Stagger-tuning of FM stereo valve IF stages

[regenfreak]

No, as your drawing implies you're damping one half of the filter you're trying to align.

I am a bit confused. Isn't it the same by strapping a 2k2 across the primary? Why a resistance of 2.2K?

It's important that you don't in any way couple onto the pair you're trying to align. Drive the sweep signal into the grid of the preceding pentode. Monitor the output from the anode of the following pentode. This output is damped as you suggest to prevent the next pair altering the response, but all you need is the 2k2 across the primary (as the coil is s/c DC) and say 1nf @500V from the "hot" end of your resistor to feed whatever device you're using to measure RF voltage.
John

Thanks for that. I think, what I did was to drive the sweep signal into the grid of the preceding pentode. Monitor the output from the grid of the following pentode with a 100k isolation resistor...I was trying to measure IF amp stage gain, fo and BW.

I thought "monitoring the anode of the following pentode" would include the effect of the primary IF of the following IF stage. I would draw picture to see if we are thinking on the same page

[Radio Wrangler]

Valves tend to act as current sources, so the signal voltage on the anode of the following stage is a strong function of the resonance of the next IFT. So if you want to see the response of one IFT, then the grid of the following stage is the place to look.

David

[regenfreak]

Originally Posted by Radio Wrangler

Valves tend to act as current sources, so the signal voltage on the anode of the following stage is a strong function of the resonance of the next IFT. So if you want to see the response of one IFT, then the grid of the following stage is the place to look.

David

Thanks! I was doing it right then. At MW I can be a bit slack in probing technique but the FM measurement is always tricky.

[John_BS]

I'm withdrawing from this discussion as you're now receiving inconsistent advice, and that can only add to the confusion.
John

[regenfreak]

I will try probing both grid and anode of the following stage to see what the differences they will make on the visual sweep. I always learn something new by digging deeper.

The temptation of any project is that once it is working after the construction/repair, we tend to get done with it. The fun part is when probing technique looks simple on the surface, it isn't simple underneath. When the measured waveforms look weird on the scope and then wait a minute...Houston we have a problem.

For sure, the the stage gain can only be measured by the grid of the following stage. Otherwise you include or multiply the gain of valve amplification of the following stage.

[Radio Wrangler]

What you can di is to fit a resistor load for the anode of the following stage then that following stage acts as a buffer for your probing and you don't therefore detune the IFT you're interested in because the load pulling it is the grid of the stage after, which even with the resistive anode load fitter will be close to what the IFT has to handle in proper use.

There are many ways of doing these things, but on the whole it saves trouble to look at things in as close to their natural environment as you can.

Always be aware that filtery things are heavily dependent on the loads applied to their input and output.

For example, if someone gave you a box of ideal lossless capacitors and inductors, you'd think you could make some pretty spiffing filters, right?

Right!

Now, that energy you put into the finished ideal filter, there's stuff on the wanted frequency and on unwanted frequencies. Where does the unwanted frequency energy go? It isn't allowed out of the output, and all the filter components are lossless, so there is only one place for it to go - reflected back out the input. Even with lossless components your filter will have a transition region where the reflection complements the passband.

For coupled resonator filters, what about Q? it's kind of important for positioning your poles to get a chaosen response, but your lossless parts all have infinite Q. B*gger! your filter is going to be infinitely narrow. Nope, it relies on the resistive component of both the source and load impedance to bring down the Q of the first and last resonators respectively. The inner resonators have their Qs set by coupling to the outer ones.

So source and load impedance are critically important.

David

[regenfreak]

What you can di is to fit a resistor load for the anode of the following stage then that following stage acts as a buffer for your probing and you don't therefore detune the IFT you're interested in because the load pulling it is the grid of the stage after, which even with the resistive anode load fitter will be close to what the IFT has to handle in proper use.

There are many ways of doing these things, but on the whole it saves trouble to look at things in as close to their natural environment as you can.

Always be aware that filtery things are heavily dependent on the loads applied to their input and output.

This is an interesting idea. I doubt many people would take the trouble to try to solder and desolder a resistor load into the anode.. There is no 100% certainty that detuning problem at the grid does not occur in spite of the use of large value isolation resistor attached to the probe tip.

So source and load impedance are critically important.

Speaking of source and load impedance, the previous VNA photos only show CRUDE resonance measurement (probably incorrect/misleading results) using S21 insertion loss(blue curves) on the NanoVNA v2.

Instead of building a test jig with LC input and output impedance matching networks, a magnetic sniffer probe ( see attached photos)can be used to measure the resonance frequencies of the primary and secondary IF. This contactless measurement gives more accurate resonance frequencies of double-tuned IF LC filter eliminating the detuning effect, impedance mismatch etc with S21 measurement. As you move the probe along the longitudinal axis of the IF coils, you can see the shift of dominance of the dips: f1 and f2. The yellow line is S11 return loss.
In theory, you can measure the Q using the contactless probe but in practice the S11 dip is too small to allow accurate 3db readings.

[Radio Wrangler]

Q is difficult to measure, the very act of getting some signal out to measure the growth of peak voltage damps it, the very act of injecting current etc, ditto.

What you do is make a measurement, then try again with lighter coupling? Did the result change? If so lighten the coupling yet more until you start getting pairs giving the same result.

David

[regenfreak]

Originally Posted by Radio Wrangler

Q is difficult to measure, the very act of getting some signal out to measure the growth of peak voltage damps it, the very act of injecting current etc, ditto.

What you do is make a measurement, then try again with lighter coupling? Did the result change? If so lighten the coupling yet more until you start getting pairs giving the same result.

David

Yes, absolutely. It works like a grip dip meter. When the coupling is too tight, you get misleading readings. So the coupling should be as light as possible but then you get a very small dip. That's why grid dip meter is not an accurate way of finding Q.

The VNA works like a weak transmitter. The amount of reflected RF power becomes less when the VNA sweep frequency coincides with the resonant frequency of the LC resonator that it "draws" the energy.

[regenfreak]

I could not resist it in the name of science. It is difficult to get a steady dip reading...very fiddly.