sous vide thread


 
Bryan Mayland said:
Yeah it is still 12V but I'm saying that when it is at, say 2%, it is running at 10% 490Hz PWM for 2 seconds out of every 10. The PWM is too fast for the SSR because it is still PWMing right?
Click to expand...

10% * 2/10 = .02 as expected. What I'm saying though is that even 1% PWM at 490Hz triggers the positive edge trigger in the SCR that keeps it on for one cycle. So, a PWM Duty Cycle of 1% is the same as a Duty Cycle of 100% as far as the SSR/SCR is concerned. Every AC Cycle it sees a positive DC trigger that turns it on for the duration of that cycle. That is why the long SRTP mode is needed when controlling SCR/SSR. Duty cycle of the pin at 490Hz is basically irrelevant.

I forgot to mention that I really like the menu system on the AVR. Very intuitive and the timers are just right. Well done. Hopefully I can update this weekend and try out the new features.
 
for all these "issues" I can say that for my crock pot the temperature was maintained with .2-.3 degrees for the 6hours I cooked. I should have saved the cook, as excel is quite good at giving you the statistical values for your cooks when you read them into it.

I'll have to save my next one,
dave
 
On a side note, here is a really good animation via wikipedia of SCR being triggered by a positive edge:
Regulated_rectifier.gif


The difference is that the trigger in this pic is phase aligned to the AC cycle and offset by a certain time to provide proportional control using an SCR. I believe this is called "phase angle modulation". I do not think we have a need to do that with the heatermeter, but still a good animation none the less. Since in our case the first positive trigger at 1% duty at 490Hz comes in the first few mS of the AC cycle, it just stays on the whole rest of that cycle.
 
D Peart said:
for all these "issues" I can say that for my crock pot the temperature was maintained with .2-.3 degrees for the 6hours I cooked. I should have saved the cook, as excel is quite good at giving you the statistical values for your cooks when you read them into it.

I'll have to save my next one,
dave
Click to expand...

I recall reading that on the other pages, and it is great news for anyone who wants to get started ASAP. This is one of those "its the journey not the destination" sorts of things. In reality On/Off (0/100%) control of heating elements is plenty appropriate, but your turkey cooker already does that right out of the box. I'm no sous vide expert, but I imagine the precision of the temperature control is what leads most people to use PID controllers and heating elements instead of temperature controlled crock pots and turkey cookers.

5 levels of power is better than 1, and 100 levels is better than 5. Then of course there is the tuning...
 
Last edited:
Evan MJ said:
10% * 2/10 = .02 as expected. What I'm saying though is that even 1% PWM at 490Hz triggers the positive edge trigger in the SCR that keeps it on for one cycle. So, a PWM Duty Cycle of 1% is the same as a Duty Cycle of 100% as far as the SSR/SCR is concerned. Every AC Cycle it sees a positive DC trigger that turns it on for the duration of that cycle. That is why the long SRTP mode is needed when controlling SCR/SSR. Duty cycle of the pin at 490Hz is basically irrelevant.
Click to expand...
Oh now I think I get it. I was thinking that the SSR possibly wouldn't trigger at all because the duration the output is active is less than one period of the AC wave. Makes sense now that it would be positive edge triggered to activate for the whole cycle.

And Dave: that's pretty awesome that it can hold temperature so well given how coarse our control is. I'd like to add some of those statistical functions into the website as well at some point... you know, just like all the other stuff I want to do!
 
I'm all for making it more accurate and possibly helping keep the SSR life span longer. I just don't want people to think it is broken. :)

I've seen that the SSR does a better job on both my electric smoker and the crock pot, than I ever get out of my charcoal smoker.

My next batch of smoked sausage will be poach in the turkey roaster using the HM after I pull it from the smoker. Should help make better sausage.

dave
 
Let us know how it comes out. I'm just hoping to do a few pre-packaged pre-marinated pork loins to get started, and pan fry them for a bit of color when they are done. (So the wife will eat them).

