Test Results for Linarity of the AD8307

The results for the linearity test are detailed below. The test circuit did not have any frequency compensation components installed in the front end.  It has been stated in articles that these components are not required for HF frequencies.  However I believe there is enough variation in the results to warrant some work on the front end circuitry in order to further decrease the variation between the different readings at the various frequencies.  I believe the adc in the ATMEGA 328 can detect some of those variations.  In the case of the -70dBm level the difference between the min and max readings for only the HF frequencies is 63mv.

While the end result after processing by the adc in the ATMEGA 328 may result in only a portion of a dBm, I think it’s important to get the best results you can in the hardware first.  Then this should result in minimal compensation or "fudge factor" in the software. i.e. calibration factors. 

Next up will be connecting the circuit to an Arduino and re-run the tests and see how the 8307 behaves when connected to the processor. Should be interesting.  Experiments done a few years ago I was able to get an accuracy of within 1 dBm and better in most cases.  Another thing that will be interesting will be comparing the behavior of the 8307 with signals not as clean as the sine wave I'm using now. 

73, Peter

 

Spectrum Analyzer Detector Board Update

I managed to get the prototype board up and running.  Naturally nothing I do is perfect to start off. Missed drawing a trace between pins 6 and 7 for the AD8307 chip. Explains I suppose why the 1st round of testing was behaving erratically.  Pin 7 is the ENB pin which should be high.  So since it was floating, I guess that's the erratic readings I was getting.  I was ready to blame the signal generator for the problem. Gawd forbid it was my fault. hihi

Here's the results. As predicted, I was able to get down to -70 -75dbm. Apply a correction factor and -80 comes into play.  Not bad I'd say. Reading the datasheet hints at getting a little further but I'm not sure its worth the effort. However it can't hurt to have another read and try to understand some of the finer details of the 8307.

73, Peter

Spectrum Analyzer Detector Board

I’ve managed to complete the 1^st version of the detector board for the spectrum analyzer project. It didn’t go to bad, but naturally not all things go as planned. No matter how carefully the planning process went.

The idea of putting the board into an Altoids tin I believe to be sound.  Effective shielding completed in one fell swoop.  

My thought on offsetting the connectors so as to be able to insert the board with the two connectors installed was not successful. So one connector was soldered into place with the board already installed. Previous post shows the initial design of the board which is quite different from the one installed.

Utilizing a power jack also proved to be not a good idea.  The profile of the jack proved to be too high.  So back to a direct connection and I fed the 12v wire through two holes in the case. I didn't have any grommets on hand so I stripped some insulation off a larger wire and used that as grommets.

To be able to carry on with some testing, I had to temporarily solder in some axial components. The 47uf cap, 5v regulator, and a 10Ω resistor.  Also I tacked in an 8 pin dip socket for the AD8307 chip. Hopefully I can do some testing later today.  Previously I documented breadboard measurements, so it will be interesting to make some comparisons with the board in the Altoids tin version.

I expect to receive the smd versions of those components early next month.  Already I’ve ran into issues doing this and pulled part of a trace in two areas when the now sticking up into the air components got hit.  Installing those smd components might prove to be now a little difficult.  Worry about it later.  I suppose I could re do and improve the board for a 2^nd version and that would be the

pertinent thing to do.

Already just into this build and I find myself thinking about ground loop problems.  I don’t want a repeat of the issues I had with the LBS project.  At least I’m well aware at the beginning this time round.

I find myself thinking more and more about cases since starting this project. And since shielding is critical, what will not only be the most effective, but also what will be the most visually appealing. After all the work going into this project it should look good as well. A wander through the hobby store or Lowes looking at the metal on hand may lead way to the direction to take.

73, Peter