B15-inspired Amp Head Build

[Jus Lukin]

-

Edited March 15, 2022 by Jus Lukin

[JapanAxe]

We have a name! I considered various one-syllable words starting with a 'bee' sound, before settling on something... Spanish. My previous build was a Champ-inspired circuit in a 1x12 combo from a deceased Line 6 Spider, which I dubbed the Araña Muerta - Spanish for Dead Spider.

 

So this one is the ABEJA 15. 'Abeja' is Spanish for 'bee'.

 

I sorted the troublesome treble pot with a blast of switch cleaner. I might still try cooling the bias down a bit - it really does get rather hot above the 6L6s!

[JapanAxe]

I'm still having some issues with this amp on the fuse front. When I switch out of standby, the T2A mains fuse often blows but not the T500mA HT fuse. On one occasion the mains fuse blew some minutes after switch-on, I suspect because it had been weakened.

 

With just the rectifier valve fitted, I tried a (temporary) T4A mains fuse - now when I switch the warmed-up amp out of standby, the mains fuse doesn't blow, but if I do the same thing with the 6L6 valves also installed, the 6.3V pilot dims momentarily but again the mains fuse remains intact.

 

I suspect the problem may be related to one or both of the following:

• Inrush current to the first filter cap;
• Sudden flow of current caused by the heater CT being 'elevated' to the output valve cathodes.

Any ideas folks?

[Beer of the Bass]

Where does the standby switch come in relation to the first filter cap? I've seen some articles suggesting placing it after the first cap, so that it charges slowly as the rectifier warms up. That's meant to be less taxing on the rectifier too.

http://valvewizard.co.uk/standby.html

[JapanAxe]

6 minutes ago, Beer of the Bass said:

Where does the standby switch come in relation to the first filter cap? I've seen some articles suggesting placing it after the first cap, so that it charges slowly as the rectifier warms up. That's meant to be less taxing on the rectifier too.

http://valvewizard.co.uk/standby.html

As per the original schematics, the HT current goes: PT secondary > Standby switch (double pole) > GZ34 rectifier > HT fuse > first filter cap

 

I'm aware of the alternative of the approach of having the standby switch after the first filter cap - there are pros and cons to both approaches.

 

I'm just about to move the heater CT to ground to see what effect that has. Thinking about it, elevating the heater CT is the biggest departure from the original circuit that I have made...

[JapanAxe]

Moving the heater CT didn't fix the problem, in fact switching from standby almost immediately blew the 3A fuse in the mains plug. I'm going to put this to bed for now and come back to it in a day or two, otherwise my sleep will be spoiled by my brain trying to problem-solve when it should be resting!

[Stub Mandrel]

The mains plug fuse will be quick blow. Perhaps you should be using a 5A there? 3A is more what you use for a bedside lamp.

 

Is it possible that you have a shorted turn or winding on your transformer, easy to do if you have a toroidal mounted by a central screw.

[Stub Mandrel]

Can you put a bleed resistor (say 100K - 1M) across the standby switch so that the filter cap can partially charge?

[JapanAxe]

1 minute ago, Stub Mandrel said:

Can you put a bleed resistor (say 100K - 1M) across the standby switch so that the filter cap can partially charge?

The standby is before the rectifier so not sure whether that would work.

 

Since my last post I remembered I have just got another GZ34 in case that’s the problem - it’s one of the few components between the mains fuse and the HT fuse.

[Hellzero]

Try a Solid State Rectifier instead of the GZ34 as these valve rectifiers pump a LOT when switched on. Something like the good old Sovtek SSR or any other (or make one yourself).

[JapanAxe]

10 hours ago, Hellzero said:

Try a Solid State Rectifier instead of the GZ34 as these valve rectifiers pump a LOT when switched on. Something like the good old Sovtek SSR or any other (or make one yourself).

A SS rectifier would give a higher B+ but I’m not sure it would help otherwise.

 

I have a full day today but when I next have time my plan is to try the new GZ34. I have heard of rectifier valves developing faults which eventually blow the PT - clearly I don’t want this to happen!

[Hellzero]

Indeed more B+ and way less current absorbed at starting by the valve rectifier (and loads of them, especially original vintage ones, are often faulty). I've had so many fuses blown by these valve rectifiers that the first thing I ever did was replace them by an SS one and the problem was solved forever.

 

Really worth giving it a try. 😉

[Chienmortbb]

Silicon in a valve head? Blasphemy. Actually while we can argue the merits of the soiund of an all valve head, it is beyond doubt that solid state rectifiers are the best. 

[JapanAxe]

1 hour ago, Chienmortbb said:

Silicon in a valve head? Blasphemy. Actually while we can argue the merits of the soiund of an all valve head, it is beyond doubt that solid state rectifiers are the best. 

 

2 hours ago, Hellzero said:

Indeed more B+ and way less current absorbed at starting by the valve rectifier (and loads of them, especially original vintage ones, are often faulty). I've had so many fuses blown by these valve rectifiers that the first thing I ever did was replace them by an SS one and the problem was solved forever.

 

Really worth giving it a try. 😉

I’ve built valve amps with SS rectifiers but with this one I was trying to recreate the vintage Ampeg as far as possible. The current GZ34 is a JJ that I had as a spare in my gig box for many years. The replacement is a brand new JJ.

