly had a set level of hysteresis designed into it already ie 33k resistor.
I think you'll find that the fridge itself, being an inductive load, is causing a power supply spike that is reflecting into the mains and into your little circuit. One easy way of confirming this would be to connect the circuit to a simple incandescent light bulb. The light bulb is a resistive load and wont induce any spikes at all.

Yep, exactly. I suggested unplugging the fridge to see what happens.
The fact that the circuit does not behave the same when the fridge is unplugged, proves hysteresis is not the problem here.

I suspect the surge current from the fridge motor may be switching the triac off, then the triac tries to switch on again and so on. MAYBE. This problem is mentioned in the article, and protection is provided. I suspect this protection may not be enough for your fridge. But can't say for sure.
A more unlikely theory, there might be noise generated by the fridge motor that the temp sensor is picking up and triggering the circuit. The comparator is comparing very small voltages, a swing of no more than 10 or 20 thousandths of a volt, any noise induced in the very long sensor lead(acting like an antenna) could quite easily be this large. You may need to shield the sensor wire somehow, or even buy two-core screened cable(approx 80c/metre at Jaycar catWB-1504).
When I got that issue of SC and read the article, I was surprised at the long length of unshielded cable.
Having said that, I don't know exactly what's happening, but It's more likely to be the triac and surge.


Stephen,
I just read the datasheet for the LM311, and it confirms my circuit analysis a few posts back.
The 33k feedback resistor generates the hysteresis, not VR2.
The inputs at pins5 and 6 are an 'offset ballance'. This means that the comparator will trigger when the pin 2+3 inputs are of a different voltage, as determined by pins 5+6. This is not hysteresis. If the 33k resistor is not in the circuit, the offset is still there, but the hysteresis is gone, the circuit will turn on at say, 12degree's, and turn off at 12degree's... oscillation.
The 33k resistor is changing the offset only when the fridge is on and cooling.
Adjusting VR2 will change the temperature the circuit triggers at, but will not change the hysteresis bandwidth.
It seems that VR2 is only included here because this is a copy of an example circuit provided by National Semiconductor in their datasheet, not the whole circuit, just the support componentts for the comparator.
According to the datasheet, a 5k VR2, fed by a 3k resistor will only adjust the offset by approx 8mV, this is a span of around 0.8 degree's! A handy 'trimming'

Simo

 

[recharge]

Well i bought a new one today though now it is called a Tempmaster, it has a slight design change allowing you to set up for heating OR cooling.
I will add a couple of small switches i think to enable the change.
I will also add an rca plug wired to pin 4 and point b in the link position to measure set voltage from outside the case. (maybe a little screwdriver epoxy'd to vr1 for easy adjustment.
Anyway i will see if this one does the same when finished. (Next night shift)

Cheers for now :beer:

Richard

PS i run it with a 400mA plugpack

 

[stephen]

Recharge, DO NOT adjust VR2. It DOES NOT adjust the hysteresis. All it will do is change the temperature(very slightly) that you spent lots of time getting VR1 spot-on 12 degree's!

I think your problem is not hystersis. The circuit has clearly had a set level of hysteresis designed into it already ie 33k resistor.
I think you'll find that the fridge itself, being an inductive load, is causing a power supply spike that is reflecting into the mains and into your little circuit. One easy way of confirming this would be to connect the circuit to a simple incandescent light bulb. The light bulb is a resistive load and wont induce any spikes at all.

Yep, exactly. I suggested unplugging the fridge to see what happens.
The fact that the circuit does not behave the same when the fridge is unplugged, proves hysteresis is not the problem here.

I suspect the surge current from the fridge motor may be switching the triac off, then the triac tries to switch on again and so on. MAYBE. This problem is mentioned in the article, and protection is provided. I suspect this protection may not be enough for your fridge. But can't say for sure.
A more unlikely theory, there might be noise generated by the fridge motor that the temp sensor is picking up and triggering the circuit. The comparator is comparing very small voltages, a swing of no more than 10 or 20 thousandths of a volt, any noise induced in the very long sensor lead(acting like an antenna) could quite easily be this large. You may need to shield the sensor wire somehow, or even buy two-core screened cable(approx 80c/metre at Jaycar catWB-1504).
When I got that issue of SC and read the article, I was surprised at the long length of unshielded cable.
Having said that, I don't know exactly what's happening, but It's more likely to be the triac and surge.


Stephen,
I just read the datasheet for the LM311, and it confirms my circuit analysis a few posts back.
The 33k feedback resistor generates the hysteresis, not VR2.
The inputs at pins5 and 6 are an 'offset ballance'. This means that the comparator will trigger when the pin 2+3 inputs are of a different voltage, as determined by pins 5+6. This is not hysteresis. If the 33k resistor is not in the circuit, the offset is still there, but the hysteresis is gone, the circuit will turn on at say, 12degree's, and turn off at 12degree's... oscillation.
The 33k resistor is changing the offset only when the fridge is on and cooling.
Adjusting VR2 will change the temperature the circuit triggers at, but will not change the hysteresis bandwidth.
It seems that VR2 is only included here because this is a copy of an example circuit provided by National Semiconductor in their datasheet, not the whole circuit, just the support componentts for the comparator.
According to the datasheet, a 5k VR2, fed by a 3k resistor will only adjust the offset by approx 8mV, this is a span of around 0.8 degree's! A handy 'trimming'

Simo

[post="103057"][/post]​

Simo

I stand corrected. I did not have the luxury of any datasheets or the circuit analysis. I was relying on my memory remembering circuit theory taught to me nearly 25 years ago as a trainee Radio Technician in the RAAF. From what I have learnt from this forum is that I will do away with the Triac and IC2 and the 240V side and look at a low current relay between the 100 ohm resistor and LED1 to switch my fridge on and off. I may have to recalculate the value of a replacement for the 100 ohm resistor dependant on the current draw of the relay coil.

