CB and Ham Radio Techs Love Their Bird Wattmeters

[xrunner]

Ok so let me get to it. I need to give this little rig a few checks to see if it even works to any degree. Does the red LED mean it works like it should if it lights up? These and many more questions will be answered in the coming hours. 

If it passes preliminary tests today (and doesn't burn out), then I'll proceed to make some graphs this week.

As the Terminator said "I'll be back."

[bd139]

The red LED lighting only denotes that the supplier of the red LED and the series resistor were not ripping the board assembler off 

[xrunner]

I have the preliminary results. I can now confirm - when the red LED is lit it does indeed mean the board is working properly! All kidding aside, as bd139 said, it does look like the AD8310 part works well.

Here's the initial data.

First to calculate dBm you need to measure Vo and use a simple equation to get power. I rearranged the equation they had in the datasheet which was using their scale of dBV, added a correction to give answers in dBm, and came up with this:

Pin (dBm) = (Vo - 2.28) / .024

I made an initial power measurement using the known accurate output power of the hp 437b power meter reference (0.0 dBm, 1 mW @ 50 MHz).

I measured a Vo of 2.295 V which results in a calculated input power of 0.625 dBm. I then set my Agilent 8648A to the same settings (50 MHz, 0.0 dBm) and connected the board to the Agilent. I did another measurement with the board and I got the exact same answer. So the error at that point is 0.625 dBm. This shows using the 8648A is good to go forward.

I then proceeded to check other data points @ 50 MHz as follows:
-80, -40, -20, +10 (dBm).

Here are all the results:

-80 dBm

Vo = 0.371, calculated input power = -79.5 dBm [error 0.5 dBm]

-40 dBm

Vo = 1.316, calculated input power = -40.08 dBm [error -0.08 dBm]

-20 dBm

Vo = 1.87, calculated input power = -19.6 dBm [error 0.4 dBm]

0 dBm

Vo = 2.295, calculated input power = 0.625 dBm [error 0.625 dBm]

+10 dBm

Vo = 2.54, calculated input power = +10.8 dBm [error 0.8 dBm]

Instruments used: hp 437 power meter, Agilent 8648A RF signal generator, hp 34401A DMM
Note: The stated power output accuracy of the Agilent 8648A is +/- 1 dB

Initial Conclusions

It works well, better than I thought it would. I don't see any reason why you couldn't design a useful power meter for hobbyist use around the AD8310. I don't think I made any mistakes. I never have before ... well I thought I did once before, but I was wrong about that time.

Next I want to make a set of graphs to better check the performance at different frequencies, which I'll work on in the next few days.

After that, I want to toss in a microcontroller with display to get real-time measurements properly displayed as dBm and watts. With a clock ... no, no clock I have too many clocks even my soldering station has a damn clock.

[joeqsmith]

You're off to a good start. 

My parts arrived but I need to sort out how to test them.  Outside of checking the LED,  I am far too lazy to collect any data by hand.        Without any sort of reference, I am at the mercy of my old equipment.   

[xrunner]

You're off to a good start. 

My parts arrived but I need to sort out how to test them.  Outside of checking the LED,  I am far too lazy to collect any data by hand.        Without any sort of reference, I am at the mercy of my old equipment.

I've got yard work to do today but I'll be on the graphs soon. I was just going to go ahead and make a power meter that simply displayed the correct power in dBm, but then I thought, well why not display both dBm and watts? What if you want to measure 100W? You'd need a 40 to 50 dB attenuator out front. So for corrected display numbers maybe add a selectable correction for any attenuation you might need to add in front using a button. I got more ideas too.

Featuritis - the death of what would have been a decent simple product. 

[joeqsmith]

Seems I saw a similar 3M board with gray wires and alum elec cap recently.      
https://www.vintage-radio.net/forum/showthread.php?t=144088

One option is to use a separate AD8307s for the forward/reflected and ignore the phase.  I'll try and power it up over the weekend. 
2X1206 105ohms to get us close to the 52.3.   

[bd139]

Yeah that was me. I actually left that forum because it was full of miserable old bastards. This one is just full of old bastards. 

Edit: oh god reading that thread reminded me about that poor HP sensor I killed  . Was playing around with frequency multipliers to try and get up to VHF. Yes I am responsible for another orphan power meter 

[joeqsmith]

I'm always impressed with the ham and CBer's ability to think outside the box.   

https://www.ebay.com/itm/225031794210?epid=1101443944&hash=item3464f0ae22:g:aOMAAOSw089iqhjh

[xrunner]

Seems I saw a similar 3M board with gray wires and alum elec cap recently.      
https://www.vintage-radio.net/forum/showthread.php?t=144088

That was bd's lashup? He's got a red LED on his setup so I'm in good company. 

