advertisement


Leak TL/12 Plus / Point One Plus

I got a champagne one when NVCF was still at the NEC:

leak1.jpg

leak3.jpg



The big cap was shot so I replaced it along with a few others and the 100R, and used as a home cinema centre channel amp for a while.

Eventually I fancied stereo and got a bronze one which was restored. The colours upset my OCD, so I swapped the champagne one with a bronze one with a fella, who was most put out with the manner in which I had fixed it (I only wanted to make it work). He got the last laugh, as my champagne one had Mullard valves and the bronze one he gave me had no-name tat. It also didn't work - dodgy caps probably, so I went from two working amps of different colours to two similar coloured amps with one duff channel. Oh well, but only 260 quid total spend twenty something years ago. I now have 4 Golden Dragons for the EL34s.

I hope to get around to fixing the dodgy one eventually.
 
The metal-base Mullard GZ34 is worth a bloody fortune, annoying Hi-Fi Hanger, where I bought the second TL12+, know this and clearly swapped it out to sell separately (£345!) swapping in an obviously non-stock Tungsram. Thankfully I have a close pair of late-60s NOS (when I bought them) Mullards, so I’ll not need to go pair hunting!

My task for the next couple of days is to produce a bill of materials and then start deciding on brands/sources. My intention is a totally OCD-grade neat restoration, e.g. I’ll figure out how to gut and re-stuff the big top mounted can caps. My aim is to add, not subtract value in case I want to move them on later, so I’ll do it all as neatly as it can possibly be done using the most sympathetic components. I do plan to rebuild using the later values so as not to run the EL84s so hard (470k rather than 1m on R16 & R17), but I need to better understand how the different transformers impact things. These early version champagne TL12+ have the 300V mains and 1st version output transformers. The annoyance is they aren’t as well documented online as the Stereo 20, so I may need a little help just getting to the bill of materials stage. I have two circuit diagrams, one official Leak that applies to the version of amps I have (that drive the 84s real hard) and a later ‘reverse engineered’ one from 44bx.com that has more moderate values for some components, e.g. the 470k R16 & R17, plus C13 & C14 (the big top-mounted can cap) is 32/32uF here (which is possible to buy) whereas the original is 100/60uF (which isn’t!). My gut feeling is to go for the 32/32, as one of those works perfectly in the Stereo 20, and that’s bigger! I also don’t understand the original 100/60uF spec as to my understanding that would over-stress the GZ34 as it is over its spec.

Anyway, I’ll come back to this with a spreadsheet sometime later as there is a lot I don’t understand so I’ll need some folk more knowledgeable to do a bit of double-checking for me! Basically once I can reduce it to a basic kit I know I can rebuild it, but I need to get a really tightly defined BoM to be able to achieve that.

I also need to figure out exactly how to gut and re-pot a capacitor as chances are I’ll be fitting a F&T 32/32uF inside the big can. Obviously I need to work out how to do this without scratching or staining the casework in the process, but it is certainly possible as others have done it many times. Once empty fitting the smaller new one inside is pretty simple and could probably be done with a card spacer and some bathroom sealant. I’d like to make it reversible so I won’t be using epoxy or anything. It just needs to hold in there well enough, maybe an ‘O’ ring or two would work? I’m far more concerned about getting the old crap out as I don’t know how to do that or what exactly to expect inside, how to clean it up etc.
 
This site contains affiliate links for which pink fish media may be compensated.
Ok, here is what I have so far:

51057673383_fc417fa4fe_b.jpg


Sources: original Leak documentation, and 44bx.com. I have attempted to document the differences, though there are also different transformer versions (it appears the 44bx schematic is for the 280V mains and later output transformer). I don’t think this ‘matters’ as such as, AIUI, Leak sold/rebuilt Stereo 20s with the earlier transformers with the later component values. My BoM is incomplete as I don’t know the correct voltage/watt ratings for each component at this point.

PS C7 isn’t a typo by me, there is an extra zero in play according to the schematics!
 
Highly amused to find the images of the pair of TL12+ amps I worked on 8ish years ago still the first result on google for "Leak TL12+" :D
 
I read that over at WigWam. A nice neat job. Any tips as to excavating the big can cap? What value did you use there, I suspect I’ll only be able to find 32/32uF? Also why did you change C5 & C6 from 16/8uF to 22/22uf? I can’t see enough detail to see what else is different, though I assume you went with 470k for R16/17? Annoyingly my colour ability isn’t up to accurately reading resistor bands so I need to be told some things!

PS Google knows you, so your search results will be different to mine, which put this thread first!
 
