A fully tunable all valve VSB VHF modulator for bands i and iii

I remember Trevor ( Murphyv310) did something very similar to your proposal on our old forum. It was I believe only Band I and used an EH90. He built it and it worked.

At the time I was interested in building it and purchased the sound (41.5Mhz) and vision crystals (45 Mhz) which I still have in stock.

Katie Bush wrote:  "It must be all valve. Able to modulate both sound and vision (lovely old references there). Be tunable across the entire band " (Band 1 only?)

Here's the circuit diagram of the tunable modulator employed in the Murphy TPG11.  Tunes to all five BBC channels. All double sideband though.  A VSB filter would be difficult to realise in a tunable modulator, but not impossible. A superhet type of circuit would have to be employed where the vision bandpass characteristics are determined at a specific frequency. This signal along with the sound signal is supplied to a frequency changer, operating with a adjustable local oscillator the resultant output from the mixer will be at the required System A channel frequency.   However, if you are going to use a superhet circuit the problem of harmonics arises.  

Till Eulenspiegel.

The Mazda 6F33 is a special valve developed for use as a modulator. To prevent Grid 3 from going positive a diode is connected internally to that grid. The 6F33 must have been considered an important valve  in industry because it remained in production well into the seventies. Later marketed by ST&C and ITT.

http://www.r-type.org/exhib/aaa0063.htm

Okay, a tad ambitious I agree, but how about...

All valve, VSB, fixed frequency, and design two modulators, one for band i, and one for band iii? At channels 2 and 10, that would satisfy most of my sets, and if we can make the thing(s) relatively easy to set up at other channels, even better!

That does away with the complications of suppressing the sidebands at varying frequencies, and perhaps as a nod toward modernity (and I'm sure the broadcasters will have done it this way) crystal controlled oscillator(s), as Chris suggested?

You build it, and we will follow?!

Hi Marion, the best plan is to get the Murphy TPG11 working and then assess the performance of it's modulator.  A modulator can be made to operate at a much lower carrier frequency than those of the System A allocations. The output from a variable oscillator will determine to the chosen channel  frequency at the output from the mixer.  

Till Eulenspiegel 

Till Eulenspiegel said
The 6F33 must have been considered an important valve  in industry because it remained in production well into the seventies. Later marketed by ST&C and ITT. 

Hi Till,

The 6F33/CV4064 was used in some versions of the famous Racal RA17 receiver as a mixer in the Wadley first conversion circuitry (analogue synthesis in 1MHz steps with quite a lot going on at low VHF frequencies), as a great many of these sets were used by the British military for many years and they liked to keep in-depth spares stock for a long time into the future courtesy of blank cheques backed by thee and me, there are probably quite a lot of NOS still to be found. I believe that it was developed from an earlier MO base dual control pentode but not sure, I think member Synchrodyne mentioned something along these lines on UKVRR. The approximately similar function 6AS6 was developed from the 6AK5, this was used in US market RA17s.

The predecessor to the Mazda 6F33 was the 6F32.   Both valves have the special suppressor grid characteristics for use as modulators etc.  There was a discussion about the valve in 2011 on the UKVRRR:

https://www.vintage-radio.net/forum/showthread.php?t=74640

A special power supply unit will have to be build for the TPG11. We know about the demands of the valve heaters is >13amps and the HT requirement is only 125 volts @ 100mA.   To make the power unit more versatile what about a series regulator for the HT?

Info about the 6AS6:   http://www.r-type.org/exhib/aaa0065.htm

Till Eulenspiegel.

Posted by: Katie Bush

One thing I've always fancied is a "proper" VHF modulator  ... so I reckon VSB to mimic the way it used to be done.

Does anyone fancy the challenge? - It's way beyond my humble ability I'm afraid, but if someone fancies designing one, perhaps we could all, or at least those who want to, could build it?

Who's willing to go for it?

I only stumbled on this thread by accident a couple of days after posting a design idea for a VSB modulator on the Golborne forum.

