advertisement


Mike's Stereo 20 refurbishment thread.

The Mullards=soft thing is a spectacular over-simplification IMO. They made five or six different designs over the years and the only ones I’d say fit that description are very late-production I63 types with nickel rods, and even then some can sound pretty decent. I could go into it in extraordinarily geeky length if anyone actually wants with model codes, date codes, factory codes etc, but I’d certainly argue they made some exceptionally good sounding ECC83s.

I have my amp fitted with the type it would have shipped from Leak with (Blackburn-made I61 types dating from the very early ‘60s) and it sounds exceptionally good, not “soft” at all, just natural, open and spacious with great bass weight and grip. I have a pretty good tube stash so have tried many types before concluding I like this original voicing best.

If on a budget the Sovtek 12AX7LPS long-plates (which Jez may have been recommending) are very good IMO and my favourite current production 83s. They have great bass grip and a nice even balance, but lack the walk-around soundstage depth and detail of the better Mullards etc. Avoid short-plate Sovteks, they just aren’t remotely in the same ballpark as the LPS.

PS The Leak seems very kind to ECC83s so I’d expect a genuinely good set to last decades. I’m pretty sure two of the three I have plugged in at present (the phase splitters) are actually original to my amp. They came with it when I bought it, are the same date as the amp, and still measure good with perfectly matched triodes.
 
The Sovtek's I'm recommending are the ones Tony says are crap.... they are vastly better than Mullards I63 ones which have the dynamic snap and crispness of a 3 day old British Rail sandwich!
 
I bet the ones you have are Mullards. I bet they are fine too. I get on well with them short or long plate...
I'm OK with Sovtek 12AX7LPS, but prefer the old (no longer available new) 12AX7WXT.
Current in my amp are old production Ei ECC83s (from a fellow fishy) and they are excellent with a nice top end that maybe the Mullards miss (?).
Further down my list are new Electro Harmonix and JJs.
In the end though, only you can tell what is 'best' for you or what you can afford. Any way after 15 or 20 minutes you will not notice a big difference.
 
snowman_al said:
I'm OK with Sovtek 12AX7LPS, but prefer the old (no longer available new) 12AX7WXT.
Click to expand...

That’s interesting as I guess they are what Jez is on about. I have a pair somewhere unless, as I suspect I might have, I gave them away. I thought they were crap; bright, thin, flat and a bit grainy. Just lacking in any refinement to my ears. If I’ve still got them I’ll dig them out for another go. Shows personal taste and system synergy is everything, but by saying that I find I prefer the same valves with both the tiny JR149s and huge Lockwood studio monitors.

PS The EI smooth-plates can be very good, but they seem to go bad (microphonic or noisy) very fast compared to decent UK or USA valves. The silver-plate variant sounds way better than the grey plate to my ears (greys are far too bright for me). I’d not pay good money for them though as I’ve just not found them reliable.
 
snowman_al said:
I bet the ones you have are Mullards. I bet they are fine too. I get on well with them short or long plate...
I'm OK with Sovtek 12AX7LPS, but prefer the old (no longer available new) 12AX7WXT.
Current in my amp are old production Ei ECC83s (from a fellow fishy) and they are excellent with a nice top end that maybe the Mullards miss (?).
Further down my list are new Electro Harmonix and JJs.
In the end though, only you can tell what is 'best' for you or what you can afford. Any way after 15 or 20 minutes you will not notice a big difference.
Click to expand...

Electro Harmonix is Sovtek. The ones I'm using are actually marked as Electro Harmonix 12AX7EH.
 
One for @Tony L I think.... I've just got my St20 on the bench and it seems I have very early 3921 OPT's! The mains TX is 8162 and I'm not sure if this is correct for those OPT's? I can ahem, make it so though if not. The board on mine is from yet another amp and it seems I've used the 1M grid resistors that "should" go with those transformers, which I'll probably change for 470K.
 
