Please Help If You Can?

[The Ballzz]

Hi Folks.
I've built an amp that is basically Marshall JCM800/2203, but with a cathode bias 6AQ5 power section. Everything up to the output of the PI is the same as per this schematic.

I have a very odd problem. The amp seems to sound quite nice, but is absolutely buried by the rest of the band, until they stop playing and then it is almost deafeningly loud! One clue is that the amp seems fairly bright/trebly.

I have and oscilloscope and a signal generator, but very little knowledge/experience at using them. I know I should/could do a bunch of studying, but I'm really just an old guitarist who wants to play and not trying to grow into being an amp builder/technician. I can follow schematics, instructions and direction and would so appreciate some one leading me by the hand here! I suspect that what is going on is some sort of high frequency (or even low frequency) oscillation, above or below the range of hearing and this oscillation is cancelling out a swath or swaths of the audible frequencies! I just need to learn how to find it/see it on my scope!

Thanks Folks,
Gene

[SoulFetish]

If you can draw up a schematic of the PI and complete output stage that would be helpful. If you designed the output stage, you could post expected and measured voltages on all the relevant nodes in the output stage as well (either way, voltage readings would be great).
A couple of questions:
Do you have a dummy load so you can scope the output?
Are there any other problems you are experiencing other than perceived lack of output?

One clue is that the amp seems fairly bright/trebly.

Sounds like you are describing a Marshall amp to me.

[frankdrebin]

what speakers are you playing through?
JCM800 are very bright and thin if not played full gain.
If you think there is an oscillation,just put your scope probe across the speaker,respecting ground connection,and you'll see any HF oscillation.
Cathode bias is not the best for punch,it has the tendency to bias the amp colder when you are clipping the power tubes,that lowers overall power.
Anyway an AC30 is a hell of an amp to play,but usually has 2 very efficient speakers on mid frequencies.

[Stevem]

Tell me your not running more than 250 volts on the plate of those 6AQ5 tubes?

[10thTx]

The amp seems to sound quite nice, but is absolutely buried by the rest of the band, until they stop playing and then it is almost deafeningly loud!

Does that mean the band is SO loud that it is drowning out your amp that is deafening loud?

with respect, 10thtx

[Phil_S]

Tell me your not running more than 250 volts on the plate of those 6AQ5 tubes?

6AQ5's are basically 6V6's in a smaller bottle. IMO, you might squeeze 40W out of them, but more likely it is 30W. That is probably not enough oomph to get over the band volume. By itself, dimed, 30W will sound loud. I'm thinking you need 6 or 8 6AQ5's to get that amp where you want it. If you are doing that, what's the point? You could just run a pair of EL34's. Your choice of speaker will also affect volume.

As to what Steve says, I'm not convinced you need to live with 250V. The spec sheet for 6V6 or 6AQ5 says the design Va max = 250V. I have a c. 1963 Gibson amp that runs about 300V on a pair of 6AQ5's and it actually sounds quite good. If we assume that amp was meant for 110VAC and I'm seeing 120VAC or a little more today, that suggests they designed it to run something between about 260-275V. The point is, there is some wiggle room on the voltage. If you are running those tubes at EL34 voltage, that's going to be a problem sooner rather than later.

[pdf64]

I get the impression that the power amp doesn't have sufficient power output to punch through the mix. I guess that the HT voltage may be too high for cathode bias to work to best effect, and power tubes will necessarily be biased far from class A, so the power output may be only ~20W and its overdriven tone thin and reedy.
If the HT voltage isn't too high, eg much over 400V, changing to fixed bias could fix both of those issues.

With regard to the 250V plate limit, bear in mind that back in the 50s, Fender didn't seem to take much notice of limiting voltages, eg ~410V on the plate and g2 of 6L6G, ~50V over the plate limit, and 130V over the g2 limit (as if no current limiting resistors there wasn't bad enough)
https://el34world.com/charts/Schematics ... _schem.pdf
https://tubedata.altanatubes.com.br/she ... 6/6L6G.pdf
And where credible info is available for Russian tubes, they often have a 250V plate and g2 limit, yet folks swear they are the best option for 600V Ampegs etc.

