Lost left channel after shunt regulator install

[blue360cuda]

Hey everyone,
I finished the stock StereomourII build last week and all resistance / voltage measurements checked fine. It sounds awesome - the bass did pick up after a few days of breaking in. 

So last night I put in the shunt regulator mod- all voltages were within spec but when I fired it up there's no sound out of the left channel. I swapped tubes, RCAs, sources - no change.  Rechecked resistance and voltage measurements:
Resistance's all good
Voltage also good but 2 were suspect -  T14 was 304 V and T15 is 21 V.

I checked the 41-45 but the manual didn't include reference ranges,  they are: 
41- O
42- 208
43- 0
44- 415
45- 208

I backtracked my work referencing the manual expecting to find some silly miss wire but everything looks good.  Solder joints are good. Not sure what I'm missing here . . .

Any other ideas to narrow down the problem further?

The only other change I made was originally I had a mundorf silver/oil 0.1uf installed and swapped that for the original kit 0.1uf cap to see if that would impact bass response.  Both caps measured good.

 

[Paul Birkeland]

I'd post some build photos.

21V at terminal 15 means the 2A3 on that side is drawing very, very little current.  Typically that would be a disturbance in the wiring leading from the power transformer through the resistors to pins 1 and 4 on that 2A3 socket.  You can use your meter to measure the AC voltage between pins 1 and 4 on that 2A3 socket (it should be close to 2.5V AC).

 

[blue360cuda]

Ok, think I found the culprit . . . disconnected the wires and removed the shunt and C4S board.  Found a loose solder joint on the 365 ohm resistor at C center to C3. Remembered when I did the bridge from C4/5 I accidentally did it C3 to C4 and had to un solder the connection. Must have forgot to reflow C3 in all the excitement.  Will rewire everything after the kids are asleep and hopefully that'll be it.

Do I need to worry about roasting any components or tubes?

 

[blue360cuda]

Well that did it!  She's singing again in stereo

 

[Paul Birkeland]

None of that will cause the 21V at terminal 15.  I suspect there's still a flaky joint around that 4 pin socket.

 

[blue360cuda]

This is what I got after reassembly.    D1 to D4 is 2.5ac    Maybe there was something else I unknowingly "fixed" during disassembly and reassembly?

1- 366 Vdc
2- 395
6- 62
7- 201
14- 161 --only one slightly out of spec (170-230) what do you make of this?
15- 61
19- 395
20- 367
21- 62
22- 62
24- 62
25- 62
27- 396
30- 405
31- 405
34- 396
36- 61
37- 61
39- 61
40- 61

 

[Paul Birkeland]

The 161 isn't concerning.  I'd say that there's probably another joint in there that may give you trouble down the road around that 2A3 socket.

 

[blue360cuda]

Thanks PB,  I'll take another look under the hood and reflow everything around that side for good measure.

BTW, I read on an old post the possibility of running Kaiju chokes / OT on the Stereomour II . . . is that a possible upgrade in the future?

 

[Paul Joppa]

...
BTW, I read on an old post the possibility of running Kaiju chokes / OT on the Stereomour II . . . is that a possible upgrade in the future?

The Stereomour II was designed to leave that possibility open, but so far the market is thought to be too small to justify the effort to make it an upgrade kit.

The design idea was implemented in the (now discontinued, for reasons unconnected to performance) MonAmour. The problem is that to work right, the plate current would have to increase, exceeding the rated limit for traditional 2A3s by 20%. It works fine with modern single-plate versions which have higher dissipation. The MonAmour had a switch to choose the current, making it easy to compare the effect, and the lower current was clearly inferior. MonAmour also used the 5670 instead of a 12AT7, with a different bias arrangement. Given these complicated caveats, it just doesn't seem practical.

 

[blue360cuda]

PJ - If someone wanted to go down the iron upgrade path in the future what would you recommend?  It seems like Isotango is one of the top options out there, would this be a good option without having to tweak too much of the original circuit? I see the OT2 is 4k and this is 2.5k
https://www.tubesusa.com/FC-30-2k5S.jpg

Another company I keep seeing pop up is Thermioniclab. They seem to be essentially made to order with lots of customization options. Has anyone had experience with these guys?

 

[Paul Birkeland]

The Tango stuff is generally a downgrade from our iron, especially compared to what's in the Kaiju, and you'd have a tough time getting them to fit.  I've used Thermonic Labs stuff a couple of times and so far so good, but he doesn't use keeper lams, so they can come apart a bit in shipping. 

Lundahl can make a 70H/50mA plate choke that you could use with the stock output transformers we provide (along with an ~8uF parallel feed cap), but again, getting this stuff to fit is precarious, and you're basically just buying deeper low frequency extension. 

Sowter does have an appropriate parallel feed transformer in their catalog, but recent orders after Brian's retirement don't leave me feeling super confident about recommending them right now.

 

[Paul Joppa]

PB beat me to the punch, so some of this is redundant  :^)

I would not recommend any transformer with a design impedance below 4K, for an unmodified Stereomour.

The linked transformer is far too large to fit on the chassis plate, and weighs ~9 pounds each, yet it only achieves 20 henries inductance, the same as the stock PC-3. The Kaiju's PC-5 provides 60 henries in MonAmour, or 40 henries in Kaiju.

Running a modern single-plate 2A3 at 18 watts plate dissipation, as in MonAmour, would require only a change in the cathode bias resistors.  You night be able to  change the four power supply resistors  and change the bias resistor to get good operation with the Kaiju iron at 15 watts dissipation.

