Reading OZ2CPU’s blog is like sex or cocaine… you quickly become an addict.
A friend of mine sent me some weeks ago a full-fledged TXCO (a temperature controlled oscillator… in other words, a stuff that gives you a very precise time). This TCXO was itself calibrated with a rubidium frequency standard.
For this reason I decided to include this item in my frequency meter, a small PIC bases DFD4 sold by AADE.
If you are looking for a simple, cheap and very efficient frequency meter, have a look on AADE’s web site
This gear costs 50 bucks only and is very robust… but it also is definitely deaf. If you intend to measure signals below -10 dBm, forget it.
That’s why OZ2CPU decided to build two preamp “front end” to boost the dynamic of hos own meter. Everything is described on his website @
A first frontend using a ERA-2SM monolithic hybrid amp is used for UHF signals from 50MHz to 3 gigs, and a second one for low frequencies (100 kHz to 50 MHz).
As Thomas OZ2CPU didn’t made any PCB of the stuff, I decided to create a new design of those two boards. I you are interested in building this “add-in”, I will be glad to send you all the .SCH and .BRD (or gerber) files you want. All credit MUST be given to Thomas. Any commercial use of these files or schematics are forbidden without authorization of the author… and the author is Thomas… not me !
Enough chat, some pictures :
First of all, the HF frontend. Four boards have been etched :
– one for the DFD4,
– another for my kid’s frequency meter based on Doug’s N3ZI instrument (http://www.pongrance.com/ddfc-cc-super-manual-20.pdf)
– a third one for a friend of mine
– a fourth one “just in case”.
all components are “SMT” mounted… even the 4 BSX89 transistors. One strap is mandatory. As usual, each board has it’s own 5 V LDO regulator. Sensitivity of the board after tuning is given at –50 dBm, and it can withstand a + 20 dBm signal… total dynamique is 70 dBs. A little bit better than the original PIC :- )
The “four in a row” modules
As usual, the front plate of the enclosure is made of PCB. From left to right : the power IN connector, the UHF BNC input (Radiall, 8 GHz specified), the UHF/HF inverter, the HF BNC input and the 10 MHz 0 dBm TCXO calibrated output. And a rectangular hole for the “biggy display” sold by AADE.
Rear view of the front plate with the AADE main board, the LCD display and all the hardware around
Exceptionnaly, I’ve found in my junkbox the perfect enclosure for the meter : a former IBM external DC600 cartridge backup. Some painting, and it definitely doesn’t looks like a computer part anymore.
Pre-wirering of the front plate.
notice the way coax cables have been fixed. The central inverter (HF/UHF) needs a special hardness (a ferule) to maintain the three coax cables without mechanical constraint and with a perfect grounding
A close up of the ferule shows the way coax are soldered. All coax are PTFE/Ag 3mm cables.
First pre-assembly of the meter and first HF tests.In the foreground, the “printed” frontplate protected with a glossy laminated plastic foil. The big aluminium brick in the background is the TCXO itself.
another view of the project. All components are fixed on a large PCB tined plate. The strange module with five polarized capacitors is a 12/24 V converter dedicated to the TCXO.
The UHF frontend is using a ERA-2 MMIC. The board is also supporting a MB506 prescaler (dividing the input frequency by 128). The original prescaler has QRT several monthes ago for unknown reason.
As the HF frontend directly sends its signal to the PIC and as the UHF frontend includes the prescaler, the original 74HC4046 and the MC12079 are no longer needed on the DFD4 original board.
Two straps can be seen on this proto board, due to mistakes during the design process : I had to disable Kicad’s DRC to “create” a quick and dirty “double footprint” for the prescaler (SOIC and DIL). But without DRC, risks are higher to make mistakes. These errors have been corrected on the new BRD/Gerber file. Only one strap is needed now and could be avoided with a 2 layer pcb.
Power input is fed thru a by-pass capacitor (lower side of the picture). A SMT 220 mH blocks any RFI going to or trying to escape from the MMIC (right side of the bypass cap), delivering a 13,6 V voltage to the MMIC (before the Bias resistor, of course) . The prescaler is powered by a fixed 5 V LDO regulator –left side of the bypass-. Several vias have been distributed all along the grounding plane, particularly near the 50 Ohms microstrip input . The “strange” shielding is in fact a piece (a slice ?) of Hammond enclosure that was cut in half many years ago to create a “half height N2PK directive coupler (https://f6itu.wordpress.com/2008/12/11/sma-tres-sm/). I was keeping this “metal ring” just in case… The rest of the schematic doesn’t need further comment.
The whole familly is present. All power routing are “choked” with ferrite beads or toroids
This is the “final picture” of the gear with an inductive probe on a UHF marker. input signal is at –28 dBm