Plexi noisy when playing bass strings

[rexindigo]

hi friends from this helpful forum,

my homebrew plexi clone starts to produce weird sounds: when I play a bass string or hit the upper strings hard there is a background hiss and crackle thats gone when I mute the strings. without any played notes there is almost no sound at all.
a (hifi) tube amp builder friend of mine suggested the nfb, but disconnecting didn't change anything (except the Presence control).
Coud it be Capacitors or the power tubes or the OT?
Any suggestions are appreciated, thanks in advance!

[mac1amps]

Hmmm, my educated guess is that you may have parasitic oscillation issues (an ultra-audio oscillation that "rides" on the guitar signal, resulting in a NASTY sounding interference, usually while the [guitar] signal is peaking). The fact that it goes away when the signal subsides is the biggest hint. Parasitics can be tricky to find, and oftentimes they may not even show-up on a 'scope unless you're using a "live" load (i.e. an actual speaker instead of a "dummy" load). When I've run into this in the past, it has many times turned out to be the result of poor lead layout (one of the reasons why many SF Fenders will have this issue IF the .002µF ceramic caps [installed at the power tube grid going to ground] have been removed-frankly many of the later 60's SF Fenders were wired pretty sloppy at the factory). I realize that you're talking about a plexi[clone], but the same principles apply. FWIW, I've re-built a number of SF Fenders, for the purpose of upgrading the workmanship, NOT so much the "design" (that's how I "Blackface" a later Fender), and whenever I've done this there's never been any parasitic issues.

[pdf64]

Agree that parasitics are likely diagnosis and that lead dress is probable cause. What schematic have you used? The 5F6A which Marshall copied didn't have grid stoppers on the power tubes. Best practice would be to fit them, 1k5 is the time honoured value, as they help stop this type of thing.
Also the 68k grid stoppers at the input stage are best fitted on to the tube socket terminals rather than on the circuit board.
Lead dress involves laying out the wiring while giving due consideration to what the wires are connected to, not just bunching them together to look tidy, or leaving them a mess just because you can. Don't run grid wires alongside other wires carrying large voltages or currents, and keep them well away from wires in the plate circuits of subsequent stages. Voltage gain and 2+ inverting stages = potential oscillator. Peter.

[rexindigo]

Thank you mac1amps and Peter,

I used this schematic:
http://www.drtube.com/schematics/marshall/1987pljp.gif
and combined it with this layout:
http://www.el34world.com/Hoffman/images/Plexi50.gif

Now as far as i can see there is no grid stopper resistors in my build.
I will try this with a 1k5 at grid1.

This will work as a very high frequency low-pass filter in conjunction with the input capacitance of the triode, am i right?


The 68k resistor in the input stage is directly soldered to the tube socket.

[pdf64]

Yes, the grid stoppers act as low pass filters in conjunction with the grid capacitence.
The hoffman layout shows 5k6 grid stoppers between terminals 5 and 6 on the el34 sockets. I would take that as good advice and do it like it's shown, use 5k6 as 1k5 is more of a 6L6 thing.
Try and find internal photos of other builders work, the hoffman layout is more of a general overview, whereas you may need to see the actual 'how to' detail.
It might be something else. If those grid stoppers don't work, check through postings from enzo for great trouble shooting tips. Spare known good tubes are a necessity. Peter.