Marconi TF2002 Signal generator repairs.


"On the Bench"


13th October 2003.

This started out as a quick re-alignment of my BC-348-Q, but rapidly went downhill!

The Marconi TF2002 is a signal generator that covers from 10KHz to 72 MHz, with an internal variable frequency AM modulation source, and the provision for an external voltage to be applied which will either generate an FM signal, or swep the generator over a small portion of the selected band. The generator also has an electrical incremental frequency control. The incrementral control, and the sweep facility make this unit ideal for aligning communications receivers. There is also a convenient seperate ouput for driving a frequency counter or a seperate digital synchroniser.

The symptoms of the fault(s) are -

No or very low level output from the RF out socket.

No indication on the "Set Carrier" meter.

Incremental frequency control does not work.

Calibration is off on a couple of bands.

Crystal Calibrator is not working correctly.

Where do I start?

Well, luckily I have the manual, so after taking off the outer cover, I can check the power supply voltages, which require a slight adjustment. All the RF circuitry (except the output attenuator) is in a large screened box, which is connected to the main chassis by a large multi-way plug and two BNC connectors, disconnecting these allows the unit to be removed from the chassis, along with the band switch and dial unit, once it comes out, the screening box can be removed, to reveal............

(picture shows the top of the RF unit, with the outer screening cover removed, and the ALC board cover removed. Boxes to the left are oscillators, and to the right are RF filters. The box at the top contains the supply filters and crystal calibrator.)

a load more screened boxes! At this point, reconnection of the multi way plug allows me to check the two outputs (RF and Counter), the RF output is about 70mV (it should be 1V), but the counter output gives 30mV ( the manual says >10mvV). A quick look at the block diagram suggests that the fault lies within the Automatic Level Control (ALC) or Wide Band Amplifier units, as the counter output is taken off befor this point.

Strangely, the incremental frequency control has now started to work, so a load of switch cleaner is applied to the multi-way plug and socket, as this is the most likely cause!

(Multi way plug an socket on main chassis)

Removing another layer of screening reveals the ALC board, and the edge of the WB amp (it is mounted vertically behind the filter unit). Trying the set up procedure on the LAC bord doesn't improve matters, so I move on to the WB amp. To get at this, you have to partly remove the filter box from the back of the RF chassis, revealing a small board with a few transistors, and a couple of Zener diodes. A few checks with the meter, and it appears that the last pair of transistors are missing their bias supply, so suspicion has fallen on the Zener diode MR2, which provides about 4V of bias to the output stage of the amplifier.

Tommorrows job - change the Zener!


14th October 2003.

Well I changed the Zener, and the output came back! There is currently a 4.7V unit fitted in place of the 4.3V specified, I'm hoping that a 4.3V diode will arrive tommorrow.

(Underside of the RF unit, Wideband Amplifier board is at left of picture)

The lack of indication from the carrier meter has been traced to an open circuit inductor in the filter unit, it is a stock value, so can be easily replaced - another order for Farnell I suppose! Interestingly, this is not the first time I have had problems with filter inductors in Marconi instruments, I have a TF2016 signal generator, which came as a faulty unit, the symptom being unstable output, I traced this to an intermittent inductor in the ALC line.

(Underside an rear of RF unit. The filter box cover has been removed, faulty inductor was on PCB to the left of the picture.)

During the dismantling procedure, one of the Nylon shaft couplers on the bandswitch disintegrated - these join the four parts of the 1mm x 4mm flat section shaft. I have looked for suitable alternatives, but can't find anything, so I will probably have to make some out of brass. I shall use brass, as the parts don't need to be insulated, I have some 1/4" square section I can use, and it is easier to machine (with my limited skills), so I think I will have to spend a couple of lunchtimes with the works milling machine.

22nd October 2003.

The homemade shaft couplers are complete, and pictured below, alongside an original coupler, and a section of the shaft.

To give an idea of size, the brass coupler is 0.25" (6mm) square.

Fitting the couplers has improved the action of the bandswitch, and eliminated a slight intermitancy that was evident in changeing band.


23rd October 2003.

Fitted a replacement Inductor in the DC filter unit, and then fitted a second, as the first was faulty out of the packet! I then ran through the setting up procedure. The generator is within spec, except for a slight tracking error on the low frequency end of band C (about 4KHz out), which cann't be corrected by adjustment of the variable inductor. To correct this requires fitting of fixed padder capacitors within the oscillator unit, and is not worth the effort, as I intend to use the generator with a frequency counter attached.

Reassembly of the generator is straightforward, the only problem being the weight when it is all back together!

That's the end of this project, now to go and put it to use!

Home "On the Bench" Backroom A-Z Index

Copyright J.Beacon & M.Wroe-Parker 2003