AB763 style build - tweaking and trouble shooting

[Fiat_cc]

Hi all.
I haven't posted here in ages, but I've got a build that is driving me nuts, so I thought I'd put the feelers out.

I've attached a schematic.  It's more or less AB763, with a slightly tweaked tone stack to hopefully give a bigger range of options, preamp bias wiggle trem lifted from the Vibrochamp circuit, a presence control 5F6-A style, and a foot switchable boost. So far I've solved an issue I had with relay hum by adding a separate 6.3V transformer, gotten rid of the low frequency speaker pumping caused by the tremolo by adding high pass filter networks in a couple of spots, fixed a problem I had where the presence control in certain positions caused the master volume to become scratchy, and toned down the overdrive. I think I'm going to go even further and replace the boost control with a 500K pot, because even the 1M resistor into the 1M pot is yielding too much gain. I originally started with a 330K into a 1M pot, and you could only use up to about 3 on the dial.
Anyway, it sounds really nice. The tremolo works sweetly, the reverb sounds great, and the amp itself sounds great, but...
I have a lot of hiss.  More hiss than I feel I should have. with V4 removed there is zero hiss.  Put V4 back in and there is a fair bit of hiss.  Wind up the preamp volume and there is a lot of hiss.  Remove V2 and we're back to a fair bit of hiss.  I am plugged into a JBL E120, so its a very sensitive speaker, and is probably accentuating it a bit more than other speakers would, but I'd like to reduce it.  It has been suggested I could try replacing grid stoppers with metal film, and I may try that, but I've built quite a few amps using these Xicon carbon film, and those amps haven't been this hissy.
The other issue I have is microphonics.  Several capacitors seem microphonic, and several grid leads are microphonic.  The grid lead to V4A is exceptionally microphonic, and when I probed it to check for DC the bang through the speaker was incredible.  No DC voltage though.  I'm used to some popping when probing, but this was monstrous.
So it feels like I'm really close, but just not quite over the line. Any thoughts and suggestions?
The chassis is not as neat as it started, since I have been making so many changes, but it has been extensively chop sticked, and I can't find any obvious solutions through layout.

[labb]

The microphonics could be a bad solder joint. That and/or poor grounding.

[OrganicEffects]

I've got nothing, besides to say this seems like a super cool soop'd up AB763. Hope you get it working!

[Fiat_cc]

The microphonics could be a bad solder joint. That and/or poor grounding.

I’ve done a fair bit since I posted this. I replaced the most microphonic wire with shielded cable, which help a bit. My biggest issue seems to be hiss getting into V2A. I can ground the input to my preamp volume control, and the pot will still introduce hiss as it’s turned up, and crackles like there’s dc on it, but I can’t measure any dc on the grid, and there is no difference in cathode current between volume at zero (grid shorted to ground) and volume at full (essentially a 1M grid leak). Shorting the output from the wiper shuts it up though. The tone stack effects the hiss, but only really the treble pot (possibly because of the frequency of the hiss), and the treble pot also crackles as you turn it up or down.
I wondered about grounding too, but I’ve never known of grounding to create hiss. More buzz and hum, which this definitely isn’t. It’s real ‘wind in the trees’ hiss.
With V2 and V4 removed the amp is virtually silent. With V4 in its slightly hissy, but probably acceptable. With V2 in, the hiss is rude at high volume. It almost sounds like there is too much gain somewhere, but I can’t find an incorrect component value or connection. Could a bad cathode bypass cap cause weird issues like this? They’re all nice new Nichicons, but I guess it’s not unheard of to get a dud. Likewise I was beginning to suspect one of the time stack caps, but if that were the case, shorting the input to the volume pot should shut it up.

[Fiat_cc]

I've got nothing, besides to say this seems like a super cool soop'd up AB763. Hope you get it working!

Thanks. I will get it working eventually! I don’t give up easily (plus it has a buyer so I have to get it done).
It sounds great! All the features work and sound good, but the hiss is not so good!

[ginger]

If this is an AB763 , did you accommodate the circuit for the missing 50k in the trem circuit ?

[jjasilli]

Thoughts:


bad tube
leaky signal caps
bad solder connections - I've never heard of leads, caps or resistors being microphonic.  To me this suggest poor physical lead connections / solder joints.  I would resolder all these particular solder joints.


