Proper wiring for HM Case with Switch.


 

Dominick A

TVWBB Fan
I have a TC HM board with a switch with the RDTC box. So I am crystal clear on the switch wiring, can someone please confirm these steps as I work from top to bottom on the image above:

1) Wire ground from the 12V jack to pin 8 of the Cat5 jack
2) Wire pin 1 of the Cat5 jack to the offset pin of the food probe adjacent to the TC probe
3) Wire pin 2 of the Cat5 jack to the offset pin of the middle food probe
4) Wire pin 7 of the Cat5 jack to the top leg of the black side of the switch
5) Wire the middle leg of the switch to the resistor side of the 100K resistor behind the TC jack
6) Wire the bottom leg of the switch the wire side of the 100K resistor behind the TC jack

Assuming this is correct, does it make a difference which side of the switch I wire to? I just want to make sure as I look at the orientation of the switch in relation to the pins I need to wire from as I do a test fit in the case to see what the easiest route is to wire this up.

I finally wired everything up as I planned in the previous thread quoted above. After plugging in the wall wart, the pit probes are either showing "No Pit Probe" both when a TC probe is plugged in and when it is not, or, it is showing a temperature that is steadily going up, sometimes starting at 300, 400, or even room temperature, but continues moving, never settles. This is the same even when I toggle the switch, or remove and then re-insert the TC probe.

Initially, when I had the HM/Pi seated firmly in the case, my Probe 2 was showing a steady 44F. When I plugged a probe in, it would show the proper temp. Probes 1 and 3 were fine. I later realized that the compression of the wire connected to probe 2 while seated in the case was the cause. It seemed that the wire was pushing into some of the pointed solder joints on the board while compressed. While this problem went away when unseating from the case, the wild pit probe readings, regardless of the toggling of the switch, remained.

One other possible issue I noticed happened while soldering the middle leg of the switch. While heating the middle leg of the switch to solder, I must have overheated it as it rose up out of the plastic housing it is connected to. I tried pushing it back in and got it nearly back to its origin. I'm wondering if I broke the switch since the fluctuations and readings (or non-readings) are happening on both sides of the swtich??

What else can I do to troubleshoot this?

Would anyone wire this up differently to avoid possibly puncturing the wires while seating in the case? Given my current wiring, I also can't fully seat the HM in the case. The Cat5 jack sits up about an 1/8 of an inch or so.

Switch connections - https://drive.google.com/file/d/0BzSBmZGY0VqYdVAwWmpFTDdGWUk/view?usp=sharing
Wire runs - https://drive.google.com/file/d/0BzSBmZGY0VqYREZScjNvVjFzUlU/view?usp=sharing
Connections Close up - https://drive.google.com/file/d/0BzSBmZGY0VqYQzBYNkE5QUlRbkE/view?usp=sharing
Not fully seated - https://drive.google.com/file/d/0BzSBmZGY0VqYWS10bnE2cWlXQUE/view?usp=sharing
 
Either resolder everything a little flatter down on the board using thinner flexible wire and less globby solder points.... Or melt out a little depression in your case to accomodate the protruding wires.

What diameter solder are you using?
 
First of all, you used stranded wire for the jumpers. That is most likely to fray and cause shorts, or just be too bulky to solder where you need it to make the connection without shorting out on other connections. Thin SOLID CORE wire, like telephone wire or the wire inside a solid core CAT5 cable will work out better.

The TC rising on it's own (Not following actual temperatures) sounds like the TC amp is not connected to the ATMega. Most likely from what you describe you melted the switch and destroyed the contacts. You are not welding, you are soldering! Crimp the wires in place then solder them quickly and get that soldering iron off the pin before you melt things. To test the switch just short out the center lug to whichever TC amp you have the TC plugged into. If you get a proper TC reading then but not with the switch in either position then you melted the switch. The good news is the switch has two halves, so you can just move the wires to the other half (if the switch isn't melted too much to click into each position). This time solder more quickly so you dont melt your switch.
As for the board not seating in the case. Make sure your probe jacks are all mounted FLAT on the HM board, if either end of the probe jack is sitting up off the board the jack with sit cockeyed and prevent the board from fitting well. You can try trimming up the plastic a bit around the probe jack holes on the case to let the board push a little further in that direction, which will make the board seat properly on the other end.
 
