LinkMeter v2 Homebrew BBQ Controller - Part 1


 
Status
Not open for further replies.
Originally posted by Ben Lawson:
I'm in the process of building my 3.1 board, where does TR1 go? Cant find it for the life of me
It is on the external schematic, connected to probe jack JP6 (FOOD3/AMB) between pins 1 and 2.

I don't have one built this way because I only have the 3 jacks, but if you look at this picture of the probe jacks you'd solder it to your 4th jack between the ground leg and the unconnected leg on the side of the jack.

What this does is connects the thermistor if no probe is plugged in, and the probe if a probe is plugged in. You can also omit the jack completely and just solder the thermistor between pin 5 of JP2 and ground.
 
Originally posted by Bryan Mayland:
It is on the external schematic, connected to probe jack JP6 (FOOD3/AMB) between pins 1 and 2.

I don't have one built this way because I only have the 3 jacks, but if you look at this picture of the probe jacks you'd solder it to your 4th jack between the ground leg and the unconnected leg on the side of the jack.

What this does is connects the thermistor if no probe is plugged in, and the probe if a probe is plugged in. You can also omit the jack completely and just solder the thermistor between pin 5 of JP2 and ground.

Ah perfect, thanks!
 
Got my v3.2 boards today. Does the 74HC595 latched shift register on the SPI bus work?


I'd say it does. The real question was if it was going to work with the RFM12B chip operating at the same time, a scenario I didn't have the parts to be able to test. It works perfectly! I've got a couple minor silkscreen tweaks to make to the v3.2 board in git, and change a dimension line so don't anyone be sending one to the fab until I've got it and the parts list updated.

To see what sort of problems folks might be running into when trying to install the bootloader, I decided to install Optiboot using the "Program from any ATmega" method described in the wiki. I took a couple pictures of the wiring and posted them in my HeaterMeter picasa (does that URL work?). I just ran the 6 wires (5 from the host Arduino/ATmega, 1 to GND) and then pressed my finger in the middle of the 6 wires in the ICSP header on the HeaterMeter board so each side's 3 wires bent a bit outward and pressed against the edges of their through hole. Hit the RESET button on the host chip and 5 seconds later I was bootloaded. I had a few more components on there when I did it but all you need is: ATmega chip, 16MHz resonator, 10k reset pullup resistor (R20).
 
So I got my MiniLink v1.0 back from Dorkbot today. Works out of box so that is great! This is my non-LCD version of the Linkmeter v3.1. The board is designed to sit upside down in non-WRT54GL routers (i.e. the non-L-shaped boards). Fits perfect in my WRT54G v3, WRT54GS v1, and WRT150N.

Link%26Mini.jpg


As you can see the transistors are backwards. That is the only major error and really not a big deal. I also was not happy with the silkscreen. They printed more than just my silkscreen layers for some reason...I need to fix that.

More Photos:
MiniLinkside.jpg


MiniLinkopen.jpg


I am working on fixing a couple errors on this board and then I will make it available. I am also working on an LCD/button version that will fit in these non-WRT54GL cases. They are all strictly based on Bryan's schematics.

UPDATE: Looks like with the board configured in this manner that the lidopen mode get triggered every couple seconds. I have looked through the heatermeter code a bit and haven't yet found what would cause this. Bryan...any ideas? The only difference between v3.1 and my board is that I removed the LCD and button components and added an LED to indicate the board is powered.
 
Originally posted by Dave S (GeoDave):
UPDATE: Looks like with the board configured in this manner that the lidopen mode get triggered every couple seconds. I have looked through the heatermeter code a bit and haven't yet found what would cause this. Bryan...any ideas? The only difference between v3.1 and my board is that I removed the LCD and button components and added an LED to indicate the board is powered.
That looks awesome Dave! I had done a few layouts myself to try to fit something into a smaller case, by building it around the serial connector. I didn't consider mounting it upside down though, what a great idea!

What you're probably seeing is due to the fact that the button pin is floating so every once in a while it floats into the range where it thinks you pressed the left button to manually enable Lid Mode. Hook analog pin 0 (ATmega pin 23) to ground.

If you're looking to do something even smaller, the only reason we use 5V is because the LCD needs it. Because you have no LCD, you can get rid of all the 5V circuitry and just run on the router's 3.3V. That also means you don't need any of the 4.7k/10k divider resistors.
 
Originally posted by Bryan Mayland:
I'd say it does. The real question was if it was going to work with the RFM12B chip operating at the same time, a scenario I didn't have the parts to be able to test. It works perfectly!
I take that back. It actually appears to hang after running for an hour or two. This will be a fun one to debug...
 
Originally posted by Bryan Mayland:
Try flashing lcdtest.hex (right-click, save as) to the avr firmware to see if it makes a difference for you. Also, if you have the LCD's D0-D3 connected try disconnecting them.

