Intermodulation Distortion on new build

[goldstache]

I built up a PP 6L6 amp with 2 6SL7's in the preamp.  The OT is in Ultra Linear operation utilizing a deluxe OT's screen taps.


Amp sounds great.  So great that I built one up with a GZ34 instead of the first builds SS rectifier.
I changed the Reservoir cap to 40uF instead of the 100uF I had in the SS version. 


Im not using a choke but instead a 1k5 25W resistor after reservoir on both builds. 


Here's the rub:


The SS rectified version was less noticeable, but there is a bit of "ghost noting" happening.  I didn't really obsess about it at first, hearing it as a bit of the harmonic complexity of the amp. 
Then, when I built the Tube rectified version with lower value filtering, I noticed it significantly.  Especially around the D note on the 7th fret of the G string and above.  Still sounds lovely, just the Ghosting is annoying and just not right.


There's a lot of gain in this preamp.  I will post a schematic once I get it scanned off.


I would like to get rid of the "Ghosting" but keep the tube rectifier.........


My B+ topology currently:
rectifier > 40uF (OT CT) > 1k5 (25W) > 40uF (LTP plates) > 20K (2W) > 20uF (V1 A&B plates)


-Should I use a 40uF reservoir and then implement different RC filters and larger filter caps to help the ghost notes?


-Supplant a choke in place of the 1k5?


-Is any of this stemming from the ultra linear operation ya think?


My OT is a fender deluxe UL rated for 25W with specs on how it can handle 30W, which is where I have the 6L6's Cathode Biased. Floating the Filaments from that cathode resistor voltage.


Its pretty quiet in terms of noise floor.  Could my lead dress effect this phenomenon?


Really love the sag, and could handle the increased low end stiffness with larger caps in filter section. 


I could use a hand, thanks everyone!

[Willabe]

Im not using a choke but instead a 1k5 25W resistor after reservoir on both builds.

Why such a big wattage dropping R?

[goldstache]

I had a pile at arms reach!

[goldstache]

Here are pics of the schematics as my scanner isn't working tonight

That goes for the 2-10K's in series.  I didn't have a 20k around.

The schematic reflects the SS build. 
I changed the 100uF to a 40uF when I went to the GZ34


Thanks

[Willabe]

I had a pile at arms reach!

  Good enough!   

[John]

I like using 6SL7s myself. One thing I've noticed is they do indeed tend to be microphonic, and not always obviously. I chased a barely noticeable buzz quite similar to what you have, and on the same strings and above. Finally thought of putting one of those silicone rings on the tubes and that was it. I guess the larger plate structure is more sensitive to frequency?


I bought a bunch of old, some used some NOS from a fellow in Tennessee, and have not had any trouble with those. And I'm  one that usually rolls my eyes at the "older is always better" thing. In this case I'd say it's right though -- in my limited experience with them anyway.

[goldstache]

John, thanks for the reply.  I too have noticed microphonics with 6SL7's.  I've tried several brands and to the same end.  Also it's built into a head configuration and is away from the speaker cabinet.  It has no audible sound when "chopsticked".  I don't think any vibrations are coupling to the amp via the cab due to it not being on top of it.  I will try to get a washer around it and see if it helps. 
Still feeling like its in my filtering section. 
thanks

[HotBluePlates]

I built up a PP 6L6 amp ... The OT is in Ultra Linear operation utilizing a deluxe OT's screen taps. ... I changed the Reservoir cap to 40uF instead of the 100uF I had in the SS version.  ... The SS rectified version was less noticeable, but there is a bit of "ghost noting" happening.  ...

If you look carefully at the Bassman 135 schematic, you'll see one extra stage of filtering than you're using. It's a lot like having a screen supply node that's not connected to the screens, because those are connected at the OT.

I'm thinking you have a higher level of ripple and/or less decoupling in the preamp. The ripple might have been hidden somewhat in the SS version due to the bigger first filter.

