6550 SE Amplifier

[Paul Barker]

The power supply has more voltage available and may get used for other projects. Its a nice quiet variac, so its a handy bemch supply aswell. The rectifier heaters are not via the variac, they have full mains. Thats why I put a mains out takeoff in the variac.

[Paul Barker]

I’m very glad to report only the grid choke was faulty this time.

Also the valve which was subjected to +75 grid volts has recovered. The bias difference between channels isnt significant.

[Max N]

Great news!

[Cressy Snr]

Seconded

[Paul Barker]

The first picture is B+ not anode volts to arrive at anode volts take 5 off for transformer 50 off for bias, 472v







I just wanted tobring to you’re attention a naughty but nice experiment. I wanted to see whoat happens when I raise the b+ until bias is 50v inceasing the current and watts. Nominal cathode resister is 618v as amp is live I’m not measuring exact value right now.

So nominally we are pulling 50/618 @ 472 anode volts. = 81mA 38 watts . Max allowed combined anode g2 is 40 watts. Not so bad. In class B push pull you're permitted 600 anode volts. I realise thats because each phase in turn one anode is off. But in A1 Triode max is 450.

It sounds better enough to shoten life of bottles a little to enjoy the extra life of music. The anode is so not cherry red I cant get a picture of any orange, just shows dark. I can see orange with human eye in dark.

But it sounds like a dht, and surpasses 6EM7 triode but with power.

[Paul Barker]

I could put a base plate on as the top, inverted bottles can sit in an oil bath to exceed ratings. Just saying. My friend who used to maintain the 4 or so PYE line amplifiers in the cold war which sent the radio round scarborough. The amps were about the size of an estate car with parallel push pull V1505’s four bottles. Their anodes were glowing white hot.

At the radio listening station at Irton More , part of gchq, they used to test valves to destruction. They got way past the published data before destruction occured. The guys that maintained the equipment broke the official secrets act to tell me. No names provided. But some were members of Scarborough Amateur Radio club with me.

[Nick]

can sit in an oil bath to exceed ratings.

Exactly how much difference would that make? I would have expected that most of the heat from the anode was dissipated as IR (it being in a vacuum and all that) so how directly is the envelope temperature related to the max dissipation? I suppose it would avoid the glass getting soft, but not sure how it would help the internal structures.

From my time with forging presses, white hot is around 1300C, and glass (according to google) starts to soften at 1000C. Would need a tungsten anode and grid I would have thought to run like that, steel is starting to soften at the same sorts of temperature, nickel not much higher. But given that, I guess that keeping the glass cooler would be the first thing to do. Stopping the grids sagging would be a problem as well I would think, especially with a multi grid valve.

Didn't liquid cooled valves use ceramics (non glass) and metal construction?

[Paul Barker]

I’ll take you’re experiential knowledge.

Meanwhile the 11uF caps arrived, so a few changes a foot. But not today.

[Paul Barker]

OK mounted the new choke and caps. Souldered everything.

Now the two smaller caps are an old idea from a Triodeel email forum. The fraction of the mu of the valve of the value of the cathode bypass cap. From cathode to b+. Similar idea to WE ultrapath but injects b+ phase correct to where the signal is antiphase to cancel any power supply noise. Unlikely to make a fifference due to shared b+ in both valves which do that job also. But nothing ventured…..

Next power up voltmeter ready!

[Paul Barker]

Firstly, no more cockups, and no failed parts.

Erm…. Brought b+ down to 505. Taken Nick’s points on board.

Oh wow it is so good. Yes the cap to b+ contibutes greatly to the project. I highly recommend it. It is cleaning something up, whatever it is. I’ll bring it to the next meet definately!

I highly recommend this 8/9 watt design.

I’ll drawer curcuit agan and put actual voltages measured in.
Once again dont miss out the cap to b+ from cathode it is remarcably cleaner. Yep as T said its another coupling cap.

Picture voltages are b+, 6550 bias x2, e88cc bias x2 e88cc Anode voltage x2














print screen windows 7

[Cressy Snr]

Excellent stuff!

[Paul Barker]

Im moving the amp up to the Quad 57’s, which gave me a chance to measure ohms. It turns out the opt’s are 260 ohm, so to achieve the 450v Anode Voltage I require 520v B+. So I can go back to what I thought was a little over where it came to life. Its wasnt over, it was bang on.

I’ll drawer it all up but just for now

E88cc anode load 3 X 27k 20 watt in series measure 80k both sides.
E88cc cathode resistors unbypassed 2 X 220 ohm in series measure 464 ohm and 463 ohm 1/2 watt resistors carbon film.

6550 cathode resistors 680 ohm // 6 k 8 ohm measure 617 ohm Left 613 ohm Right (the valve that was subjected to + 75 grid volts)

So all in all, I think the RH valve has survived pretty good, and I wasn’t exceeding the Maximum 450v Anode.

It might make more sense when I’ve drawn it up.

Very happy with where Ive ended up.

There is just one final modification. I have two small hammond chokes that can take 8ma dc. The ecc88’s are drawing 3.8 ma. So I could add a choke each in series with the 80k load. There have been a few mentions over the years of a choke added to resistor load had sounded better.

Nothing ventured, nothing gained

But not until after Ive enjoyed it upstairs. Its still on dining table. Ive earned a few weeks rest and listen.

[Nick]

Just remember the max anode voltage (cold) of a ecc88 is 550v.

[Paul Barker]

Just remember the max anode voltage (cold) of a ecc88 is 550v.

I manually turn heater supplies up then b+. Two outputs on the variac, all heaters first, then b+. The chokes arent instead of the 80k it is added. There have been a number of circuits especially vintage ones where it has been positively reported. One particular amp was a gm70 Russian design, subject if a Glass Audio article 24 years ago, in which 3 ideas were compared and the choke with resistor was found best sound. Ive lost a number of glass audios though, but I dont forget these details.

[Paul Barker]