Cossor 1324 FM generator.
In those days when 'scopes tended to bring the output of the sawtooth generator out to a terminal, it meant that manufacturers like Cossor could use this to drive the reactance valve. It also meant that the output frequency was kept in perfect synchronisation with the 'scope, and it saved on a few valves too. Same principle used on the 1320 TV alignment generator and the earlier 343 sweep generator.
Rather different methods for creating the sweep were used by Samwell and Hutton and other manufacturers.
Thanks Mike, I thought it would have been self contained.
Ok on the X out for sweep gens, I used to use a Samwell and Hutton at work in the 70’s, forgotten most of how I used to use it though.
One capacitor was found to be leaky, it was a tubular 0.02uf TCC (C6) that couples the output from the 1 khz audio oscillator. It had a healthy 25vdc passing from the anode of the oscillator valve, V2 6AM6.
The X output from the 'scope causes the carrier output to sweep in accordance with the set timebase frequency and level as expected. The 1khz audio (when selected) also causes the carrier to be frequency modulated, but as seen previously in the photo's, the waveform contains many unusual components.
Leaky TCC "Metalmite" C6 pictured centre.....
Now, if you know me by now you will probably guess that merely replacing that capacitor with a modern equivalent is not enough. I re-stuffed it with a modern replacement. Yes, that's right...completely bonkers, but hey....why not? ?
I have also purchased the manual from Mauritron for a nominal £7.50 in order to enable me to fully understand the operation and how to adjust/set-up and use the generator correctly.
the crystal probe had never been used, maybe the owners forgot they were there ?
Quite correct, it is in fact a crystal detector probe. It does indeed look as if it has never been disturbed on mine either.
One capacitor was found to be leaky, it was a tubular 0.02uf TCC (C6)…..
Correction, it is in fact C8, not C6.
The reason for some of the rather jumbled waveforms I obtained previously is probably due to the fact that the voltages from two oscillators are mixed together. The frequency difference provides IF coverage of 7.5-27.5 MHz, whereas the sum provides RF between 87.5-107.5 mhz. No wonder the results shown on a frequency counter connected directly to the RF output proved less than stable. ?
If you put some frequency selection in the circuit to the scope, I.e 10.7 MHz or say 90Mhz does it look any cleaner?
If the scope wont work at 90Mhz then a receiver tuner to 90Mhz and check the IF waveform would do the same job.
A reasonably frequency selective device was found and the Leader FM generator set to 95 MHz and the output loosely coupled to the radio's antenna. With radio duly tuned in to a 400hz tone at 95 MHz (radio tuning dial not very accurate), the tone was heard. I then reduced the Leader signal strength and tuned in the Cossor to 95 MHz. I was rather pleased to discover the 1 KHz tone from the Cossor at exactly 95 MHz. The same was tried at various FM frequencies and levels, I was quite surprised at the accuracy of the Cossor.
The Leader generator is spot on with it's output frequency, so it served as a good reference for the Cossor.
Later, I will test the IF frequency of 10.7 MHz.
I think it is fair to say that the obviously good storage conditions of the Cossor over all these years has helped to preserve it's original calibration. There are some more test's, like deviation and sweep bandwidth plus of course the IF frequencies to check, but I am pretty confident these shall be ok.
The above test was only a basic check of a general nature, but it proved the Cossor is outputting something sensible for an FM radio.
I have a Cossor 1322 Telecheck generator, a Cossor 1099 miniature O'scope, and a Cossor 1324 FM alignment generator on order. I also have several Cossor radio and television service manuals.
The 1322 and 1324 are wobbulators which work with the O'scope and receiver's detector circuits to trace bode plots onto the scope screen at about 50 Hz to 100 Hz.
The Bode plots trace the response curve for the frequency range of interest.
The Cossor 1099 puts out 250 V .. 0 V sawtooth wave which is connected to the X and Gnd inputs on the wobbulator. The sweep frequency is set to between 50 Hz and 100 Hz, or could be selected to mains frequency.
To get a response curve signal that can be fed to the scope Y input you disable the automatic gain control on the receiver and connect the scope's Y input to the detector output.
The frequency swept response curve is called a Bode plot.
You also use a second signal generator to feed a marker frequency to the sweep frequency and this will appear as a blip at the corresponding frequency on the Bode plot. This marker frequency is how you calibrate your Bode plot display.
The Cossor Telecheck 1422 contains the reactance valve, sweep generator, mixer, a second marker oscillator that can be set to a known frequency, and a crystal oscillator that outputs a signal every 2 MHz that together with the magic eye is used to calibrate the marker oscillator.
Where exactly you connect, and what to expect of the stage of interest, are detailed in the service manuals for the Cossor receivers.
The service manuals for the Cossor wobbulators detail how to repair and calibrate the wobbulators rather than how to use them.
My task now that the Cossor scope and Telecheck Wobbulators are repaired, will be to build some device that can replace the receiver detector circuitry, and then I would be able to play around with tuned circuits and response curves, and get them to display on my vintage Cossor oscillograph.
To help the process along I am designing and building helper circuitry.
I used sweep generator, scope and marker gen to align TV’s etc, forgotten most of the setup, didn’t know the display was called bode plots though.
The detector circuit should be straightforward, it a DC plot that’s displayed, I don’t know the output level of the sweep gen you are going to use but it will likely require amplification at the frequency in use before the tuned circuit on test.
If the output of the sweep gen is say 200mv, it hardly enough to turn on a germanium detector diode.
Hope that makes sense. 🙂
Thank you. A scope usually displays a waveform, Y amplitude, against X time. With a bode plot you are looking at Y frequency response against frequency along the X axis.
I am exploring the advantages of these old machines and how they managed to fulfil functions that were way beyond their basic specification...
Both the wobbulator and the scope were isolated from earth - meant to be used for working on mains-live chassis. That is a facility worth having available in the workshop.
Modern scopes on X10 lead setting contrive to have 1M ohm in parallel with 20pF "looking in", and a switching attenuator with up to 20 positions. So I am contriving that for my helper box, plus battery amplification so that it registers on a millivolt or so at most sensitive. The Japanese O'scopes, e.g. Hitachi of the 1980s - their user manuals give full circuit details.
Digital scopes do not show fast-changing signals (e.g. oscilloclock), CRTs do.
Electrically noisy high voltage signals can damage a modern scope, but with old scopes you can switch them directly to the deflection plates with relative impunity.
Basically I am "playing"... Enjoying myself with electronic experiments that use these fascinating old instruments. The old TV sets, radios and radars that they had been used to repair are no longer available, but exploring electronics... 🙂