New build - No power


 
I've got good news and bad news. The good news is that I'm pretty sure that glob on my OKI was the problem. I removed it and measured readings at 5.30 (meter set to 20K Ohms) on the 5V legs, compared to all zeroes before. The bad news is that I fried off some metal contacts on the board and also broke the middle leg of the OKI. What happened was that after I removed the OKI, the holes on the board were still filled with solder. I tried wicking the solder out and just couldn't get it all out so that I could fit the OKI back in. The contacts came off, and I decided to use some left over bare pins from the LCD part of the build to punch through the solder by squeezing with a needle nose pliers. That worked fine. However, when I went to put the OKI back in, there must have been some solder left on the middle pin, because that one buckled under as I applied plier pressure. I tried to straighten it out but it broke off. I think I can salvage the part by just desoldering that pin and adding another. If not, I'll just order a new one. However, am I screwed with the board having no metal contacts on either side for that one hole? the middle hole at least has a contact on one side, but I have no idea if that is any good. Please check out pics. Literally 1 step forward, 2 steps back....


Board Front - https://drive.google.com/file/d/0BzSBmZGY0VqYbHZIdFpQM3EyYjQ/view?usp=sharing
Board Back - https://drive.google.com/file/d/0BzSBmZGY0VqYNWF6LUh0WGlTcHc/view?usp=sharing
OKI - https://drive.google.com/file/d/0BzSBmZGY0VqYOEFwV0Y0NVVvZFU/view?usp=sharing
 
Well damn... I would have just took the soldering iron and solder sucker or solder wick and heated up the one leg and sucked out the blob of solder and moved on once the short went away, rather than attempting to remove the OKI completely for no good reason.

The good news is as long as the OKI board is still OK you can fix things up with jumper wires.

The center pin is GND, it is the only pin that has a connection on the top side of the board. Check for continuity from the center pad (on the top side of the board) to any ground, if that's good that's all that matters. There is no connection on that missing pad on the bottom of the board. If you do not get connection to ground then add a jumper wire to ground. With that mangled center pin on the OKI, you might just solder a wire to it and run the other end of the wire either to the original spot or another ground junction on the board.

The pin toward the CAT5 jack is the 12V input, it connects to the capacitor and diode right next to it. The trace looks ok there, but if you do not have continuity from the cap next to it then add a short jumper wire there

The pin toward the ATMega is the 5V output. I has two traces coming off of it. One goes over to the capacitor right next to it, and the other goes to the CAT5 jack to power the servo. Check continuity to both locations, if not present then add jumper wire(s)
 
Last edited:
Needle nose pliers and electronics never go together and have a happy ending.

All you needed to do was use a soldering iron to heat up the solder and while using the Wick, the solder would have been wicked away.

Or you could have heated all 3 pins together with the oki upside down and it would have dropped out, then you could remove excess solder with the Wick.
 
RalphTrimble said:
Well damn... I would have just took the soldering iron and solder sucker or solder wick and heated up the one leg and sucked out the blob of solder and moved on once the short went away, rather than attempting to remove the OKI completely for no good reason.

The good news is as long as the OKI board is still OK you can fix things up with jumper wires.

The center pin is GND, it is the only pin that has a connection on the top side of the board. Check for continuity from the center pad (on the top side of the board) to any ground, if that's good that's all that matters. There is no connection on that missing pad on the bottom of the board. If you do not get connection to ground then add a jumper wire to ground. With that mangled center pin on the OKI, you might just solder a wire to it and run the other end of the wire either to the original spot or another ground junction on the board.

The pin toward the CAT5 jack is the 12V input, it connects to the capacitor and diode right next to it. The trace looks ok there, but if you do not have continuity from the cap next to it then add a short jumper wire there

The pin toward the ATMega is the 5V output. I has two traces coming off of it. One goes over to the capacitor right next to it, and the other goes to the CAT5 jack to power the servo. Check continuity to both locations, if not present then add jumper wire(s)
Click to expand...

Yeah, I'm a bumbling fool. I tried heating just those pins on the OKI with the wick and it didn't seem like I was getting any of the blob out of it. I couldn't see past the front of the pins so thought it would be a good idea to remove it all together to make sure I got all of the solder that may have been there. I should have used the sucker. I think I know what to do at this point.
 
Yah, I have seen this a few times now, people tend to want to reset a bad situation by removing a part and holding it in their hand, then starting over and doing things "right". Not realizing that removing parts, particularly multi-leg parts, is much harder than installing them.... specially without damaging the part or the board in the process.