I read that over at WigWam. A nice neat job. Any tips as to excavating the big can cap? What value did you use there, I suspect I’ll only be able to find 32/32uF? Also why did you change C5 & C6 from 16/8uF to 22/22uf? I can’t see enough detail to see what else is different, though I assume you went with 470k for R16/17? Annoyingly my colour ability isn’t up to accurately reading resistor bands so I need to be told some things!

PS Google knows you, so your search results will be different to mine, which put this thread first!

For the big cap I just hacksawed it off at the ridge near the end, I think it was put in a vice with rubber grips so it didn't mark the casing. I have a feeling the guts were softened with a heat gun and yanked out with long nose pliers. This was done outside as the fumes are quite epic. It's an oddly pleasant smell but probably not ideal to inhale!

I can't remember the exact value I used - I have a feeling it was to the upper end of the GZ34's handling (probably a bit of an oversight but hey it was 8 years ago). C5 and C6 were 22uF just due to convenience - I had lots of them and the value in this position isn't massively critical - going too small may lead to LF stability problems but the stock values should be large enough to avoid that so going larger still moves the already non-problem pole even further out (or should that be in?)
 
Oh yeah, forgot to answer your other Q: Looking at the pics, originally one amp had 1Meg grid resistors for the output stage, the other had 470Ks. I of course went with 470Ks on both on the rebuild due to the well known issue.
 
Any tips as to excavating the big can cap?

Long ago observation suggests that if you replace the 5 volt rectifier valve with silicon diodes and point the cap base outwards a few hours later it will discharge it`s entire contents up the wall. (old Bush TV)
 
To add yet another post! I would use a 32/32uF for the main cap. An entire STA15 gets away with a 40/40 (though with a choke which does make a big difference). In a pinch an STA15 actually works fine with as little as 32/32.
 
Oh yeah, forgot to answer your other Q: Looking at the pics, originally one amp had 1Meg grid resistors for the output stage, the other had 470Ks. I of course went with 470Ks on both on the rebuild due to the well known issue.

Thanks. I’m pretty sure I’ll go with 32/32uF for the smoothing cap as it’s a safe value for the GZ34, works fine in the S20 and I’ll never be hooking any preamps, tuners etc to the supply. I’ll definitely go with 470k too, I want to end up with a stable reliable amp!

I suspect I’ll struggle to get some values, e.g. I don’t think 50uF exists these days, but 47uF should be in spec. I was using someone else’s brain when rebuilding my Stereo 20, they chose the parts for me, but I suspect I can cross-reference that to some degree, e.g. I know the Russian K40-Y9 0.22uF PIOs work great and I have some so I’ll not go hunting 0.25uF. I’m still very much a beginner at electronics so I’m really trying to reduce it to ‘kit form’ at this stage. Once I know exactly what to put where I can do it and do it neatly, but I don’t have the knowledge to understand what values are absolutely critical. As such this BoM is essential to get right!

I’ll double-check both BoMs with what is actually in there too as far as the capacitors are concerned as some of the discrepancies between the Leak and 44BX schematics are confusing, e.g. C7, C9 & C15. If both the amps have the same value there I’ll likely go with that rather than the BoM unless anyone tells me not to! I think the one that is coming is totally stock and my existing one has had a minor fix at some point (C11).
 
47uF in place of 50uF and 0.22 in place of 0.25 is absolutely fine. In fact in many ways more sound and sensible than hunting down snowflake 50uF / 0.25uF caps.

With capacitors in amplifiers, problems arise roughly when you get to the point of doubling or halving values - this can change the open loop response of the amplifier such that it is no longer stable with feedback, or simply degraded performance in various aspects. This is less apparent in power supply or smoothing applications but it still makes a difference.
 
The method I use to empty the cap is to cut the base off as far up the can as far away from the crimp as possible so you are left with a good section of the reduced diameter part on the base. Once you have the two halves you can de-burr the inside of the can so the base fits snugly back inside. I normally use hot melt glue to hold them together. But you could use the clamp and have the cap just a tiny bit higher on the chassis.
I cannot remember what is inside, wax I think, but drop it into boiling water for a few minutes before trying to extract it usually works for wax and pitch.







This example is from a Heathkit MA-12, but the principal is the same.

This cap was 60uF and 250uF so I re-stuffed it with a 47uF and 220uF replacements 'stacked' on top of each other.







You have plenty of room for a 47uF and 100uF 450 volt caps in the original.

But before you do empty the cans, check the originals are not still serviceable. Of all the Leak amps these do appear to have a long life.
 
