Version 2 of my homebrew controller


 
So mounting. that is something that i was giving some thought to..

i was thinking about building something like the bbq guru 20 dollar mount..

http://www.thebbqguru.com/pdf/..._Smokey_Mountain.pdf

i was thinking about taking a big washer and cutting it about 1/3 off (or grinding), then pounding the edges down, and maybe like making a half bowl then running a bolt through it.

then for the outside piece i thought i could figure out some piece of metal tubing (copper pipe maybe?) that i could build a mount on.. then i was think that i could use lead free solder and mount a nut in the middle of it..

that would give me easily 425 degrees (up to about 450).. i don't know if that would be enough? or i could have my cousin take it to school and weld it on.

to mount the copy pipe part to the blower, i was thinking that i could cut the end of the pipe into 4 sections and bend each part into the side of the opening of the blower... or something like that..

thoughts?
 
i didn't mean to imply i would run it that hot.... i mean i can see myself running at 350 to do turkey or something?.

Brian, what is the limitation with heatermeter at ~400F? just curious?

i was thinking of using something like

http://www.amazon.com/Lead-Fre...id=1297802235&sr=8-9

95% tin - 5% antimony. Most commonly used where frequent and extreme temperature changes and vibrations occur. Melting range: 450 - 464 Fahrenheit. ( i have some at home)

i guess i should measure how hot the bottom vents are when i have the WSM fired up and hot for a while, to see how close it is on the outside.. of course every time the fan turns on the temp of those bonds will fall, then heat and fall, so i can see that it would cause the joints lots of stress..


i don't have the equipment to braze but maybe one of my uncles/cousins do... i can always check with them. they have all kinds of tools.

braze/hard soldering with silver is in the 800F range right?
 
Originally posted by Cole Ferrier:
Bryan, what is the limitation with heatermeter at ~400F? just curious?
Click to expand...
The original Maverick probes are only rated to 400F, which is pretty accurate because I took one up to ~500F and ever since then it stopped working reliably (would give all sorts of false readings, I believe from something shorting out inside the probe itself).

The high heat probes work fine but around 300F or so you're out of the linear range of an NTC thermistor. From there the temperature difference between two adjacent ADC values grows so large the temperature seems to jump all over. Check the graph here of the ADC value vs the temperature. We use the green line, Maverick 22k
 
Originally posted by Bryan Mayland:

The high heat probes work fine but around 300F or so you're out of the linear range of an NTC thermistor. From there the temperature difference between two adjacent ADC values grows so large the temperature seems to jump all over. Check the graph here of the ADC value vs the temperature. We use the green line, Maverick 22k
Click to expand...

What is the red line Maverick 6.8k? is that just with a smaller resister? the problem with this is? (just curious)

Is the problem primarily with the chamber probe or the food probe or both?

If i wanted to get to 350, can i manage to do that? any experience in that temp range?
 
Originally posted by Ed Pinnell:
Use a propane torch (like a Bernz-o-matic), spread the flux on, and when the flux melts put a little bead of silver solder on. When the bead begins to flatten and flow out, try to maintain that temp by moving the torch around. Try to keep the solder in a near-plastic state...if you get it too hot it runs all over the place. If it gets to a dull red it's getting too hot. The hardest part is getting the flux off when you are done. Get a spray bottle with water and heat the part up a bit and spritz it with the spray to remove the glass-like flux.
Click to expand...

Thanks for the explanation.

Now the real hard part.. The hard sale to the wife.
 
Originally posted by Cole Ferrier:
What is the red line Maverick 6.8k? is that just with a smaller resister? the problem with this is? (just curious)

If i wanted to get to 350, can i manage to do that? any experience in that temp range?
Click to expand...
Because I had already built it with a 22k based on Bob's design. It would seem that 6.8k would be a better solution but I think the graph might be wrong. Looking at the graph it seems that an ADC reading of 100 wound be about 225F and a reading of 99 should be 0.6F higher, but empirically there appears to be 0.3F difference between ADC steps.

It is hard to tell what the step is at around 350F, where the device should start becoming wildly inaccurate, because 8 measurements are averaged so it gets smoothed out in the logs.

If you want to go with a 6.8k give it a try, it is only a constant in the code.
 