Disconnect trem.  Hiss?
Disconnect reverb.  Hiss?


Use a listening amp to pinpoint origination of hiss.

[Fiat_cc]

Thoughts:


bad tube
leaky signal caps
bad solder connections - I've never heard of leads, caps or resistors being microphonic.  To me this suggest poor physical lead connections / solder joints.  I would resolder all these particular solder joints.


Disconnect trem.  Hiss?
Disconnect reverb.  Hiss?


Use a listening amp to pinpoint origination of hiss.

I need to build a listen probe, and that may well be my job for today.  I don't really have an amp to plug it into, although I suppose I could plug it into my audio interface for my laptop.  Seems less than ideal though.

I have changed tubes, and no difference. Removing V4 stops it altogether though, and removing V2A reduces it to almost imperceptible. The tremolo and reverb don't appear to affect it so far as I can tell, the noise originates in V2A.  Grounding the grid directly will stop the hiss, but oddly, grounding the input to the volume pot doesn't, but grounding the wiper does.  I've moved grounds all over the place, and no perceptible change. I've disconnected the tone stack from the volume pot, no difference.  Lifting the cathode bypass cap on V2A reduces the gain of the stage, so the volume of hiss is less, but it is still very much there, and still responds to the volume pot.  Is it possible for a bad pot to create noise?  This is a brand new build, and the volume pot is a CTS 1M Log with a push pull switch for the bright switch.  I've tried disconnecting the brught cap from the pot lugs too.  No difference. Pulling the reverb driver tube makes no difference.  I'll try pulling the tremolo tube just for the hell of it.  You never know.  Being preamp bias wiggle, the trem circuit is directly connected to V2B, so I could try disconnecting the lead running from intensity pot to cathode I guess.  I have my doubts though given that V2B is the first gain stage, and the noise is still there with the input to V2A (second gain stage) disconnected.

[jjasilli]

Looking again at the schematic:


It isn't clear to me that the preamp tubes have grid leak resistors.


V1 looks like a parallel boost stage.  This will increase hiss & noise.  How loud is the hiss compared to signal volume?  (IOW, is there really a problem?)


Don't know your layout, but the schematic suggests that everything -- relay, HT, etc. -- shares a common ground at the signal input side of the amp.  If so, this shouldn't be.  In any event, there is maybe a grounding scheme issue.


Well-filtered DC filament supply, at least to V1 & V2 might help.

[Fiat_cc]

Looking again at the schematic:


It isn't clear to me that the preamp tubes have grid leak resistors.


V1 looks like a parallel boost stage.  This will increase hiss & noise.  How loud is the hiss compared to signal volume?  (IOW, is there really a problem?)


Don't know your layout, but the schematic suggests that everything -- relay, HT, etc. -- shares a common ground at the signal input side of the amp.  If so, this shouldn't be.  In any event, there is maybe a grounding scheme issue.


Well-filtered DC filament supply, at least to V1 & V2 might help.

Thanks for the reply.
I need to get better at drawing schematics, as you're not the first person who has mentioned grounding scheme.  I'm not entirely sure yet how to better represent my intentions in my drawings.  Anyway, I'll talk about my ground scheme in a minute.
So, yes, V1 is a parallel boost stage, when it is engaged, it has the standard Fender style input jack arrangement, and when disengaged, the grid is grounded directly.  V2B (first standard Fender input stage) is either hooked up to the standard Fender input jack arrangement when there is no boost, or it uses the boost pot as a grid leak (some would say not best practice, but many production amps do this). V2A (tone stack recovery) uses the preamp Volume pot as it's grid leak.  This is where is seems the bulk of noise is getting into the circuit. From here it goes off into a standard Fender reverb circuit, and the standard Fender mixing gain stage before the phase inverter.  More noise appears to get into the circuit through this triode (V4A), but it has a 220K grid leak. Reverb and tremolo both work as expected, as does the boost (although it's a little over zealous, and I might change the voltage divider ratio to make the range of the pot more useful), but removing reverb driver, tremolo driver, or boost tubes doesn't appear to have any effect on the hiss.  Pulling V4 stops it dead, and pulling V2 reduces (with V4 replaced) reduces it significantly, but not entirely.
The hiss is louder than I would expect from a freshly built, lowish gain amp.  Using my dad's original '63 blackface Bandmaster as comparison, this amp is noisier.  Yes, I am listening through a JBL E120 which is very high efficiency, but I have also put it into a Celestion which also had prominent hiss.  Playing through it, the signal is definitely louder than the hiss, but the hiss is more audible than it should be, and the fact I have some scratchiness on the volume pot and in the tone stack, from brand new pots suggests to me that there is some wort of DC issue, or a grounding problem.  V2 and V4 both share a filter cap node, and I am beginning to wonder if that cap is suspect.  It is one half of a F&T 16+16 450V can.

