Advice needed...

[TriodeLuvr]

You know, this circuit has an interesting feature I didn't notice before. Do you see where the 470K resistor R8 is connected to the junction of R6 and R7? I saw that this morning and assumed it was an error in your reverse engineering of the amp. In a post I have now deleted, I advised you to ground that end of R8 instead, which is how the grid resistor of the lower output tube would normally be treated. However, I ran the circuit through a SPICE simulator a little later and discovered something unusual.

With the grid resistor grounded, output of the lower triode is much higher than it should be, roughly 10X as much as the upper triode. Changing the ratio of R6/R8 can correct that, but it doesn't explain what's going on. IOW, if the connection to R8 shown in the schematic is an error, why are the two triodes so badly unbalanced when its "corrected"? I'm sure you didn't get the R8 connection AND the values of R6/R7 wrong. I eventually realized that R8 is applying negative feedback at the lower triode. When R8 is connected as shown in the schematic, gain of the lower triode decreases and the PI becomes very well balanced with the values used for R6/R7.

So why didn't the factory simply use a different ratio for R6/R7 and ground the grid resistor as is usually done? As it turns out, this is a very effective technique to compensate for differences in gain within the two sections of the 12AX7. I proved this by temporarily substituting a 12AT7 for the lower triode. Amplification factor of the 12AT7 is 60, whereas the 12AX7 is 100. In a standard grounded cathode circuit, this creates approximately 5 dB difference in gain. That's also what should happen when the lower triode in this PI is changed to a 12AT7. Instead, because of the feedback circuit, the gain of the lower triode only decreased about 3/4 dB. The two sections of a 12AX7 would of course be much better matched than 60/100 even in a worst case scenario. That means this circuit really works well under any normal circumstance at improving PI balance. Between this circuit and the cathode feedback windings on the output transformer, this is (was) a pretty good hi-fi amp.

Anyway, I just thought this was worthy of mention so we both understand the circuit a little better. It could also be important to know in case you decide to unbalance the PI for an experiment at some future date. If you ever decide to remove this feature and ground R8, output of the two triodes can be reasonably well balanced by reducing the value of R6 from 47K to about 7.5K or 8.2K. That will put the two triode outputs within one dB of each other, and you can experiment from there.

Incidentally, I'm sorry I didn't see that upside-down wiring problem earlier. I did review your drawings, don't know how I missed it.

Jack

[zaphod_phil]

What schematic do these references come from? Could you please post it?

Thanks!

[TriodeLuvr]

It's the schematic the OP posted early in the thread. I think there's also a partial located one or two pages back that shows these parts.

Jack

[The4thWatcher13]

OK guys, I'm fried. I had no idea how much work building an amp, any amp, actually is. It's my first time building one and it's been a real learning experience in almost every area. I'm eternally grateful to all of you who have taken the time to assist, guide and critique this project. So now it's all wired up and the first thing I noticed was that it wasn't pulling snot for current. I had a 120VAC lamp in series with the mains. I removed it hence the hole next to the switch on the front. With the lamp gone I've got current all right..looks like it's pulling almost 2A at startup and the 1.25A fuse that was installed during the mock-up went south. I don't think it was the lamp!
So now, as Jack so kindly forewarned me about, begins the de-bugging stage. I'll start after dinner tonight but for now pictures, updated schematic and and a wacky wiring diagram!
(..don't ask about the reverse labeling on certain parts, it's a trauma I'm still dealing with to this day!) Back in a few!

[The4thWatcher13]

OK, already found one mistake: C6/C7 are supposed to be 0.01uF and I grabbed 0.001uF. ...lovely.

[The4thWatcher13]

C6 & C7 have been replaced with the correct value caps.
I also edited the previous post about the pilot lamp. I'll revisit that later. For now it's remains removed and the amp is still toasting 1.25A fuses. Therefore I'll continue inspecting the build.

[TriodeLuvr]

Congrats, that's nice clean point-to-point wiring!

I looked over the connections, but I don't see any glaring errors. Is that a 3AG (fast blow) fuse or a slow blow? Does the fuse blow as soon as you turn it on? Or does it blow after the tubes heat up?

