OK…so I’m on wi’ summat

[andrew Ivimey]

Where shoddy meets dreadful I'd have thought!

[Cressy Snr]

There's a lot more room in the chassis than in the Mk1 version of the amp, so, copying Paul, I purged the power supply of electrolytic caps by installing a nice pair of LCR motor run caps in place of the electrolytics that decoupled the input stages.
Didn't think it would make a lot of difference, but a layer of hash has been taken out. It's one of those hash layers you don't notice until it isn't there... interesting:

LCPSUAllMotorRuns.png (87.25 KiB) Viewed 1049 times

[Paul Barker]

There were a lot of good SE circuits with the oil cap philosophy. This one below for example. This circuit also was the first to introduce the 6EM7. Eric Barbour was the US rep for Svetlana for a while, but disappeared off teh scene early doors, and the Svetlana website vanished too.



I posted it here to demonstrate the oil cap preference.

- I shall extended other thoughts from the Eric Barbour circuit on to my 6550 thread since they don't belong here, and only wanted to show the trend to use oil caps came from a long way back.

But dont miss something not always shown in circuits the 6em7 heater supply is biased up to + 75v.

[Paul Barker]

Circuits:

1. Audio Circuit with measured voltages:
EL34Audiocct.png


2. Choke input power supply:
EL34LCinputPSU.png

Sorry to be a pain in you’re proverbial Stephen but I looked at youre blokes circuit to see if I could use his design of the use of the 6sl7. Having tied myself in knots following a table in the vade mecum.

There is only one small thing I’d like to point out which may be a misprint in his source. The cathode bypass makes sense at 47 uF which -3 db is 3.8 hz whereas 470uF is 0.5 hz

Out of interest teh gain of the stage (unaffected by this) is a very useful 43. Good input headroom at 2v peak turned into 86v peak drive. So youre amp is sensetive. Its about right for me to drive the 1619 directly through the EL84 cathode follower. Half expecting it not to sound as good as the EL34 and only 5 watts. Gives me a diversion from the madness at home though. Im turning to the 6SL7 to get the extra gain and benefit from maybe higher 2nd as Nick suggested to counter the third of A2. However after calculating distortion at youre operating point 2nd h is only 3.9%. Back to the drawing board for me. The ecc 88 does better distortion than that Lol, but not enough gain any longer, suppose I could bypass the cathode resistor. Or better still add a stage, such as keep the E88cc which is acceptably distorted, and multiply it with the 6EM7 cap coupled driving the CF. That should better the distortion profile.

[Nick]

However after calculating distortion at youre operating point 2nd h is only 3.9%.

Just to put this in perspective, when I talked about high 2nd I was thinking 0.5% at most. Beyond that I would start to consider it a guitar effects box.

[Cressy Snr]

I don't know what the hell has been going on with this amp, but it was putting out some kind of electromagnetic pulse every time it was powered down. It has been doing this ever since I built it with the choke input supply.

It got that bad on two occasions, that on power down, accompanied by a crack through the speakers, the TV threw a hissy fit, the screen went black for several seconds before coming back up.

Most of the time, rather than the screen going black, three wide, white bars would appear for an instant when the off switch was thrown, with the same crack through the speakers.

Anyway I've had quite enough of these shenanigans and the potential for damage, not only to the amp itself but also to other equipment connected to the same power strip, and having reverted to a cap-input supply, off a lower voltage secondary tap on the power transformer, the problem has vanished. WTF?

[Cressy Snr]

So here are the mods:

RC filters after the first cap now have 270R resistors as opposed to the previous 470R, which have been shifted onto the cathodes of the EL34s.

EL34CinputPSU.png (84.71 KiB) Viewed 774 times

Because of the higher HT voltage, the OPT is now on the 5K tap, cathode resistor on the output valve has been changed to 470R. The 470uF input stage cathode bypass cap has also been changed to 50uF.