Originally posted by Cole Ferrier:
Any thoughts if the new probes that come with

http://www.maverickhousewares.com/et732.htm

have the same range of actually detecting temperature as the high heat ones on the list of parts sheet?
Click to expand...
I bought 2 of these probes (one to replace the one I burned out by running too high). I have yet to try them at high heat but at 212F they read within a degree of the old probes, and a quick check of my HeaterMeter sitting around at home says it is within 1 degree at 70F too, so I assume they use the same internals, just a sturdier design. They also have like a 6ft cable which I'm sure is designed so you trip on them all the time.

I can't vouch for their accuracy at higher temps but I will check it this weekend. I don't have an ET-732 but I'll test them on the 'Meter.
 
Originally posted by Bryan Mayland:
I bought 2 of these probes (one to replace the one I burned out by running too high).

I don't have an ET-732 but I'll test them on the 'Meter.
Click to expand...

Bryan, how did you get them? i just emailed the company and they said they don't sell replacement probes for the ET-732?

Edit:

Did you get these?
http://www.bigpoppasmokers.com...uper-food-probe.html
http://www.bigpoppasmokers.com...super-pit-probe.html


Edit:

Further thoughts, if we need more resolution in our ADC we could in theory use something like
MCP3204 (or MCP3208)
http://search.digikey.com/scri...name=MCP3204-CI/P-ND
http://www.arduino.cc/playground/Code/MCP3208

with 4k steps of data verse 1k, couldn't we use a smaller resister, and therefore have a steeper curve? but still get decent data out of it?
 
@Ed:
The graph is generated from this gnuplot script. You may want to check it for accuracy. The Y axis for the top graph is degrees F and the bottom graph is the degree delta per 1 value of ADC. It doesn't look right but I think it is based on the math.

The probes they use in the ET-732 I believe are just their standard "high heat" probes, or at least they have the exact same specs as far as temperature ranges. I got them from their parts order page, all the "6ft" probes are high heat probes.

I don't think 4k vs 1k steps would make much difference, the issue is that the curve just starts going to the moon up there. The tradeoff of using the smaller biasing resistor in the divider is that you can measure higher temperatures more accurately at the price of slightly less resolution across the entire range. For example, the 22k resistor at 220F has a resolution of 0.27F where the 6.8k resistor is 0.33F. I wouldn't try to use the HeaterMeter with a 22k for anything above 400F, which is where the resolution is about 1F per tick.

This all might be making a mountain out of a molehill though, considering this is still probably more accurate than a kitchen oven which may have 10-20F swings in temperature over a cycle. EDIT: Regarding thermocouples, they're great and all but the problem is their cost and the work of fabricating my own probes.

Speaking of cycles, what do you guys think about changing the HeaterMeter cycle from updating once per second to once every two seconds? I am thinking this to reduce the amount of data coming out by half, and make the fan speed seem less spastic if you're watching instantaneous reads.
 
Originally posted by Ed Pinnell:
BUT...I'm also one of those kind of guys that can't stop futzing with the friggin' thing until it stops working altogether ("it used to work until I fixed it!")
Click to expand...
Agreed! I've spent more money and time on this project after I had it working than before. It is a fun little project that I can incrementally upgrade as I have time. Swapping out parts for better ones, modifying the algorithms a bit, building the same thing out of different parts.
 
Thanks again guys for all your efforts. Bryan, when do you think you may have some updated downloads. I was never able to get the your regular version to run on mine without the flash.
 
Originally posted by RJ Riememsnider:
Thanks again guys for all your efforts. Bryan, when do you think you may have some updated downloads. I was never able to get the your regular version to run on mine without the flash.
Click to expand...
Errrr tomorrow maybe? I haven't gotten around to finishing/testing the alarm code so I'll just take it out of the menus for now. I just placed another **** Mouser order and completely forgot I wanted to order a few different piezo speakers to test with. I see another order in my future!
 
