Adding a frequency counter to a valve set

[Alf]

I'm thinking of connecting a frequency display with i.f. offset to equipment which uses valves, looking at the circuit, I suppose the best place would be to connect it to the earthy end of L4 via a capacitor. I'd be interested in any other suggestions.

[Radio Wrangler]

Frequency counters are noisy beasts and there are lots of things going on in them at various frequencies. The waveforms are fast edged pulses, so all those frequencies have harmonics.

A small amount of all this muck comes out of the counter's input and will quite happily travel backwards to any receiver it's connected to.

So to avoid infesting your receiver with lots of sproggies, good practice is to choose your pick-off point carefully and to have a buffer amplifier with good isolation in the reverse direction.

The same sort of thing happens with modern digital oscilloscopes, there is quite a bit of RFI comes out of their inputs from the sampler chips and you can hear the effect on radios when you're probing around.

David

[G6Tanuki]

As Radio Wrangler has observed, counters can generate noise [though the modern types with LCD displays are a lot quieter].

The issue you're likely to come across is the additional capacitance of any screened/coaxial cable between a counter and the local-oscillator of your radio can cause the frequency of the oscillator to be 'pulled' away from the originally aligned one, and the additional load of the cable/counter can affect the stability and amplitude of the oscillations too.

What I'd suggest is building an "isolation/buffer amplifier" - either a cathode-follower [a simple triode valve like one half of a 12AT7] or a source-follower [a FET like a 2N3819] in a very small box that can be fitted close to the local-oscillator in the radio, very 'loosely' coupled to it (a capacitor of a few pF) and the signal from this isolation amplifier to the counter can then be fed down quite a length of coaxial cable to the counter without problems.

The cathode/source-follower has a very high input impedance (so will minimally affect the oscillator function) and a low output impedance (typically 100 Ohms or so) so will match well with the input impedance of most counters.

[Alf]

Thanks David and G6Tanuki, I'd completely forgot about needing a buffer stage. I came across this from G3YNH, hopefully, this should be ok.

[G6Tanuki]

Hi Alf.

Yes, the circuit shown will work just fine - the BF256 FET gives the necessary high impedance to prevent the counter and its coax-feed from unduly loading the receiver's local-oscillator.

When I last retro-fitted a counter (in the 1980s) I used a little module called, from memory, the "DFM7" from Ambit/Cirkit which used the "FC177" counter-display module along with a Plessey prescaler chip to let it count up to 39.999MHz. This module had a built-in buffer-amp using a BF256 FET.

[Biggles]

I have used the FCM177 module before albeit a long time ago and didn't encounter any spurious problems with it inside an HF set. You have to choose the IF offset by hard wire links so it's worth checking if the module you eventually use can support your particular set. I seem to remember you could extend the count range by adding a prescaler which enabled use for the FM broadcast band, although the resolution obviously suffered in this mode. I know there are modern versions of these around but I have no experience of how noisy they are.
Alan

In the good old days at Labgear we made the 'vox pop' TV monitoring kit for audience figures. To decide upon which channel the viewer wanted a bit of coax was poked into the tuner to pick up the local oscillator, no contact, merrily a bit near.

[turretslug]

http://www.vintage-radio.net/forum/s...d.php?t=110931

gives some useful general insight and musing into buffers with a very light touch- a classical valve HF receiver of typical 2.5-3:1 frequency span over each range with free-running ganged LO depends on minimal stray capacitance at range HF end- at 30MHz, even an apparently insignificant extra few pF may shift it down by MHz!

Beware also of meaninglessly high resolution- a set with LC IF filtering may be plus/minus a couple of kilohertz off the nominal frequency- contemporary workshop signal generators used by previous owners may not have been terribly precise or accurate. If you find that, say, Radio 5 Live sounds off-tune at an indicated 909kHz but fine at 911kHz read-out, there's a clue! Even multi-element ceramic fixed filters may have a tolerance of several hundred hertz. Unless you can access traceable, lab-grade certainty re.IF frequency, going higher than about 500Hz, even 1kHz resolution doesn't really mean much.

[turretslug]

Given that the LO probably produces several volts and counters typically need tens of millivolts, you may find that not only can the buffer be lightly coupled, it might also be possible to use a pick-off pad on the buffer output to further reduce digital hash finding its way back, favouring both the drive capability of the buffer output and matching the output coax appropriately and largely resistively. The classic high-level free-running valve LO can be very clean spectrally (harmonics aside), it would be a shame to introduce any more broadband hash/spot spurioes to the mixer than necessary!

[Bazz4CQJ]

On some older receivers, notably the HRO, it would be exaggerating only slightly to say that a few inches of wire wrapped around the power lead would suffice. In its original form, the harmonics generated when listening on 80m with my HRO would cause significant interference to the Band 3 TV set in use on the same mains circuit!

In practice, a lot will depend on the sensitivity of the particular frequency counter and I would think it best to approach the task by starting work at the highest frequency involved.

B

[Bazz4CQJ]

Something has been bothering me about this thread; I think it's to do with the issue of "matching" the output level of the actual LO with the drive required by the particular counter, which addresses the challenge beyond that of simply avoiding loading the oscillator. Most counter modules made for this purpose will only work happily over a limited input range, say 50-150mV.

This subject has come up previously https://www.vintage-radio.net/forum/...ad.php?t=77674 and I recall that Bob Green published the circuit he had successfully used in his HRO and I think it had 3 or 4 transistors. Unfortunately, the old link to his schematic no longer works, but I've PM'd him asking if he could re-publish it on this thread as it would be useful to see that proven circuit again.

B

[turretslug]

Re-reading Alf's original post (relevancy check doesn't go amiss!), I think the point suggested would be as good as any, that is the end of the feedback winding remote from the grid- bearing in mind previous points, this will have lower amplitude and be less affected by stray capacitance than goung across the tuned circuit itself. A combination of inductive and capacitive (padder C-derived) feedback coupling has been used, presumably in the interest of keeping amplitude more constant across the band. Is this an HF variant of a command receiver?- there's something vaguely familiar about it.

[Alf]

Yes, it's the 3 - 6MHz version, the original tuning drive shaft is missing, I've tried to get hold of a replacement drive shaft but no luck, otherwise, it's working very well.

[turretslug]

Proof that US designers weren't entirely averse to using the triode-hexode- though even later versions of the BC348 used a 6SA7 in self-oscillating mode at 18MHz.

[BobGreen]

As Bazz suggested, here is the DFM I built for my HRO. The amplifier sits adjacent to the Local Oscillator so the "aerial" picks up from the grid lead to the top cap. Power is fed down the coax from the DFM module that has the programmable IF offset set to 455KHz. It work from 100KHz to 30MHz and is powered from rectified 6.3VAC heater supply

[Bazz4CQJ]

Thanks for re-posting that Bob; I'm still working on the re-build of my HRO but have the counter module sitting on the shelf. The idea of using an antenna rather than direct capacitive-coupling sounds like it may be a good way to go.

B

[BobGreen]

The HRO is know to have a aggressive LO to the point that it is strongly radiated via the RX antenna. You can imagine the Nazis listening to all the HRO LOs to work out which frequencies we were monitoring. Probably then having a laugh is we could never crack their gobbledegook Enigma codes!

[Alf]

Thanks Bob, I think I'll have to build that circuit.