[bikerhifinut]

Some will remember the headache I had with a pair of mono power amplifiers I built based on the Leak TL25+ circuit with a couple of tweaks and mods to suit modern sources.
I eventually cut my losses a while back and sold off the mechanically noisy mains transformers as an expensive dead loss.
Enter Joe Bog who kindly donated me one of his own wound Toroid mains transformers. Massively over specified for this job as it will power 8 x EL34 and associated driver/preamp valves.
As the heater secondaries were wound for 9V and I didnt really want the extra parts, complication and heat from a rectifier and regulator to bring em down to 6.3V I managed with advice from Joe and Ed Dinning to rewind one of the secondaries to get me 6.5V AC. Pleanty close enough bearing in mind our mains usually runs at over 250V right on the allowable limit.
Anyway, silicon rectified the 325V secondary gives me 431VDC under load, which after reservoir capacitor aand chokes on each side gets me 410V on the output Transformer CT, again enough for the leak circuit as although the TL25+ ran at 450VHT, the Stereo50 variant ran at 385V on the anodes so I'm nicely in the ball park there, and what's more my 500V capacitpors are running inside their limits.
I settled for 100uF reservoir which is split into two HT rails via a 5H 150mA choke on each side into 50uF each and fed to the output transformer. the subsequent supplies to the phase splitters and preamp valve are dealt with by 22k and 100k resistors and decoupled by 50uF. I guess its a sort of Pseudo "Double mono" design with the split rails.
Anyway the output stage is more 5_20 as I couldnt easily find 440R wirewound resistors and rather than bung a pair of 220R in series I settled for 10W 470r as they would still be inside the 10% tolerance of resistors in the day.
The rest is pretty much by HJL except the input EF86 is configured as a triode and the feedback was tweaked to take that into account, my back of a fag packet calculation/flog took me to the use of a 2.7k resistor with 800pF compensation cap in parallel, thats the loss of gain with triode and also no 16ohm tap so forced to use the 8ohm output. I think i could increase that and reduce the feedback a bit with a 3.3k or even 3.9k but it seems stable but as yet I havent had it on the scope to see if theres any HF issues.
I'm so far very pleased with it, had a couple of teething issues, the usual dry joint stuff and a couple of schoolboy errors like forgetting to wire up the cathodes of an EL34! and connecting one end of the coupling capacitor to the anode of an EF86!
It sounds like a nice hi fi amplifier, which is to say it hasn't got any particular valve "flavour" if such a thing exists.
Theres a couple of issues I'd like to iron out, not serious but there's a low level hum that I can't identify the cause of yet, baffling in as much as with a shorting plug in the input socket both channels are deathly silent, very impressive for a valve amp of this kind of power, but as soon as a source is connected I get a low level hum that I think is 50Hz, This ghappens with a "passive" volume control (50k log pot in a box) also with a good quality commercial solid state preamp (Rega Cursa3 set to unity gain) and with my own homebrew Cathode follower buffer stage with active load which also is quiet as a mouse with my other kit. It's not loud enough to spoil my enjoyment as its inaudible even 5 feet away and I guess I could just ignore it.
I paid attention to the heater wiring, all properly referenced to ground and the small valves are DC heated via an LM338 based regulator.
I am minded to junk the earth lift of 10ohms and take the signal star earth to the mains earth potential as i think its not the issue here and my feeling is if I dont need to lift the signal earth then its better that way. I did pay attention to the earthing runs to the star tag and my feeling is it paid off as the amp is hum free with a shorting plug in it which if there had been a loop would not be the case?
I didnt use screened wires from RCA sockets to the gridstopper of the EF86 as it was only a couple of inches if that. I have subsequently twisted the hot wire with the earth wire to see if that helps any. It doesnt. I may try that tweak as its an easy job to sort.
ok I'll attach some pics of my scruffy handiwork.
Andy.