Recently I suggested someone remove the ATMega to reseat it and they removed the whole socket! YOW! I have since made sure to say "remove the ATMega from its socket" if I suggest that move! It's really hard to anticipate what someone is going to do and say the right things to prevent a situation like this. It would be better to just ask the question if you are unsure, "should I remove the whole socket?"....

This should have been a 5 second fix, a quick tap on that blob of solder with solder wick or solder sucker and that blob of solder would have been gone along with the short. I guess in the future we need to stress the concept of "touch-up" solder work and hope that people get the picture more clearly that you don't need to rip a part off the board to clean up the soldering....
 
What's the best type/size of wire to use in my situation if I need to? I have some stranded wire that is in the neighborhood of 18-22 gauge. I have that left over from wiring a switch on our doorbell chime box for when the kids are napping. That project definitely made the success column! This one will too, just taking longer than expected :)
 
Yah, you'll get there with the HM, seems you are pretty close at this point...

I would shy away from stranded wire because it can easily frey and cause short circuits with work like this. Stranded wire is good for carrying more current and being more flexible, neither attribute is a plus for this situation.

I think the doorbell wire should be what you want if I recall properly the wire type. You want a small solid core wire like the wire inside a solid core CAT5 cable, or old school telephone wire (for wiring telephone jacks in houses), I think doorbell wire is pretty much the same as old telephone wire?
 
RalphTrimble said:
Yah, you'll get there with the HM, seems you are pretty close at this point...

I would shy away from stranded wire because it can easily frey and cause short circuits with work like this. Stranded wire is good for carrying more current and being more flexible, neither attribute is a plus for this situation.

I think the doorbell wire should be what you want if I recall properly the wire type. You want a small solid core wire like the wire inside a solid core CAT5 cable, or old school telephone wire (for wiring telephone jacks in houses), I think doorbell wire is pretty much the same as old telephone wire?
Click to expand...

Wait, so you're saying there is actual doorbell wire that I didn't use in my doorbell project!? LOL.

I'll be able to pick some small solid core wire up this afternoon.
 
I thought that was what you were saying... at any rate, thin solid core wire is what you want. It is the easiest wire to piggyback solder to a solder joint to remake a broken trace.

The traces you may need to fix are very short runs, just a few mm's each, so you could even use the cut off legs from resistors or whatever from your build (or from spare resistors). You could, for instance, solder a spare resistor into the center hole on the board (from the top where the trace is), leaving the majority of the bare lead on the top side of the board. Then put the OKI in the board, solder in the outer two legs, then solder the protruding resistor leg from the center hole to the damaged center leg of the OKI and then cut off the resistor. Or do the same thing with just the cut off leg from a resistor or diode or whatever.

If the traces on the outer two legs end up being broken and not making the proper connections you can do the same. The easiest thing to do is use a spare resistor (if available) as described above, that gives you something to hold on to. Bend the leg so it is easy to hold the resistor while making the short jump from the OKI pin to the capacitor leg beside it, solder the resistor leg in both places and then cut off the resistor... or, again, just use a cutoff leg from a component....
 
RalphTrimble said:
Yah, you'll get there with the HM, seems you are pretty close at this point...

I would shy away from stranded wire because it can easily frey and cause short circuits with work like this. Stranded wire is good for carrying more current and being more flexible, neither attribute is a plus for this situation.

I think the doorbell wire should be what you want if I recall properly the wire type. You want a small solid core wire like the wire inside a solid core CAT5 cable, or old school telephone wire (for wiring telephone jacks in houses), I think doorbell wire is pretty much the same as old telephone wire?
Click to expand...

Actually solid wire carries slightly more current for a given wire gauge (unless you are talking high frequency) because stranded wire has a slightly higher resistance than solid for a given gauge. At high frequencies the "skin effect" comes into play where the current is only conducted on the very outside layer of the wire so in this case stranded wire does carry more current because there is much more surface area of copper. This is the whole concept behind waveguide for microwave frequencies...they are made of tubing because any material other than the wall of the tubing is useless for signal transmission. Given all that, unless it is being used for RF frequencies and above the choice between stranded and solid boils down to flexibility issues with stranded being more flexible as Ralph said.
 
Last edited:
Completed the fix, it was actually pretty easy, thanks for the how-to. I had saved my clipped resistor ends so that was the perfect solution.

Fix - https://drive.google.com/file/d/0BzSBmZGY0VqYOUhaOXlsVFdJUFk/view?usp=sharing

After making sure I had continuity from the OKI, I mated the HM to the RPi and voila, it powered up! However, I am seeing my LCD show 22 degrees F for the pit probe when it should read "-No Pit Probe-". I can see both devices on my network. How should I proceed from here? New thread, or is this a power problem still?

