Power problem?


 

John Case

TVWBB Member
I've sorted out the issues I was having with the SD card not formatting to load a new image.

Now when I power up the HM to the usb power port all is golden.

When I power up using 12v the unit goes nuts. All LEDs blink on the daughter board as well as the LEDs on the Rpi
The LCD blinks as well

I wired up the power supply, its 12v DC, 1.5amps, center connector positive.

Tried another power supply 12v 1000mA from a linksys router, same thing.

Any suggestions?

Thanks,
John
 
Last edited:
John,

I had similar issues. I found that I could power via the rPI but not the 12v socket. First make sure that you do not try to apply power via the rPI and the 12v connection. I used the diagram Bryan posted to check voltages.
http://tvwbb.com/showthread.php?43606-Dead-Heater-Meter-I-hate-the-rain&highlight=Voltage

I found that it was much easier to disconnect the rPI and install the MCP1700-33 so that it is able to run stand alone without the raspberry pi. Once that was installed i checked my voltages with all probes disconnected. I found that some of them were not getting the 3v or 5v. Once I found that I went through the whole board and added a bit more solder in places and made sure it flowed just slightly to the underside of the solder joint. You will need to be careful and not get it too hot. Once I did that I was able to meter the voltages and boot the device standalone. I then removed the MCP1700-33, connected the Raspberry Pi and it booted up and ran normally.

I think I had a bad solder joint for the OKI-78SR-05H. I would start there and reflow that solder connected to the PCB and potentially the pins connecting to that board. Seems like when I touched that up it was able to start right up. It could have also been the power getting to that board. So I would follow the power out of the 12v adapter and check compared to the diagram posted by Bryan.

I had mine working, then I put it in the case and it started flashing and wouldn't boot with the power from the 12v adapter. I knew it was either a bad solder joint or I shorted something. After inspection I didn't see any nasty short welds so I started to go over all of my solder joints, and check voltage. I now have it installed in the case and it has at least a good 7 days of continuous running. 48 hours cooking food, the rest of the time I have just left it one rebooting here and there to ensure stability.

Nick
 
Nick,
Thank you I'll go over the solder joints. I may have to order the MCP1700-33? But before I do that I'll double check for cold solder joints.

Thanks again,
 
I've gone over my solder joints. All seem ok.
The HM starts up OK, within a short time it starts blinking the LED's and the LCD, the Rpi continues to function up to the point the power starts cycling.
The voltage regulator gets very hot.

It only does this when powered up with the 12v power supply, when powered by the usb connection to the Rpi it works fine, except the blower will not work. (is this normal??)

I shot a video so you can see what is going on

I looked up the specs for OKI-78SR-05H, IC1 and found the following in the data sheet

Output Short Circuit Condition
When a converter is in current-limit mode, the output voltage will drop as the
output current demand increases. If the output voltage drops too low (approximately 98% of nominal output voltage for most models), the bias voltage may
shut down the PWM controller. Following a time-out period, the PWM will
restart, causing the output voltage to begin rising to its appropriate value. If the
short-circuit condition persists, another shutdown cycle will initiate. This rapid
on/off cycling is called “hiccup mode”. The hiccup cycling reduces the average
output current, thereby preventing excessive internal temperatures and/or
component damage.
The “hiccup” system differs from older latching short circuit systems
because you do not have to power down the converter to make it restart. The
system will automatically restore operation as soon as the short circuit condition is removed


I'm going to look for a short, does anyone have any ideas where to start?

Thanks,
John
 
Check that your 2 diodes are installed in the correct orientation I suppose? The one on the fan part of the circuit could be shorting when the MOSFET turns on. Also make sure the soldering around the blower jack isn't bridging multiple pins, and that your RCA jack doesn't have continuity when nothing is inserted (that it's not internally shorted)
 
Check that your 2 diodes are installed in the correct orientation I suppose? The one on the fan part of the circuit could be shorting when the MOSFET turns on. Also make sure the soldering around the blower jack isn't bridging multiple pins, and that your RCA jack doesn't have continuity when nothing is inserted (that it's not internally shorted)

Bryan,
I previously checked the diodes and did again with your suggestion. They are installed correctly. I also checked continuity in circuit and they only flow 1 direction.

I removed the RCA blower connector completely -- same problem

Looking at the schematic I see 12v enters into the power supply block thru D3, C4, IC1, C5, to IC4
Is IC4 required? I thought it was optional unless using HM4 standalone without the Rpi??