Arkless Electronics said:
One for @Tony L I think.... I've just got my St20 on the bench and it seems I have very early 3921 OPT's! The mains TX is 8162 and I'm not sure if this is correct for those OPT's? I can ahem, make it so though if not. The board on mine is from yet another amp and it seems I've used the 1M grid resistors that "should" go with those transformers, which I'll probably change for 470K.
Click to expand...

Confusingly the mains transformer number never changed even though the voltage did! Try measuring the input to the GZ34, if its around 300V its the early, if 280V the later one. As I understand it Leak rebuilt all returned amps with the later resistor values regardless of the transformers, but check this. The change of OP transformer as well as the resistor values came as they were going through EL84s too fast as Leak were running right on the edge of the Mullard spec. I know some of the Leak cognoscenti feel the earlier transformers sound better, but I’ve never heard that. I do have a lone unrestored TL12+ with the original transformers though, so if I ever find it a mate I’ll get to hear to some degree.

Has your amp got a little round sticker inside with a serial number and date code? It would be interesting to try and establish when the output transformers etc changed. My bronze one is ‘61 and seems to be the late spec all round.

PS Given your stated GZ34 frying experiences I’m betting on 300V!
 
Tony L said:
Confusingly the mains transformer number never changed even though the voltage did! Try measuring the input to the GZ34, if its around 300V its the early, if 280V the later one. As I understand it Leak rebuilt all returned amps with the later resistor values regardless of the transformers, but check this. The change of OP transformer as well as the resistor values came as they were going through EL84s too fast as Leak were running right on the edge of the Mullard spec. I know some of the Leak cognoscenti feel the earlier transformers sound better, but I’ve never heard that. I do have a lone unrestored TL12+ with the original transformers though, so if I ever find it a mate I’ll get to hear to some degree.

Has your amp got a little round sticker inside with a serial number and date code? It would be interesting to try and establish when the output transformers etc changed. My bronze one is ‘61 and seems to be the late spec all round.

PS Given your stated GZ34 frying experiences I’m betting on 300V!
Click to expand...

No change in transformer number eh? Interesting... Yeah I know all about the thermal runaway problems with the EL84's. 470K or 1M will make no difference to the sound and I'm surprised I fitted the 1M's... it was years ago and I probably hadn't heard about the issue at the time I rebuilt it and so just went with the 1M's as that was what had originally been fitted. If mine is not a 300V TX I'll reduce dropper resistor to get it up there.. I have a choke fitted to mine so the resistor is not really needed for smoothing...

Mine is a bitza I made with OPT's, mains TX, board and chassis from different units so any serial numbers etc are useless!
 
snowman_al said:
I bet the ones you have are Mullards. I bet they are fine too. I get on well with them short or long plate...
I'm OK with Sovtek 12AX7LPS, but prefer the old (no longer available new) 12AX7WXT.
Current in my amp are old production Ei ECC83s (from a fellow fishy) and they are excellent with a nice top end that maybe the Mullards miss (?).
Further down my list are new Electro Harmonix and JJs.
In the end though, only you can tell what is 'best' for you or what you can afford. Any way after 15 or 20 minutes you will not notice a big difference.
Click to expand...

Yes my amp came with a full set of Mullards and my best guess is that they could be the original valves judging by the low hours the rest of the system seems to have had. I don't have a valve tester but they certainly still sound great.

I want to put together a spare set of valves and logically it makes sense to buy them now if I want any NOS valves as they'll only get rarer and consequently more expensive as time goes on.
 
Tony L said:
Confusingly the mains transformer number never changed even though the voltage did! Try measuring the input to the GZ34, if its around 300V its the early, if 280V the later one. As I understand it Leak rebuilt all returned amps with the later resistor values regardless of the transformers, but check this. The change of OP transformer as well as the resistor values came as they were going through EL84s too fast as Leak were running right on the edge of the Mullard spec. I know some of the Leak cognoscenti feel the earlier transformers sound better, but I’ve never heard that. I do have a lone unrestored TL12+ with the original transformers though, so if I ever find it a mate I’ll get to hear to some degree.

