School me on triode-strapped pentodes

[jacobarber]

The context here is that I have an Ampeg GVT-52, which I have struggled to get good drive sounds from in full 50w mode. No matter of fiddling with the baxandall EQ or gain results in a pleasing tone. It is always grainy, gritty, harsh, etc to some degree. I have a hard time describing it otherwise. It is almost like you can hear individual clipped peaks, like static.

I switch to half 25w mode and the drive channel becomes usable. It breaks up in a much smoother fashion, can even get closer to modern high gain (though not quite there), and just sounds so much nicer. That graininess might still be there a little bit, but it vastly overpowered by what generally sounds like pleasant overdrive.

I learned that this switch apparently changes the output of the 6L6GCs between pentode and triode modes. And digging further, I learn that this means it connects the screen grid to the plate. In other words, it straps the screen to the plate. Triode-strapping.

But what I don't really get is how this so drastically seems to change the distortion characteristics.

Now also understand that for my situation, I am keeping it at a loud bedroom level using the master volume. Not hush quiet: the neighbors might hear a little even with windows closed.

If interested, here's the schematic: https://ampeg.com/data/6/0a0005091d0c66204d39a3114/application/pdf/

Thoughts?

[shooter]

how old are the PA tubes?
what are they biased at?
how many sets of speakers did you try?


got a scope?


triode mode essentially puts the PA in "linear mode"  more like Class A, or Audiophile vs Class A/B guitarist, can't recall if it changes the harmonic content from odd to even.


by using the MV, most of your "sound" is produced in the PRE, not the PA.  a MV limits the drive signal to the PA

[jacobarber]

how old are the PA tubes?
what are they biased at?
how many sets of speakers did you try?

got a scope?

triode mode essentially puts the PA in "linear mode"  more like Class A, or Audiophile vs Class A/B guitarist, can't recall if it changes the harmonic content from odd to even.

by using the MV, most of your "sound" is produced in the PRE, not the PA.  a MV limits the drive signal to the PA

PA tubes are practically brand new. They have just a few hours on. I bought this amp used from GC, it arrived with extremely shattered power amp tubes. I refused to accept it and they had it sent to highly regarded local amp repair shop to have power tubes replaced (JJ 6L6GC) and biased, and the amp fully checked over. They biased according to the instructions described on the schematic sheet I posted but I have not double checked anything, as I see this shop as very trustworthy. I could definitely check though, and bias is extremely straighforward to set on this amp.

The amp originally came with the not well regarded Celestion 70/80. I did some A/B testing and swapped it out for an Eminence Cannabis Rex, which is whats in there now. I do think the speaker is still breaking in a bit, but it overall has sounded better than the Celestion.

And about using the MV, that's actually what sort of confuses me. Since I do have it relatively low I'm not pushing the power tubes all that much yet there is a SIGNIFICANT difference in the drive channel.

[Merlin]

But what I don't really get is how this so drastically seems to change the distortion characteristics.

Triodes have different-shaped grid curves from pentodes which makes them behave differently during overdrive, than pentodes.

• Triode mode will have heavier grid current which should increases the 'hardness' of overdrive clipping, and will probably produce more bias shift.
• Triode mode won't generate flyback spikes across the OT.
• Triode have more Miller capacitance, so HF filtering will be different.
• In triode mode you will be running closer to the class-A than in pentode mode.

I would suggest adding catching diodes to the OT, and increasing the PA grid stoppers (R95 + R97) to 10k, as this might bring your pentode tone closer to the tone tone.

[HotBluePlates]

... I have an Ampeg GVT-52 ... in full 50w mode. ... It is always grainy, gritty, harsh, etc to some degree. ...

I switch to half 25w mode and the drive channel becomes usable. It breaks up in a much smoother fashion ...

I learned that this switch apparently changes the output of the 6L6GCs between pentode and triode modes ...

Is it correct to say you do not have the Master Volume full-up in these situations?  And that you're using the preamp distortion generated by the Drive channel with the Master Volume turned down somewhat?

If "Yes" then the distortion is not "coming from" the output tubes being in Triode mode or Pentode mode.  You're simply not driving them hard enough for that to be the case.

