VHF / UHF down converter



Some time ago I helped someone to dispose of an old and broken TV set.  Before I went with it to the communal disposal plant I stripped it of the most interesting components.  The price component was a digital hyperband tuner UV916S by Philips (see front and back picture).  Though I could not find the actual datasheet of this module, I found enough related information on the Internet to engage it in a design (descriptions of similar tuners, the datasheet of its components, …). 

To understand the meaning of the pins (8 in total), I also opened carefully the lids and had a look at the circuitry.  The UV916S contains numerous SMD circuits, one of its main components is the TDA5736 VHF/UHF oscillator/mixer chip.  It does it all for band A :  45...168MHz, band B : 168...448MHz and band C : 448...860MHz.  TV signals (5.5MHz wide) at these frequencies are converted to a 1st IF of 34...39MHz.  One of the three oscillators is active at any time and is stabilised by a PLL circuit based on the MC44829 chip, it contains a crystal oscillator (4MHz crystal) and a loop filter, controlled by commands via an I2C bus.

To be able to receive continuously from 45 to 860MHz was a great attraction, so I decided to build a VHF-UHF down converter as a front end for the SoftRock-40 receiver.  
A SoftRock receiver for the 10m band (version 5, I believe) modified to 36MHz (an other crystal and modified phase shift components), should have done the trick easily.  But since I don’t have this version of the SoftRock receiver, I decided that a 2nd mixer to meet the 7.06MHz of my SoftRock-40 version 6 would not be too cumbersome.  


The circuit

See here for the circuit diagram.   The antenna signal goes straight into the UV916S, where it is filtered, amplified and down converted.  Its output is first filtered by a SAW filter (G3962M, Epcos, also salvaged from the above TV set). 
It is then down converted again to 7.06MHz (the frequency of the SoftRock-40), by a NE612 oscillator/mixer chip. A crystal of 13875kHz (which I had in my junk box) is used in the 3rd overtone mode to stabilise the 2nd LO at 41630kHz.   The 1st IF becomes now 34570kHz, which is a bit exocentric in the pass band of the SAW filter, but still workable.

The UV916S needs 33V to polarise its tuning diodes, a simple DC/DC up converter followed by a regulator (based on an LM317L) takes care of that.  It is running at 40kHz generated by a 555 chip.
The AGC voltage of the UV916S should be somewhere between 0…12V, it is put at +5V for now (the only regulated voltage available).  There should be system gain enough, we will see about that later.
The rest of the circuitry is really straightforward: level shifters for the I2C signals (SCK and SDA) and a regulator (7805) for the +5V.  The entire circuit takes some 250mA.


Dialling application in Visual Basic

I wrote a basic in Visual Basic 6.0 to control the frequency of the down converter, see vhfdcnvt_gui
The parallel port of the PC is used to emulate the I2C bus, only bits 0 and 1 are used.  Note that the 9th bit (the acknowledge bit) of the I2C bus protocol is not read back by the PC, I thought it was too cumbersome to implement it and I could do without it (we will see about that).
This small program features:
* a window for the frequency in kHz, including buttons for increasing/decreasing the frequency,
* a window for a directory of stations (still very much work in progress),
* windows with the actual messages over the I2C bus.

The really annoying thing is that the frequency can only be set in increments of 62.5kHz, this is due to the PLL set up used in the UV916S.  This is not so bad as it seems, as it is well within the 96kHz band when using the SoftRock-40 in conjunction with a sound card sampled at 96kS/s.  Having had an API for e.g. the M0KGK-SDR application to set the frequency of the scale would have been very handy here.

Those interested can download this program including its source here (zipped file).  Don’t forget to install “inpout32.dll” as the program uses it to access the parallel port.


Tests and results so far

The board (double sided 100*75 mm) contains everything, regular through hole components are used.  The top (component side) is not etched and serves as ground plane, pins are simply being soldered on the top side for a ground connection, see a picture of the component side and one of the solder side.
A 9 pin D-shell connector is used for the connection to the parallel port of the PC, the same special cable as for the SDRtwo project is used.

The board was home made, see elsewhere on this Web site for some comments on the “manufacturing” process.
The schematics were drawn using the TinyCad freeware, the artwork was draw using the CirCad (free evaluation).  Find here a zipped file with the design files containing the drawings in their original format and here a large zipped file with the PCB artwork.

Having most of the components for free (already in my junk box or salvaged from the TV set), I spend less than 10 euro on this project.

I had trouble having the 13875kHz crystal (which is actually meant for fundamental oscillation) to oscillate in the 3rd overtone mode.  Finally I didn't use the oscillator inside the NE612, but build an external Colpits oscillator based on a BRY90 transistor. 
The free running DC/DC converter was a bit too efficient, it gave more than 100V.  I had to put a load resistor of 22kOhm to reduce its output to a value more acceptable for the regulator behind it.

I was not able to perform the typical receiver measurements and tests (e.g. the sensitivity) due to the lack of anything that goes beyond 15MHz.  I had nevertheless a series of listening evenings on the lower and mid VHF bands, scanning for aircraft, amateur and maritime traffic.  
I was even able to decode ACARS messages (
Aircraft Communications Addressing and Reporting System.) receiving on 131.725MHz in AM and feeding the signal back in the second sound card while running WACARS 0.7 in parallel with M0KGK-SDR on the same PC.
Attempts to use ShipPlotter 10.7 to decode maritime AIS
messages (Automatic Information System, I live nearby a seaport) failed however (though I could hear the chirps quite clearly) due to the rather marginal NBFM detector implementation of M0KGK-SDR (the author has promised some improvement in the future).

As an afterthought, I tend to believe that building a Soft Rock version at 36MHz (with a modified 10m version of the SR40) and connecting it right behind the SAW filter would have yielded a cleaner signal.
What is however really missing is a way to set the frequency of the scale of M0KGK-SDR directly from the dialling application.  This would be much more convenient than to calculate the actual receiving frequency out of one’s head.