YAMA (Yet Another Mystery Amplifier)

[Richardgr]

This was bought because I think it is very elegant, and I took a punt that those OPTs, that are 10cm x 12cm, will be something special.

It is PP 6550 with150C2 regulated screens (1 per valve) and 6 5814As for input and phase splitting duty.

It was stated as being built by Jan Lodström sometime in the 60s or 70s. Having read a bit more, and from examination of the amplifier, I would put it at early 90's, and supposedly there is a magazine review out there somewhere. He was the designer of the Remus preamplifier, a well known design here in Sweden.

At 20kg it is quite the boat anchor.

Unfortunately it has an issue that was described in the advert when I bought it, so I have also taken a gamble that the issue will be localised and can be addressed.

The person who was selling it was disposing of somebody's estate, so there is not a lot more information than above. He said that there was some smoke when it was switched on and the transformer got hot. (Sounds like that troublesome Quad II that is getting an autopsy on another thread here).

Looking around, I can see some warning signs. The leads from the power transformer (also a massive 12cm x 10cm) for the heaters for one channel have got very hot (the red and black wires in the penultimate picture), and it looks like they have fused over time. I am presuming that this shorted one of the heater windings, and the transformer overheated before it was switched off. The last picture attached shows some scorching on the power transformer. I am assuming that the heater winding is one of the top two windings, so considering the options from here.

The circuit is a bit of a mystery, and there would seem to be a paucity of components under there. I need to study the pictures a bit more. I do not want to do too much to it for now.

What would be the correct way forward with that power transformer?

[Diabolical Artificer]

What would be the correct way forward with that power transformer?

power it up with a variac & LL with everything disconnected. The tfmr looks easy enough to pull apart in needed.

Intriguing amp, is that a choke or low voltage tfmr? Don't like the construction underneath, that floating bus bar! Looks rushed. 6 5814's! Not cheap & hopefully the OPT's are decent, it has potential.

Andy.

[Richardgr]

... and not just any old 5814's, Sylvania gold 5814A's with a life of 10,000hrs.

From first inspection it looks like there are 3 heater circuits. The 5814A's have a rectified and smoothed supply from that extra transformer at the back. Then the power transformer has two windings, and unless I am mistaken one of them appears to be 12.6 volts so the output valves on one side have serial heaters, and the other side has conventional parallel heating from the other 6.3V winding.

I am thinking that the transformer did not have the poke for the heaters for 4 6550's (6.4A) so this was a bit of a compromise, so it was always being stressed somewhat with the tails from the secondary not being specced for the 1.6A at 12.6V.

I will have another pry this evening.

I agree with you it does not look like a refined build, but then I did some reading about its creator, and he was meant to be a bit chaotic and bohemian, so maybe a reflection of that.

I am unclear how the small jacks and switches are meant to be used for checking bias, but saying that I see now that they are not jacks at all but tiny little potentiometers. One toggles the switch and adjusts the bias with the meter.

[m0cemdave]

For something with such a nice appearance, the build quality looks disappointing. One would have expected professional quality soldering, and holes to be deburred properly. It looks like something home-made, on good quality bought-in metalwork.

With careful restoration, you should end up with a really nice amplifier.

I've seen this with Heathkit products (usually sold as kits) - they can be anything between factory built, to basket cases needing stripping back to the kit and rebuilding from scratch with a lot of new components.

[Richardgr]

I am thinking it might have been a prototype, so nicely finished for a show, but a bit rough and ready under the covers. That would explain the power transformer if indeed one pair of output valves are serially heated.

[Diabolical Artificer]

What are the laminations on the burnt tfmr - E114/120? If you measure the tfmr l w & d you can work out the cores power capability.

I'd love to see it's schematic, it's could be audio exotic & might set the cat amongst the boffins : ) Are there any transistors?

Andy.

[Techman]

Originally Posted by Richardgr

so it was always being stressed somewhat with the tails from the secondary not being specced for the 1.6A at 12.6V.

