Scratch build amp wiring issue

[RogerLLL]

I wired the HT line of a scratch build single ended 6SL7GT + 6V6GT + 5Y3GT amp to the AC circuit instead of the DC on the rectifier. I've wired rectifier tubes correctly before, brain just wasn't switched on this time! I was using a dim bulb current limiter which glowed brightly, no dimming indicating a short, so disconnected power immediately. I rewired the rectifier correctly and powered on. Bulb dimmed as it should. Voltages all seem fine, but no audio output from signal into the grid of the 6SL7. I fed a line level audio signal to the grid of the 6V6, and there was audio output, albeit quieter than I would expect. Nothing when I did the same for both stages of the 6SL7. I tried replacements of both 6SL7 and 6V6, and the same behaviour so thinking tubes are probably ok.
Could this wiring mistake have damaged any components? Wondering about the output transformer which I know worked previously. Like I say voltages are all as expected, so thinking power transformer & rectifier are probably both ok.
Is there anything I can try to test it further?

[Chrispy57]

Hi Roger - perhaps you hit on the problem without realising it in your "wiring mistake" text? Having made that original error, you may be focusing on the possible consequences, and missing the bigger picture fact that as a new build it has never actually worked yet, and there may be an undetected simple wiring error. A thorough checking of the early stages should uncover such an error.

Good luck
Chris

[kalee20]

If you put AC on the HT line, it's unlikely to have hurt anything, especially as you used a lamp limiter.

The HT electrolytic capacitors will have taken a hammering, but for a moment or so, they'll recover (hopefully they weren't on long enough to have got hot).

It's very unlikely the output transformer will have suffered. You can put 240V AC mains directly on the primary of many an output transformer, without causing harm. (A loudspeaker connected the secondary will let you know, in no uncertain terms, anyway).

As Chris says, this amplifier has yet to work - it is possible that there is another fault! Measure the anode voltages of all the valves, and let us know.

[RogerLLL]

Many thanks for the replies! All good advice!
I have checked the wiring fairly thoroughly and all looks good, but I will check again! Could be something I've missed, and I will record all the voltages and report back.
It's encouraging to know that the AC is unlikely to have caused any damage. I will check the filter caps as you say. I also have 4 6SN7GT tubes which have the same pin out as the 6SL7s just in case both the 6SL7s I have are bad. After I've re-checked the wiring of course.

[Radio Wrangler]

Earlier stages will likely have their HT fed through dropper resistors with decoupling capacitors on the lower HT sections. With AC applied, the decouplers will look like shorts and the full AC voltage gets applied across the droppers. You may have fried a few resistors. Similarly with series resistors for screen feeds to pentodes and tetrodes.

You'll just have to go round and check all the components. It isn't off the cards for some electrolytic HT decoupler to have failed. Just diligent component by component foot-soldiering is needed to get confidence that there isn't something lurking as a result. Then you need to contend with the unknowns of a newly constructed unit.

An American colleague used to train new graduate design engineers, and amongst his slide shows was always a simple one saying; "Experience is what you got when you didn't get what you wanted". We've all done this sort of thing!

David

[RogerLLL]

That all makes sense. Coupling capacitors block DC but allow full AC to pass, normally just signal but in this case mains voltage! I'll check the dropping resistors, and electrolytics. In fact everything!

[Silicon]

After you have measured the HT, anode and screen voltages, measure the voltage across the cathode resistors and calculate the cathode current for each valve.

[kalee20]

Originally Posted by Radio Wrangler

Earlier stages will likely have their HT fed through dropper resistors with decoupling capacitors on the lower HT sections. With AC applied, the decouplers will look like shorts and the full AC voltage gets applied across the droppers. You may have fried a few resistors. ... It isn't off the cards for some electrolytic HT decoupler to have failed.

Though, Roger used a lamp limiter, which he says glowed quite brightly, so the 'full AC voltage' won't have been very much.

I'd be pretty optimistic that little or nothing will have been fried due to the miswiring, and the lack of performance is due to another fault.

Originally Posted by RogerLLL

I fed a line level audio signal to the grid of the 6V6, and there was audio output, albeit quieter than I would expect. Nothing when I did the same for both stages of the 6SL7.

Getting something from the output stage says that this is working (though there may be a fault here nonetheless), but nothing from the previous stage points to a fault in the driver or the initial amplifier stage.

Originally Posted by RogerLLL

I also have 4 6SN7GT tubes which have the same pin out as the 6SL7s just in case both the 6SL7s I have are bad. After I've re-checked the wiring of course.

They do have the same pinout, and in a simple amplifier they can be swapped. The 6SN7 is lower gain, and it also sucks twice the heater current, but it'll be fine for fault-finding.

Measuring anode voltages with a DMM will be insightful, let us know!

[RogerLLL]

So, "diligent component by component foot-soldiering" is what was needed! I checked all components and realised that I'd made a fundamental error in the two 150k dropping resistors - I'd fitted 150 ohm, so quite a big mistake on my part. The two main filter caps had both taken a big hit and were damaged. Nominally 8uF but reading 17uF and with elevated ESR and Vloss values. Once all these had been addressed the amp is fully working.
The circuit is loosely based on a Fender Champ 5C1 modified for a 6SL7GT rather than the 6SJ7 as per the original. Also with a Princeton 5C2 tone control and a gain control between the two 6SL7 stages. I've attached the schematic.
The tone control works brilliantly, perfect for calming down what is quite a harsh tone. I'll check the 6V6 voltages and look at possibly changing the 500 ohm cathode resistor value to maximise output volume. Also look at trying a variable negative feedback to calm the gain down a little for a cleaner tone, as well as trying different bypass cap values.