Grounding wise, I have a preamp buss, which everything up to and including V5 is connected to.  It grounds at a lug right near the input jacks.  V6 (phase inverter) and the power section ground at a lug right near the PT, that includes both HT and filament centre taps, ground reference from OT secondaries, bias supply etc.  Basically everything that is not part of the preamp grounds at this lug.  I tried isolating the preamp lug, and running a lead back to this main lug at the PT, to create a kind of bus/star hybrid ground scheme, but it made zero difference.  I have tried with relay supply disconnected, moving the relay supply ground point etc, and no discernible difference.  I've rerouted various grid leads, replaced some with shielded cable, and even mucked around with different locations for grounding the shielding (only ever at one end of each run), and no discernible difference.

[Fiat_cc]

Hit character limit...

As far as DC filaments, I don't think in this instance that is going to be my issue.  Hum is very low in this amp, and when I was messing with grid wire placement, I could clearly hear the effect of grid wires too close to filament wiring. The sound I am battling is very much a hiss, like resistor hiss, or that hiss you get sometimes with old power tubes.  I would actually suspect the power tubes, except that they are silent with V4 removed, just the slightest amount of transformer hum with the master volume wide open (barely even perceptible except in a quiet room).

I'm getting frustrated. I was wondering about DC leakage from coupling caps, but they are all brand new Mallory 150's and whilst being new doesn't make them immune to failure, it seems unlikely given the noise seems to be originating in more than one place (or at least is being picked up and amplified in more than one place).  I'm kind of down to ground bus, and filter cap as being the only two components that are common to the two stages exhibiting the issue (unless you have something I haven't thought of...).  It's worth noting (if I haven't mentioned it already) that grounding V2A grid directly kills a large portion of the hum, however disconnecting the input to the volume control doesn't have the same effect, and in fact grounding the pot input lug with a lead doesn't have the same effect either.  Turning the pot away from chassis ground towards the input lug still creates hiss and pot scratchiness.

Sorry for the massive essay, but the more info I can give, hopefully the more I can narrow down the possibilities and have a constructive conversation.  I really want to solve this one so I can get it to it's customer.

[bmccowan]

I'll add my 2 cents for a signal tracer/listening amp. I delayed making one, but once I did, I kicked my own ass for spending so much time guessing as to where probems were in the circuit. There are many variations, but I use an iPad with a tone generator app, and a cheap little Dan Electro mini-amp. Works great. I know Doug recommends strumming guitar through it, but for most issues, a variable Hz signal generator is much more convenient than asking you wife to strum the guitar while you probe. You could ask your girlfriend, but that'll piss off your wife even more.

[Fiat_cc]

I'll add my 2 cents for a signal tracer/listening amp. I delayed making one, but once I did, I kicked my own ass for spending so much time guessing as to where probems were in the circuit. There are many variations, but I use an iPad with a tone generator app, and a cheap little Dan Electro mini-amp. Works great. I know Doug recommends strumming guitar through it, but for most issues, a variable Hz signal generator is much more convenient than asking you wife to strum the guitar while you probe. You could ask your girlfriend, but that'll piss off your wife even more.

I am going to build one this afternoon. My only question is, did you wire the 1M pot as a variable resistor, or potentiometer? If it’s wired as a pot, what did you use for ground reference?

[Fiat_cc]

Built myself a listen probe. Not sure it got me much closer. As anticipated, I can hear a load of hiss after V2A. There is some hiss in the tone stack prior to V2A, but it is at a very low level compared to after (but then it probably gets amplified by a factor of 30+ in that gain stage). There is no discernible noise on the input jack, or the grid of V2B. No discernible noise on either of V2’s cathodes.
I’ve had to put tools down to cook dinner for my kid, but I’ll pull V2 after dinner, and see what I can hear later in the circuit.

