LinkMeter v2 Homebrew BBQ Controller - Part 1


 
Status
Not open for further replies.
(kyle@violane) bin/avrdude -C etc/avrdude.conf -p m328p -P /dev/tty.usbserial-A9007Wrk -c stk500v1 -b 9600 -F
avrdude: AVR device initialized and ready to accept instructions
Reading | ################################################## | 100% 0.13s
avrdude: Device signature = 0x1e950f
avrdude: safemode: Fuses OK
avrdude done. Thank you.

Hot ****! The 390 ohm resistor didn't work, but the 22 uF cap + stk500v1 did.
icon_biggrin.gif
 
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 Jay Bremner:
Where do I find the probes themselves?
Click to expand...
1) Yes you only need one RFM12B on the receiver, but you also need one for each lmremote (wireless temp transmitter device). Each lmremote can have up to 6 probes plugged into it. HeaterMeter is still limited to displaying only 4 probes though, so even if you have 3 transmitters with 6 probes each, you'll still only be able to map 4 of them to display.

1A) There's no schematic for the lmremote board yet. My version is still on breadboard. I've been wrestling with the question of how to power it. I use a single AA and a JeeLabs AA power board, or just using 3 AA batteries (which seems like overkill considering 1 AA lasts like a month running continuously). It is a pretty basic design other than that-- just a digital line connected via resistor to an analog in and then from there the probe goes to ground. Iiiiii'll get to it soon!

2) The HeaterMeter wiki should be the best place to start and maybe the first post of this thread. If you have any specific questions you can ask and I can try to fill out the wiki with the answers. As for where to start, I'm really not sure! I'd probably jump right in with getting parts and a LinkMeter v3.1 board, an FTDI cable, and either a pre-bootloaded ATmega chip or an USBtinyISP. While you're waiting for that to arrive, you might want to play with the HeaterMeter build process using the Arduino IDE. Even though that shouldn't be necessary, it seems to be working out that just installing the prebuilt binaries don't work. You definitely want to keep your eyes open for a WRT54GL router to put it all in on eBay / Craigslist. </div></BLOCKQUOTE>

I understand about building the LMRemote, but I am trying to get a grasp on what the wireless probes themselves would look like and where to source them. I am imagining something that just pokes into the meat with no wires attached so that (in the case of a rotisserie) it could turn with the meat without having to worry about wires getting tangled up.
 
Originally posted by Jay Bremner:
I understand about building the LMRemote, but I am trying to get a grasp on what the wireless probes themselves would look like and where to source them. I am imagining something that just pokes into the meat with no wires attached so that (in the case of a rotisserie) it could turn with the meat without having to worry about wires getting tangled up.
Click to expand...
Oh, gotcha! Yeah the idea was conceived by Ed Pinnell as a way to attach probes to his rotisserie. The design idea is like a 1"x3"x1" box that has one or more probe jacks in it. It would run off a battery, but it wouldn't be "high heat" (>80C) tolerant. Because it had no power cord the idea was that you could attach it to a rotisserie or something.
 
So, is there such a thing as an affordable pair of crimping pliers for these molexy things? I've found these:

http://www.sparkfun.com/products/10219

Which I assume are the correct thing, but $60 is too rich for my blood given how infrequently I'll use them. I was trying to crimp some by hand last night when I was fiddling with this bootstrap stuff and they are are real PITA.
 
Originally posted by Kyle Christensen:
So, is there such a thing as an affordable pair of crimping pliers for these molexy things? I've found these:
Click to expand...
Yeah no joke. In the mouser catalog I was flipping through, some of the strangest **** costs a lot of money. Why the heck are there $500 pliers? Has China not caught on to our untapped specialty tool market?

I use these crimpers from Pololu which aren't quite wide enough to crimp the whole connector in one squeeze, but for $30 I can't complain too much. Using needlenose pliers was killing me.

Physical connectors are like the hardest part of making electronics apparently.
 
How tolerant is the web server to additional information being sent back via the serial connection?

I've been using to send debug information back, like calculated resistance and the probe configuration information.

