If Ain't Broke Don't Fix It Part 2

Well problems continued. Inserted the TFT LCD off by one pin. Poof.  No smoke but poof non the less. Good thing I had another screen that was going to be used for a project that never really got going...yet.

Re did the shield and much happier with the results. Noise floor levels went from -64dBm no shield to -71.8 with the shield installed.  Success with a little pain.  Over all a good mod and the cost was a junkbox resistor, $20usd LCD, and couple of bucks for the copper.  Ok can't really count the resistor and lcd.

Here's a couple of pictures and a graph of the power meter response to various inputs.

If it ain't broke, then don't fix it.

The term, “If it ain’t broke, then don’t fix it” reared its ugly head the last couple of days.  In an effort to improve the noise floor readings on my SNA, I installed a shield around the AD8307 circuits.

Problem number 1:

After installing the shield, I could not make the appropriate adjustments to calibrate the power meter. After several tries, considerable reading and thought, I removed the shield as a last ditch effort.  Since that did not solve the issue, no surprise, I was advised, “Don’t forget to press the alt key when making the measurements” Duh!! Problem #1 solved.

Problem number 2:

Shield interferes with the touchscreen and sd card. Not so easy of a fix, but trimmed the shield down, trimmed the L1 loop a little to lower its profile.

This is where I decided to put it all back to original, make some baseline measurements and then have another go at the shield mod. This now takes us to problem #3.

Problem #3:

I noticed that a resistor and a capacitor were not fixed to the pcb properly.  In fact they were off kilter by a fair bit so I knew I had done something. Must have happened while de-soldering the shield.

No problem, while they are 805 smd’s I was able to reattach easily. No pinging either.  Well, this lead to problem #4.

Problem #4:

Trying to calibrate the SNA was going much better now until I got to the part of the 3dbm level settings confirmation.  The readings I was getting looked like the AD8307 circuit was not working. Troubleshooting was easy enough with the screen removed and I quickly determined that the 6.8 ohm 805 resistor was toast.  This leads to problem #5.

Problem #5:

While I have a pretty good supply of on-hand parts, what were the odds of 6.8 ohm in the 805 package? Well as it turns out I have 4.7 and 8.2 in the 805 package. And as luck will have it, the designer told me either value would suffice.

No problem # 6, yet, but I’ve done enough for today and tomorrow can only get better.  I’m sure I’ll quickly forget the if it ain’t broke saying.  After all, the breaking of things offer a good learning opportunity.

73, Peter

SNA Modification

I decided to forgo the measurement receiver in order to finish of the sna calibrations.  That way something can come off the list.  I had been spending a fair bit of time calibrating my sna the last few weeks.  There is nothing like having someone else with the same equipment so you can make some comparisons.  Once that happens well the gloves come off. 

In an effort to improve the dynamic range and following down the same path as a fellow homebrewer, it was time to install a shield around the AD8307 and associated components.  Purchasing .025 copper sheet was the thinnest I could find and in the end was not all that difficult to work with. Especially if you don’t mind really ugly.

First picture shows where the shield will be installed and there is already an outline in the board to facilitate this modification.  I scraped a few strategic areas for the shield to be soldered to the board.

The first piece was a lone side in picture 2.  The 2^nd side on the left was installed the same way.  The remaining sides were in one piece and folded appropriately.

I do wonder if the shield going over top will affect the characteristics of the wire loop or not, but we will find out shortly I suppose when calibrating.  Either way I doubt too much of an issue.

So the install is complete, just need to power up.  Wasn't to difficult to avoid components and cause shorts as there was a fair bit of space around the circuit as you can see in the first picture.

Power up and calibrations I think best be left until tomorrow.  Then I can start the measurement receiver boards.

73, Peter