Is it safe to probe the B+ side of plate load resistors with a listen probe? My gut feeling is leaning towards noise on the B+ (although I’m not certain why).

[bmccowan]

Is it safe to probe the B+ side of plate load resistors with a listen probe?

The capacitor in your probe (make sure you have one and that its rated for 600V) will block DC from travelling the path to the listening amp.

I am going to build one this afternoon. My only question is, did you wire the 1M pot as a variable resistor, or potentiometer?

My setup is an iPad with tone generator app to the amp input for the signal end. For the probe end, I used the body of a gutted plastic ball point pen to house a 0.1 600V cap wired to a phone plug which goes to the little portable Dano amp. An alligator clip to ground. No pot. There are many variations on the Forum and around the Web. The latest here was interesting: https://el34world.com/Forum/index.php?topic=15893.0
Another https://el34world.com/Forum/index.php?topic=16126.0
But let me increase my insurance before you take my advice on safety.

 

[Fiat_cc]

I taped (heavily with insulation tape) a 630V 0.1uF cap to the end of a wooden skewer. Ran some microphone cable back to a plastic jiffy box. I’m the box I added a 1M pot and output jack.i run a guitar lead into my laptop audio interface. Ground is connected both to the output jack, and also back down the probe lead to an alligator clip.
Sounds like should be fine to listen in on the b+ nodes then.
It’s difficult to a handle on exactly where the hiss is injected. The earliest in the circuit it is easily audible is after V2A plate, but if I crank the volume more, I can hear it in the tone stack, however, if noise after V2A still present with tone stack disconnected.
Driving me absolutely nuts!!!

[bmccowan]

Laptop - that's an expensive tool! You will find it. Shot in the dark, but did you try replacing the plate resistors on V1?
Anyways the signal tracer should help narrow it down. Curious; if you compensate/adjust for the volume, does the hiss sound similar comparing into the tone stack and out of 2A? As you said, the signal is being amplified, but I can usually tell if the hiss has the same characteristics. (this is where the scope guys say, "get a scope")
The moderators here are much more expert than me, so one or more will likely jump back in and help. Last time I had a similar hiss issue, I had to add a grid leak resistor, as somebody suggested in an earlier post.

[ac427v]

Two years agoI went crazy looking for noise in a new build. Turned out to be a bad CTS 1 meg pot with a push/pull switch. Maybe a coincidence but since you mentioned using one...

[jjasilli]

For grounding you might want to try ground input jacks, and cathodes of V1 & V2b together at one point the far input side.  TS could also be grounded there; or maybe tied to cathode ground of V2a; further down the bus. 


Your amp is not simple, but rather complex.  You really need all the tools of the trade.  Ultra modern test equipment presents yet another challenge in the tube world.


Because there's hiss before V2 in the TS, his must be originating in or before the TS.  Unless the TS hiss is a false result of working with an open chassis.  With the amp chassis open, you may get false results.  Even quiet amps may be noisy with an open chassis. 


I'm suspicious of the input relay.  It's a coil and may be inducing noise into the incoming signal. The noise would become continually re-amplified.  Maybe disconnect the relay at the PS end.  If the relay is a culprit, maybe it can be powered with filtered DC.

[Fiat_cc]

Two years agoI went crazy looking for noise in a new build. Turned out to be a bad CTS 1 meg pot with a push/pull switch. Maybe a coincidence but since you mentioned using one...

I got all excited, but sadly no win there.
In the process of reflowing every solder joint on the board and tune sockets etc. I found a cold solder joint on the filter cap that feeds v2 and v4, but sadly
That still didn’t solve it. Grrr!

For grounding you might want to try ground input jacks, and cathodes of V1 & V2b together at one point the far input side.  TS could also be grounded there; or maybe tied to cathode ground of V2a; further down the bus. 

That is essentially how I’ve done it. I’ve tries shuffling grounds around a bit too and no change. I’ve never been this frustrated by an amp yet, and I’ve had a few challenging builds!