Maybe, but 1.6A isn't that high in the grand scheme of things regarding heater current and they look like they should have been sufficient. If they've overheated like they seem to have done, then I suspect a short circuit somewhere on a valve base, possibly due to rough soldering and the 'pull' of the heavy gauge single core wire pulling on the tags, but that's only a guess and careful examination and testing should soon isolate the problem. Let's just hope the transformer has survived.

[PJL]

First impression is that mains transformer looks far too small for a stereo PP 6550 setup. I would recommend working out the circuit then you can estimate the power requirements for a replacement transformer but don't expect it to fit in that case.

[Richardgr]

Just checked the dimensions of a suitable Hammond replacement, e.g. 370LX which is rated at 550VCT@460mA, and also has the right heater winding, 6.3V@9A, and yes, you are correct, would be the next size up from 10cmx12cm.

There is no tube rectifier, the 6550s could run a bit colder and I could fit a toroid underneath for the heaters, so it could be marginally OK in that case; I would need to see if it gets hot in operation and what percentage of voltage drop there is when loaded (10% would seem to be the goal).

I suppose Plan B could be to use less powerful output tubes, 6L6 for example.

It is quite enigmatic. The construction is not great, but the circuit looks to have the touches of an expert (e.g. a single capacitor on one side of the OPT which was presumably added to suppress some ringing). It has that 'impressionist artist meets tube guru' vibe underneath.

Two of the power supply caps are not connected to anything, and the dead-bug style wiring had two junctions 1mm apart, no chassis earth (but it is not that old), the heater wiring issue (that makes me think it has never really had any service life) ... all in all a mixd bag.

What are your views on using 150C2 on each screen grid, to the UL connection on the OPT? The max signal g2 current can be over 40mA according to the valve data, and a 150C2 is specced at 5-30mA. They are wired so they see the 5mA so they strike OK, but 25mA headroom feels like there is a risk using them with this particular valve.

[Richardgr]

Here is a first stab at the schematic. It has me scratching my head somewhat. Those 1/4W 5band resistors are impossible to read, and a pet gripe is when components are fitted with the values hidden, when with a bit of planning they can be read in situ.

That is a long tail pair as the input, with some sort of feedback I am unfamiliar with going on anode to anode. Then after that I am a bit nonplussed. It looks like an SRPP circuit. The 150C2/0A2 is represented by the zener symbol, but it is a valid model. I will sort out the OPT with a saved model, then hoprfuly I can simulate it.

I haven't looked at the power supply yet. It looks like it must be a centre tapped power transformer, and 2 banks of electrolytics for B1+ and B2+.

There is a sooty patch by the base of V7, and the shadow of something in the soot. It looks like there was an issue before, confidence is sinking.

[Radio Wrangler]

Maybe someone decided to make themselves a valve amplifier. They chose push pull and a type of output valve they had handy. Found some output transformers, made a chassis. Drilled for the transformers they had and then found the output transformers they got had ratios and ratings better suited to something bigger in output bottles, but fortunately the output device sockets were octals, so output valves were bought. The drive circuitry still hadn't been decided on, but no matter, drill holes for a number of B9G bottles and different drives can be tried out to see which seemed best. So the drive circuitry left when playing stopped was just a bodge up. Shame it would have been better with a big enough mains transformer....

I think you could have someone's playground - an incomplete project/experiment. Development stopped when they liked the sound and they just started using it. The undersized mains transformer burned out as they tended to do.

There are various explanations for how this could be arrived at, but being based on available parts is a common starting point.

David

[Richardgr]

I tend to agree with your summation. However, it does not chime with the effort that went into the chassis. From the exterior it looks like there was a lot of attention to detail. I am still thinking it might have been used as a prototype, something for the shop window to garner interest. I will have to get the soldering iron out soon to remove the wiring to the fused heater winding. I will see if I can get a date code from something at the same time. It looks to be a lot later then the 'provenance' from the advert.

