I take the opportunity to investigate the phenomenon of not being able to reduce the volume to zero. I've encountered this with many old radios but I've never investigated it until now.
Thank you Jim for your great efforts to post these videos. Interesting to read the in-depth and detailed views by other experts too. Being a beginner, it’s an exciting journey for me, I’ve been faithfully following your work and all the interesting stuff that you do, including the discussions kicking out of this exercise. Keep it up! Best greetings from Vienna. simon
37:59 this whole explanation as interesting and logical as it sounds would only make sense if the diode plates and triode section were sitting exactly opposite to the same piece of cathode and sharing the same spot of emissive material. The physical layout of the 12SQ7 however is different and puts this explanation into the realm of urban legend. Looking at the internal structure of the 12SQ7 the systems of diodes and triode are stacked with the triode on top and the diodes on the bottom. They wrap around the same cathode tube but at severly different heights. So in Jims words, that would mean triode and diodes draw not from a common pool, but from their own reservoirs of electrons. A possible explanation of the experienced crosstalk is that there is a small resistance in the path of the common cathode to ground. Any cathode current caused by the diode demodulating AM would cause a tiny voltage swing of the cathode line with respect to ground, which in turn also acts as audio signal to the triode cathode; that is what is called cathode modulation, as opposed to the usual grid modulation. Or aka common grid amplfier circuit, a circuit often found in UHF band amplifiers (UHF TV tuners) The more probable explanation is however a RF crosstalk between the diode to the triode due to inter system capacitances. The RF that is meant to be rectified by the diode section capacitively couples into the triode section, where it gets not only amplified but also rectified by subsequent tubes until the RF is blocked at the audio output transformer. Just think of it in the same way an FM radio IF is aligned; by simply blasting the 10,7MHz onto the glass envelope of the FM frontend tube, and hoping for the FM IF to capacitively crosstalk into it.
47:58 That is nonsense. All through the 50s and 60s german manufacturers used the EABC80 tube which also has two diodes and the triode electrically sharing the same cathode connection. But there has never been any issue with this crosstalk you see. It is just because the tube was constructed correctly using physically separated cathode tubes and individual, parallelled heaters. And most importantly circuit layout was done with care to ensure a direct zero run length ground connection of the cathode pin, and shielding(!!!) of the triode grid by means of both shielded wire and a shielded(!!!) coupling capacitor, to prevent not only hum but also IF pickup (which subsequently would get demodulated). None of this; neither careful wire dressing nor a shortmost path from the 12SQ7 cathode to ground, nor shielding of the triode grid coupling cap is seen on your radio. This radio is RF wise a crosstalk mess. It is you cheap radio design that is at fault, not the tube. Plus you did not even observe right orientation of the grid coupling cap with respect to the outside foil going toward the volume pot not grid, in order to at least make a small effort to reduce signal pickup on the super sensitive grid of the 12SQ7 triode.
21:22 The negative voltage on the grid is caused by both the proximity of the grid to the cathode and the flow of electrons from cathode toward plate. There does not need to be any audio. It is a static effect. And -600mV seems low for a grid biasing tube. An EABC80 triode in this configuration and on 10Meg grid bleeder resistor would be at -1,3V grid bias IIRC. Maybe that triode section of the tube is weak or plate voltage is too low.
9:00 Wrong assumption. The 10Meg grid bleed of resistor is essential for this type of self grid biasing tubes. Its value is EXPERIMENTALLY chosen by the designer to set the correct equilibrium between the negative grid charge that the tube develops and the amount that is bled off to ground, with regard to a compromise between max amplification and min distortion. And the bias point that the tube eventually settles on is extremely dependant on its age, make, model and even the load down on its plate circuit which a following tube presents. Any attempt to measure the grid bias potential on grid biasing tube is bound to fail unless an extremely high impedance meter (>100 meg) or a compensation technique is used. You may however get away with removing the 10Meg resistor and placing a DMM with a known 10Meg impedance in its place. Another trick is to put a 470nF to 1uF polystyrene low leakage cap across the 10M grid biasing resistor. Letting the tube run for several minutes the cap in parallel to the grid bleeder resistor will charge up to the grid bias voltage across the 10M, and if you then - and only then - touch the cap with your voltmeter or better yet digital scope, the initial(!!!) reading will show u the actual grid bias voltage that the tube had settled on. But that being said, and going back to the efforts of our blind tinkerer, to anyone with even an ounce of technical knowledge it becomes more than evident that any capacitor leakage in the coupling cap going to the grid of a grid leak biased tube will be extremely disturbing the bias point. As a rule of thumb it was said that the leakage resistance of the coupling cap between the volume pot and the triode grid of an EABC80 that which was grid biased with 10meg, should be above 200Meg!!! Saba who biased the EABC80 with as high as a 22Meg grid bleeder even spoke of 500Meg maximum leakage. Both these leakage resistances are way way way higher than Jims capacitor checker can resolve. Just to put this into perspective; if you put such a cap on a 200V (500V respectively) supply the leakage current thru it must not exceed 1uA! So much for Jim and his "not that bad, still serviceable" comments when using his crapacitor checker
I was going to comment on his measurements but have grown tired of it and look at it as a lost cause. Clearly he does not understand the measurements or the test equipment he uses. Thanks for your comment.