[Glenn]

Hi fiat_cc
I don't know if you checked the silver mica's - a few years ago I built a Trainwreck express and had scratchiness and DC on my volume pot, which interacted with the tone stack
I found that a had a crappy Silver mica cap, even though it was brand new, replaced it and it solved the problem
hope you find your problem

[Fiat_cc]

Hi fiat_cc
I don't know if you checked the silver mica's - a few years ago I built a Trainwreck express and had scratchiness and DC on my volume pot, which interacted with the tone stack
I found that a had a crappy Silver mica cap, even though it was brand new, replaced it and it solved the problem
hope you find your problem

Thanks Glenn.
Where in circuit was it? I have one silver Mica in the tone stack, but currently have the tone stack disconnected. There are one or two more further down stream, but they’re not really associated with the noisy stage. I’m not ruling anything out though, especially some sort of weird interaction.

[Fiat_cc]

Ok I have made what I think is a significant discovery.

Using my signal generator to inject 0.2v at 200Hz into the inputs, whilst probing around etc, I can clearly hear the tone from the speaker, even with preamp volume pot down. I can see the trace at the input jack, through the tone stack, then there is nothing at the tone stack recovery grid. It is however present on the plate (well after the coupling cap anyway, I can’t seem to get a meaningful reading of the plate with my crappy old scope). It still looks relatively clean on that plate, but if I probe after the coupling cap on V4A I have the signal, just slightly stronger, but dirty as can be. Not distortion so much as hiss/hash all over the wave form.
Is the signal somehow getting into the B+? I don’t see how that would work, but how else is signal magically getting from v2b to both V2A and v4a grids?
I feel like I’m on the cusp of getting this, but I don’t quite understand what I’m seeing here.

[Fiat_cc]

So I replaced the filter cap. It seems to have improved things, but not solved. I’ve realised that some cross talk is inevitable when in-phase stages share a b+ filter node. There is less of it now though. There is also less noise, but still more than I would like. It is not an issue at higher volumes, as guitar signal easily drowns it out, but at lower volume settings it still seems a bit excessive.
The only other amp
I have here only has a single gain stage before the phase inverter, so it isn’t really comparable. It is far quieter noise wise, but that is quite possibly just because it’s lower gain. I might try metal film plate load resistors, cleaning all the pots, 15K grid stoppers on V2A and V4A, and see where that gets me. Maybe try a 33uF snubber in V4A, or even in the phase inverter as is seen on some Marshall’s etc.
I still need to investigate a couple of microphonic wires, to see if they’re tube related, or something else related, but progress has been made.

[Fiat_cc]

So...
I found a diode in my rectifier setup that was backwards...
How the amp worked at all I don’t know. B+ is a little higher now (not much), amp is rebiased.  Aaaand...
It still behaves exactly as it did. Hissssssss.
Gah...!

[shooter]

the tone stack, then there is nothing at the tone stack recovery grid

probing "inside" the TS is problematic even with good equipment and experience.
Grids can be problematic also.  I typically probe plate R - left side coupling cap stage to stage.
so going into TS good, coming out of TS recovery bad, time/money calcs say just replace parts, 10 parts, 1hr, happy amp.  If you want deep dives and mathematics read a book 

[Fiat_cc]

the tone stack, then there is nothing at the tone stack recovery grid

probing "inside" the TS is problematic even with good equipment and experience.
Grids can be problematic also.  I typically probe plate R - left side coupling cap stage to stage.
so going into TS good, coming out of TS recovery bad, time/money calcs say just replace parts, 10 parts, 1hr, happy amp.  If you want deep dives and mathematics read a book 

I see where you’re coming from, but because it’s a new amp, and a prototype of a design I’d like to refine and sell a few if, I need to figure out where I’ve f***ed up, or what component is bad, so I can build it without issue second time around.

As an aside (musing out loud), could high pass filters, with high value resistors introduce hiss? That’s probably one of the only major departures from AB763 in the outputs of the two stages that seem to be noisy, with the purpose of trying to filter out the trem oscillation from making it to the speakers. I succeeded in doing that perfectly, but I wonder if those filter networks are upsetting other things I didn’t foresee.

[shooter]

since your "base" was AB763, that's where you start, when it works great, sounds good, then tweak, otherwise it's one rabbit hole to the next.

I've built many amps, and only 1 "canned" amp 5F2.  what I lern't, build 1 piece at a time, test, verify, add next piece, otherwise you just scrap and move on

[PRR]

> some cross talk is inevitable when in-phase stages share a b+ filter node.

"Some" should be "very very small".