Jack

[The4thWatcher13]

Congrats, that's nice clean point-to-point wiring!
I looked over the connections, but I don't see any glaring errors. Is that a 3AG (fast blow) fuse or a slow blow? Does the fuse blow as soon as you turn it on? Or does it blow after the tubes heat up? Jack

Thanks, Jack. It's a 20mm fast blow fuse. And it blows pretty much right away, the tubes don't even get to the glow point. I've been inspecting for solder bridges and other physical issues but haven't found any as of yet. I'll be going over the entirety of it again today checking for any mistakes in component values or wiring. The only thing different about this build compared to the last one is that the on/off switch on the tone pot isn't being used and the elevated ground for the filament CT to ground. As I mentioned before I had 1.25A fuse in it at mock-up, would there be any good reason that it'd pull just a little more current due to the elevated ground or just the new build?

[TriodeLuvr]

There are a number of tests you could perform to narrow this down. If you're sure the wiring is identical to the prototype, try removing the two output tubes. Don't let it run like this, just throw the switch to see if it still blows the fuse. If it's simply on the edge, this reduction in filament surge should be enough to tell you. Incidentally, I only use slow-blow fuses for mains fuses in tube gear. This lets you get closer to a safe rating without having to worry so much about turn on surge.

[JMPGuitars]

I also only use slow blow fuses for tube amp mains. Inrush current can easily blow fast blow fuses on otherwise fine devices.

[zaphod_phil]

Those three electrolytic caps, strapped together in a group, are a real no-no for hum control. They each need to be grounded to the main ground point of the specific stage that they are filtering power for - see Josh's advice on "Modern Grounding" viewtopic.php?f=4&t=25372
It also appears that your OT and PT are lined close up, in parallel, which will likely cause the OT to pick up hum from the PT. Search online for "the headphone trick", to help you optimize the position of the OT relative to the PT

[The4thWatcher13]

There are a number of tests you could perform to narrow this down. If you're sure the wiring is identical to the prototype, try removing the two output tubes. Don't let it run like this, just throw the switch to see if it still blows the fuse. If it's simply on the edge, this reduction in filament surge should be enough to tell you. Incidentally, I only use slow-blow fuses for mains fuses in tube gear. This lets you get closer to a safe rating without having to worry so much about turn on surge.

Thanks , Jack. I'll give it a go but I may not have any slo-blo 1.25A fuses. I can probably pick some up tomorrow at the HW store. Hopefully...

[The4thWatcher13]

Those three electrolytic caps, strapped together in a group, are a real no-no for hum control. They each need to be grounded to the main ground point of the specific stage that they are filtering power for - see Josh's advice on "Modern Grounding" viewtopic.php?f=4&t=25372 It also appears that your OT and PT are lined close up, in parallel, which will likely cause the OT to pick up hum from the PT. Search online for "the headphone trick", to help you optimize the position of the OT relative to the PT

I think I get what you mean, Phil. Each cap get's it's own lead to the "star" ground. Thanks for mentioning Josh's post on grounding. I can also use the info there to make a change to the input jack/pot(s) grounding which needs to be "as near to the input jack(s) as possible". I forgot about that. In regard to moving the OPT I'll have to come back to that after this other stuff. Thanks, Phil. I appreciate it.

[TriodeLuvr]

I think the two transformers are far enough apart so it won't be problem. The 45º angle will help somewhat too. If it's an issue, there are ways to deal with it other than moving the transformer.

Jack

[JMPGuitars]

Each cap get's it's own lead to the "star" ground.

No, that's not what he means. Only the cap(s) for the power amp section go to the main star ground (a shared lead is fine if they're going to the same place). The other cap(s) go where they're filtering. For example, it looks like two of the caps are filtering the power amplifier, and one is filtering the preamp. That preamp filter cap should be grounded in the preamp section, along with the preamp, pots, and input jack.

Thanks,
Josh

[TriodeLuvr]

No, that's not what he means. Only the cap(s) for the power amp section go to the main star ground (a shared lead is fine if they're going to the same place). The other cap(s) go where they're filtering. For example, it looks like two of the caps are filtering the power amplifier, and one is filtering the preamp. That preamp filter cap should be grounded in the preamp section, along with the preamp, pots, and input jack.