LCD - https://drive.google.com/file/d/0BzSBmZGY0VqYSFVLMFFBOVh0a28/view?usp=sharing

Edit: After about 10-15 minutes plugged in the Pi side felt pretty warm, wouldn't call it hot. Not sure that's typical or not, but I have unplugged it for the night.
 
Last edited:
David Crawford said:
I would go look at the Thermocouple circuitry and look for solder bridges. When you first built this did you go thru this procedure?:

https://github.com/CapnBry/HeaterMeter/wiki/Thermocouple-Amplifier-Testing

Just be sure to read the cautions about testing after the rest of the HM is built. Good progress though!
Click to expand...

I did test before moving on with the rest of my build. Here's what I wound up with:
http://tvwbb.com/showthread.php?62120-Testing-the-Thermocouple-Amp

I'll double check for post-test solder bridge tomorrow evening.
 
I thought about this more and the only way that you can be getting a temp reading with no thermocouple connected is that you must have a solder bridge somewhere that shorts pins 1 and 8 on the AD8495 or between there and the thermocouple connector pins. The thing that is baffling is that the way this circuit is wired it should not be able to put out a value that would equate to a temp lower than 41 degrees F unless you have a negative offset value loaded for the pit probe on your configuration page.

I just duplicated your reading on my HeaterMeter by shorting the thermocouple connector pins and entering a -58 offset on the configuration page for the pit probe.

I would also make sure that pin 2 on the AD8495 is measuring as a solid ground (actually pins 2 & 3 should be grounded). Pin 2 is the reference input pin, pin 3 is ground for the amp and I am not sure what would happen if either was just "floating".
 
Last edited:
Did you make sure you have thermocouple selected in the GUI instead of another type of probe? Something interesting about this whole thing. I added a RDTC board to my setup and I get a temp reading without a TC probe plugged in. whether or not the RDTC is actually connected to the HM or not. I assure you there are no solder bridges on my setup, just odd behavior. Not sure if it has something to do with the resistor that is disconnected for the RDTC board, but when I put things back together it works correctly.

Charles
 
Yes, the first thing you need to do is ensure that Thermocouple mode is the type selected in the configuration GUI. Shorting the probe inputs should force the probe to real ambient / room temperature.
 
I just did some more experimenting with my HM. With nothing attached to any probe input, I selected each probe type in the GUI and in each case the result was "No Pit Probe" just as you would expect (I do not have a thermistor internal, I use an external one). I then shorted the TC connector pins and selected each probe type and the results were: TC 80, Internal 318, RF "No Pit Probe". I did select Disabled also and with nothing connected got the "No Pit Probe" display. I then decided to test the mode where the HM finds the first available probe to use for the Pit Probe and connected a probe on each of the probe inputs (other than TC). This function is available if you have loaded the LinkMeter Snapshot and it worked just as expected.
 
Dominick A said:
Completed the fix, it was actually pretty easy, thanks for the how-to. I had saved my clipped resistor ends so that was the perfect solution.

Fix - https://drive.google.com/file/d/0BzSBmZGY0VqYOUhaOXlsVFdJUFk/view?usp=sharing

After making sure I had continuity from the OKI, I mated the HM to the RPi and voila, it powered up! However, I am seeing my LCD show 22 degrees F for the pit probe when it should read "-No Pit Probe-". I can see both devices on my network. How should I proceed from here? New thread, or is this a power problem still?

LCD - https://drive.google.com/file/d/0BzSBmZGY0VqYSFVLMFFBOVh0a28/view?usp=sharing

Edit: After about 10-15 minutes plugged in the Pi side felt pretty warm, wouldn't call it hot. Not sure that's typical or not, but I have unplugged it for the night.
Click to expand...

Congrats man, hurdle cleared!

ON the TC, I have to ask if you selected Thermocouple for the probe type on Probe 0?
 
Thanks guys! I have not touched ANYTHING on the GUI yet. I paused when I thought I should have seen "no probe" but saw a temp reading. I will get into the GUI tonight and adjust settings as needed. I should note that I had this solder blob on the middle food probe jack prior to fixing my earlier solder bridge problems. I later wicked out the blob, but perhaps I didn't get it all. Would this cause a bad reading on the pit probe at this point?

https://drive.google.com/file/d/0BzSBmZGY0VqYXzhrVDY0V2xJTkE/view
 
It's been a while since I did a fresh build so I dont recall the default setting for PROBE 0 (pit probe), but I think by default it is set to "Internal" rather than "Thermocouple". You can get a false reading if you have the wrong probe type selected... So I would first go in and check your HM config and make sure you have TC selected before I go messing with the board further.
 
You must log in or register to reply here.

 

Back
Top