Is there a way to read 32 bit floating point 3110 & 3111 as 2 seperate 16 Bit values (PM5110)

[casey_skillen_s]

I am using looking at ION setup using the Modbus Tester Interface and pulling back the below values for Frequency

3110 - 16967 or 4247 in Hex

3111 - 60918 or EDF6 in Hex

I know that you can use a IEEE 754 converter to convert the Hex to a useful decimal value however i was looking for a formula to do this or another way around,

We come across a lot of problems in Technical Support where customers are using Modscan or other Freeware tools and the figures appear to them to be in correct because they would see 16967 & 60918 instead of 50 Hz so looking for a way of simplifying this

Could anyone comment?

Regards

Casey

[sesa186702]

Hi Jon,

Please describe more detail what you want . I don't understand what you're saying although I read many times your post

Sorry for this

[casey_skillen_s]

register 3110 on the PM5000 meters is for Frequency in 32 Bit Float

When reading on ION setup you get

3110 -16967

3111 - 60918

Is there a way to convert these numbers to display 50 which is what it should read?

[sesa186702]

Hi Jon,

For reading Float32 Bit, you can choose it easily as below

But unfortunately, Floating point in PM5000 is Most Significant Register First, but the Modbus Interface Testing in IONSetup doesn't support this display, so, you also can not display 50 Hz .

You can use Modscan or some Software which support this

Finally, I just recommend that Modbus Tester Interface just a tool for testing the Healthy of Communication, not to view real Data of Power Meter

If you would like to view Data, yoiu can choose View/ Data Screen, click on Real time item on the left window to view

Thanks and Best Regards

[casey_skillen_s]

Thanks for your reply Tri DANG , very helpful

Have you ever heard of having to input the register before the one you are using?

For example 3026 which is L-L Avg Voltage? i have had to input 3025 to get the figure to work

Is this something you have come across before?

Regards

Casey

[Robert_Lee]

The register offset is defined in the Modbus Application Protocol specification (available from www.modbus.org) and states for the Read Holding Register function code (0x03):

6.3 03 (0x03) Read Holding Registers

This function code is used to read the contents of a contiguous block of holding registers in a remote device. The Request PDU specifies the starting register address and the number of registers. In the PDU Registers are addressed starting at zero. Therefore registers numbered 1-16 are addressed as 0-15.

MODBUS Application Protocol Specification V1.1b3 Modbus

April 26, 2012 http://www.modbus.org 15/50

[Robert_Lee]

A very simple (and slightly inaccurate) way to do the conversion is to use the following formula (from Wikipedia on single precision):

where

S = the sign bit (i.e. whether the first register is greater than 32767, then it is negative)

m = the fractional portion

x = the exponent

First take the first register and remove the sign bit (if its > 32767, subtract 32768, otherwise proceed) and then divide by 128.  This will give you a decimal number (let's call it x.B where x is the exponent and B is the decimal portion).  Next you would  take the decimal portion (B) and then multiple by 8388608 and add the value of the second register to obtain the fractional portion.

So given you example of the first register having 16967 and the second 60918, we get:

s = 0 since the first register is not greater than 32767 (thus the value is positive)

next we take the first register and divide by 128 to yield: 16967 / 128 = 132.5546875

thus x = 132

we then take the decimal portion (.5546875) and multiply by 8388608 and get 4653056 and then add the second register to get the fractional portion (m): 4653056 + 60918 = 4713974

Finally putting everything thru the formula yields:

1 x (1 + 4713974 x 2^-23) x 2 ^ (132-127)

which solves to

1 x (1 + 0.561949491) x 2 ^ 5

1 x (1.561949491) x 32 =  49.98238373