12v to D1 & blower, to Q1
From Q1 to R3 to Pin 5 on ATmega328p (source voltage from ATmega328p to Q1 to turn on 12v to blower)

I also checked the orientation of the capacitors
Could there be a problem with Q1 the MOSFET?

Any other suggestions?

Thanks for your help!!!!!

John
 
No, IC4 isn't required unless you're operating HeaterMeter without the rPi. I don't think it is a problem with the MOSFET. It it was shorting from gate to source, there's a resistor limiting current there to <1mA. If it was shorting from drain to source, that's what it is supposed to do anyway so even if it were broken and fused on at all times, everything could handle that.

Have you tried firing up the HeaterMeter disconnected from the Pi via 12V? The atmega won't be powered or anything but at least you can see if you get the same results. The 5V power supply shouldn't even get warm at all. You can then check all the 5V power areas and make sure they're getting 5V and you should see something on the order of 11.3V at the input to the 5V regulator.
 
No, IC4 isn't required unless you're operating HeaterMeter without the rPi. I don't think it is a problem with the MOSFET. It it was shorting from gate to source, there's a resistor limiting current there to <1mA. If it was shorting from drain to source, that's what it is supposed to do anyway so even if it were broken and fused on at all times, everything could handle that.

Have you tried firing up the HeaterMeter disconnected from the Pi via 12V? The atmega won't be powered or anything but at least you can see if you get the same results. The 5V power supply shouldn't even get warm at all. You can then check all the 5V power areas and make sure they're getting 5V and you should see something on the order of 11.3V at the input to the 5V regulator.

Bryan,
It appears (correct me if I'm wrong) that the 5v power supply has died?

Pin 1 & 3 are both showing the input voltage of 16v
Pin on header to Rpi connector is also showing 16v

I'm going to pull the 5v regulator out now and test out of circuit.

If the 5v regulator is dead, what could have it taken out with it by applying the 12 volts where 5 should have been?

Your thoughts?
 
Last edited:
Well, if the input and output of the 5V voltage regulator are both at 16V you've got a problem there for sure. First off I would ask why 16V? The power supply is supposed to be 12VDC, 16V is kinda high to start with? Is your PS running at 16VDC? Or is that just a typo you made twice, as later you said "applying 12V where there should be 5"? Seems you would be applying 16V, not 12V? A little over 12 from your PS is ok, but 16?
I would check the orientation and value of all the components in the power supply section, being D3, C4, C5, C6, IC1, IC4. If IC1 (5v regulator) is installed in the proper orientation and the voltage on pin 3 is high like that then I would say it's toast. So I would replace IC1 first off, then I would try a power supply that is actually putting out 12V DC and see how it works. If at that point you have +5V coming out of IC1 and your HM is still acting up then I would start looking down the road at what else could have been damaged...
 
Well, if the input and output of the 5V voltage regulator are both at 16V you've got a problem there for sure. First off I would ask why 16V? The power supply is supposed to be 12VDC, 16V is kinda high to start with? Is your PS running at 16VDC? Or is that just a typo you made twice, as later you said "applying 12V where there should be 5"? Seems you would be applying 16V, not 12V? A little over 12 from your PS is ok, but 16?
I would check the orientation and value of all the components in the power supply section, being D3, C4, C5, C6, IC1, IC4. If IC1 (5v regulator) is installed in the proper orientation and the voltage on pin 3 is high like that then I would say it's toast. So I would replace IC1 first off, then I would try a power supply that is actually putting out 12V DC and see how it works. If at that point you have +5V coming out of IC1 and your HM is still acting up then I would start looking down the road at what else could have been damaged...

Ralph,
Thanks, I already figured the 5v regulator was toast.

I was actually getting close to 16v (15.8) out of the transformer. Under load it should settle down to 12v. I will replace the transformer though. The 5v regulator should have been able to handle the 16v, the specs state operating range is 7.5 - 32v, 16v nominal. My understanding of the circuit is only 12v goes to the blower, so 16v should not effect it. Correct me if I'm wrong please.

I double checked the orientation of all the other parts and they are correct.

While looking over the board with a magnifying glass, I noticed a small trace of solder between 2 pins of the MOSFET, I suspect that may have been what took out the 5v regulator.

I'll be placing an order and trying again.

Thank you both for your help and support.

