JTM 45/100 build, pics, clips and some questions on ghosting

[VintageCharlie]

I just uploaded another clip of the JTM 45/100, this time rather a demo of the amp than a ghosting demo
This time the amp was set with almost all EQ's at noon, except the bass, which was at 0. Bright volume at 10. Sounds a bit smoother than the previous clip and i think i prefer this tone a bit. Recorded with an Epi Les Paul:

http://soundcloud.com/vintagecharlie/jt ... i-les-paul

The balancing resistors are the same what can be seen on dual filter caps in later Marshalls? - 10k 1w on the preamp dual 32uF cp and 47k - 56k 2w on the dual output cans, or is it something different?

Yes, the output filtering is as you guessed, but i'd resort to increasing the filtering or henries of the choke as last option - i love the squishiness of the amp, when it's dimed and higher filtering would probably make it a different animal.

[Structo]

That makes sense because bass frequencies take much more energy to produce.
That is why you see the car audio guys put huge caps in their car systems, to sustain the bass notes because the power supply (alternator and battery) can't keep up.

You can probably fix your problem by replacing the filter caps.

From the way those caps are setup in series parallel you only wind up with 32uF on the power tubes right?

Most 100w amps have twice to three times that value.

So I would expect that amp to sag pretty good but with four caps even though the total is only 32uF it may have a pretty good reservoir of current.

The problem with that setup is that if all four of those caps are not close to each other in value, that one or two may be taxed more than they should.

That is why the more modern amps used balancing resistors across each cap.
This equals out the voltage causing the caps to work as one large cap easier.

Values of 100K-220K are common for this use.

[pdf64]

The balancing resistors are 100k or 220k 1 watt in parallel with each 'stacked' cap; they help to set the Vdc across each cap, otherwise one may have 400V, the other 100V, across them. You may get away without them, but it's good practice to fit them.

[VintageCharlie]

I remember reading that for Marshall circuits something at around 220k would be best - i wonder why Marshall used 47k and 56k (2w) for their output dual filter caps - this would be quite a bit below the 100k that you guys mentioned as the lower recommended margin. This was also something that some guys at the metro forum were wondering about.

[overtone]

My old 1970s KT77 amps all use 100k resistors in that location. That is for a 220uF pair in series.
You could read up more at the bottom of this page:
http://www.freewebs.com/valvewizard1/smoothing.html

[VintageCharlie]

Great, thanks for the link!
Do you guys have a recommendations for a halfway idiot-friendly introduction to tube amps? At the moment i'm just at the level of monkey-see, monkey-do. I'd like to understand that stuff. I opened up the Radiotrons book and closed it very soon - out of my league. I'd be after something good, but that explains how things work from the very basics up to some more advanced stuff.

[overtone]

That web site is a good place to start. Merlin has written two good books, but you may find this one more useful at a beginning stage:
Design and Construction of Tube Guitar Amplifiers, Robert Megnatz, TacTec Press.
It has good clear language, a no nonsense overview. It takes you through load lines, designing stages, through to testing, kind of everything you could need. A good one to start with, and good value. From there you could progress to more in depth publications when you know what you want to pursue further.
Best, tony

[VintageCharlie]

Thanks for the tip! I had already looked into tube amp books (most of which have either absurd prices or have rather crappy reviews), but missed that one. Seems to be pretty much perfect for me, especially due to the included electronics basics, as i haven't had to do much with physics in the last 10 years.

[VintageCharlie]

One last question on ghosting (i suppose there is not much sense in discussing this any further before i have taken a dive into the amp) - is there a way to test if the ghosting starts in the preamp or the output section? Like taking the preamp signal into oscillator and look for the "ghost rider" frequency? This way i could find out fast if it is the NOS Erie cap in the preamp or something else.

[pdf64]

To my mind, ghost noting is signal related, ie when an AB power amp draws more current, 120Hz ripple on the B+ increases and modulates signal.
120Hz at the power tube screen grid, main signal at the control grid - the tube mixes them and all sort of weird frequencies are created at the power tube plate.
So it can't normally appear in the pre-amp, although power amp grounding could induce something signal amplitude related in the pre-amp.
Generally though, bad pre-amp B+ decoupling caps result in hum / motorboating. Pete.

[VintageCharlie]

Thanks, pdf64!

I better order the "Design and Construction of Tube Guitar Amplifiers" book soon and read up to get some understanding on all this.

Thank you very much for all suggestions and info guys! I'll report back as soon as i have teamed up with the guy who built this and made first attempts at solving the ghosting issue.

Best regards,

VintageCharlie