Regards and beers

Steve

BTW TRIAC1 is only rated at 8 amps - the switch on current draw of a compressor can easily exceed this, though for a brief momentr

 

[stephen]

To add further to the above, if you suspect any interferance from external sources messing up your signal from the sensor, just wrap the lead with alfoil for the entire length of the sensor lead, leaving only the sensor exposed and link the end with the earth, say pin 4 or 1 of IC1 and see how it performs. The alfoil only has to cover the lead, ie. a strip about 1-1.5" wide and wrapped completely around the lead is more than ample. You may have to use multiple strips to achieve this.

More beer to drink

Steve

 

[recharge]

Stephen,
Could you post your mods when finished and let us know how they go.
Cheers

Maybe we should start another thread just for modifications as there are others posted in different threads, we could get them all in one place for future reference/users.

Cheers all
:beer:

Richard

 

[stephen]

Recharge

More than happy to post any "mutilation" I do to any project that I do, however, I hope that when I do post them that anyone reading them and trying to emulate me does so on their own back. I will not be held responsible for any "oops!" (My disclaimer)

Steve

 

[recharge]

Stephen,
I would like to think that would go without saying, but you never know i guess, best to cover your ass.
As for me i can put things together and tho i dont understand it all i know that 240 volts hurts.
As for stuffing one up well at 39 bucks i think they were designed for us to play with, im already learning stuff just from this one kit. This electronics stuff is pretty cool, why didnt i pay more attention in school.

Thanks again for your input

Cheers :beer:

Richard

 

[Simon W]

Ooh yeah, 240v hurts like hell, and a good jolt leaves you well shaken up for a day or two. One thing I've noticed is that you realise how stupid whatever it was you were doing was AFTER you did it....... 'I'll just adjust thaBZZZZZT!!'
Electronics is great fun, I've been playing with it since I was 6, but the main thing I've learnt after doing 3 years of electronic engineering at TAFE, this is definatelly one of those things where the more you learn, the less you know! It can get rather frustrating at times! Ohm's law helps ya get through most things though.

Stephen, great idea with the alfoil. I'm not entirely comfortable that the sensor would be picking up noise, as the capacitor across the comparator's inputs will short any AC, but a transient would probably get through. Alfoil is a cheap way of eleminating this possible cause tho. Nice.

 

[stephen]

Simon

I am very familiar with the "kiss" of 240V.

The foil "fix" I suggested will only work if all pieces are joined together to ensure continuity. If this proves to be the fix then I suggest, like you said, to use some shielded, insulated pair cable (twisted will suffice) with the sheild connected to circuit earth, ie. pins 1 or 4.

I think that if this a problem, I would look at, "Is there another fridge, freezer, clothes dryer etc" nearby that could be inducing the false signals.

For those out there that probably don't know what I'm talking about, I apologise. For those you who learnt something, then I have taught. For those of you who want to know more, feel free to ask and I wiil endeavour to answer your questions as best I can.

More Beer

Steve

 

[recharge]

Well it definately isnt picking up interference from another appliance as my brew room is about 50+ meters from any other fridge etc.
Might try making a new sensor with some cat5 being twisted pairs i think it should suffice and would fit better into provided plugs etc. Might be a bugger to get under the seal.

Might look into building a digital thermometer next, a little mini project to pass some quality time at work.

And yes i know the sting of 240v too [ "I thought you switched that fu#$%r off"] always check myself now.

I have a mate who is an electrician who once told me that a little knowledge is far more dangerous than none at all.

Richard

 

[Simon W]

Actually, I was thinking of possible noise from the fridge, as the motor tries to start up
It's a long shot, but it's another thing we can try to eliminate.
the more things we know it's not, the more likely we are of finding the real culprit, but my money is still on the triac false triggering.

Simo

 

[recharge]

Ok built a second unit and it does the same as the first. Starting to think they're a peice of crap.

I got to watch it do its thing a couple of times today and they both flicker etc even with nothing plugged in so i assume this would rule out the freezer being the problem.

I did make a few small additions to the second unit so i can switch it between heat and cool option and also read set and fridge temps(voltage) via an rca plug that goes to my DMM.

Oh well they do at least work

Richard :angry:

 

[ruserious]

Ooh yeah, 240v hurts like hell, and a good jolt leaves you well shaken up for a day or two.

[post="103134"][/post]​

i got a zzzzzap off 415V once at work. (Air Arc Gouger).

gave me heart palps for two days.

gggood nnow thoughhhh....... :blink: :blink: :blink:

be careful.

imserious

 

[recharge]

i got a zzzzzap off 415V once at work

Dont be doin that again in a hurry

:beer:
Richard