If you want to save your original Mullard GZ34s, TL12+s work fine with the Russian 5U4M/5C4M in their place. (Not in your Stereo 20 though.)
 
That is remarkably neat work. I’d not contemplated going that far, my plan was just to hollow it out and stick a new F&T 32/32uF in there, but that is food for thought.

Good to know the TL12 Plus is more gentle on GZ34s. Thankfully I’ve no shortage of Mullards here, I’ve actually got 4 of them. Two are a near NOS late ‘60s pair, I think a year apart, so I’ll use those once I’m sure I’ve built them right. I’ll use the supplied Tungsram and Russian 6p14p-EV for testing. I’ve got loads of ECC81s and I think some decent EF86s (I’m hoping there is another good pair in the preamp), so thankfully there is no valve expense here.

It should be an interesting project. I don’t know where it will end up, but it will make sense of the existing TL12+ which was just an ornament up to this point. I guess I’ll end up having a shoot-out between them and the S20 and keeping the ‘loser’ as a spare/investment. I definitely want to get them working well whatever the end-point.
 

I watched this last night. It is an interesting, neat and very modern rebuild of a initially cosmetically very poor pair. Obviously not the way I’d approach it as I’m coming at it with a close to mint pair in the most desirable finish, so I plan to retain as much authenticity as I possibly can. I’ll not even be changing the crappy speaker connectors, I’ll just make up some new speaker cables with little spades and use the octal plugs for input etc. If I ever sell them later on I want top dollar, and that’s the way to get it!

The thing I found interesting was his use of two solid state diodes to feed the GZ34. Can someone explain the logic here and why it would make the amp safer/more stable as claimed? Does it protect the mains transformer in a fail condition? If so I’d maybe consider it as it is neat and so easily reversible, yet should retain the sonic traits of valve rectification (I know haters gonna hate). I’ve been real lucky with the S20 so far, it seems very stable and well behaved since my rebuild, but I understand when these things go wrong they can go real wrong real fast, so a bit more protection might be welcome. The fuse is tempting too.
 
... The thing I found interesting was his use of two solid state diodes to feed the GZ34. Can someone explain the logic here and why it would make the amp safer/more stable as claimed? Does it protect the mains transformer in a fail condition? If so I’d maybe consider it as it is neat and so easily reversible, yet should retain the sonic traits of valve rectification (I know haters gonna hate). I’ve been real lucky with the S20 so far, it seems very stable and well behaved since my rebuild, but I understand when these things go wrong they can go real wrong real fast, so a bit more protection might be welcome. The fuse is tempting too.

The diode mod started a while ago on the Dynaco forum. (https://dynacotubeaudio.forumotion.com/t1006-tube-rectifier-yellow-sheet-diode-mod) It reduces the chances of rectifier flash over when using new build GZ34s (not the same quality as the Mullards...). It also stops AC getting to the caps if a flash over does occur. I fit them as a matter of course, but use UF4007s now.

Yes the HT fuses are a good idea.

Edit, just found it on the video and it could be better implemented, better to use the Dynaco method and tag both ends of the diodes...
 
That looks good. I think I’ll implement it, and do the S20 too. I assume this uses tags of internally unconnected pins of the GZ34 just as a mount-point? I’ll add some UF4007s to the shopping list.
 
Yes move the AC HT wires off pin 4 to pin 3 and off pin 6 to pin 5. There are no (should not be) pins 3 and 5 on the valve itself. Then fit the UF/1N4007 from pin 3 to pin 4 and pin 5 to pin 6. The anodes, the silver bands, go to the pins 4 and 6. Same for all GZ34 amps. Stereo 20 example:

 
Excellent, it is so cheap, easy to do and easily reversible is seems crazy not to. Any slight voltage drop is likely a benefit to as I’m in a very high voltage area (I seem to get about 245V!). I’ve ordered some and I’ll do the S20 first.
 
Yes move the AC HT wires off pin 4 to pin 3 and off pin 6 to pin 5. There are no (should not be) pins 3 and 5 on the valve itself. Then fit the UF/1N4007 from pin 3 to pin 4 and pin 5 to pin 6. The anodes, the silver bands, go to the pins 4 and 6. Same for all GZ34 amps. Stereo 20 example:


On the Dynaco GZ34 diode mod, I seem to recall the diodes were between pins 4 & 5, and 6 &7, with each HT wire moved to pin 5 and pin 7. Guess it makes no operational difference so long as the unused tags are just that. Want to make sure I get it right on my Stereo 20 too.
 


advertisement


Back
Top