One option is to use a separate AD8307s for the forward/reflected and ignore the phase.  I'll try and power it up over the weekend. 
2X1206 105ohms to get us close to the 52.3.

Yea sure, use as many as you can that's my advice. 

I'm taking data now and will post at least one graph today. I know you guys are super-excited waiting to see them. 

[bd139]

I blame China. Arduino mini had the LED on it. 

[bd139]

I'm always impressed with the ham and CBer's ability to think outside the box.   

https://www.ebay.com/itm/225031794210?epid=1101443944&hash=item3464f0ae22:g:aOMAAOSw089iqhjh

I would love to see the smith chart for that termination 

Edit: this is my favourite thread on her for a while (apart from TEA of course) 

[xrunner]

Edit: this is my favourite thread on her for a while (apart from TEA of course) 

Well we're trying to keep you entertained. The fun is just beginning. 

[xrunner]

I'm always impressed with the ham and CBer's ability to think outside the box.   

https://www.ebay.com/itm/225031794210?epid=1101443944&hash=item3464f0ae22:g:aOMAAOSw089iqhjh

Oh dear ... 

"Hello, hello Mars? This is Earth calling - do you copy over?" 

[CJay]

I'm always impressed with the ham and CBer's ability to think outside the box.   

https://www.ebay.com/itm/225031794210?epid=1101443944&hash=item3464f0ae22:g:aOMAAOSw089iqhjh

WTAF?

They've got the nerve to call the WW a 'dummy load'?

[bd139]

Perhaps it makes the Bird read higher

[joeqsmith]

Perhaps it makes the Bird read higher

Higher Watts = More $$$    Who ever buys it should send you the terminator to measure. 

One lead appears stuffed into the T's female pin but the other looks like it may be soldered.  The string sells it.   

You think with that twin BNC T setup, that extra port was attached to their scope? 

[xrunner]

Here's a preliminary set of plots for the lower frequencies of 10, 50, and 100 MHz comparing to the datasheet plots on p. 6.

(always say your data is "preliminary" in case anyone wants to criticize it.)

Looks pretty good. The "Data Sheet" line was estimated by me so it might not be an exact match to the "average" in the data sheet graph.

I did notice something in the datasheet. On p.9 it states -

"VSLOPE is the logarithmic slope, now expressed in V/dB (25 mV/dB for the AD8310).

However, everywhere else it says 24mV/dB. 

[xrunner]

Here's a preliminary set of plots for the higher frequencies of 200, 300, and 440 MHz compared to the datasheet plots on p. 6. On this plot I decided to include a set of data that represents the lower frequencies (since they mostly track the same) for comparison. The thicker black line represents them.

The good news is the Vo response is pretty linear for a given frequency, we just need to correct for the error to get the right answer (Po). For example @200 MHz the error is a pretty consistent 2.5 dB. If the system knew the frequency, that error would be easy to correct given we'll have software to work with.

The question will be how best and simplest for the operator. TBD at the moment ...

[bd139]

Worth reading about the compensation network on the input of the AD8307 version here: https://www.qsl.net/sz1a/download/build%20an%20rf%20power%20meter.pdf

The error is a mismatch by the looks.

[xrunner]

Worth reading about the compensation network on the input of the AD8307 version here: https://www.qsl.net/sz1a/download/build%20an%20rf%20power%20meter.pdf

The error is a mismatch by the looks.

Thanks bd, I was looking at your thread on the Vintage Radio forum last night and I followed the yellow brick road to that article. That solution does look promising. Did you make that input compensation circuit back then?

I plan on getting into that eventually. I think I'm going to work on my software a little now since that's the fun part to me and see if I can get some things going on a little OLED display.

[bd139]

I did. I could only test it to 100MHz at the time and it was flat to 0.5dB across that range.

[xrunner]

I did. I could only test it to 100MHz at the time and it was flat to 0.5dB across that range.

Great. One more question, what do you think the inductance of that one-turn coil was? Just curious.

[bd139]

Not a lot. I used a couple of loops around a 2mm screwdriver. Couldn’t measure it realistically at the time due to lack of a RF generator with calibrated output.

[joeqsmith]

You may be interested in the following:
https://www.nutsvolts.com/magazine/article/a-discussion-on-dbs-and-rf-power-meters

First eval board appears to work, or at least the LED turns on. 

[bd139]

Have read that one before.

As always frequency compensation is always more difficult than it looks. See attachment (tek 485 vertical)