9-bit Signed Temperature

[caahsc]

Hi,

I tried to do a 9-bit 2's complement to determine the signed bit. I wrote in text-based, and it was work fine; however, I tried to convert LabVIEW. It seem that the digital thermometer is not taking a negative temperature. Could anybody help me look through this code? Here is the text-based code.

if ((tempdata & 256) == 256

   tempdata = ((((tempdata & 255) ^ 255) + 1) * -1);

else

   tempdata;

When I used the "Read Field" module to read a 16-bit temperature data, but I only care bit 7-15. However, the "Read Field" module takes care extracting the data into 9-bit. The data that I read from the "Read Field" module is returned to numeric U32. So I used the returned data determine if the MSB is equal to one.

[RavensFan]

I don't understand the ^255 within your text based code.

However the AND 255 makes sense to get the 8 bits.  However you aren't doing that in your LabVIEW code.  The Number to Boolean is going to give you a 16-element boolean array (assuming the blue wire coming from the read subVI is 16 bit.)  And you are never masking off those 8 bits.  I'm not so sure you even need to do any conversions from number to boolean array and back again.

[caahsc]

Hi,

Thank you for your quick input. I also confused myself. Maybe I should write in this sequence.

1. The module "Read Field" will read the 16 bit temperature data from a device, returned to the output as U32 data type. The module automatically takes care extracting the bit D15-D7.

2. So I used the unsigned 32 bit decimal number to determine the MSB bit (I'm not sure if I do need to convert back to binary).

3. Do a 2's complement mathematical operation to get the negative sign.

4. Multiply by 0.5C because the device temp (C) is LSB = 0.5C.

I have attached the file. Please advise if this is correct. I checked by entering the number. Would you please show me how to implement two complement in LabVIEW. Please help! Thanks.

[caahsc]

Should I use the FXP?

[bsvare · Accepted solution]

When you are doing a bit operation, you shouldn't forget to verify position still. Your case statement with "1" isn't valid. It needs to be "256".

This should be the result you're looking for.

[caahsc]

Hi bsvare,

Indeed, that's what I am looking for. Thank you so much for your help.

[DavidBoyd]

Perhaps it's simpler to use the functions built-in for this. (See Numeric->Fixed-point palette)

(I screenshotted the properties dialog setup for the cast and added it to the BD for illustration only.)

Best regards,

Dave