Removing D0-D3 helped. Once I disconnected those I started seeing gibberish text on the display. Not sure if that's due to a bad flash, mis-wired component or operator. Still haven't had a chance to flash the firmware you suggested. Hoping to have some time tonight to keep plugging away at this. Got my probes on Friday and getting anxious to finish this project up. Thank you so much for all your efforts and support on this Bryan.
 
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 Dave S (GeoDave):
UPDATE: Looks like with the board configured in this manner that the lidopen mode get triggered every couple seconds. I have looked through the heatermeter code a bit and haven't yet found what would cause this. Bryan...any ideas? The only difference between v3.1 and my board is that I removed the LCD and button components and added an LED to indicate the board is powered.
That looks awesome Dave! I had done a few layouts myself to try to fit something into a smaller case, by building it around the serial connector. I didn't consider mounting it upside down though, what a great idea!

What you're probably seeing is due to the fact that the button pin is floating so every once in a while it floats into the range where it thinks you pressed the left button to manually enable Lid Mode. Hook analog pin 0 (ATmega pin 23) to ground.

If you're looking to do something even smaller, the only reason we use 5V is because the LCD needs it. Because you have no LCD, you can get rid of all the 5V circuitry and just run on the router's 3.3V. That also means you don't need any of the 4.7k/10k divider resistors. </div></BLOCKQUOTE>

Smallest that will fit into any Linksys router is the goal (non-smd). The MiniLink v1.0 boards were $6 a piece from Dorkbot. I learned a lot on this one which I am using to improve the MiniLink v1.1. I currently have it down to about $5.30 per board after some tweaks last night. Thanks for the suggestions to use the routers 3.3v...didn't think of that! This will eliminate a couple parts and might get it below $5 a board.

I will let you know if grounding out the pin works...I am assuming it will. I guess I could set the button left to zero instead of the value of atmega pin 23...at least for these boards. Do you want me to send you a board Bryan? Didn't know if you had any use for it...I know you probably have a bunch of boards by now.

UPDATE: I jumpered the wires and all works great! I am almost done with MiniLink v1.1 which includes a couple changes including grounding atmega pin 23 and using 3.3v from the router. I think after this we will be looking at $4.50 a board and it should work in any Linksys router. If you noticed on the image above there are two spots to solder the female pin headers. This is because while the WRT54G v3 has ground on the right side header pin, the WRT54GS v1 has the ground on the left side. This give you some flexibility in mounting the board.

P.S. http://capnbry.net seems to be down at least since lastnight...I can't get my snapshot!
icon_frown.gif
 
Originally posted by Dave S (GeoDave):
UPDATE: I jumpered the wires and all works great! I am almost done with MiniLink v1.1 which includes a couple changes including grounding atmega pin 23 and using 3.3v from the router. I think after this we will be looking at $4.50 a board and it should work in any Linksys router. If you noticed on the image above there are two spots to solder the female pin headers. This is because while the WRT54G v3 has ground on the right side header pin, the WRT54GS v1 has the ground on the left side. This give you some flexibility in mounting the board.

P.S. http://capnbry.net seems to be down at least since lastnight...I can't get my snapshot!
icon_frown.gif
Huzzah! That board is shrinking by leaps and bounds. If you want to send me the schematic before it goes to fab, I can check and see if there's any other superfluous components that aren't needed if there's no LCD/Buttons. Thanks for the offer of a board but you're right, I have a ton of these things sitting around, from prototype to alpha to beta to production.

As for capnbry.net being down... son of a *****. We *just* had the ISP come in last week and replace equipment which was supposedly causing our outages. Guess they're coming back!
 
The ET72 is definitely the model number that I used to order my probes from Maverick, and I believe the ET73 are the same (with the exception of the probe and cable lengths). The ET732 we've found is completely different, but we have coefficients for those now too so what could possibli go wrong?

I picked up 2 of those IKEA probes while I was there buying a shelf for HeaterMeter Labs and they seem to work ok as at room temperature so far. Will have to pop them in the oven later.
 
Originally posted by Bryan Mayland:
If you want to send me the schematic before it goes to fab, I can check and see if there's any other superfluous components that aren't needed if there's no LCD/Buttons.

Here are the links to the board and schematic:

https://docs.google.com/open?i...YzYtMDRhYWFlYmYwOGFl

https://docs.google.com/open?i...MTMtMzhjNDE2YzEzOWUx

Not happy with the routing yet so I haven't finished that (i'm at 98.8% with no vias but can't seem to get 0 vias). I also need to change the CAM job to match my needs.

By the way...the long female header pin is just for support and to prevent the components from touching the routers PCB. I thought I would make the remaining pins available through the support. Got any better ideas?
 
Originally posted by Dave S (GeoDave):
Here are the links to the board and schematic:

By the way...the long female header pin is just for support and to prevent the components from touching the routers PCB. I thought I would make the remaining pins available through the support. Got any better ideas?
Wow that is super compact. Biggest thing is R2 and R21 should be removed. They were to bring the 5V TX output down to around 3.4V for the router, but since you're already at 3.3V you don't want to take down the TX voltage. Two less components! In their place you might want to add a small 22u/6.3V capacitor to the 3.3V power line just as a small buffer for the board.