Thanks,
Josh

That's a little confusing. All those caps must somehow have their negative terminals connected to the negative output of the power supply. That's the return path for the DC differential. A ground loop can be created if the caps share the same conductor, whether it's a single wire or the chassis. However, running a separate wire from each cap's negative terminal back to the star prevents that. Once that's done, the negative terminal of the preamp bypass cap serves as a local ground. All grounds within the preamp circuit should be connected there. I think the OP has done this correctly and it won't be a problem.

Jack

[JMPGuitars]

That's a little confusing. All those caps must somehow have their negative terminals connected to the negative output of the power supply. That's the return path for the DC differential. A ground loop can be created if the caps share the same conductor, whether it's a single wire or the chassis. However, running a separate wire from each cap's negative terminal back to the star prevents that. Once that's done, the negative terminal of the preamp bypass cap serves as a local ground. All grounds within the preamp circuit should be connected there. I think the OP has done this correctly and it won't be a problem.

Jack

Yeah, it can be confusing discussing it in text, that's why I have the diagram linked in the grounding thread.

[The4thWatcher13]

\Each cap get's it's own lead to the "star" ground.

No, that's not what he means. Only the cap(s) for the power amp section go to the main star ground (a shared lead is fine if they're going to the same place). The other cap(s) go where they're filtering. For example, it looks like two of the caps are filtering the power amplifier, and one is filtering the preamp. That preamp filter cap should be grounded in the preamp section, along with the preamp, pots, and input jack. Thanks, Josh

[My page loaded slowly so I missed Jack's replay to Josh.
Below is my reply to Josh's correction of my interpretation.]

I think I get what you're saying. I've got three caps;
C3-A is a smoothing cap and only connects to the next two caps, C3-B goes to the power amp section (Plate winding CT, Screens & filament CT for ground elevation)
C3-C is both preamp tubes plate supply. So C3-C gets grounded with the pre tubes and input on a separate ground "up front".
C3-B (power amp)goes to the main star ground so would the smoothing cap also go to the main star ground or can it be strapped to C3-B?

That's a little confusing. All those caps must somehow have their negative terminals connected to the negative output of the power supply. That's the return path for the DC differential. A ground loop can be created if the caps share the same conductor, whether it's a single wire or the chassis. However, running a separate wire from each cap's negative terminal back to the star prevents that. Once that's done, the negative terminal of the preamp bypass cap serves as a local ground. All grounds within the preamp circuit should be connected there. I think the OP has done this correctly and it won't be a problem. Jack

The way Jack describes it is how I actually reconfigured the PS ground scheme last night. All three caps have a separate lead to the star ground and the input jack is to be grounded at a separate point. But Apparently, from Josh's viewpoint, I made a mistake in my interpretation, and therefore now I'm rather confused. So I defer to my reply and question, to Josh, about what he was describing in his post above. [The schematic below hasn't been changed: it's just there for your convenience]

[JMPGuitars]

But Apparently, from Josh's viewpoint, I made a mistake in my interpretation, and therefore now I'm rather confused. So I defer to my reply and question, to Josh, about what he was describing in his post above. [The schematic below hasn't been changed: it's just there for your convenience]

The schematic doesn't need to change, it's simply the location of the preamp filter cap ground that moves, which is a layout change only.

As Zaphod_Phil mentioned previously, look at my grounding scheme thread (link in my signature).

Jack mentioned the grounds all returning to the OT negative terminal (which connects to the star ground). But this only applies to the power amp section. Since the preamp filter cap is filtering the preamp, and the preamp is grounded elsewhere, you're creating a potential ground potential by not grounding that filter cap with the preamp. This means that you put your amp at risk for parasitic oscillations and unwanted noise. Sometimes you can get away with the star ground as you have it, but it's not a good idea. A PTP build like this, you need to take precautions to avoid noise, as PTP can be much harder to deal with than using a circuit board.

This is the layout referenced in my ground thread: files/JMPGuitars_18_Watt_Lite_2b_Layout.pdf

You can see how the preamp grounds all go to the input jack ground terminal, which is then connected to it's own grounding point.

Thanks,
Josh

[TriodeLuvr]

Since the preamp filter cap is filtering the preamp, and the preamp is grounded elsewhere...

Grounded elsewhere? It shouldn't be. All the preamp circuitry, including the sleeve of the input jack, should be grounded only to the preamp bypass cap negative terminal. Grounding in more than one location is practically the definition of a ground loop.

Jack