Hi Ned,
Uncle Ned wrote:I take it you're roughly knocking off the HF87 circuit into the ST70?
Exactly. The LM334Z CCS is the key, as it permits usage in circuits that would otherwise need much higher B+ voltages to compensate for the voltage drop across the tail of the long tail pair phase splitter. Of course, then there's the well known benefits of a CCS tail for high phase splitter performance. But these benefits would mean nothing if the modified Mullard topology itself wasn't excellent - a perfect mix of performance and the classic 'tube sound'. Thermion has used half this circuit for the last year in a Mark III prototype PCB I made with amazing results.
Uncle Ned wrote:Any reason that 12BH7 or ECC99 couldn't sub for 12AU7? There ought to be enough filament current available on even an original ST70 power xfmr to support that. Here's an idea: stick a filament jumper in there so 6CG7 or 6H30 could be used.
Thanks for the confirmation on that, Ned. I decided against the jumper because I was afraid of stressing the power transformer any more. In my mind, I derated the PA-060 10% since some of the AC mains voltages in the US are way off these days, for example 128VAC in the town I work!
On my Mark III board I did have jumpers for switching the filaments for 6CG7s and indeed, Thermion and I have tested 12AU7s, 6CG7s and 12AV7s in this circuit. Thermion has told me he will cut-n-jumper his ST70 driver to run 6CG7s. FWIW, I've got the best test numbers using 12AU7s but kind of prefer the 12AV7s musically - maybe computer tubes are the best for processing digital music. Yellow_Light_Colorz_PDT_04 I just don't have a quality stash of 6CG7s. I'll have to test the 12BH7s.
Shannon