[The Ballzz]

Well Folks,

It's taken a while to gather up the info to answer the many thoughtful questions. I'm going to reply in the same order as the questions/comments came:

@SoulFetish,
Below is a schematic with measured voltages to each of 5 nodes: A=OT Center Tap, B=Screens, C=PI, D=Cathode Follower & E Is V1. Amp was fully on through a 16Ω (yes, I have) dummy load. I don't know how to use my scope to look at what I want to see. There is not really a lack of output of the amp, but something about it's tonality just won't cut through unless it is blazing loud! And no, the band doesn't play very loud. Typically a 5 watt Class 5 is loud enough and my 5E3 Tweed Deluxe needs an attenuator to not be overbearing. I also have pin for pin for each tube, but nothing looks weird, so I left it off. Although I don't have quite a whole schematic for the whole amp, everything before the master volume is identical to that Weber, 2203 I posted earlier.

The GR amps 6AQ5 TMB schematic copy.jpg

@frankdrebin,
2 different Marshall cabs. 1965B-4x10/G10-35L/8Ω & 1960B-4x12/greenbacks/16Ω I play many cathode bias amps (Tweed Deluxe, DSL20, etc.) and rarely lack any punch. This is very different.

FWIW, I'm 64 years old, been a steadily gigging guitarist/singer since 13, do the math. For all that time, I've earned my living either playing/singing or as a multi faceted sound engineer, including repairs, maintenance, fabrication of gear, etc, et c, as well as hauling it all around, setting it up and using it! It's only been the last 8 or 9 years that I've had the time to try my hand at building my own guitars and amps. While I can follow schematics, instructions, etc, and have good bench/soldering skills, I am not a really knowledgeable technician, by any means. I'm simply trying to pick up a few "useful to me" skills to help me achieve the goals of my tonal quest!

@pdf64 & All,
I tried an experiment. Given that I had just installed a Granger effEcts loop (already tried taking it out of circuit, no difference), I could try driving the effects return/power amp (just before the Master volume) from the pre-amp out/effects send of my DSL20. The power section performed well and sounded amazing! I then tried the pre-amp out/effects send of the 10 watt 2203 into the effects return/power amp in of the DSL20. The over trebly, non hearable artifact had returned. At least I seem to have narrowed it down to the preamp!

Can anyone lead me by the hand, step by step, thru the course of sending the proper signal into my amp and where to hook up and set my scope to view different frequency wave forms. Can I assume that I want to send full range/white noise thru the amp's input at a specific voltage level/signal strength and then set my scope to the expected/measured voltages at various locations throughout the amp?

Thanks Folks!
Gene

[pdf64]

Great response, shot my assumptions down in flames!
Your partial schematic show 5 HT nodes, but the Weber only has 4?
If it's built right, the Weber shouldn't be anywhere near as trebly as a real Marshall 2203, as it doesn't have the 1nF bright cap on the gain control.
If it's too trebly, try lifting C15, the 470pF treble peaker.

Regarding testing, I get the impression you need to develop your skills and confidence first. I suggest feeding a sine wave signal in the range 200Hz to 1kHz at about 100mVrms to 200mVrms to the input of the amp, with all controls at minimum.
Then see what the plate signal at the input stage looks like. Expect a sine wave about 3 - 6Vrms, set the scope input to ac coupled.
Then move to the 2nd stage plate and turn the gain control up a little.
And so on.
You'll probably need a 10:1 attenuation probe for the scope after the 1st stage.

[wpaulvogel]

Shouldn’t the shared cathode resistor be about 135 ohms?