It seems to crash the webserver more frequently with this debug info streaming back. I was hoping that LM could safely ignore this info.

I'm also surprised to see how many errors there are on the serial connection. Not a very reliable interface, I imagine you have to drop a lot of bad data on the server side.

dave
 
Originally posted by D Peart:
How tolerant is the web server to additional information being sent back via the serial connection?

I've been using to send debug information back, like calculated resistance and the probe configuration information.
Click to expand...
I actually spit out debug stuff via Serial.print all the time and haven't had any problems with it, the serial interface rarely has any errors.

The linkmeterd code isn't written very defensively though, it expects the data to be reasonably formed. It looks at the first 5 characters of every line and sees if it matches a known segment type then pulls it apart and looks for the minimum number of fields to form a complete bit of data and discards everything else.

Are you trying to cat /dev/ttyS1 while linkmeterd is still running? I can see that causing problems because both processes are trying to read the serial port and each will get some of the data. I've actually done a little work in creating a new segment HMDB which will allow yo to write debug stuff that can be streamed to lmclient. If you can disable option 'daemonize' (set to 0) in /etc/config/lucid then start it up, you can see whatever error crashes it and maybe I can fix it.

Note that if you leave this set to 0, your router won't boot completely because the init process waits for lucid to daemonize before starting whatever follows it, such as sshd, the dhcp server, the led blinker, etc.
 
Originally posted by Bryan Mayland:
Yeah no joke. In the mouser catalog I was flipping through, some of the strangest **** costs a lot of money. Why the heck are there $500 pliers? Has China not caught on to our untapped specialty tool market?

I use these crimpers from Pololu which aren't quite wide enough to crimp the whole connector in one squeeze, but for $30 I can't complain too much. Using needlenose pliers was killing me.

Physical connectors are like the hardest part of making electronics apparently.
Click to expand...

Thanks for the heads up. I think I'm going to get the $35 version that is a bit wider (it's only $5, more, but it's out of stock until early feb).

http://www.pololu.com/catalog/product/1928/pictures

Crappy cell picture, but mine is starting to come together.

http://dl.dropbox.com/u/119708...25%2C-6-42-30-PM.jpg

Spent entirely too long today with a dremel and file getting that LCD fit. I just need to wire up the LCD and jacks, build a button board, and figure out how to fab brackets for the blower and pray that it all works.
 
It looks great!

Originally posted by Kyle Christensen:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Bryan Mayland:
Yeah no joke. In the mouser catalog I was flipping through, some of the strangest **** costs a lot of money. Why the heck are there $500 pliers? Has China not caught on to our untapped specialty tool market?

I use these crimpers from Pololu which aren't quite wide enough to crimp the whole connector in one squeeze, but for $30 I can't complain too much. Using needlenose pliers was killing me.

Physical connectors are like the hardest part of making electronics apparently.
Click to expand...

Thanks for the heads up. I think I'm going to get the $35 version that is a bit wider (it's only $5, more, but it's out of stock until early feb).

http://www.pololu.com/catalog/product/1928/pictures

Crappy cell picture, but mine is starting to come together.

http://dl.dropbox.com/u/119708...25%2C-6-42-30-PM.jpg

Spent entirely too long today with a dremel and file getting that LCD fit. I just need to wire up the LCD and jacks, build a button board, and figure out how to fab brackets for the blower and pray that it all works. </div></BLOCKQUOTE>
 
Upgraded to the latest/greatest HM and LM.

For HM I had to download to my computer and upload through the webpage, as the wget; avrupdate way didn't work.

LM updated via opkg install method. One little reboot later and it all looks good.

I had issues with LM freezing on me yesterday during my rib cook. I ended up just using cat /dev/ttyS1 and monitoring the output to see see what was happening.

For the cook today, hopefully it will work without issue.

I have also seen one other issue on all the cooks I've done. When I start the cook everything looks good, then at some point my pit temp gets off. It usually takes me a while to figure it out because it happens gradually.