The plate signal on the send source goes to B+ via say a 100k resistor. There it hits, say, a 33uFd cap. The cap impedance is 300r at 17Hz, 30r at 170Hz. (Eventually we run into ESR but not till the top of the audio band.) Since we don't hear small deep sounds, let's look at 170Hz. 100k plate resistor to 30r of capacitor is 3000:1 of reduction. To get to the other channel it passes through another plate resistor shunted by a plate resistance, which in triodes is typically PSRR of about 2:1. So signal is 6,000 times weaker. Much weaker in midrange where we would hear it best. That's very small.

Round-the-bush grounding or old dry filter caps can spoil this.

[ginger]

Just try the 47k resistor to ground, before the 220k going into the PI

[Joel]

I seem to remember reading somewhere that bypassing the filter caps with something like 100nF caps can help remove high frequency hash on the B+. 

[Fiat_cc]

I think I’m going to pull the board, and go through the whole chassis from scratch. I also think I’ll move my under board links to above board. Suggestions for running b+ links above board in terms of routing? Keep it low and pressed against the board, and 90 degrees to all components?

I don’t really want to pull it out, but I’m seriously running out of ideas. I’ve found and fixed a couple of errors, but nothing so far has made an impact on the hiss.

[shooter]

pull the board, and go through the whole chassis from scratch
I linked one of Sluckeys build as a study guide.  it's NOT your ab763, but the "complexity" knobs jacks switches etc.  Once I did this one thing (plagiarize :), I went from a C- builder to B+ 

http://sluckeyamps.com/revibe/revibe.htm

take it to original 763 TMB amp only, no trem or verb or FX or..........
make it "your standard" after you already like it 

[bmccowan]

First, I'll second Shooter's suggestions. Build a standard AB763 first, and get it working right. We all know that circuit works, and works without a lot of excess hiss. And then make your additions and changes incrementally. The advice to follow Sluckey's layout strategy is good advice too.
I do not think you need to avoid going under the board with some of the leads, but you certainly can. I do run some under the board, but all the connections are above the board where I can access them. I just drill some 1/4" holes in the board near the connection location and pop the wire up through. If building a fender circuit I follow their layouts pretty closely.

[Gnobuddy]

In a good design, the hiss (thermal noise) for the entire amp will be dominated by noise from the input tube (triode), because that's where the signal is smallest. That isn't the case for this design, though.

In this case, R1, R2, and R7 are the dominant sources of noise in the entire amp, and that's where I'd start. Even if you'd used metal film resistors in all three places, the resistor values are so big that thermal noise from these three will dominate the noise of the entire amp - the resistors are noisier than V2B, which is just a terrible shame. (Yes, Leonidas used the 68k resistors, because he didn't understand thermal noise.)

So I would suggest replacing R7 with a 10k metal film resistor - this will still work as a grid stopper, but now it hisses no more than V2B, so at least it doesn't make the amp much hissier than it needs to be. As for R1 and R2, do you ever actually use the "Low" input? If not, get rid of them. Replace R2 with a wire jumper or zero-ohm metal film resistor, and remove R1 entirely.

The design of the boost circuit is suspect, too. With your 1M RV1 "Boost" pot at maximum, the 1M pot and 1M resistor (R6) combine to put an enormous source impedance of 500 kilo ohms in series with the grid of V2B. This is a lot of resistance, and it will add a lot of hiss. I would suggest lowering both R6 and RV1 quite a bit; maybe 100k and 100k respectively, and R6 should be a metal-film resistor. That shouldn't load down your boost stage too much, as they present a 200k load to V1 A/B, and since those two triodes are in parallel, they should have a lower-than-usual output impedance anyway (the internal plate resistances of the two triodes are in parallel, lowering the output impedance.)

Whether or not this solves your current hiss problem, it will ultimately make your amp less hissy, when you do find and fix whatever other problem might exist.

If it were me, I would also replace R5, R9, and R10 with metal film resistors. The signal is still pretty small at these locations, and it just makes sense to use the quietest resistors you can.

If you get this far, you've eliminated unnecessary thermal noise where it matters most, and the amp should be much quieter. If it's still too hissy, the next obvious culprit is very large-value resistors in series with the signal: in this case, R14, R25,R26, the mix network for the reverb. These should definitely be metal film; better, they should also be lower in value. (I know, Leo used 3.3M...again, he, or his tech, didn't understand thermal noise.)

