As I got a second set of log amps from Analog Devices, I’ve decided to build a second, more compact milliwattmeter. This time avoiding some mistakes I did on the first run :
– No more HE10 connectors for every wires (they generate too much stray RF)
– Internal probes (for portable operations)
– Smaller housing (less pcb… small is beautiful)
– Shorter connections between subsystems
– Better (far better) grounding plane, an important detail in measurement and instrumentation.
The two probe are left on the same pbc, for a better grounding (compensation coils are not yet installed. A “U” shape shield has been cut to isolate each input stage and will cover the input stage)
The box has been cut in a 20x30cm single side pcb. The microcontroler board is exactly the same as the one used in the former version (same kicad .brd file, same “toner transfert” technique). One of the first task is to place precisely the different boards to know exactly where will be located the different connectors, buttons and the encoder. On this model, I’ve decided not to use the voltmeter function neither the serial output (they will probably be added later).
This “blank” assembly shows the amount of space on the front plate… everything fits, and error margin are very low : less than 2 to 3 mm for the pcb’s. Precision of the holes for the SMA connectors is 1/10th of mm.
Once location of the boards have been decided, the probe board is ajusted to the SMA output. As usual, a silvered coper wire has been used to creat vias between upper and lower ground plane. NO HOLE, even for interconnection wires.
Amp board is fixed to the front plate with small drops of solder. Once the position is OK, a solder joint is made all along the junction
Same stuff on the other side. SMA’s ground connection will be soldered on this side.
The regulator is added then. Every board has it’s own 5 V LDO regulator. But as the main Lab’s power supply is a full rated 13,8 V gear, I choose to add a frontend regulator to “drop down” the input voltage from 13,8 to 7 V. This way, the temperature due to the voltage difference bewtween input and output is distributed among the enclosure. This “overkilling” drop down stage gives another advantage : it protects the milliwattmeter in cas of an overvoltage or voltage inversion (yes, I made this mistake : blue smoke and nice odor :- )) The LM317 survived, and I only had to change 2 schottky diodes and two tantalum caps. That’s all)
A piece of pcb is added on the right side of the plate to support the needed nuts and bolt of the enclosure.
The two milliwattmeters side by side : the shrinking process is obvious
Next step : the wirering and testing
The probe shielding , power plug and encoder have been added
The Front plate design is made with the Schaffner software, printed on a photo “deluxe” paper and protected with a laminated plastic foil. The second push-button is integrated in the “select” encoder. The matt laminating of the front plate gives a better et sober look, but forbid to use the foil as a “display protection layer” : the display becomes blurred and unreadable.