Last edited:
First of all, you used stranded wire for the jumpers. That is most likely to fray and cause shorts, or just be too bulky to solder where you need it to make the connection without shorting out on other connections. Thin SOLID CORE wire, like telephone wire or the wire inside a solid core CAT5 cable will work out better.

The TC rising on it's own (Not following actual temperatures) sounds like the TC amp is not connected to the ATMega. Most likely from what you describe you melted the switch and destroyed the contacts. You are not welding, you are soldering! Crimp the wires in place then solder them quickly and get that soldering iron off the pin before you melt things. To test the switch just short out the center lug to whichever TC amp you have the TC plugged into. If you get a proper TC reading then but not with the switch in either position then you melted the switch. The good news is the switch has two halves, so you can just move the wires to the other half (if the switch isn't melted too much to click into each position). This time solder more quickly so you dont melt your switch.
As for the board not seating in the case. Make sure your probe jacks are all mounted FLAT on the HM board, if either end of the probe jack is sitting up off the board the jack with sit cockeyed and prevent the board from fitting well. You can try trimming up the plastic a bit around the probe jack holes on the case to let the board push a little further in that direction, which will make the board seat properly on the other end.

Thanks Ralph, I think I have some spare Cat5 cable that I can open up and split out. Yes, my soldering skills still need some work, felt like it was a quick touch, but obviously need to make it quicker! I should be able to test the amp/lug tonight.

Jas E - I'm using .032 diameter solder.
 
You want solid core wire, some CAT5 cable is solid, other is stranded... Stranded is better in that it wont break as easy if bent back and forth, but can fray and short out easily when used for jumpers like this. Since the wires don't move once the case is together solid core wire is fine, and much easier to work with. Just don't bend it back and forth a whole bunch or it will break.
On the switch put the wire through the lug, bend it over, crimp it with some pliers so it makes a good physical connection to the lug, then solder it in place with a quick application of solder. Tin the soldering iron tip, touch the wire/lug, immediately feed solder onto the iron/wire and pull the soldering iron off the lug as soon as solder flows to it... Even if you have to go back and touch up, better to solder lightly twice and let the part cool in between then solder the hell out of it once and melt your part.
 
Good news - I picked up a foot of solid wire 24 gauge cat5 cable at Home Depot for $0.31.

Bad news - In trying to replace the wire that is in this step, "6) Wire the bottom leg of the switch to the wire side of the 100K resistor behind the TC jack", I pulled off the trace. Initially, it was showing an ever increasing temp, but I think I may have really killed it when I accidentally started to resolder that joint while it was plugged in to the 12v wall wart. My HM now shows No Pit Probe when the probe is plugged in. None of the food probes are registering temps anymore either.

I am not wired to the switch here. I took the wire from the resistor and connected directly to the wire that I was replacing next to it to eliminate the switch as the variable here.

Did I really kill this thing? If not, where would I need to jumper this wire to where I lost the trace?
 
Last edited:
IDK why you would solder on a board that is powered, the soldering iron is metal and will conduct and can short things out easily. That is just not a wise move, just pull the power cord and solder, then plug back in, not very difficult.
Each probe has an independent connection to the ATMega, so off the top of my head I would guess that you shorted out the 3.3v to gnd and may have killed the 3.3v regulator. Start back at zero, measure your 12v, 5v and 3.3v and go from there.... I
 
IDK why you would solder on a board that is powered, the soldering iron is metal and will conduct and can short things out easily. That is just not a wise move, just pull the power cord and solder, then plug back in, not very difficult.
Each probe has an independent connection to the ATMega, so off the top of my head I would guess that you shorted out the 3.3v to gnd and may have killed the 3.3v regulator. Start back at zero, measure your 12v, 5v and 3.3v and go from there.... I

It was totally an accident, I had flipped it over and after a short time forgot I still had it plugged in. I tested the voltages and everything passes from 12v through 3.3 on the regulator. Any other thoughts?
 
I am not going to dig out the board and schematic at this hour to look, but my first guess would be there is a trace where 3.3v is delivered to the TC amp and the pullup resistors that has been severed and 3.3v isn't making it over to that side of the board to provide that power which is required.
 
I am not going to dig out the board and schematic at this hour to look, but my first guess would be there is a trace where 3.3v is delivered to the TC amp and the pullup resistors that has been severed and 3.3v isn't making it over to that side of the board to provide that power which is required.