I'd suggest keeping the smaller filter for your 1st filter cap, and adding a cap between the 10kΩ resistors between nodes 2 & 3. Then shift the phase inverter power node to the "new B+ 2" at this added filter cap.

[goldstache]

tried the above.  helped but still noticeable.  Anybody got any other ideas? Seems more notable with a les paul rather than strat.  should i start tweaking preamp?

[shooter]

helped but still noticeable

did you leave it in?

try knocking your input .1uF cap down to .02 - .047. 
try moving your V1A 500pf from plate/cathode to plate/B+, also try changing it's value in both locations.

Is it in a narrow *note*/*freq* range or all over the neck?
move your filament CT from the cathode to a main chassis ground.

[goldstache]

Its currently wired with all mods in place.


I don't have a cap before V1A's grid anymore.


The ghosting happens on g string and above.


Ill try the pf cap in the places you described tomorrow!
And I'm a tad unsure about the suggestions moving the float voltage off the CT of filaments and attaching the CT to ground.  Would i just send that voltage divided through 100E resistors to the filament secondaries????????


Thanks

[shooter]

Would i just send that voltage divided through 100E resistors

your schematic shows a CT wire for filaments so no, you don't need the resistors, they're only used if you have no CT. 
Do you have any cap on the grids input?  you mentioned the problem was more noticeable with different guitar, that's where that suggestion came from.

[goldstache]

Im sorry, the schematic definitely implies that I do.  But i ditched it a while ago.  So no caps on the grid of V1A.


Im gonna try and wire it not in ultra linear mode tomorrow and see if that effects anything.


Or may have to re design the HT supply.  Pretty bummed its not working and I don't really want to over filter it.  The tube rectifier is less noticeable the more filtering I add.  So id like to track down the problem and deal with it elsewhere!


Thanks everybody

[goldstache]

I ended up with

gz34 > 60uF (OTCT) >1k5 > 40uF > 10k > 20uF (PI) > 40uF (V1a/b)

Its still there on single notes on the g string above 5th fret.

Tried the cap across the plate resistor of V1, different values. 

Seems like a lot of filtering to me. 
 
I also have 100k plate resistors in V1a/b.  1k5 on both cathodes.  V1a decoupled with 25uF and V1b decoupled with 10uF

theres got to be another way. 
 
changed the screen resistors to 1k.  same voltage on screens and plates, still in UL configuration

any other ideas?

[shooter]

any other ideas?

Is your preamp grounding *separated* from your PA grounding?

my last amp had what I called *tremolo*, found AC hiding in the pre-amp ground, my *fix* was a bigger bypass cap on V1A which you already have.  for fun you could gator clip another 25uf on V1a's cathode for testing, it will affect tone though.

[goldstache]

Well I'm using a JJ 40/20/20/20 cap can that share a ground tab, so I'm not able to use the Hoffman grounding.  Everything is local star, with OT grounded at speaker out jacks.

I pulled the 10uf cap on V1b and it's less noticeable.  It's still there but the gain lowered so of course the audible ghosting did as well. 

Kinda bummed.  Wish I could find the issue and solve it. 
Thanks for all the help! 
If anyone has any thoughts what so ever, please let me know!

[sluckey]

Try this... Use the JJ cap can for B+1 and B+2 nodes only. And connect the can ground lug and HT center tap to a PT bolt. Now use a separate cap for B+3. Mount it near the preamp circuit that it feeds. Separate your preamp grounds (including the cap for B+3 node) from the ground at the PT bolt. Connect the preamp grounds and the filter cap ground to a dedicated chassis bolt near the input jacks.

[goldstache]

Well I ditched the UL taps and went normal screen supply.  Problem gone. 
Sluckey thanks for the great idea.  I'm gonna play it a bit as is and see if I dig it.  Then if UL is indeed my design/sound goal.  I'll rewire the HT as you have prescribed!