EL34SE_HigherV.png (52.49 KiB) Viewed 774 times

[Paul Barker]

Good move Stephen. When I had a 212 amp filling the dining room I could never get rid of a hum. One late night I was listening and with lights off I saw a spark from the potted military choke input choke. It was keeping perfect time with the noise. Obviously it was replaced. Problem solved. Just because a choke is potten and painted shiney black doesnt say whats happening inside.

To go choke input you have to be very certain about youre components.

Anyway dont worry through my journey to keep up with you I’ve had an output transformer short terminally, a grid choke go open circuit a capacitor short a resistor smell like a barbequeue. I think thats all.

Another thought you could persue is a series pass regulator after the boost of capacitor input voltage. In the past you have studied series pass regulators extensively. You may find a history of youre discoveries in this forum somewhere.

Stephies page on regulators demonstrate how choke input is the least effective regulator. Even the first level series pass is more effective. I just like to choke input for simplicity plus I live on the memories of my discoveries in the early days. Jono would mail me and say, omg you must try choke input and so on. I would really enjoy the discoveries that were simple but effective. You hear a better way and never forget it. But youre transformer and choke have to be top dollar, and even then the strains they are under can cause failures.

By the way, the EL34 was my first ever introduction to how simple and how beautiful a simple valve amp sounded in the days before MOFOS. I live in those days, and thoroughly enjoy my experiences as my old fashioned life carries on.

[Cressy Snr]

Thanks for that Paul.
I’ve started to realise that inadequately rated components and choke-input supplies don’t mix. It seems self-evident from the wobbler throwing telly, that the iron I have is not up to the job. A cap-input supply is perfectly adequate for my needs.

The amp sounds great with the CLCRCRC supply and the higher HT seems to have put the EL34s in a sweet spot. I’m dissipating about 22W per EL34, plus screen in Ultra-Linear. Seems to work very well.

I’m living in the old days too. At my time of life, there’s something comforting about the glow of valves and an episode of “The Footage Detectives” on a Sunday afternoon. I’ll be eating peach slices and Carnation milk, followed by a bath at 6 o’clock before long.

[Paul Barker]

[Nick]

Yep its often ignored that the entire AC component of the rectified sine wave is applied directly across the choke (the coupling cap being a short at AC). The effect of that instantaneously being stopped at some point in the 100Hz cycle when its turned off is never going to be trivial. Brian Sowter always distinguished between choke and cap input when you were specifying a choke. It always confused me that most other choke suppliers never make that a question.

[Paul Barker]

Wound components these days are not wound for 243 volts as theyre imported and we said to the rest of europe our voltage is 230. But it isnt. No wound components from cheap sources have any reserves. More true for chokes you want to use as choke input. Mr Senior Sowter is right.

[Nick]

Wound components these days are not wound for 243 volts as theyre imported and we said to the rest of europe our voltage is 230. But it isnt. No wound components from cheap sources have any reserves. More true for chokes you want to use as choke input. Mr Senior Sowter is right.

Well, to be picky, we have never said our voltage was 230v, all we said was the UK mains supply was still 240v which aligned with the CE spec of a nominal 230v -6% +10%. Which it did and does.

And I would argue that the difference between choke and cap use for chokes is not a matter of just reserve, but an actual specification difference.

Also, specifications for electrical components have not had "reserves" for decades now. A part that is rated for 400v can and should be expected to fail at 401v but work without fail at 400v. Whats the point of a specification if its not true? Why would I want to pay for a part that can handle 20% more of something that I need it to. The only reason for what you may call a reserve in the past is wider manufacturing tolerances meant that the target value had to be higher to cover the low performers. I remember years ago being told that the only definition of quality that mattered was consistency. At the time that didn't seem right, but over time I realize it was exactly right. What I had previously regarded as "quality" was a product of marketing not engineering.

[Nick]

Oh, and I was quoting Mr Sowter "Junior". I used the past tense as he has now retired and Sowter Transformers sold.

[Paul Barker]

Ah, I didnt know. Luck then they still had a record of my transformer attenuater wiring diagram which I’d misplaced.