Well,

Since I got my linear amplifier that I told you about here:

http://g0mgx.blogspot.co.uk/2012/09/how-much-power.html

I've been wondering about power measurement. Now, I have the power meter on the linear, a power meter made by Diawa, one made by some other manufacturer.... the point being that they all read different values for the same output!

So, this got me wondering about how to really measure power.

So, in simple terms if we measure the Peak-to-Peak voltage of a waveform, we need to calculate the RMS voltage by:

(P-to-P / 2) * (Sqrt (2) / 2) which is:

(P-to-P / 2) * 0.707

So, for example, lets say we have 100 V peak-to-peak, thats the same as 35.35 V RMS.

Now, to get power in Watts we need to:

(V RMS ^ 2) / R where R is the load impedance (in my case nearly always 50 ohms)

So, my 100V P-to-P or 35.35V RMS is:

(35.35 ^ 2) / 50

which is equal to, or as near as damn it, 25 Watts.

So, making a simple plot of P-to-P vs. RMS vs. Power (Watts) is quite simple, and it looks like this:

So, hopefully you will be able to see that the range of power or voltage we plot doesn't matter, the line will always be the same shape.

So, theory says, that if I connect an RF generator (perhaps one made by Yaesu) to a dummy load, then connect my x10 scope probe across the dummy load, I should be able to measure an accurate P-to-P voltage and quickly calculate the Power in Watts. This didn't seem to be the case when I tried it.

I used the FT-857 I have here and set it to its 20 Watt output setting, I was reading about 124V P-to-P which is more like 38 Watts, whilst I expected the radio setting to be an approximation, I never expected it to be this far out!

So, another experiment was called for. This time I connected the 'scope in the same way, but plotted different power and frequencies. Here is what I found:

You will notice that the reading for 40 Watts at 50 MHz is missing from the graph above. Just as I applied the power was exactly the start of the smoke emanating from the 'scope probe. That probe is RIP.

Now, whilst I don't claim to understand why, it is clear from the measurements above that at low power & low frequency my measurements are close to as expected. As the power increases and certainly as the frequency increases then the measurements become wildly inaccurate!

So, this made me rummage on the interweb for some kind of accurate power measurement project. This is what I found:

http://www.arrl.org/files/file/Product%20Notes/2012%20Handbook/KAUNE.pdf

Now this looks like just the ticket!

So, today I have started to make the directional coupler - this is the bit that sniffs the RF in a feeder and produces a forward and reflected waveform... mine looks like this:

Far Circuits in the USA:

http://www.farcircuits.net/

do the circuit boards for this power meter project so I've ordered a set. I've bought from them before and the service is usually excellent. So, when I have the boards and all the other bits, I'll post some more on my power measurement conundrum.

Good egh?

Since I got my linear amplifier that I told you about here:

http://g0mgx.blogspot.co.uk/2012/09/how-much-power.html

I've been wondering about power measurement. Now, I have the power meter on the linear, a power meter made by Diawa, one made by some other manufacturer.... the point being that they all read different values for the same output!

So, this got me wondering about how to really measure power.

So, in simple terms if we measure the Peak-to-Peak voltage of a waveform, we need to calculate the RMS voltage by:

(P-to-P / 2) * (Sqrt (2) / 2) which is:

(P-to-P / 2) * 0.707

So, for example, lets say we have 100 V peak-to-peak, thats the same as 35.35 V RMS.

Now, to get power in Watts we need to:

(V RMS ^ 2) / R where R is the load impedance (in my case nearly always 50 ohms)

So, my 100V P-to-P or 35.35V RMS is:

(35.35 ^ 2) / 50

which is equal to, or as near as damn it, 25 Watts.

So, making a simple plot of P-to-P vs. RMS vs. Power (Watts) is quite simple, and it looks like this:

So, hopefully you will be able to see that the range of power or voltage we plot doesn't matter, the line will always be the same shape.

So, theory says, that if I connect an RF generator (perhaps one made by Yaesu) to a dummy load, then connect my x10 scope probe across the dummy load, I should be able to measure an accurate P-to-P voltage and quickly calculate the Power in Watts. This didn't seem to be the case when I tried it.

I used the FT-857 I have here and set it to its 20 Watt output setting, I was reading about 124V P-to-P which is more like 38 Watts, whilst I expected the radio setting to be an approximation, I never expected it to be this far out!

So, another experiment was called for. This time I connected the 'scope in the same way, but plotted different power and frequencies. Here is what I found:

You will notice that the reading for 40 Watts at 50 MHz is missing from the graph above. Just as I applied the power was exactly the start of the smoke emanating from the 'scope probe. That probe is RIP.

Now, whilst I don't claim to understand why, it is clear from the measurements above that at low power & low frequency my measurements are close to as expected. As the power increases and certainly as the frequency increases then the measurements become wildly inaccurate!

So, this made me rummage on the interweb for some kind of accurate power measurement project. This is what I found:

http://www.arrl.org/files/file/Product%20Notes/2012%20Handbook/KAUNE.pdf

Now this looks like just the ticket!

So, today I have started to make the directional coupler - this is the bit that sniffs the RF in a feeder and produces a forward and reflected waveform... mine looks like this:

Far Circuits in the USA:

http://www.farcircuits.net/

do the circuit boards for this power meter project so I've ordered a set. I've bought from them before and the service is usually excellent. So, when I have the boards and all the other bits, I'll post some more on my power measurement conundrum.

Good egh?

## No comments:

## Post a comment