[Chrispy57]

Well done Roger, mystery solved! Easy mistake to make and difficult to realise you've made a mistake when glancing at the colours on a resistor that you think is what it should be. Perhaps that 50k resistor saved your bacon a bit?

BTW are you sure about the accuracy of your circuit diagram of the rectifier section?

Cheers
Chris

[RogerLLL]

Missing a ground connection on the rectifier taps? I also need to add a switch and a fuse, and new filter cap values which are now 22uF.

[RogerLLL]

Here's an updated schematic.
Can anyone shed any light as to what the oscillation type sound is at high volume as in the attached clip (which I had to zip)?
It sounds almost like a noisy tremolo in the background, but just affecting background hum & hiss, not evident in the actual audio from a signal. Only there when both gain and master volume are at the highest setting, disappears when either is reduced.
Wondering if this is motorboating? I tried "chopsticking" with internal wiring to see if there was any obvious cable routing issues, but nothing had any impact on the sound.
Any clues as to what this might be would be very much appreciated! I get the feeling the power transformer might be a little under-spec'd, as the valve amp "sag" is extremely pronounced at higher volumes.

[Radio Wrangler]

I don't know the amp that's based on, but I was a bit surprised seeing 6L6 G2 floating and G3 tied to dropped HT.

As you can use a 6L6 as a power amp on some of the lower amateur bands, I'd expect a need for some parasitic stoppers and an HF load on the anode just to keep it tame. RF oscillation can squeg, thus making audible sound effects.

But the sensitivity to volume control position, suggests some involvement of the earlier stages likely some coupling from the output.

David

[kalee20]

It's motorboating-type oscillation.

You have 3 stages, so feedback from the last stage anode to the first stage anode, via the common impedance of the HT line, will be positive.

You might think that the decoupling filter formed by the 50k (which incidentally seems high!) and the 22uF capacitor would eliminate that - but the 6V6 g2 (? Shown as g3!) behaves as a small anode for the purposes of this, and it's connected to the 'wrong' end of the decoupling filter.

Just add further decoupling - suggest 10k and 47uF - to the first stage only, and you'll find it's gone. Or, to save even doing that to prove it - stick a 47uF capacitor across your 22uF one, and note how the frequency of 'pulsing' drops.

One other comment - adding to the g2 / g3 anomaly of the 6V6 - I really don't like the idea of a 1M volume control on the 6V6 grid. Several reasons:

(1) if the wiper leaves the track - which some day it will, as the control wears out - the grid will 'float' and the 6V6 could pass a high current.

(2) 1M seems a bit high for a power valve such as the 6V6 - residual gas causing parasitic grid current is more significant. I'd use no more than 470k for its grid leak.

(3) the driver stage, which is where noticeable distortion can take place owing to the high signal levels, is running at full whack always. I can see the advantage, that noise and hum arising from earlier stages will decreas as the volume is turned down, but it shouldn't be much anyway from the driver. So it would be better to put the level controls at the input to the first and second valves.

[Richardgr]

Could you turn that 1M volume control around so that the wiper is from the 22nF cap, and the grid sees a constant load from the grid leak? Or move the wiper it to the other side of the 22nF coupling cap, and use a better grid leak option.

[Sideband]

Yes that volume control is asking for trouble. At the very least connect the track between grid and chassis and the wiper to the coupling cap.

[kalee20]

That would give a changing load on the driver, leading to driver distortion at low volume levels (with the wiper nearly at 0V). And moreover, with the coupling capacitor, it would give a progressively worse bass response as the volume is reduced.

No, much better to have the two controls at input of first and second stages. If either wiper leaves its track then, the valve's current is limited by its anode load resistance.

[RogerLLL]

I took the basic schematic from the 5C1 Fender Champ, which uses the master volume as I've represented it. Also for the 5C2 Princeton. That said, I've just noticed that the other Champ and Princeton versions do as Kalee20 suggests putting the volume on the second pre-amp stage, so I'll try that.
My bad btw regarding the way I've represented the 6V6! Corrected in the attached schematic, albeit G3 is connected to the cathode internally which I couldn't represent.
I'll try a larger filter cap on the first stage - not sure where you're suggesting using the 10k, is that in place of the 150k?
Here's a link to the Fender schematics I've been referring to:
https://vintagefenderamprepair.com/pages/library-schematics-layouts

[kalee20]

Originally Posted by RogerLLL

I'll try a larger filter cap on the first stage - not sure where you're suggesting using the 10k, is that in place of the 150k?

Nope - here! As extra decoupling in the HT rail, to the first stage, to avoid any 'wiggles' that g2 of the 6V6 puts on the HT rail from finding their way to the first-stage anode.

I'd be 95% confident that the oscillation you're getting (yes I've listened to the sound clip in your earlier post today) is due to this.

[RogerLLL]

Gotcha, should be an easy mod. Many thanks for that! I'll report back.