[The Ballzz]

Great response, shot my assumptions down in flames!
Your partial schematic show 5 HT nodes, but the Weber only has 4?
If it's built right, the Weber shouldn't be anywhere near as trebly as a real Marshall 2203, as it doesn't have the 1nF bright cap on the gain control.
If it's too trebly, try lifting C15, the 470pF treble peaker.

Regarding testing, I get the impression you need to develop your skills and confidence first. I suggest feeding a sine wave signal in the range 200Hz to 1kHz at about 100mVrms to 200mVrms to the input of the amp, with all controls at minimum.
Then see what the plate signal at the input stage looks like. Expect a sine wave about 3 - 6Vrms, set the scope input to ac coupled.
Then move to the 2nd stage plate and turn the gain control up a little.
And so on.
You'll probably need a 10:1 attenuation probe for the scope after the 1st stage.

@pdf64,
Thanks for the testing tips, that will likely get me started! Please help me understand why full range, white noise would not be better for input signal injection. My uneducated thought was that an even level, full ranges signal would produce a display that might pinpoint some errant, transient frequency?

The extra node (easily removed) was part of an attempt to keep all other voltages fairly high, while running the 6AQ5 power tubes so low. Again as mentioned, during testing with my DSL20 preamp driving the PI this 6AQ5 power section sounded absolutely glorious! Much clearer and more harmonically rich than the cathode bias EL34 power amp section of the stock DSL20.

Please note that this actual Marshall schematic is nearly identical to the Weber, except for that one mentioned cap (unreadable value) near the "Preamp Volume" pot, just much harder to read. It is very handy & informative that it shows the various tested voltages for the main models of both 50 watt and 100 watt models. Please note also (to protect your sanity) that Marshall swapped the "physical" positioning of V1 & V2 between the Standard preamp and Master Volume preamp parts of the drawing, yet at least they kept the reference numbers (1 thru 4) for the voltages straight!

Thanks Again,
Gene

jcm800 50 and 100 std-page-001 2.jpg

[pdf64]

...Please help me understand why full range, white noise would not be better for input signal injection. My uneducated thought was that an even level, full ranges signal would produce a display that might pinpoint some errant, transient frequency?...

A scope's forte is displaying waveforms; yes, other signal analysis systems can process impulse responses or noise signals and get info out of them, but good luck trying to use a scope for that.

[The Ballzz]

...Please help me understand why full range, white noise would not be better for input signal injection. My uneducated thought was that an even level, full ranges signal would produce a display that might pinpoint some errant, transient frequency?...

A scope's forte is displaying waveforms; yes, other signal analysis systems can process impulse responses or noise signals and get info out of them, but good luck trying to use a scope for that.

Ahhhh, I see!! "Right tool for the job" kinda deal. Likely need something like a "Real Time Analyzer" or similar?

Hmmmm... I am a tool junkie!

I have had fun looking at wave forms at 800hz.

Thanks Again.
Gene

[The Ballzz]

So I'm able to inject a signal (for now I've chosen 400hz and varying the level between 0.1v & 1.0v) and with a 0.3v I have just enough level to push the second triode into clipping with the gain pot at about 8 to 9 of it's 0 to 10 travel. Funny thing is that the wave form only flattens out on the top and not the bottom. I'm kind of guessing that I simply continue through the signal path of the amp, adjusting the scope for higher voltage levels as I go along, as well as being mindful of where there may be DC present. What am I looking for?

Thanks?
Gene

[pompeiisneaks]

You want to keep it to maybe 400mV and say 1kHz. Then look at the signal throughout the amp and ensure at each stage it has gain to it. I.e. it goes up in AC Voltages. Try first to keep it a clean sine wave. Then see what you get at the speaker jack etc. Try pushing it into compression/heavy drive/gain. Then see what you get as well. See if somewhere there seems to be no gain or loss. If so that's potentially the stage with a problem. The only place you might see loss is at the tone stack, most amps have a recovery stage post tone stack to bring the levels back up again.

~Phil