Thanks for reiterating that the SSR control does indeed work as is. I would not worry about the life of the SSR, it is just doing its job. Heat is about the only thing that should kill it... and of course over voltage on either side.
 
Good news! As of CapnBry-HeaterMeter-cb1762d (maybe a bit earlier), the AVR code supports 10 levels of "proportional" control for SSR (Solid State Relay) based heaters.

Set your minimum and maximum fan speed to 100% in the web interface when using a relay instead of a fan. This will force it to give 100% duty cycle on the fan output, and always be in "long pwm" mode.

10% control power should be on for 1 second of every 10, 50% is 5 seconds of every 10, and so on.

For instance, if you run a heater that is 1000W, you can now control it down to 100W increments with a ~$10 SSR.
 
Last edited:
My first sous-vide cook (and last like this)....

So, my kiddo got sick, which gave me some weekday time to stare at a computer screen for hours. I set up my HeatermeterPi prototype setup with my SSR detailed on page 2.

I ran the extra probe I have as a heatsink monitor, since I was especially worried that it would overheat and melt since I'm pushing 12A through a 10A relay and power cable. Note that I would not leave this un-attended, and I made sure to put it on a metal pan so any melting plastic would not mess up my faux granite counter top.

Settings: MIN and MAX fan speed both set to 100% to force "Long PWM" or "SRTP" mode. PID gains started as stock, and I tuned them a bit to reduce oscillation and overshoot. The gains I ended up with are P=10, I=0.007, D=20. I have not tested those gains from a cold setup to see how much they overshoot, but steady state I got it easily within +/- 0.5 degree F, and even for a while within 0.1 degree of the setpoint. Most blips in the temp were due to me opening the lid to flip the pork around since I have no setup to mix the water up and the bubble in the bag made it float a bit.

Next time I run it I'll start with I=0, D=0, and tune in P, then minimize overshoot/oscillation with D, and then decrease steady state error with I. I did it out of order thinking the stock gains would be close. The gains can be much better. Oh, and the heat sink got really hot. At around 185F, I found an old RCA splitter and hooked my blower fan up to blow on the blue box. A properly rated relay or a hole in the case, or heat sink compound are all things that would have kept me from needing active cooling, but I was working with what I have. The kill-a-watt claimed 1.5KWh for the whole cook, so maybe 25 cents in electricity? Cheaper than a bag of charcoal for sure.

Some photos:

c0.png


c1.jpg


c2.jpg


c3.jpg

^ Just showing off the HeatermeterPi's versatility :P ^

c4.jpg



Overall, I think the cook went very well. I do not own a vacuum sealer, so I though I would be clever and use pre seasoned tenderloin that we eat a lot at my house... I had a small amount of air in it that puffed up the bag, but cooking went well and I measured an exact 140F in my bbq thermometer in the center of the pork.

It was juicy, but the flavor, which I typically "enhance", was actually a bit too strong. I guess a lot of it falls off when I grill or oven roast these things... lots of extra salt flavor, probably from the fat not being allowed to escape. The fat (as seen in the picture) did not really fully render out... it was soft, but I still chose to pull it off for the most part. I'm not giving up on sous-vide yet, but I think I prefer the grilled flavor on these tenderloins. I did attempt to sear it, but I used too much oil, and I could not find my cast iron skillet, so I failed at that, but it still looked decent enough.
 
Last edited:
That's really cool, love all the pictures. That SSD really started getting hot at the beginning there, eh?

I'm not surprised the default PID constants didn't do a very good job. Although both a BBQ and a water heater are slowly changing systems, the input dynamics are so different. I've got a sketch combining the PID Autotune Library with the HeaterMeter configuration and output code sort of Frankensteined together I've been wanting to try. Maybe I'll get to that this weekend.
 
