Um, that doesn't affect the drive to the head! 
I found the problem, the Bias/Erase oscillator circuit on the EL84 was a marginal design. There are no voltage taps, so the supply will be lower here than UK, EL84 may be slightly aged. The R&TVS had the clue as mentioned above:
I always clean the heads, capstan and pinch roller.
I had a stranger problem on the TK18 which I did next. But sorted. I was close to scrapping it for parts and valves.
Yesterday I sorted a McMichael 235 I've had for years. Still a big backlog on the Round Tuit pile.
I found the problem, the Bias/Erase oscillator circuit on the EL84 was a marginal design. There are no voltage taps, so the supply will be lower here than UK, EL84 may be slightly aged. The R&TVS had the clue as mentioned above:
Quote: But ALSO later versions have a 6.8k Ohms in PARALLEL with the R16, 22K. Why do that rather than a 5.6K, or just a 6.8K? My 6.8K resistors look either stupidly tiny or about x3 bigger. So I thought I'd try 2x 15k in parallel, soldered to the track side of the PCB across R16, 22K. There was then massive erase AC on the erase head.
I always clean the heads, capstan and pinch roller.
I had a stranger problem on the TK18 which I did next. But sorted. I was close to scrapping it for parts and valves.
Yesterday I sorted a McMichael 235 I've had for years. Still a big backlog on the Round Tuit pile.