After I figure it out, the fix is to reposition the pit probe, wiggle the wire, make sure that there are not sharp bends in the wire. After doing this, the temp always corrects itself.

It seems that the ET-732 probes are very susceptible to reading an increase in resistance due to outside influences. Has anyone else seen issues like this with the ET-73 probes?

thanks,
dave
 
One other problem I've had off and one, but has been really bad of late.

When I change the temperature on the LM it just resets back to what was originally on HM. I actually have to make the change on HM to get it to stick.

If I manually set it from a telnet window like:

echo /set?sp=250 > /dev/ttyS1

it works fine as well. It seems that its just the web page that doesn't work.

If I do it manually on the web server like:
http://192.168.1.141/luci/;sto...e737bb/lm/set?sp=225

It works.

But if I click on the main page and change the temp, it doesn't work.

Actually nothing on the main control page will change the HM, I can't change probe names any longer either. Something is amiss.

dave
 
I'm going to need some help debugging why my LM will only run for a few minutes before hanging.

Here is what I know.

1. it will run for days without issue if it is in the room below my router.
2. it will run for about 1 minute if it is outside, even though the signal strength is good and the telnet session has no issues.
3. I can cat /dev/ttyS1 for ever and never have any issues, except if it is done simultaneously with the web server running the webserver wont update while the cat is running.

I do see data corruption on /dev/ttyS1 when the router is outside, I do not see it when inside.

Here is a sample from /dev/ttyS1
$HMSU,225.207.4,U,61.2,U,40,39,0
$LMSU,225,207.5,U,61.2,U,40,39,0
$HMSU,225,207.6,U,61.2,U,40,3N°
$HMSU,225,207.6,U,6±.2,U,40,39,0
$HMSU,225,20?.6,U,61.2,U,40,39,0
$HMSUl225,207.7,U,61.3,U,40,39,0
$HMSU,225,207.7,U,61.2,U,40,39,0
$HMSU,225,207.8,U,61.3,U,40,39,0
$HMSU,225,207.8,U¬61.3,U,40,39,0
$HMSU,225,207.9,U,61.3,U,40,39,0
$HMSU,225,20?.9,U,61.3,U,40,39,0
$HMSu,225,208.0,U,61.3,U,40,39,0
$HMSU,225,208.1,U,61.3,U,40,39,0
$HMSU,225,208>1,U,61.3,U,40,39,0
$HMRF,A,3300,128,1310
$HMSU,225,208.2,U,61.3,U,40,39,<
$HMSU,225,208.3,U,61.?,U,40,39,0
$xMSU,225,208.3,U,61.3,U,40,39,0
$HMSU,225,208.4,U,61.5,U,40,39,0

I'm not sure if this is corrupted between HM and LM, or LM -> telnet -> putty

But because the LM is hanging, and requires a restart to get it going again, I'm thinking that the corruption is between HM and LM.

Maybe I'm way off base though
icon_smile.gif
I just want the webserver to run
icon_smile.gif


Currently I'm doing my cooks via cat /dev/ttyS1

Let me know what I can do to help debug.

thanks,
dave

P.S. - HM and LM are up to date as of 1/25/12
 
Originally posted by D Peart:
I do see data corruption on /dev/ttyS1 when the router is outside, I do not see it when inside.
Click to expand...
TCP/IP guarantees that there will be no corrupted data received over the network, so what you're seeing is definitely coming through the serial port. Have you tried doing your cat /dev/ttyS1 inside, but with the fan turned on? I'm wondering if power fluctuations are causing the serial port to get wonky. If set a negative setpoint it will set the fan to be manual mode so you can test by like setting sp=-40 and see if you get corrupted data on the inside too. Try some other numbers too like -100, -10.

If it is corrupted, maybe increasing the power caps might help or something, we can cross that bridge when we come to it. If it is not corrupted, is there something different about the power outside? Is it really cold maybe? You can also try recompiling HeaterMeter for 57600 baud and setting LinkMeter to match in /etc/config/lucid.
 