The SSD did get hot, but was fine with the fan on it. Also, I could have easily bypassed the SSD and got the water up near temp, then hooked it back in line.

I'm interested to see how the auto tune works out. I can try it on my setup sometime soon.

I'm hoping to get my fan tubing welded to my smoker this evening to do my first real fire smoke on saturday with the HmPi.
 
The autotune doesn't work for crap on my BBQ because it takes so little fan to keep it going. The autotune library takes your initial temperature (tempI) and fan speed (fanI). Then, if the temperature is higher/lower than your initial temperature by X it changes the fan speed to fanI +/- Y. Once it has found an oscillation of a certain mystery size within the time period Z, it counts the time between the two and spits out coefficients.

The problem is that even for an X of 2 degrees, fanI = 10%, Y = 10%, it takes my grill about 30 minutes to make a cycle, about 3 hours to come up with values, and then it spit out some crazy numbers like PID=0.3,0.00034,68.85. That just make my grill get hotter and hotter over time.
 
Hi,

A quick post to resurrect the thread.

Evan, have you had any further experience with the Sous Vide adaptations?

I'm about to venture into Sous Vide... I have the water bath and am starting to contemplate rigging up a SSR to maintain the temperature.

I've read through the thread and note the high-ish temperature of the relay - I planned to attach a heat-sink to the SSR to try and dissipate.

Are there any other notable 'gotchas' which I should consider?

Thanks

Nick
 
It might, but only for a short period of time. Regular relays are rated for a number of switching cycles on the order of 100,000 times. If you're switching a couple of times an hour, that's like 10 years. But HeaterMeter switches 500 times a second which would expire the life cycle of a normal relay in a matter of minutes. SSRs have a number of other enhancements such as phase matching and opto-isolation which make them a lot better for fast switching applications.
 
Okay, SSR it is then for Sous Vide / Mash tun project ;-)

I guess a regular relay would work for something like regulating the temps of a fridge/freezer. This new low-cost way of automating things is awesome ;)
 
Sous Vide / Mash Tun you say?


This is actually my 15.5 gallon sous vide boil kettle which I use a stand to keep the food off the heating element. I have a separate hot liquor tank with a heat exchange coil in it that heats the mash tun. Both use SSRs to pulse the heating elements on and off to hold temperature. There's no web control though... yet!
 
Hi,

Seasons greetings to everyone... with the UK holiday season giving me a little more time to get my sous video setup working I have successfully boiled and egg!

It was as hoped... soft with a 'custardy' yolk.

I used a 25a SSD relay which seems to cope without generating too much heat (although a fan built into the case is there just in case).

Bryan, one question though... is there a simple adjustment I can make to the heater meter code to recognise single decimal place set temperatures? Ideally I'd like to set it to 64.5C... but setting via the web interface always rounds the value to the nearest whole number.

Nick
 
Hi,

Seasons greetings to everyone... with the UK holiday season giving me a little more time to get my sous video setup working I have successfully boiled and egg!

It was as hoped... soft with a 'custardy' yolk.

I used a 25a SSD relay which seems to cope without generating too much heat (although a fan built into the case is there just in case).

Bryan, one question though... is there a simple adjustment I can make to the heater meter code to recognise single decimal place set temperatures? Ideally I'd like to set it to 64.5C... but setting via the web interface always rounds the value to the nearest whole number.

Nick
 
N Waring said:
Bryan, one question though... is there a simple adjustment I can make to the heater meter code to recognise single decimal place set temperatures? Ideally I'd like to set it to 64.5C... but setting via the web interface always rounds the value to the nearest whole number.
Click to expand...
There is not. Internally the setpoint is an integer and stored in only half the number of bytes needed to store an floating point value. All of the places the data is passed around would need to be modified as well as create an incompatible EEPROM storage structure.

Comedy option: you could switch to using the goofy US measurement system, which would roughly double the precision of the setpoint.
 
You must log in or register to reply here.

 

Back
Top