Originally posted by Bryan Mayland:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by RJ Riememsnider:
Thanks again guys for all your efforts. Bryan, when do you think you may have some updated downloads. I was never able to get the your regular version to run on mine without the flash.
Click to expand...
Errrr tomorrow maybe? I haven't gotten around to finishing/testing the alarm code so I'll just take it out of the menus for now. I just placed another **** Mouser order and completely forgot I wanted to order a few different piezo speakers to test with. I see another order in my future! </div></BLOCKQUOTE>


Thanks Bryan. I have a few laying around I'd be happy to donate to the cause? How are you implementing the alerts? We're pretty low on IO pin count. You thinking of making a generic driver circuit for pezio or LED or car horn, etc.
 
Thanks but I'll probably just head to the Dollar store and find something that has one in it and rip it out. I had an old MODEM with a piezo speaker on it that was perfect but my boss stabbed it with a screwdriver the other day and I'm not sure exactly why that happened.

I was thinking just a small PWM running at a certain frequency. It would take too much code space to actually make it go beep beep beep beep instead of just beep. I've got plenty of pins still available!

I'd also like to confirm that the High Temperature Probes to stand up to some pretty high temperatures. I stuck one in the grill last night as I was heating it up for steaks and I came back out to see HeaterMeter reading 806F. After it cooled back down the probe seems to be working just fine still so they're not kidding about the high heat bit. At that temperature the measurement is also around +/-20F so I'd not recommend attempting to control it.
 
Originally posted by Bryan Mayland:
Thanks but I'll probably just head to the Dollar store and find something that has one in it and rip it out. I had an old MODEM with a piezo speaker on it that was perfect but my boss stabbed it with a screwdriver the other day and I'm not sure exactly why that happened.

I was thinking just a small PWM running at a certain frequency. It would take too much code space to actually make it go beep beep beep beep instead of just beep. I've got plenty of pins still available!

I'd also like to confirm that the High Temperature Probes to stand up to some pretty high temperatures. I stuck one in the grill last night as I was heating it up for steaks and I came back out to see HeaterMeter reading 806F. After it cooled back down the probe seems to be working just fine still so they're not kidding about the high heat bit. At that temperature the measurement is also around +/-20F so I'd not recommend attempting to control it.
Click to expand...

either that or just a pin that goes high with alarm and we could take it from there. One could use a simple oscillator for tone or add a transistor for high current for flasher, etc.

Good news on the probes.
 
HeaterMeter r62

Changes:
-- Web index much improved and rewritten. Temperature change indicators, last update indicator
-- Added HEATERMETER_SERIAL status update and config urls
-- FIX: Manually entering Lid mode would allow error sum to build and artificially raise setpoint
-- Change default PID params to (BPID) 5, 4, 0.002, 2.5
-- FIX: Bug in antiwindup prevented decay unless error was [0, -1]
-- Remove some unused code, optimize some data types in grillpid
-- Split main file into hmmenus/hmcore
-- Added README.txt
-- FIX: Blank display when no FOOD or AMB probe
-- FIX: No display for first 1s if no probes plugged in

Check out README.txt, there are both instructions and a section on some of the #defines in the new code. For no WiFi module you probably want to disable HEATERMETER_NETWORKING.
 
Thank you sir! I still can't get to compile though, I get the following in hmcore.cpp.

I am running Arduino 0022
WiShield v1.3.0
DataFlash v1.0.0

Any ideas??????

hmcore.cpp: In function 'void sendFlashFile(const flash_file_t*)':
hmcore.cpp:366: error: 'struct webserver_state' has no member named 'ackedCount'
hmcore.cpp:370: error: 'struct webserver_state' has no member named 'ackedCount'
hmcore.cpp:372: error: 'struct webserver_state' has no member named 'cursor'
hmcore.cpp:384: error: 'class Dataflash' has no member named 'Cont_Flash_Read_Enable'
hmcore.cpp:386: error: 'class Dataflash' has no member named 'Cont_Flash_Read'
C:\Data\Dev\arduino\libraries\dataflash/dataflash.h:97: error: 'void Dataflash:
icon_biggrin.gif
F_CS_inactive()' is private
hmcore.cpp:387: error: within this context
hmcore.cpp:390: error: 'struct webserver_state' has no member named 'cursor'
 
Oye. Half the errors a from my changes to dataflash and the others are from... some... wiserver differences.

Maybe try with my libraries r24.
 
You must log in or register to reply here.

 

Back
Top