The problem, as I see it, is the VSB filter. For a tuneable modulator, that would would require 13 individual filters unless the filtering is done at IF and the modulator is built as an upconverter  - essentially a tuner unit in reverse. As I said, it is only an idea but I see no reason why I shouldn't also post it here as it is my own work so here goes.

Every now and again on forums such as this there is a request for adjacent channel working. Usually because one set has been acquired that is only capable of receiving on the adjacent channel to that used by the rest of the collection.

Unfortunately, all of the modulators usually available are Double Sideband so that, with the exception of channels 1 & 2, adjacent channel operation is out.

I don't have a fully worked solution but an idea that someone interested in the subject might like to develop into a working model.

Some time ago I stumbled on the response curve of a SAW capable of filtering out a single 8MHz channel and presumably intended for use in digital set top boxes. It wasn't what I was looking for at the time but saved a copy to study later. (I also found the curves for a switchable 7MHz/8MHz SAW, presumably for continental VHF/UHF transmissions).

This is what I found:

It occurred to me that, if a vision carrier were to be placed at exactly the right place in the passband, an IF signal with the lower sideband suppressed could be generated. After up-conversion to the required channel - basically the reverse of what happens in a TV tuner unit - the IF signal will undergo frequency inversion which will place the suppressed sideband above the vision carrier, which is exactly what is required.

Here's the first attempt:

Residual sideband suppression in the adjacent channel which, in signal terms, is the upper adjacent channel, is >40dB down at the adjacent sound carrier frequency and +>45dB down at the vision carrier. The unwanted image of the sound carrier is down by a similar amount.

There is. of course, no attenuation of products which fall in the majority of the vision passband of the lower adjacent channel, again in signal frequency terms, so the figures here will depend on how well the vision modulation signal is filtered above 3MHz.

I've placed the limit of the Lower sideband at the 3db down point but it doesn't reach the -30dB point until 1.75MHz below vision carrier. Is this realistic? What is the practical response of a VSB filter expected to be - it obviously can't be the text book square response!

I've played around with the Vision Carrier frequency to see the effect on the VSB of -250kHz increments - see what you think.

Hi Terry,

I couldn't even begin to pretend to know how you do it, but from those graphs, I like what I see and would say it is worthy of greater development.

In any case, I suspect my exact requirements can be modified if the results are equal to that desired - ergo, it might not be achievable entirely with thermionics, and it may be that 'fully tunable' can't be done. These features would be nice, as would the use of appropriate technology of the time, but as long as the correct transmission specifications can be met, it would meet the primary object.

As you have correctly surmised, adjacent channel working is very much what I had in mind. As a goodly few of my sets are drawn from all over the UK, I have a lot of channels to fill, and once again, you are correct, most of them have only two sets of biscuits in their tuners. I fear there must be a finite limit to the number of times one can reset the channel selector on an Aurora before it breaks or wears out, and in any event, it would very very desirable to allow the use of several sets at the same time.

Hi Marion,

I can't take any credit for the SAW or its response! It was my idea to use it in a lop-sided manner to get the VSB response, though. In addition, by displacing the vision IF above the centre frequency it could produce a 625-line VSB IF response but I don't think anybody needs adjacent channel working at UHF!

Remembering back over 45 years to a small room containing 22(!) valve modulators which, even with air conditioning, was decidedly 'cosy', I think dropping your thermionic stipulation might be a good idea for a multi-channel set-up, particularly when you add the heat produced by the sets themselves! Anyway, think of the savings on electricity bills with a solid state solution!

It is obvious from your reply that you are only interested in distributing two programmes but each on multiple channels. This might be much easier than you think!

One IF carrier generator can feed two VSB modulators, then each IF signal can feed a number of up-converters to generate the individual output channels. Crystal controlled oscillators would be preferable unless you have a frequency counter to accurately set all the output frequencies - if the channel spacing isn't right, the adjacent channel rejectors in the receiver IF strips won't be able to do their job properly.

Taking things to extremes, in a not to large a box you could have two IF generators, one feeding five up-converters, the other eight, to produce all 13 channels from B1 to B13!