I can't think of anything better than running the unused pins to that extra header, so what you've got there looks good to me. A0 you've got pulled to ground though so I don't know how useful that one is.

EDIT: And I just got a call saying our Internet has been restored so the snapshot repository should be back up. Bunch of tubes my ***!
 
Originally posted by Bryan Mayland:
Wow that is super compact. Biggest thing is R2 and R21 should be removed. They were to bring the 5V TX output down to around 3.4V for the router, but since you're already at 3.3V you don't want to take down the TX voltage. Two less components! In their place you might want to add a small 22u/6.3V capacitor to the 3.3V power line just as a small buffer for the board.

I can't think of anything better than running the unused pins to that extra header, so what you've got there looks good to me. A0 you've got pulled to ground though so I don't know how useful that one is.

EDIT: And I just got a call saying our Internet has been restored so the snapshot repository should be back up. Bunch of tubes my ***!

Darn...I thought I got all of the LCD stuff out of there. Well at least I am learning...I should have consulted with you first. Either way I do have a working board which I am going to put in a router and ship off to my mom.

So a capacitor will help prevent fluctuations in voltage on the routers serial port...got it.

A0 is tied to ground because of the crazy button issue...should I deal with that another way?

Yeah...it is as small as I can get and stay within the DorkbotPDX design specs. No problems with solder bridges on the current board which had a couple very close pads. Can you see any disadvantages to this?

PS The site is running SLOW
icon_frown.gif
10 hours estimated to download the bin
 
Just managed to flash my linksys and so far everything looks fine
icon_smile.gif

I saw in the presets that you even added the Ikea probes... that's really fast!

Now I'll have to make the heatermeter board, but i think i'm going to wait for the "compact version".
While waiting, I wonder if it's possible to just use my excisting Arduino and connect it to the serial port of my linksys.

Is it just connecting two pins of my arduino with the serial tx and rx of the linksys, or do I have to do something more?
 
Just wanted to check back in, I was able to reset the configuration of the router(finally got putty to work so I can ssh in) and now the pages work again. Also figured out how to burn the bootloader using the burn from any atmega method. I did notice with doing that I had to uplug the ardiuno I was using from the pc and plug it into a 12v power supply to get it to burn the bootloader. Now when the heatermeter board is plugged into the linksys router the temp readings say off instead of --- so I'm guessing its programmed correctly now. My LCD doesnt work, going to have to do some more troubleshooting on it tonight. But progress
icon_smile.gif
 
Originally posted by Arwin:
While waiting, I wonder if it's possible to just use my excisting Arduino and connect it to the serial port of my linksys.
Is it just connecting two pins of my arduino with the serial tx and rx of the linksys, or do I have to do something more?
Yeah! Sort of. The RX line can connect directly, but if you're running the Arduino at 5V (what you'll have if it is plugged in to USB or the barrel jack) you need to lower the TX to 3.3V by using a 4.7k/10k ohm voltage divider 4.7k goes in the Arduino TX pin, 10k goes to ground, where they meet goes to the router's RX.

I *think* you can actually flash your Arduino with HeaterMeter, then unplug it and run the router 3.3V line into the pin labeled 5V on the Arduino, GND to GND, and then connect RX/TX directly. Your Arduino should have no problem running 3.3V and it can't hurt to try. Just makes sure you don't plug it into a 5V source at the same time or you will most likely kill your router.
 
Originally posted by Bryan Mayland:

I *think* you can actually flash your Arduino with HeaterMeter, then unplug it and run the router 3.3V line into the pin labeled 5V on the Arduino, GND to GND, and then connect RX/TX directly. Your Arduino should have no problem running 3.3V and it can't hurt to try. Just makes sure you don't plug it into a 5V source at the same time or you will most likely kill your router.

I "accidentally" let 5V touch the 3.3V serial connection on the router and I can with absolute certainty say it does kill your router. No magic smoke, but the next time you turn the router on you will have lit LEDs, but no activity.
 
Bryan...made some tweaks per your suggestion. I want this to be a final design so I am trying to get it right this time.

Here are the link: MiniLinkv1.1.zip

It would be great if you could check it out.

Do you have another suggestion other than grounding A0? It could probably be fixed in software but probably not worth the extra work.

I tried the new Lid Open button...awesome! It didn't work at first on my MiniLink v1.0 board. I had to switch out the Atmega328 chip. It was hanging on toggle for some reason. Maybe a bad flash?
 
Well thanks for the advice on the voltage divider, pretty sure I would have connected them directly to each other...

Isn't it strange that there are different voltages for the serial lines? Always thought that computers were standarized..

I also have another question:

Here in europe, we use degrees Celsius instead of Fahrenheit; therefore my current controllers show Celsius. Is there an easy way to let linkmeter show Celsius? I know how to change the code in the heatermeter but I don't think that will be enough for the linkmeter...

Thanks again!
 
Status
Not open for further replies.

 

Back
Top