LOL....I didn't expect you too, I just wanted to get my thoughts down before calling it a night.

Using this schematic, I diagnosed that I am getting 3.3v in each of the orange points, except for those circled in red. The trace that came off was the top right trace in the middle circle. I can see that I would need to jump that to the spot directly left and that should take care of delivering 3.3V to the red circle on the ATMega.

It also looks like it should run to the bottom left set of 6 holes behind the food1 port, which should take care of the top circle. Beyond that, I'm a little fuzzy as to if I would need jumpers for the path of the blue lines.

hm-424.png
 
Last edited:
OK, sounds like I was right about the 3.3v trace that delivers power to the probe end of the board being broken somewhere... I was trying to give you some bread crumbs and give you a nudge to dig in yourself and try to figure this one out... Glad you took the first steps in diagnosing the issue and are providing some good information here.

Some time ago another forum member had an issue with 3.3v and I made up the following diagram. The yellow line shows the route the 3.3v takes to reach that side of the board. The yellow line does not follow the actual traces but identifies the connection points, the traces are on both sides of the board so they are a bit tough to follow. (the traces run between the ATMega pins and around to various spots, you can see the actual trace in the purple color)

hm-424 (3.3v).jpg


You can see easily the 3.3v trace from the left leg of the regulator on the board and in this pic, it goes over to the top leg of that 10K resistor and seems to dead end there... However, on the top (component) side of the board a short trace goes from that resistor to the capacitor next to it, then back on the bottom side of the board a trace goes from that capacitor leg to the ATMega pin beside it, then goes through the legs of the ATMega over to the pullup resistor for the pit probe (which is omitted for TC Pit Probe) and the other pin on the ATMega and continues on to the pullups for the rest of the probes from there.

The top right leg on the center circle is the input from the probe jack, so that is not your problem, though it is right in the vicinity of the 3.3v on the (empty hole) for the Pit Probe pullup resistor. So, at this point your job is to determine if your 3.3v is missing due to a short circuit or a broken trace... Another closeup pic of the areas in questions would help me help you identify possible shorts. You can also use your multimeter in continuity mode to check for connections from those 3.3v pins to gnd (POWER OFF ON THE HM while doing this!)

If you determine you don't have a short then it seems the easiest way to get 3.3v over to that side of the board would be to use the convenient empty hole for the pit probe pullup resistor (the one the yellow line points to) and solder a wire in that hole, then connect the other end to any point where 3.3v is present, the easiest point would probably be the empty VCC hole on the FTDI header.
 
Last edited:
I did a continuity check and none of the 3.3V pins showed continuity to ground. So, no shorts, right?

I do have continuity between the convenient empty hole for the pit probe pullup resistor and the empty VCC hole on the FTDI header. Both were measuring 3.3V when I tested for voltage across all of the 3.3V pins. Would connecting these 2 points still solve my problem? I know you said the bottom side of the board on the top right pin of the middle circle is not my problem, but to be clear, that is where I lost the trace.

For my knowledge, are the blue lines on this schematic indicating connections on the component side of the board that differ from the back side?
 
If I understand what point you are referring to, it is the 100k resistor for the RC filter circuit on the Pit Probe... (the one that is standing up on end) The only connection there is from the Pit Probe output through that 100K resistor and then to the corner pin on the ATMEGA. That is the path for the Pit Probe input, it does not pass the VCC through those traces. That circuit can only effect your Pit Probe function, if it is broken the pit probe reading will float unpredictably but it can not effect the other food probes at all, unless there is a short there to some other circuit.

Turn off the power, check for continuity on every point where that yellow line touches to see if the 3.3v is getting to all the places it needs to be.

If this problem is limited to the pit probe only then I have mistaken what you are reporting, if you can confirm that would help me give you better advice...
 
Bad news - In trying to replace the wire that is in this step, "6) Wire the bottom leg of the switch to the wire side of the 100K resistor behind the TC jack", I pulled off the trace. Initially, it was showing an ever increasing temp, but I think I may have really killed it when I accidentally started to resolder that joint while it was plugged in to the 12v wall wart. My HM now shows No Pit Probe when the probe is plugged in. None of the food probes are registering temps anymore either.

I am not wired to the switch here. I took the wire from the resistor and connected directly to the wire that I was replacing next to it to eliminate the switch as the variable here.