John
 
Oh man that sounds serious. I can tell you that when I designed the PCB and had the input and output traces reversed so I put 12V everywhere 5V was supposed to go. It burned out the 3.3V regulator in the Pi, which I replaced, but something else on the board is also burned out because when I plug it in the regulator gets really hot and it never turns on. The shift register was also blown as well but I think that was it. Your shift register seems fine though because the LEDs and LCD works.
 
You've got me more confused when you start talking about voltage "out of the transformer". Typically transformers are used to step down AC voltage before it is rectified to DC in a power supply, as opposed to being used on the output? And the transformer would be inside the DC power supply/wall wart? So this makes me wonder what kind of power supply you are using in the first place? (Perhaps you are referring to the DC wall wart as the "transformer"?) A 12V AC supply would be a transformer, however, so that is why the term "transformer" raised my interest....

Looking at the schematic I see 12V going to the mosfet, as well as a gnd and the connection to the Mega. If you shorted the 12V to gnd that will drag your whole supply down, if you shorted the 12V side to the mega side that would apply 12v through r3 to the mega, not SURE if that would cause damage but I would think it could, though Bryan has said the Mega is a pretty harty chip so perhaps it would live through it. It would seem less likely that shorting the gnd to the mega would cause any damage. (not sure what two legs of the mosfet had a solder bridge?)

Speaking of solder bridges, if somewhere on your board you have a solder bridge that has a 12V line shorting to a 5V line that would push the whole 5V buss up to 12V and make it appear as though your 5V regulator is dead, if this is the case it may not be dead at all. This seems like a likely scenario to me....

If you have a digital camera shoot a picture of the front and back of your HM board and post it, with a picture we can zoom in close and look at your soldering to try and find issue, and can inspect the components to be sure they are installed in the proper orientation...

Since you are having an issue with power and have given what I consider strange details about your PS, being at 16V and something about a transformer(?), I would suggest you examine what you are using to power your HM. What kind of power supply are you using?
 
Last edited:
It's actually common for a 12VDC wall wort to output up to 16VDC under no load. I had a bunch of Triad brand 12VDC 600mA wall worts at work outputting 16VDC. When we attached the cameras it was powering, the cameras would not work. I guess the cameras didn't place enough of a load on the wall wort.
 
A little high with no load is to be expected, but 16V seems kinda high to me, but who knows. Thing is he said pin 1 and 3 of the 5v regulator and a pin on the rPi connector is reading 16V, and the voltage regulator on the rPi is getting hot, so that is not a "no load" measurement...

I would place my bets on the 12V supply being shorted to the 5V supply line somewhere via solder bridge, and still wonder why his 12V is running at 16V with the HM and rPi loading the supply....
 
Last edited:
...Looking at the schematic I see 12v enters into the power supply block thru D3, C4, IC1, C5, to IC4
Is IC4 required? I thought it was optional unless using HM4 standalone without the Rpi??

Any other suggestions?

Thanks for your help!!!!!

John

You are right, IC4 is optional and only needed if you want to run the HM stand alone (without the rPi), but it CAN be installed along side the rPi and everything seems to function properly for Bryan and I (we both have it installed).

This seems to indicate the rPi is putting out 3.3V to the HM when they are bundled together, thus the lack of need for a 3.3v regulator on the HM board. I see +5V and 3.3V on the rPi header in the schematic, so I assume you are feeding in +5v and the rPi is feeding back 3.3v....

So it makes sense that when you power the rPi via USB everything functions besides the fan, cause the rPi is powering the HM with 3.3V but there is no 12V present to power the mosfet and turn the fan. It also makes sense that the voltage regulator on the rPi is getting hot if your +5v line on the HM has as much as 16V on it....

So it seems either your 5V regulator is dead short input to output or you have a solder bridge shorting the 12V line to 5v line somewhere like I said previously. If you pull the 5v regulator from the board and THEN measure the voltage on your 5v line (and 5v pin at the rPi) and the voltage is still high at 16v or whatever then your regulator is probably good and you have a solder bridge somewhere....

I was just pondering the situation further and thought I should comment in hope it could help you get things fixed up...
 
Last edited:
Oh man that sounds serious. I can tell you that when I designed the PCB and had the input and output traces reversed so I put 12V everywhere 5V was supposed to go. It burned out the 3.3V regulator in the Pi, which I replaced, but something else on the board is also burned out because when I plug it in the regulator gets really hot and it never turns on. The shift register was also blown as well but I think that was it. Your shift register seems fine though because the LEDs and LCD works.

