Junkbox build

[joebog1]

A quick update.
Ive mounted all but three or four components, which are all associated with the neg feedback loop. I post the pics to keep you up to speed,
Pic one is my " single earth" and as you can see, apart from the pole compensation cap ( 100pF SM ) is final.
Pic two is both channels, but without the loom or cable lacing. I still use waxed linen ( or maybe flax ) thread, that I have had for about 45 years, ex RAAF in fact.

regards
Joe

[joebog1]

Spent all day working on the amp.
I switched it on, and after about 10 seconds, I hear a bang , NOT a big bang, but enough to make me switch off as fast as I could.

Look as I might, I cannot see anything wrong, nor is there any "flash" damage.
I recheck my circuit - nothing wrong .
I switch on again, and all appears to be normal without signal.
I have both channels plugged into my CRO, and noise appears to be normal.
( There is no negative feedback connected yet).
I apply a signal and watch the CRO traces, all normal, untill I get a bit above the noise floor. One channel is clipping the bottom of the waveform ( sine at present). I grab a meter and check HT - all normal. I check cathode voltage on the output valves. One valve isnt drawing current the voltage is zero across the cathode resistor. HHMMM new valves!! one doesnt work, OK I think and remove the valve that doesnt work. Lo and behold theres a big hole blown in the side of the envelope, and the cathode pin is completely removed from whats left of the envelope. The getter is white, so I figure OK a micro fracture in the glass that has split completely as it heats up - no probs I put another valve in.
Well output is higher but still missing the bottom half of the waveform, I recheck the cathode and ? NO CURRENT .
I check HT to the plate, at least partially removing the OPT from suspicion,
NO Cathode voltage?? Must be an open circuit balance pot or cathode resistor - check, theres resistance there, but stupidly I set the AVO to ohms.
I dont bother setting exact zero ,yep about 470 ohms. I turn the amp on again and accurately measure the voltages around the valve. all normal apart from cathode which at first switchon I would expect "about" 20 volts, but its zero. I remove the 470 ohm resistor and its open circuit!! A 1.2 watt Rodestein that is robust as is available. I replace the resistor and switch on , SAME SAME. What else can it be ?? All parts are brand new, and havent been overheated or damaged in the build process. The ONLY thing left is a short cathode electrolytic. Remove it and find dead short.

Replace capacitor with ANOTHER $10 RS supplied RIFA 100uF 25 Volt cap and switch on. PERFECT .
My question to the boffins, What failed first?
Valve, resistor or capacitor? After all I spent half the day sorting this out ( dont tell me, I know Im stoopid ).
Amp works nicely without feedback, tomorrow Ill fit feedback and "tune" the square wave response - pics will follow.

Any answers to the questions?

Regards to all
Joe

[tony brady]

The valve went first, putting a much higher voltage on the cathode resistor and capacitor. The resistor would have gone open first followed by the capacitor shorting due to the high voltage across it. The failure has occurred pretty much as soon as the heater has come up enough for conduction.

The question then is did the valve die or was it pushed? It would be worth checking that the HT voltage is correct as earlier you mentioned you were using voltage doubling - also if you are using silicon diodes this would put HT onto the valve before the heater came up which is not good for valve life.

As you are expecting 20V on the cathode resistor of 470 ohms this means the resistor is dissipating 0.85W - it would be better to fit a 2W resistor for long term reliability and may be a 35V capacitor

[joebog1]

A quick update today. CRO pics of initial tests. These are taken at about 1Khz, and about 9 watts. ( measured with Fluke 8050A, 8.5 volts into 8 ohms).
I am particularly thrilled with square wave performance, as I have not done any tuning yet. At 10 Khz I am going to need pad the feedback resistor, as I have a small amount of ringing on the falling edge, which I have always found to be S.O.T ( Select On Test) I have modified the circuit slightly. The pair of 6BW6's have their own bias caps and resistors, and I have fitted a balance pot to equalise the current drawn by each valve. The rest of the circuit is pretty much as published, taking into account values have been rounded up/down to fit E12 values.
Out of the box, the Brimar 6BW6's are remarkably close in characteristics, considering they are from at least a couple of different sources.