Did I really kill this thing? If not, where would I need to jumper this wire to where I lost the trace?

Let me try to run this down clearly for you....

The 100k resistor that is standing on end, of which you have lifted the leg furthest from the ATMega out of the board, leaving that empty hole and one disconnected leg on that resistor....

The empty hole is the output of your onboard TC amp or the standard pit probe jack.

The end of the 100K resistor that is closest to the ATMega (and still soldered into the board) connects directly to the corner pin of the ATMega and is the Pit Probe Input.

So, if you insert a TC amp output "signal" from the RDTC board (CAT5 Pin7) to the disconnected end of that 100k resistor your pit probe should work (connected to the RDTC), AND the broken trace on the other side of that resistor can have no effect on this working, again, it is the output of the onboard TC amp and is not needed when using the RDTC as the Pit Probe.

As for fixing that broken trace... the output of the TC amp has a trace (on the solder side of the board) over to the output pin of the standard pit probe input jack (the pin furthest from the edge of the board). On the component side of the board there is a trace from that same pin on the Standard Pit Probe jack over to the (empty) 100K resistor hole. I believe that is the trace in question? There is no need to fix that trace really since the resistor is not in that hole anyways, you want a wire to connect there... So just connect the wire to the output pin for the standard pit probe instead (again, furthest from the edge of the board, there should be a big empty hole since you do not have a standard pit probe jack installed. The hole is under the TC Pit Probe jack so you wont see it on the component side. The only place to connect to the onboard TC amp output is the corner pin on the TC amp chip which is extremely small and delicate, or that big ol' empty probe jack hole, take your pick! LOL) The wire from the empty probe jack hole should have your onboard TC amp output "signal" on it... Connect that wire to the unconnected end of the 100K resistor and the onboard TC amp should function.

So you have there the wire from the CAT5 pin 7, that is the RDTC output...

and you have the wire from the back pin of the standard pit probe jack, that is your onboard TC output....

and you have the disconnected end of the 100K resistor, that is the Pit Probe input.....

You can connect them manually one at a time to test, and connect through the switch to select between the two of them....

I hope that explains things clearly?
 
Last edited:
Thanks Ralph - I will have to digest this later tonight. I'll be sure to be as detailed as possible on what I give you back, I really appreciate your help and patience with this.
 
As for fixing that broken trace... the output of the TC amp has a trace (on the solder side of the board)
This is the trace that is broken. It looked like you believed it was the other side of the board. Does that fact change anything? Do these symptoms indicate I might have pulled off both traces on this pin? I tried to get a good hi-res close up, but couldn't get a clear shot.

Turn off the power, check for continuity on every point where that yellow line touches to see if the 3.3v is getting to all the places it needs to be.

If this problem is limited to the pit probe only then I have mistaken what you are reporting, if you can confirm that would help me give you better advice...

All pins on the yellow line have continuity.

I do not have any additional wiring on my board at this time, either to the switch or to the RDTC. When replacing my stranded wire, I stopped after I got it back to just the TC HM board and realized I didn't have any readings. I wired the pulled up end of 100K connector back to the wire coming out of the broken trace. You can't see the connection, but here is what I have. I kept the wires long for later wiring them to the switch.

Component Side: https://drive.google.com/file/d/0BzSBmZGY0VqYUktQdUw1aUN0T0E/view?usp=sharing
Solder Side: https://drive.google.com/file/d/0BzSBmZGY0VqYMHdFcEgtdVdEYk0/view?usp=sharing
 
Last edited:
I'm not sure if you are not understanding me, or I am not understanding you...

I've looked at your most recent pic and made a few notes.

The IC there is the TC amp chip. The arrow points to the output leg, the trace over to the empty hole for the standard probe jack is visible and seems intact, the other arrow points to that hole where the trace joins it. This is the onboard TC output and where you need to attach the wire. On the other side of the board (under the TC Jack) is the trace that goes from that standard pit probe hole over to the resistor (the hole where you removed the resistor and inserted the wire), I assume that is the trace you must have damaged because there is no trace on that resistor leg on the solder side of the board. The trace does continue on from the resistor (on the component side of the board) but only to go to the probe header that is not used, so it doesn't matter a bit if that trace is broken or not. What matters is that a wire connects from the hole where the arrow points in the picture to the lifted end of that 100K resistor.