I will also test the 6550A's today, see how much life they have. There are two pairs: GE with a slightly bigger plate structure, and another two with similar etched marks but I can't see the manufacturer. The strange thing about the GE tubes is that they have the message on them "Warning: P connected to base disconnect from power before removing". Has anyone seen that before? What is the full text? I would guess it is a sign they were produced in a more safety conscious time?

Any views on the topology? Did it sound so good that the master had to lay down his soldering iron and savour the smell of smouldering insulation?

[Diabolical Artificer]

I'm not an expert and don't have as much knowledge as some members but do have quite a bit of experience building valve amps. Looking at the schematic it has as mentioned a differential/LTP IP followed by either White cathode followers/SRPP/ one of those hard to spot two valve combo's. The phrase "All mouth and no trousers comes to mind".

I think it's too busy and you could do exactly the same with 3 (or 4) double triodes for both channels, ie drive the OP valves cleanly. Don't like the non adjustable bias and the 330k grid leak resistors, and screen grid arrangement on the OP & not enamoured with all the precision resistors.

You could always test it and if you don't like it redo it. Any idea what the HT is? Andy.

[Richardgr]

Thanks for saying

one of those hard to spot two valve combo's

I thought I was the only one!

The OP valves were quite well matched, tested good not great (60mA/gm 8).

Shall measure some voltages once I have to the simulation working.

[Radio Wrangler]

If I set out to make something, even something involving some playing around and experimenting, I usually make a good job of chassis/cabinetry first, knowing that once I've got the thing working properly, I'm unlikely to go back and do the mechanicals properly. As I worked in a place where aluminium offcuts were there to use and had access to things like a Rotax turret punch, guillotine and folder, I could do some nice chassis work as easily as doing a rough job. I don't have such access to such facilities any longer, but I do have a TIG machine and I construct chassis and framing by cutting flat plates and welding into 3-dimensionality. People work around what they have.

You could turn that amplifier into something quite decent, but you'd need to replace the mains transformer with something meatier.

David

[Richardgr]

The problem is all of the holes if I elect for something simpler.

I am thinking building around EABC80's (inc 1/2 ECC83) so solve the hole issue that way.

[Diabolical Artificer]

The problem is all of the holes if I elect for something simpler

That's a tricky one, I've had that, see pic. One thing you can do is to parallel double triodes say in a LTP for the back two sockets. For the three front holes, make a plate up that will cover the L&R holes with one central hole for a B9 base.

The triodes you have would make excellent drivers if paralled, you don't need both triodes to match just use two Rk's or Ra's so any disparity in emission etc is evened out. You can then sell the spare valves to pay for components etc.

Andy.

[Richardgr]

Unloaded voltages 550V for the output valves, 450V for the LTP and SRPP. So I suppose I can knock 10% of those voltages and work with 500V and 400V for the stages.

The heater wiring that I assumed was 12.6V is 6.3V. Now I start to think that there is some connection between the sooty valve base and the mangled heater wiring. The wire gauge is 1mm, so must be 19 Swg', which ought to cope with 5A according to my googling. What could have caused an overload? Was this a 6L6 amp that was 'upgraded' to 6550? 6L6GC starts to look like a solution with the current power tranny. Max screen current is 22mA with 450V on the plate, 400V on the screen, so the figures start to add up a bit better.

Connectivity is OK for the OPT windings, secondaries low ohms, centre tap to UL 30R, CT to plate 90R.

There is a bit of scorching on the paper of the PT from the heater winding getting too hot. I would like to be mañana about it, do I need to inspect it? I am not trying to defer responsibility, but would appreciate advice from someone with first hand experience.

[Silicon]

That input stage does not look like a standard 'long tailed pair' (differential) phase splitter.

There is no shared cathode resistor.

It looks more like a Paraphase inverter.

[Richardgr]

Yes, I think you are right. Looks odd because the voltage divider is between the anodes.