Now to finalise the feedback ( I will show results )and then add my phono stage, which I have also changed from initial ideas. Instead of a cathode follower, I will try a 6CS7 double triode with dissimilar sections.
The first stage is a conventional voltage amp using the 12AU7 section of the valve, but that only fetches about 350 mV which is nowhere near what is needed. So, I will use the big triode, as another voltage amp, which Im hoping will fetch me about 2 volts out, although I would like a bit more.

I havent designed around the 6SC7 before so its a learning experience.
Regards

Joe

[joebog1]

A bit more progress.
Valve bases are in, heater wiring is done.
No heater supply fitted yet, but will still be DC for the phono stage.
HT supply for phono ( choke and main capacitor ) fitted, although I still need perhaps 3 resistors and 3 more large decoupler electros. These will be inside the phono "cage". Fitted a 1960's Ducon dual ganged volume pot ( 500K ohms logarithmic) instead of the Alps unit ( 100K ohms log.)
Power amp works perfectly, but I will have to see how well it lasts ( reliability !!! )

Regards
Joe

[joebog1]

Amp is running !! Sounds delightful - MY opinion
A few others have commented on " clarity and clearness".
Its my first ever pentode phono stage, and I must say its very nice.
Soundstage is huge!!! But then 6BW6's are 6V6's, a sound I have always liked.
You might say very well behaved even in times of stress. A pommie marvel!! ( thanks Mr Churchill )

From ZZtop, to Huddersfield choir, I cant cause the lil amp to trip up.
Overload is awful, so I tend to keep it below audible distortion level.
At first I thought bass was thin, but its highly record dependant - remember I have no tone controls, only hand selected components in the RIAA section.
Power amp is 6.3 volt AC heated with no connection to earth; as it turned out its totally ( to the ear ) silent. Phono stage is DC heated with just a bridge rectifier, 10,000 uF cap, and a .47 ohm resistor which fetches 6.4 volts. Close enough for me. I havent made any measurements yet, but lets say VERY quiet by any standards, even with phono inputs open circuit, and no base cover on the chassis.

More to come, after I do the promised tests on Neils switchmode amp
regards
Joe

[joebog1]

I should point out, the ugly cables in the background are the SME origionals :-;
I have discovered that they need replacing - Im open to suggestions.
j.b.

[Radio Wrangler]

Ah, Huddersfield Choral Society... That 4-manual Willis in the town hall certainly produces bass notes, with a good few ranks at 32-foot, but rarely gets let loose as an accompaniment.

There are a few records available with Jonathan Bielby playing, but I grew up with it when Harry Gledhill used to throw it about a bit.

The town hall was built to foundations when the organ was bought. The organ was too big, so they ripped up the foundations and a new hall was designed around the organ, with tiers upwards behind the stage for choirs.

David

[mole42uk]

The last 6V6 amplifier I owned the valve was in an octal base and a much bigger envelope. Does the 6BW6 have the same electrical characteristics?

[Herald1360]

Yes.

6AQ5 is same again and even smaller, on B7G, but slightly lower voltage ratings.

[Radio Wrangler]

Back in the day, the 6BW6 was heavily touted as an exact 6V6 in a smaller envelope.

The 6AQ5 got the reputation of being a bit iffy but the 6BW6 had a good reputation. Some HF transmitters in shortwave mag and radcom used it because of the shorter, lower inductance connections in that package.

David

[joebog1]

I have an idea for another amp, and I have the bits n pieces.
This one will be 7C5 a REAL 6V6 but with loctal base, all badged U.S. ARMY SIGNAL CORPS the order was placed in 1942.
Amp will be push pull parallel into my own hand wound transformers.
Im expecting around 20 watts.

6BW6 does have the origional specs of a 6V6, the thing I have noticed however, with the plate being so close to the envelope, is that they run "stinkin hot!!". Im hoping for a cooler running amp with the 7C5's.

Huddersfield Choir LP is one I use for testing! The bass content is sufficient to test ANY amp, and a full choir would have to be one of the most difficult sounds to reproduce faithfully.
The Martyr Mantras digital LP is also capable of tripping up a badly designed amp, this little 8.5 watter performs wonderfully, especially with the quite sensitive Tannoy's.
Ill CAD up the circus ( circuit) shortly and post them here. Its a very simple design thats probably a good beginners amp.

Joe