SMD Heatermeter on a RPI B+? Yes, it can be done......


 
Hmm, I hear you on the inductor... I was going to make suggestions but not having the board and rPi in my hands my suggestions would probably be irrelevant.

I'm sorry to hear the TC has some noise issues... I see that a lot of the power components have been pushed down the edge of the board and are now very near the TC amp (the inductor you referred from the blower feedback circuit and the large electrolytic filter caps on the power supply). I would guess this is the source of your noise, I would try to push them all down to the other end of the board if possible.

Also, I am wondering if this board has the traces running from the CAT5 jack to the probe jacks on board? If so, they could be picking up some noise somewhere too...

I would work on the TC noise issue though, because the Pit Probe is the one probe where noise is a real problem, if your pit temp is rolling up and down because of noise that will make a servo damper go nuts with constant adjustments trying to level things out.
 
Because I wasn't happy about the noise on the TC signal I decided to take some measurements to find out what the source of the noise was and to see if an easy fix was possible.

All measurements were taken at 20% output at "voltage" setting with a 6.5CFM blower attached.

Noise at output AD8495 before mod.
Pit%20probe%20noise%20before%20mod.png


Noise at 3.3V line before mod.
Noise%20on%203.3V%20line%20before%20mod.png

The frequency of the signal is the same as the switching frequency of the PWM motor driver.

Noise at the TC connector before mod.
TC%20connector%20noise%20before%20mod.png


I suspected that the position of the 220uH inductor was to close to the TC ampifier, to prove my point I moved the inductor all the way to the other side of the board just to see if this would reduce the amount of interference.

Noise at output AD8495 after the mod.
Pit%20probe%20noise%20after%20mod.png


Noise at 3.3V line after the mod.
Noise%20on%203.3V%20line%20after%20mod.png


Noise at the TC connector after the mod.
TC%20connector%20noise%20after%20mod.png


Moving the inductor has a huge impact on the signal quality, no more yellow squiggly lines on the web interface!
There is still some noise left on the 3.3V line that I would like to get rid of but this will require a resigned of the PCB.

So after the infamous "white wire" mod I'm proud to present the equally infamous "black wire" mod!

5e%20batch-6.jpg


5e%20batch-7.jpg
 
I am attempting to build some similar SMD HeaterMeters to mate with an RPI B+ (currently waiting on one diode from Mouser before they can be finished). What version of the OpenWRT image are you using? Is it the A+ snapshot image linked to in the software page in the Wiki? If so, when did you download yours? I'm trying to use the snapshot image dated 2015-05-31 and it hangs on bootup on my RPI B+. (Freezes on "Freeing init memory: 100K". I'm tried booting with and without the WiFi adapters plugged in, and it makes no difference. The WiFi adapters (Edimax EW-7811UN and EDUP N8531) were pre-tested on a working HeaterMeter with a regular RPI B, so I know they work. I took the same image and plugged it into an RPI B and it booted up just fine, with and without a WiFI adapter plugged in. I am testing the image on the RPI's without the HeaterMeter boards attached...would that be a problem?
 
Just orderd my SMD boards, anxious to try out my new hot air soldering station and solder paste. Also tempted to hack a toaster into a reflow oven. Should be fun!
 
Jas E said:
I am attempting to build some similar SMD HeaterMeters to mate with an RPI B+ (currently waiting on one diode from Mouser before they can be finished). What version of the OpenWRT image are you using? Is it the A+ snapshot image linked to in the software page in the Wiki? If so, when did you download yours? I'm trying to use the snapshot image dated 2015-05-31 and it hangs on bootup on my RPI B+. (Freezes on "Freeing init memory: 100K". I'm tried booting with and without the WiFi adapters plugged in, and it makes no difference. The WiFi adapters (Edimax EW-7811UN and EDUP N8531) were pre-tested on a working HeaterMeter with a regular RPI B, so I know they work. I took the same image and plugged it into an RPI B and it booted up just fine, with and without a WiFI adapter plugged in. I am testing the image on the RPI's without the HeaterMeter boards attached...would that be a problem?
Click to expand...

I'm using the a+ snapshot from the wiki, the file date is July 10 2015. However, this could also be the date that the file was unzipped.
 
John Bostwick said:
If you are looking for this smd version by Peter on OSH, I have it already up

https://oshpark.com/projects/hXIwbpVr
Click to expand...


So, I ordered and received those oshpark boards today. Starting work, and noticed that a lot of the components don't have any text to indicate value like all the ones near the jacks. In the above pictures of the board, the text is there. I'll have to use it as reference, but I'm assuming there have been a few changes?
 
The 9 components near the probe jacks are 3 sets of components repeating 3 times, so from top to bottom its,
10k
100K
0.1u
10k
100k
0.1u
10k
100k
0.1u

Due to lack space for the text I had to do it like this.

For some reason the + indication for the LED's is also missing. The long legs (+) of the 3 LED's have to be in the right side hole (with the board lying top side up and with the jacks on the right side)
 
Peter F said:
The 9 components near the probe jacks are 3 sets of components repeating 3 times, so from top to bottom its,
10k
100K
0.1u
10k
100k
0.1u
10k
100k
0.1u

Due to lack space for the text I had to do it like this.

For some reason the + indication for the LED's is also missing. The long legs (+) of the 3 LED's have to be in the right side hole (with the board lying top side up and with the jacks on the right side)
Click to expand...

Thanks, I'll get the rest of the parts on a board today and test it out. On two boards I realized I had my hot air gun at 350 C.. yikes. Hoping I didn't cook the Atmega and shift register. Luckily I didn't go much farther. I love hot air and solder paste.. very slick!
 
Peter F said:
The 9 components near the probe jacks are 3 sets of components repeating 3 times, so from top to bottom its,
10k
100K
0.1u
10k
100k
0.1u
10k
100k
0.1u

Due to lack space for the text I had to do it like this.

For some reason the + indication for the LED's is also missing. The long legs (+) of the 3 LED's have to be in the right side hole (with the board lying top side up and with the jacks on the right side)
Click to expand...

Near the thermocouple jack, it looks like it goes 10K 10K 100K for the resistors.. but on your boards are you omitting the middle 10K?
 
The second 10k resistor is only necessary when you build a 4 probe version. For a TC version this resistor needs to be left out , if you leave it in then it will interfere with the TC amp.
 
Thanks. Just did a powerup with a PI A+ on it and just got black bars on the lcd. Zoomed in on the atmega chip with a camera and realized I never cleaned off the leftover solder paste residue. Scrubbed with IPA and it fires right up! Totally stoked!
 
Spoke too soon. It runs for a few seconds and then glitches out on the display. I've tried resoldering the atmega, and it seems to help for a bit. I had it stable for a while stuck on No Pit Probe. Replaced the atmega, same issue. Gotta be something glitchy on the board, just have no idea where to look.
 
I have completed my build based on Peter F's SMD board. I had a little trouble because I was using the USB port for power. As soon as I used the 2.1mm DC power source, everything booted up just fine. It's working great on an rPi B+. Now all we need is working software so we can also use rPi 2 boards, which are now cheaper and more available.

Naturally, in order to make this cost effective, I had to order a few boards and components more than I needed. I have soldered up a couple of extra boards if anyone is interested. PM me.
 
You must log in or register to reply here.

 

Back
Top