Bryan,
I have to check out the Rpi, after I removed the solder trace I mentioned earlier, the LCD was not working after I powered it up. It had some vertical lines on it upon start up then went blank. Not sure if the LCD got fried or if as you mention something is now wrong with the Rpi.

Even with the LCD not working I was still able to get temp readings via the web interface.

How do I determine if the shift register is blown?
 
Bryan,
It appears (correct me if I'm wrong) that the 5v power supply has died?

Pin 1 & 3 are both showing the input voltage of 16v
Pin on header to Rpi connector is also showing 16v

I'm going to pull the 5v regulator out now and test out of circuit.

If the 5v regulator is dead, what could have it taken out with it by applying the 12 volts where 5 should have been?

Your thoughts?

I pulled the 5v regulator out and it is indeed dead, the output pin is putting out the same voltage as the input pin.
 
You've got me more confused when you start talking about voltage "out of the transformer". Typically transformers are used to step down AC voltage before it is rectified to DC in a power supply, as opposed to being used on the output? And the transformer would be inside the DC power supply/wall wart? So this makes me wonder what kind of power supply you are using in the first place? (Perhaps you are referring to the DC wall wart as the "transformer"?) A 12V AC supply would be a transformer, however, so that is why the term "transformer" raised my interest....

Looking at the schematic I see 12V going to the mosfet, as well as a gnd and the connection to the Mega. If you shorted the 12V to gnd that will drag your whole supply down, if you shorted the 12V side to the mega side that would apply 12v through r3 to the mega, not SURE if that would cause damage but I would think it could, though Bryan has said the Mega is a pretty harty chip so perhaps it would live through it. It would seem less likely that shorting the gnd to the mega would cause any damage. (not sure what two legs of the mosfet had a solder bridge?)

Speaking of solder bridges, if somewhere on your board you have a solder bridge that has a 12V line shorting to a 5V line that would push the whole 5V buss up to 12V and make it appear as though your 5V regulator is dead, if this is the case it may not be dead at all. This seems like a likely scenario to me....

If you have a digital camera shoot a picture of the front and back of your HM board and post it, with a picture we can zoom in close and look at your soldering to try and find issue, and can inspect the components to be sure they are installed in the proper orientation...

Since you are having an issue with power and have given what I consider strange details about your PS, being at 16V and something about a transformer(?), I would suggest you examine what you are using to power your HM. What kind of power supply are you using?

Ralph,
Sorry about the misunderstanding. Voltage out of the transformer is referring to the wall wort. Which is rated at 12v DC, 1000mA from an old LinkSys router. It seems to freewheel with no load at close to 16v. I will get another wall wort.

I was very surprised to find a solder trace in my soldering. Which only showed using a magnifying glass.
This is not my first foray into electronics. I've build and bread boarded stuff over the years. At 14 I was licensed as a Technician class amateur radio operator, later upgraded to General class. But alas have let my license laps some years ago. Now, I'm 45 wow! I actually still have the resistor color code memorized!

The photo suggestion is a good one!

I'll keep you posted.

Thanks again,
John
 
Oh wow John, you've probably got more experience than I do! The real key to knowing how long someone has been building electronics projects is the mnemonic they use to remember the resistor codes. Mine is "Bad boys rape our young girls but Violet gives willingly" but I'm sure the kids these days have a more politically correct version.

When I get home tonight I'll pump 16V into my HeaterMeter too, because like you've pointed out, the 5V regulator is rated for way more than that. I think some of my wall warts put out at least 13V open circuit too.
 
Oh wow John, you've probably got more experience than I do! The real key to knowing how long someone has been building electronics projects is the mnemonic they use to remember the resistor codes. Mine is "Bad boys rape our young girls but Violet gives willingly" but I'm sure the kids these days have a more politically correct version.

When I get home tonight I'll pump 16V into my HeaterMeter too, because like you've pointed out, the 5V regulator is rated for way more than that. I think some of my wall warts put out at least 13V open circuit too.

Bryan,
Thanks for the cudos, but it's been a long time. I'm getting old, you are the expert!!
I learned it that way to, but after awhile I just have the colors themselves memorized.
Power law, and ohms law to.
It's easier to remember them like this

Power law
P
___
I x E

Ohms law

E
___
I x R

Just cover the one you want to find
 

 

Back
Top