 

[blue360cuda]

Thanks Paul and Paul !

So you're really not impressed with the Tango OTs and Chokes?  I have zero experience with them but "everyone" seems to think they're one of the best.  I guess it's hard to improve on something bespoke designed for your amps like the original irons. 

The Stereomour II just keeps getting better and better     I've been playing my streamer through it as background music every evening for the past week to hopefully break things in a bit.  Just got notified my Mour Quiet  for the S2 will be here tomorrow so that's exciting!  Also pulled the trigger on one of those goldpoint 47 step attenuators for the SEX.  What can I say, the tinkering bug has got me this winter. 

Side bar -  do you guys have any good resources / books you'd recommend to learn more of the circuit design theory and other amp-related electronical wizardry?  Being a bio/chem guy this stuff is not exactly in my wheelhouse but its very interesting and has been a fun hobby to get into!

Appreciate it!

 

[Paul Birkeland]

I have built amps with Tango transformers before, as sometimes people send me what they really want for a custom build and I just go with it.  My impression is that these designs haven't changed much in 35 years, and they aren't competitive with what's currently on the market. 

You could check out the old Valve magazines:
https://forum.bottlehead.com/index.php?topic=1642.0

A lot of people recommend the Morgan Jones books, and the Valve Wizard has some good content as well:
https://valvewizard.co.uk/

Morgan Jones

 

[blue360cuda]

PB-  At the risk of beating the horse to death,  what specifically do you not like about the Iso tangos?  In some aspects things have improved for sure with audio but it seems in other ways the old 50s-70s tech and manufacturing was superior. Take this as me asking you to school me, not start an argument. I'm honestly curious.   

Thanks for the links too!  I'll check them out

 

[Paul Birkeland]

what specifically do you not like about the Iso tangos? 

Let's look at the XE-20S, which is often pushed as a great transformer for use with 300Bs at the $1000 price point.
https://tango-trans.com/wp-content/uploads/catalog/XE-20S.pdf

In the 3.5K configuration, the 18H of primary inductance means the transformer holds up to about 35Hz, then kinda dies below that (both rolled-off response and rising distortion). In the 3.5K:8 ohm configuration, there's a ton of unused copper on the secondary winding which reduces the performance of the transformer (poor coupling/treble issues).  This issue is really bad in the 5K:8 configuration where very little of the secondary winding is used. 

Perhaps the most irksome thing about the XE-20S is the "20W" rating.  I've seen measurements of this transformer at 3.5W at 30Hz when used as a 5K unit, which gets you about 5W/30Hz when used as a 3.5K unit.

Now if we compare that to the Monolith S-9:
https://monolithmagnetics.com/sites/default/files/datasheets/Single-ended-output-transformers/datasheet%20S-9%20300B%20single%20ended%20output%20transformer%20prelim.pdf
There's about double the primary inductance, and this output transformer will stay well composed to 16Hz.  Notice that they give a power rating that is power with frequency, so the 10W at 17Hz is about 6 times the power handling of the XE-20S.  This transformer is also flat to about 100kHz being driven by a 300B.  All of this performance comes at a price tag that's around $700. 

The Lundahl LL2785C gives 30H of primary inductance with about the same DCR as the Monolith and the Tango with slightly poorer high frequency response and a 16W/30Hz power rating.  These are $600...

These comparisons become even more obvious when you compare the weight of these pieces of iron, as the Tango is 2kG, the Monolith is 5kG, and the Lundhahl is 2.5kG with no steel case.

In the Kaiju iron, there's even more primary inductance in the plate choke (40H), and the power handling of the output transformer is quite high (PJ might remember where that landed, I wouldn't be surprised to see 20W/30Hz), though we don't push as hard to extend the highs beyond what's considered audible.  The issues associated with unused copper on the secondary windings are mitigated by having two tapped windings that can be put in series or parallel to get different step-down ratios rather than having a single secondary winding.

The FC-40-3.5S is a little bit better, but now you're spending $1500 to get a transformer that's still not as good as the European units that are less than half the price. 

 

[blue360cuda]

Thanks PB for the explanation!  Appreciate it 

 

[Paul Joppa]

OT-5 is 10W at 25Hz, 20W at 35Hz, 40W at 50Hz. (In Kaiju it's 8W at 22Hz).

Incidentally, I looked up OT-2 as used it Stereomour and S.E.X. As used in Stereomour (4K primary impedance ) it's about half that, and in S.E.X. it's 1/4. All three amps have the same bass extension - full power down to 22Hz.

 

[blue360cuda]

JB and PJ -  thank you both for the info.  Any desire to play with OT and chokes is pretty much a thing of the past for me with the Stereomour now. I've finished the shunt regulator, DC filament and last weekend did the stepped attenuator mod.  Holy cow, loving it more every day!  Before this amp, the best sounding I've owned was an older model of the Decware EL84 SET that's had a few tweaks. Dang, this Stereomour blows it away in every way now- I'm seriously grinning ear to ear listening to it. It's always fun when your favorite records sound new all over again  

Follow up Q-  Since I did the DC filament mod, is there any way to remove the trim pots from the circuit?  It'd be nice to clean up the top and maybe use that real estate for something else like an impedance switch. I just scored some 16ohm 604Es and my other speakers are 8ohm so going between 8-16 would be nice.

 

[Paul Birkeland]

The hum pots probably aren't necessary for most people once the DC filament supply is installed, but you'd have to fit some new 22 ohm resistors to where they land now on the terminal strip over to where the white wire/cathode bias resistors land.