Nombrex 31 RF Signal Generator - off fault with one part range

[6SN7WGTB]

I have acquired of the above RF sig gens.

It is untouched from new so far as I can see. Installing a new PP4(!) battery showed it to be 100% functional.

The only immediate fault was that the modulation is more like 900kHz than 400kHz as specified.

Notwithstanding that, it actually works on all ranges and the tracking of dial to frequency is so good I was reluctant to touch it.

What I have just noticed is that on one range only (55-165MHz), there is no apparent output from 55-70MHz (about 1/3 of range), then it pops up and is fine until the upper end.

I'm really puzzled by this as (a) the tuning is via one ganged VC applicable to all ranges, and (b) 2/3 of the range works fine.

Most resistors have wandered, as one might expect, some by a large margin (most by 20-30%, a few 50%+) but curiously it still works well and tracks well. Which is why I left it alone not spotting the one range glitch.

I'm simply not experienced enough with RF to know why this is displaying this, to me, odd fault on part of one range.

The schematic is attached. The fault is on part of Range 7.

Advice most welcome!

[Ed_Dinning]

Hi, from memory I think the top frequency range was the second harmonic of the previous range, which, if at a but of a low level may not give much output.
Have a careful look for resistors that have gone high

Ed

[6SN7WGTB]

You're right Ed re the top range, but this is the range below.

How could a resistor gone high cause part of one range to 'stop' oscillating but the remainder (2/3 of it and right through the FM band!) to be exactly spot on?

Sorry of that's a daft Q, but I am newly getting in to RF...

It was because I had used and tested it on the FM broadcast range (88-108) that I didn't spot the 'hole' from 55-70-ish.

[6SN7WGTB]

Also, my curiosity is piqued by the fact that any resistor (albeit off-spec) is common to every range.

[Ed_Dinning]

If it is the higher frequency part of the scale it may reduce the gain of the oscillator sufficiently to stop it oscillating, but could be wrong. I had also thought it was the 2nd harmonic range that was failing.
A good clean of the switches may also help

Ed

[6SN7WGTB]

It seems I may have got to the bottom of this.

Having decided to replace the nine resistors, removing the front panel knobs to dissemble showed that the VC rotor shaft was perilously close to the front aluminium faceplate (which is battery ground).

The faceplate is captive by the three other switch/VR shafts that are held to it (and the underlying PCB) by their shaft nuts.

It's not a particularly high quality device and the slight misalignment of holes punched and/or fixings means the VC shaft, which has an insulated spacer between PCB and the faceplate, can just short at certain rotations. The spacer has no collar so the shaft isn't protected from touching the edge of the hole in the faceplate.

It would appear that the combination of a certain range position and VC setting resulted in the slightest of shorts. Evidently, that one range position must have provided just the tiniest amount of twist/movement to the PCB meaning the merest of shorts could occur with an equally tiny eccentricity of the VC shaft.

Solution: can't move the positions as they're soldered/bolted to the PCB, so opened out the VC shaft hole and placed a small O-ring on the shaft into the annular space between it and faceplate, to act as an insulating collar.

[Cobaltblue]

Title updated.to make it easier to search.

Despite the poor build quality these seem to be popular little units so hopefully your discovery will help others

I have several of the Nombrex units mostly due to the local manufacturer (Exmouth) interest.

Great job fixing it.

Cheers

Mike T

[6SN7WGTB]

Thanks Mike - I couldn't work out how to go back and edit the title!!!

Used it to align a Roberts R300 and I have to say can't fault it for practicality.

It would appear to be pretty close on dial to output match!