I have also circled the point where the 100K resistor remains soldered to the board, and have drawn a line parallel with the trace that connects to the ATMega. Three solder joints there, the resistor, the capacitor and the atmega make the circuit. If all three of those points are connected (It looks like the trace is there to me) then the pit probe should work when a TC amp output is connected to the lifted end of the 100k resistor.

The X is over the hole where the resistor was lifted out of, and I assume where the broken trace is. If you do what I say above you can forget about that hole and any trace connected to it, you don't need it. Move on. If the TC doesn't work when wired like I said then maybe you blew the TC amp when you soldered there with the power applied. You can connect back the jumper wire to Pin 7 of the CAT5 jack and connect that to the lifted end of the 100k resistor and test with the RDTC instead of the onboard TC amp to confirm if the onboard amp got fried or what.

TC_Traces.jpg
 
Last edited:
After wiring the lifted end of the 100K resistor to the empty pit probe hole, I still didn't get any pit probe readings. Before plugging in my TC probe, the voltage measured 3.3V along the path from the 100K resistor circled above to the ATMega. After plugging in the probe, it measured .14V.

Rewire to empty pit probe hole: https://drive.google.com/file/d/0BzSBmZGY0VqYRmJVX1NZSEJEcU0/view?usp=sharing

I then removed the wire from the empty pit probe hole and connected it to Pin 7 of the cat5 jack. With the probe on the RDTC unplugged, I got 3.3V at the TC jack, 5V at the servo, and 12V at the blower. (this was the first time I connected to the RDTC with power - still havent wired the servo or blower yet)

With the probe plugged into the RDTC, I still got "No Pit Probe".

Pin 7 to the Lifted end of 100K: https://drive.google.com/file/d/0BzSBmZGY0VqYcG15djNjUm9tb00/view?usp=sharing

Did I wire appropriately here for what was suggested? Am I in need of the full switch wiring to complete this test?

Would this all point to a blown ATMEGA since the other food probes are not registering?
 
Last edited:
Confirm you still have Thermocouple selected as the type of probe for Probe 0

Your wire from the Pit Probe output to the 100K resistor looks right, without a TC plugged into the HM you should see NO PIT PROBE, with a TC Plugged in you should see room temp. If you are not getting that then either you damaged the TC amp or the ATMega when you shorted things out with the soldering iron. (assuming the 100K resistor is still soldered into the hole nearest the ATMega, and is connected to the leg of the capacitor next to it and the corner pin on the ATMega by that trace there. You can check continuity with a meter while power is off)


On the RDTC, do you have the ground jumper on the CAT5 jack to pin 8 as well? The probes on the RDTC all run on that dedicate ground so you need to have that ground jumper and the wire on pin 7 for the RDTC Pit Probe to work... (I should have noted that from the beginning, sorry. I was under the impression you had that all wired up already)
 
Last edited:
Confirm you still have Thermocouple selected as the type of probe for Probe 0


Your wire from the Pit Probe output to the 100K resistor looks right, without a TC plugged into the HM you should see NO PIT PROBE, with a TC Plugged in you should see room temp. If you are not getting that then either you damaged the TC amp or the ATMega when you shorted things out with the soldering iron. (assuming the 100K resistor is still soldered into the hole nearest the ATMega, and is connected to the leg of the capacitor next to it and the corner pin on the ATMega by that trace there. You can check continuity with a meter while power is off)


On the RDTC, do you have the ground jumper on the CAT5 jack to pin 8 as well? The probes on the RDTC all run on that dedicate ground so you need to have that ground jumper and the wire on pin 7 for the RDTC Pit Probe to work... (I should have noted that from the beginning, sorry. I was under the impression you had that all wired up already)

I added the jumper back from ground to pin 8. Still no output from RDTC. That is the only additional wiring that would be added to this photo to match my current build: https://drive.google.com/file/d/0BzS...ew?usp=sharing

When I plugged my HM Cat5 jack into my router in order to access the probe settings, the LCD went off and then the LCD backlight and the LEDs blinked consistently once per second or so. This is new, and likely a result of my solder bumbling I'm sure. What does this mean?

I confirmed that the 100K resistor is still soldered into the hole nearest the ATMega, and is connected to the leg of the capacitor next to it and the corner pin on the ATMega by that trace there.
 

 

Back
Top