[Stub Mandrel]

My Dad was reminiscing about the change from valve to solid state rectifiers. My grandad blew a few when fitting them to replace HT rectifiers because he didn't realise they needed a resistor to limit inrush current.

 

When a selenium rectifier went on someone's TV, he said you could diagnose the fault as soon as they opened the front door and you smelt it.

[Geek99]

On 20/08/2021 at 22:22, Stub Mandrel said:

Can you put a bleed resistor (say 100K - 1M) across the standby switch so that the filter cap can partially charge?

Yeah I was going to say that too 😳

 

Actually I’m totally amazed by all of this ….

@JapanAxe sound sample? 

Edited August 22, 2021 by Geek99

[JapanAxe]

1 hour ago, Geek99 said:

Yeah I was going to say that too 😳

 

Actually I’m totally amazed by all of this ….

@JapanAxe sound sample? 

Sorry, no sound samples until I can get this working reliably!

 

I've made some progress this morning - I fitted the new GZ34 and that doesn't blow mains fuses - yay! - but I now have a blown HT fuse (T500mA) so I am working my way through to see whether that was anything more than a fluke. The new GZ34 also produces a slightly higher unloaded B+ at the first node than the old (faulty) one (454V rather than 437V). My filter caps are all rated 500V so that's fine, and it will drop anyway when the other valves go in. Both 6L6 valves still test good. I'll keep y'all posted.

 

Incidentally, I write down everything I do as I'm doing it so I always have a record of what went wrong (or right) and what changes preceded it.

[JapanAxe]

Well that seems to be sorted. Conclusion - the rectifier valve was the problem. I've now binned it. I've done a quick playing test and confirmed it still sounds just as good. The hum/buzz is still there and persists when the PI valve is pulled, so I'm going to try altering the lead dress around the 6L6s. I'm also going to dial the bias down a bit with a higher-value cathode resistor.

[JapanAxe]

Mostly good now. With a 470R cathode resistor, plate dissipation is now just over 46% of max. This is somewhat on the cool side but I'm not hearing any crossover distortion. I'm going to order some 330R and 390R 10W resistors in case I need to warm it up a bit. Having read that current production rectifier valves don't cope as well as they might with reverse voltages, I tried installing some SS diodes in series to protect the GZ34, but the amp didn't seem to like that so I took them out again. Still might try a choke in place of the first dropping resistor, and/or a small dropping resistor before the first node...

 

Here's some nice pictures of the completed amp with its BF playmate.

 

[jimbobothy]

Lovely stuff. No idea what you’re on about most of the time but love it anyway 😁

[JapanAxe]

For several months, two items have been sitting on my work station, taunting me every day:

1. Bag of UF5408 diodes

2. 10H 90mA choke

 

Until today!

 

1. For some reason this amp didn't like my earlier attempt to ease the strain on its rectifier valve by adding SS diodes. Today I routed the incoming AC through UF5408s mounted across to the unused pins on the socket, and all is well. No change in sonic performance, just the likelihood of increased longevity for the rectifier valve.

 

2. That background buzz. I thought it might be power supply related, which is why, months ago, I bought a suitable choke, with the intention of trying it between the first and second power supply nodes.

 

The buzz was unaffected by the volume control until it was up nearly full, where there was a null point. The volume control comes just before the PI, so I pulled V2 - no buzz. I re-routed the screened cable carrying signal to the first control grid of V2, and tried moving around (with a chopstick) the wires running to V2 - no change. Then I got out my cap-shorting device, a probe with a 330R resistor inside and a croc clip on the other end that I connect to ground. When I touched it to the first V2 grid the noise disappeared. That means the only place it could be coming from was the 470k grid stopper. Eh?

 

Examining the schematic, I saw that these 470k resistors are not in fact there as grid stoppers, but to provide resistive mixing for the two channels as they feed into the PI. Going through the schematics I saw that all except the Heritage version used 270k resistors in this position. Given that I'd built a single channel amp, I arguably didn't need a resistor there at all. Then again, I'm aware that grid stoppers can prevent oscillation, so I thought it would be a good idea to have something there. I looked at 6SL7 circuits online but many didn't use any grid stoppers. In the end I tried an arbitrary 1k, and hey presto, the buzz is gone, at least until you get the amp wound up close to Patent Pending.

 

I'm not entirely sure why the 470k resistor would cause buzz, maybe someone can chip in with an explanation.

 

I didn't need the choke in the end, but no matter, it will find a cosy home in one of my boxes of Useful Bits.

[JapanAxe]

Ooh, forgot to mention. Having treated myself to an oscilloscope I fed a sine wave into the front end of the amp and hooked the test probes to an 8ohm dummy load. I worked out the amp will deliver 22.5W before it starts to clip. This is the cathode biased version with a reputed output of 25W, so allowing for the slightly reduced B+, that looks about right.

[Richard R]

I've enjoyed this thread, @JapanAxe.  I don't really understand the implications of the in-and-outs, but it's been interesting. And if I ever do need to understand valve amplifiers, I know who to ask.

[JapanAxe]

29 minutes ago, Richard R said:

I've enjoyed this thread, @JapanAxe.  I don't really understand the implications of the in-and-outs, but it's been interesting. And if I ever do need to understand valve amplifiers, I know who to ask.

 

Thanks, but you'd be better off consulting the many useful articles published by The Valve Wizard!