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 D Peart:
I do see data corruption on /dev/ttyS1 when the router is outside, I do not see it when inside.
Click to expand...
TCP/IP guarantees that there will be no corrupted data received over the network, so what you're seeing is definitely coming through the serial port. Have you tried doing your cat /dev/ttyS1 inside, but with the fan turned on? I'm wondering if power fluctuations are causing the serial port to get wonky. If set a negative setpoint it will set the fan to be manual mode so you can test by like setting sp=-40 and see if you get corrupted data on the inside too. Try some other numbers too like -100, -10.

If it is corrupted, maybe increasing the power caps might help or something, we can cross that bridge when we come to it. If it is not corrupted, is there something different about the power outside? Is it really cold maybe? You can also try recompiling HeaterMeter for 57600 baud and setting LinkMeter to match in /etc/config/lucid. </div></BLOCKQUOTE>

Bryan, spot on. If I have the fan blowing I get serial port corruption. If I have it turned off I get none.

I then used my SSR instead of the fan, and get no corruption with the use of the SSR. We are definitely getting noise back into the system.

dave
 
Ah lovely. Just a quick side note, there's a design flaw in hmdude that will prevent it from working if you flash your AVR with any other baud rate than 115200. I'll fix that tomorrow. Also, my lucid crashes if the baud rate is wrong (which would give me garbage data). I'll have to look into that later. I've also managed to somehow corrupt my EEPROM contents in a way that apparently heatermeter thinks the fan goes up to MaxFanSpeed=255. This all is happening of course ans I am trying to cook dinner!

What size fan are you running again? Is your 12V power supply up for it?
 
Originally posted by Bryan Mayland:
Ah lovely. Just a quick side note, there's a design flaw in hmdude that will prevent it from working if you flash your AVR with any other baud rate than 115200. I'll fix that tomorrow. Also, my lucid crashes if the baud rate is wrong (which would give me garbage data). I'll have to look into that later. I've also managed to somehow corrupt my EEPROM contents in a way that apparently heatermeter thinks the fan goes up to MaxFanSpeed=255. This all is happening of course ans I am trying to cook dinner!

What size fan are you running again? Is your 12V power supply up for it?
Click to expand...

My supply is a 12V 1A and the blower is 12V .3A

I'll hook it up to a PC power supply and see if that fixes it.

dave
 
Originally posted by Bryan Mayland:
The LinkMeter Home page is designed to be usable on small screens, and readable on larger screens from a distance. When I put it together, I used 240x320 and 320x480 as the target screen dimensions
Click to expand...

No it looks great on my Galaxy SII. I just have to scroll down to see the chart. No big deal really. You probably couldn't squeeze the chart down for mobile phones and make it look good.

It would be nice to see a button on the main page to allow you to toggle lid open off and on(maybe near the top left hand corner where is shows you when you in lid open).
 
Originally posted by D Peart:
It seems that the ET-732 probes are very susceptible to reading an increase in resistance due to outside influences. Has anyone else seen issues like this with the ET-73 probes?
Click to expand...

Yes, be careful with your probes. This came straight from an e-mail exchange I had with the great folks at Maverick:

"For the probe that came with the unit, if it is giving you high temperatures like you stated, then that means your probe shorted out. This is due to coming in contact with water, high heat, flames, or if the wire gets pinched or crimped. Even roughly handling or dropping the probe can cause the wires inside to dislodge and thus short out."
 
Built my breadboard tonight with my ATMEGA chip on it hooked to my Arduino Uno so that I could burn the bootloader on the chip. I followed the instructables guide linked to on page 1 and am getting the 'Heartbeat' on my breadboard, but kept having issues trying to burn the bootloader. Eventually, I dug around enough and found out that you can't use the Uno to burn the bootloader. (Found this on the Arduino site: NOTE: Currently, you cannot use an Arduino Uno as an ISP programmer because the optiboot bootloader does not support this sketch. A revision for this is in progress.)

Any ideas on how I can get the bootloader and sketch onto my ATMEGA?
 
Status
Not open for further replies.

 

Back
Top