6SJ7 Grid Leak detector and 6K6 amp 

After acquiring a few untested 6SJ7GT RF pentodes and 6K6GT power pentodes, I needed a way to test them. So I built the simplest radio with amp I could come up with, a Grid Leak detector with an output stage.

The coil is something I had left over from a project a decade ago, I don't remember the exact specs, just that I wound it for the 320pF section of the capacitor. The tuning capacitor is a 320pF+120pF ALPS I got off eBay years ago. The output transformer is a 70V PA line transformer, and the valve (tube) sockets are 8 pin octal relay sockets. All screwed onto a scrap of pine board I had lying around.

This version thrown together with random parts got listed on Hackaday, https://hackaday.com/2024/05/23/grid-leak-radio-draws-the-waves/

The circuit is as simple as I could make it. The screen grid controlled detector idea comes from https://www.angelfire.com/electronic/funwithtubes/Grid_Leak-1.html which has a better explanation of how it works. 

The single ended amp stage was initially intended to have about -14V grid bias, to keep the current down to 10mA since my HV power supply is a bit weak. It ended up measuring -11V. The screen grid of the 6K6GT can cope without a dropper resistor when the B+ is below about 180V. The 6K6G/GT is an octal based version of the 41 output pentode. Back then, being able to save a few cents on a resistor in the radio output was significant, so it was designed to cope in cheaper radios.

The next step was to transfer to a better board, and wind a new coil.

The new tuning coil is 100 turns of what I think is AWG30 wire that I scrounged from a washing machine pump motor that I replaced, wound on a 29mm cardboard tube. It's tapped at 25, 50 and 75 turns. At the top end is 30 turns 5mm away for antenna coupling purposes. At the ground end is 19 turns, for future regen experiments. Mounted on a scrap of wood from a bed slat with 3mm screws and spacers, and tag strips on the side to avoid wear and tear on the coil itself when I keep changing circuits. The tuning range isn't quite right, but it covers enough of the band for my experiments. 

The other experiment in the video was to try a tone capacitor across the speaker transformer. 4.7nF sounded reasonable to my ears. 

Of course this left me with a spare socket, and measuring the signal voltage with a meter is boring.

The 6E5C magic eye tube, a Soviet version of the original 6E5, but with an octal base instead of 6 pins. Also sometimes listed in English as the 6E5S or 6E5G. 

Before moving on, I made a few measurements of current and voltage to see what was going on. The pair of 220ohm resistors make a virtual centre-tap for the heater to reduce hum. Didn't make much difference, most of the hum appears to be coming down the aerial. The magic eye has a reasonable range of display for the stations in my area, though I do have a couple of very strong signals.