Has your amp got a little round sticker inside with a serial number and date code? It would be interesting to try and establish when the output transformers etc changed. My bronze one is ‘61 and seems to be the late spec all round.

PS Given your stated GZ34 frying experiences I’m betting on 300V!
Click to expand...

Just out of curiosity I measured the voltage at the input of the GZ34 on mine last night and it was around 284v with the amp plugged directly into the wall socket and 292v if powered from my big Airlink balanced mains transformer.

My mains voltage is normally high and was 246v at the time I did the measurements.

That's with 470K fitted.
 
I don’t understand any of this I have to admit, nor do I know what an Airlink transformer is. I am similarly in a high voltage area (which my TD-124 hates!) and my Leak seems to start at around 295-300V but settles back to 284 or so once the amp wakes up. As I understand it the GZ34 has a deliberate designed-in ‘soft start’ aspect, so I guess this is what I’m seeing as it wakes up and loads the transformer.
 
The GZ34 is an indirectly heated rectifier so it warms up at about the same rate as the rest of the valves, it should settle within a minute or two. After the valves are properly warmed up the voltage will drop slightly due to the transformer warming up and the resistance of the copper windings increasing.
 
I’m assuming the transformer values specified on the two different Leak Schematics are the warmed-up ‘running’ values, is that correct?

PS I took my measurements pretty fast as I didn’t like the idea of running the amp on its side for long in case the EL84 heat damaged my cheap Ikea shelving unit! I’d done everything within say three minutes, so that 284V I mentioned may well drop a little lower once the transformer gets properly warm.
 
That`s a reasonable assumption but It`s difficult to be sure, most readings on valve equipment are fairly "nominal"

For instance if you measure with a slightly tired but in spec GZ34 and brand new EL84s you will get a lower voltage reading than if you use a brand new rectifier with well used but in spec EL84s.

Worrying about exact readings is a waste of time.
 
OK well I replaced all the remaining crap composition resistors which I'd left in in the grid stopper positions (22K x 2, 1K x 2 and 10K x 4) as they were "unlikely" to do any harm there, with metal films and a pair of precision 1K's just for the hell of it (they are not critical even) as they were to hand (measured 1.000K and 0.998K!), reversed all the stuff I'd done to make it a specialised 4R amplifier and basically put it all back to standard other than the choke and 220uF cap in the power supply...
I can now set it to 8R to suit my Spendors... and? Well it sounds bloody marvellous! A different amplifier.... so different I may have to re-experiment with the ECC83's...
I'm a happy bunny:) I think it will be getting much more use now!
 
Arkless Electronics said:
OK well I replaced all the remaining crap composition resistors which I'd left in in the grid stopper positions (22K x 2, 1K x 2 and 10K x 4) as they were "unlikely" to do any harm there, with metal films and a pair of precision 1K's just for the hell of it (they are not critical even) as they were to hand (measured 1.000K and 0.998K!), reversed all the stuff I'd done to make it a specialised 4R amplifier and basically put it all back to standard other than the choke and 220uF cap in the power supply...
I can now set it to 8R to suit my Spendors... and? Well it sounds bloody marvellous! A different amplifier.... so different I may have to re-experiment with the ECC83's...
I'm a happy bunny:) I think it will be getting much more use now!
Click to expand...
Nice one Jez :)
 
I've now replaced all the coupling caps etc with better ones, have 2 x 220uF, 10uF low ESR and 0.47uF film cap in parallel as main reservoir cap, 150uF as cathode decoupling (takes a 12Hz pole here down to 4Hz and reduces LF phase shifts. I may go bigger here.... had 1000uF here on one St20 I had in the past!). It's sounding very nice indeed!
 
You must log in or register to reply here.


advertisement


Back
Top