My first attempt, egt meter

[JaniM]

I'm trying to do egt meter.
The code is which I'm trying to use is from PH Anderson's.

' MAX6675_1.Bas - PICAXE 28X1
'
' PICAXE 28X1                  MAX6674 / 6675
' CS (term 26) ------------ /CS (term 6)
' SCK (term 25) ----------- SCK (term 5)
' MISO (term 12) <------- SO (term 7)
'
' Uses 155 of 2048 bytes
'
' copyright, Peter H Anderson, Baltimore, MD, Jan , '04


Symbol  CS = Output5	' define terminal associated with MAX6674 / 6675
Symbol  SCK = OutPut4
Symbol  MISO = Input1

Symbol Val = W0
Symbol WholePart = W1
Symbol FractPart = B4
Symbol Digi = B5
Symbol N = B6

TOP:
     GoSub MeasTemp
     GoSub DisplayTemp
     Wait 1
     Goto TOP

MeasTemp:
     High CS	' idle
     Low SCK

     Low CS	' start transmission sequence

     Val = 0
     For N = 1 to 16
        High SCK
        Val = Val * 2 + MISO
        Low SCK
     Next

     High CS	' terminate the transmission

     Val = Val / 8	' use only the 13 most sig bits
     Return

DisplayTemp:

     WholePart = Val / 4
     FractPart = Val & $0003	' two least significant bits


     If WholePart > 999 Then DisplayTemp_4	' Four places
     If WholePart > 99 Then DisplayTemp_3	' Three places
     ' else				        ' Two or One places
     SerTxD (#WholePart)

Fract:
     SerTxD (".")
     Branch FractPart, (Fract0, Fract1, Fract2, Fract3)

Fract0:
     SerTxD ("00", 13, 10)
     Return

Fract1:
     SerTxD ("25", 13, 10)
     Return

Fract2:
     SerTxD ("50", 13, 10)
     Return

Fract3:
     SerTxD ("75", 13, 10)
     Return

DisplayTemp_4:

       Digi = WholePart / 1000	' thousands
       SerTxD (#Digi)
       WholePart = WholePart % 1000

DisplayTemp_3:
       Digi = WholePart / 100	' hundreds
       SerTxD (#Digi)
       WholePart = WholePart % 100

       Digi = WholePart / 10	' tens
       SerTxD (#Digi)
       WholePart = WholePart % 10

       Digi = WholePart 		' units
       SerTxD (#Digi)

       Goto Fract

And the schematic is in that pic.

And the "problems":

* should that MAX getting hot? It's burning finger.
* should that code display temp in terminal window correct? Now there shows only nonsense code. In simulation there is temp in lcd screen.
All wiring is ok, but i'll check again. Also I use project board 28/40 for downloading.

I'm going to next add serial lcd, axe033.

 

[manuka]

Yikes- well EGT is assumed Exhaust Gas Temperature, but the sensor itself should certainly NOT get hot alone. Turn everything off fast while we ponder your schematic & PHAnderson's ref. Herewith for those who (like me) needed an IC refresh -

The MAX6675 performs cold-junction compensation and digitizes the signal from a type-K thermocouple. The data is output in a 12-bit resolution, SPI™-compatible, read-only format.

This converter resolves temperatures to 0.25°C, allows readings as high as +1024°C, and exhibits thermocouple accuracy of 8 LSBs for temperatures ranging from 0°C to +700°C.

Why would you need such fine resolution? Isn't winter over in Finland-what is the actual application? Stan
EXTRA: It's not just a 5V instead of a 3.3V supply issue is it? Check yourself from IC data!

 

[slurp]

The data sheet was the first thing I looked at...

Supply Voltage (VCC to GND) ................................ -0.3V to +6V

Next I'd be checking connections, poping out the chips and checking supply voltages with a meter.

regards,
Colin

 

[hippy]

I wouldn't have thought the MAX should have been getting hot so there could be a hardware problem.

In your diagram you don't have the Reset pin ( leg 1 ) connected via a 4K7 to +V.

 

[manuka]

Hippy is of course correct on the reset, but I've noted your schematic doesn't follow the PHAnderson code connections. Instead of his 26-6 & 25-5 , don't you have 26-5 & 25-6? Suggest you give a reference to his actual application site so we can cross-check directly (instead of Googling). Stan

EXTRA: Thought this seemed familiar. Check the MAX6675 - PICAXE Forum postings from some months ago =>www.picaxeforum.co.uk/showthread.php?p=92683 &=> www.picaxeforum.co.uk/showthread.php?p=84179

 

[sleazyd]

No, it should not be getting warm. I have a similar circuit and there are no thermal issues with it. I'll agree with Hippy in that you need to tie reset high.

 

[JaniM]

In schematic pins 25 and 26 was wrong. But the wirings was ok.

The code was taken from here:
http://www.phanderson.com/picaxe/max667x.html

5 volts comes correctly.

The application would be in race cars exhaust gas temperature measuring and 5 degrees resolution is also fine. But now only making testboard and trying only to get working.

Picaxe is in axe022 project board alltime, so reset thing should be ok.

Without Picaxe the Max don't get any hot.

With 3.3V it don't burn, there is little warm only. When you put 5 V it getting hot.

And the winter is over here and snow is smelting. It's all almost gone.

 

[SD2100]

Is the 6675 soldered to a PCB ??, if so check for solder bridges between the pads as being a surface mount device and the pads being very close this can easily happen. The 6675 shouldn't get hot, while reading the probe temperature it's checking the chip temperature as well so it can do the cold junction compensation, if the chip gets too cold or hot I think somewhere between -20c to +85c there will be errors.

 

[JaniM]

I'll check all wiring and I don't find any errors. All solderings are ok.

One, if I take MISO (term 12) off, then max don't getting hot.