Hope this helps!

-Gnobuddy

[Fiat_cc]

In a good design, the hiss (thermal noise) for the entire amp will be dominated by noise from the input tube (triode), because that's where the signal is smallest. That isn't the case for this design, though....

Hope this helps!

-Gnobuddy

Thanks Gnobuddy.
I understand everything you’ve mentioned in your post, but I have disconnected the rest of the preamp from the input to RV5, so the boost stage and V2B are out of circuit. I have also removed V1, V3 and V5. I have disconnected R25 from the junction with R26 and R14. I have then used a wire jumper to jumper directly across R14 to effectively take it out of circuit. I have replaced R13 and R27 with metal film. None of this has reduced my hiss. This is why I am getting frustrated! I’ve built higher gain amps than this, using the same type of resistors and even with the same parallel gain stage in front of a Marshall 2203 circuit, it doesn’t hiss this much. There is something causing this that is either a mistake that I haven’t found yet, or an unusual problem/component failure, but I can’t figure out what. I’m going to try replacing all of the diodes in the rectifier, as earlier on in trouble shooting, I found one in reversed, and whilst the other 3 all tested ok on my diode tester, perhaps there is damage there, and switching noise or another issue that are particularly noticeable on V2A and V4A. I have pretty much zero ideas after that.

[Gnobuddy]

...disconnected the rest of the preamp from the input to RV5, so the boost stage and V2B are out of circuit. I have also removed V1, V3 and V5. I have disconnected R25 from the junction with R26 and R14. I have then used a wire jumper to jumper directly across R14 to effectively take it out of circuit. I have replaced R13 and R27 with metal film...

Any chance you could post your schematic, with the changes above? That will make it easier for others to come up with troubleshooting ideas. Just use MS Paint or The Gimp or Photoshop or whatever image processing software you have to put big red Xs through the bits you removed, and draw a red line where you've added jumpers.

I can only imagine how frustrating this must be to you by now. Hang in there, we'll get it sorted eventually, one way or the other!

Your hunch about the diodes is certainly worth following up. I have had transistors go hissy permanently if the collector-base junction was damaged by excessive voltage. Perhaps something like that happened to one of your diodes. (But why don't the filter caps remove the hiss, in that case? Hmm.)

I'm going to be busy the rest of today, but I'll try to take a look tomorrow and see if I can come up with any more diagnostic ideas.

-Gnobuddy

[labb]

It might be worth your time to visit the www.ax84 site and do a search for "hiss". I remember that those guys did a lot of trouble shooting for a hiss problem.

[Fiat_cc]

Today is divide and conquer day.
I am going to disconnect B+ from everything besidesoutput section, phase inverter, and V4A. Any unnecessary grounds will also be disconnected, and any input into the grid, besides the ground reference, will be disconnected. The amp will literally be a gain stage, into the phase inverter.
If that hisses, there isn’t much left to look at. If it doesn’t, I slowly add things back in one by one until it does.

[Gnobuddy]

The amp will literally be a gain stage, into the phase inverter.

Sounds like a plan!

Is the attached image about right?

-Gnobuddy

[tubeswell]

Echo what others have said about the S:N ratio, particularly suspecting V1. It’s an advantage to have full bypassing here for this reason. Also see if reducing your 1M resistances in the boost circuit might help (e.g. to 470k (and 500k pot) or 270k (and 250k pot), or increasing the power rating of these components. Signal grid shielding may also help, and decent sized (10k to 33k) grid stoppers.

[Fiat_cc]

The amp will literally be a gain stage, into the phase inverter.

Sounds like a plan!

Is the attached image about right?

-Gnobuddy

That was right.
I went even further until it was literally just V4A, phase inverter, and power section. Grounding grid of v4A equals silence (besides minimal transformer hum), v4a with 220k grid leak equals hiss. I’m going to rebuild it. Something is amiss, and the time I’ve spent trying to trouble shoot it, I could have rebuilt it twice. I’m actually going to slightly rework the layout, use all new transformers and parts. Only recycled parts will be foot switch and face plates.
If this one works, I’ll rework the other one, and see if I can identify what’s causing the issue. If this new one exhibits the same issue, I obviously have a design problem.