What accounts for the differences is the EQ imposed by the power section:

   - Merlin points out "Miller Capacitance" of Triode-mode.  Put simply, Highs will be rolled off some.

   - Pentode mode will also interact with the Speaker's Impedance (which changes at different frequencies) to have more-output as frequency rises.

   - Negative feedback around the power section of the amp can counteract the rising Treble of Pentode output tubes.  This amp doesn't use power section negative feedback (unless I missed something in the schematic).  That means Ampeg is emphasizing the sonic-difference of Triode-mode vs Pentode-mode output tubes.


Overall, I think you're reacting to "EQ" rather than an actual change of how this tube or that tube is distorting.  I suspect the output tubes are not driven hard enough that they are changing their distortion-contribution, but that they change what frequencies are fully-amplified when switching between Pentode- and Triode-mode.

[jacobarber]

I definitely do not have the MV turned up fully at the levels I'm playing in my home office. What I SHOULD be doing is controlling individual channel volume with each channel's volume control and allowing the power tubes to breath more through more gain and more MV.

Originally I compared the sound of this amp to my old VT-22, and just couldn't come to terms with how utterly different they truly are. Even though Ampeg sort of tried to sell these GVT amps as spiritual successors of the V amps, they aren't anything alike aside from just having baxandall-ish tone stacks.

I do think the GVT amps completely lack a power amp negative feedback loop, where the V4/VT-22 did have one. I have spent time reading the schematics of the GVT5, GVT15, and GVT52. The 5 and 15 are easy to see a complete lack of negative feedback, where the 52 is a little more complex to follow. I don't think I could identify one in the end.

I've never paid attention to the differences between those two topologies, but I think I am slowly beginning to understand.

I am going to check the bias myself tomorrow if I get the chance just to make sure. It is always possible that the tech who serviced it was operating at a more broad level for 6L6GC tubes instead of this particular amp, but that's a long shot. Can't hurt to check.

[shooter]

with the amount of brother-in-law stuff going on in that pre, try jacking into the EFX RETurn and bypass 90% of the "junk".  might need a boost/gain pedal between guitar n FX-RTN.


that should give you direction on "Is it better"??

[jacobarber]

with the amount of brother-in-law stuff going on in that pre, try jacking into the EFX RETurn and bypass 90% of the "junk".  might need a boost/gain pedal between guitar n FX-RTN.


that should give you direction on "Is it better"??

'brother-in-law stuff' is a curious way to put it. Yeah, the preamp section is mighty busy. I'll give this a try tomorrow for sure. Can't hurt, just to see what the story is.

[shooter]

 
from my Navy days, where my system had like 40 circuits that at 1st "seemed normal" but once you started diggin' n measuring they were all basically useless to the system.  so I concluded it's just a way to scam the tax-payers, $29.95 for this "addon".....


I gutted them, which was a felony, got about a 30% overall improvement, the Weapons Officer took notice, He was a E.E., ordered me to CIC n I explained "off the record" what I found, what I did. He ordered me to make it original, asked how long it took, I told him I could make the changes either-way in about 10 minutes, he smiled n said; "If the bad guys are coming install your patch"

[jacobarber]

Fun update.

I'd played a lot with various iterations of channel gain, channel volume, and master volume prior to testing anything else. The story continued that the drive channel just sounded gross and I could not EQ that grossness out. It'd go away if I knocked the gain down far enough, but then channel 2 just sounded exactly like channel 1 with the gain turned up.

So I plugged directly into the FX loop return. Sounded fine. It was a purely clean sound, so the gritty grossness wasn't there. I didn't turn it up loud enough to get the power tubes to break up, only because my family is in the house and they would get quite upset with me.

The next step is to check bias. Fast-forward, I didn't check the initial bias value at all. I followed the instructions on the schematic and adjusted it accordingly.

Voltage across the cathode resistor of V4-A was set exactly to 0.45v.
Voltage across the cathode resistor of V5-A was measured at 0.478v, which falls between the specified 0.4-0.5v in the schematic.

And guess what. The amp legit sounds wayyyyy better. I honestly never thought that this specific shop would have mis-adjusted bias on an amp. But I am now thinking they did set it a tad cold maybe.

I don't know. Maybe its all a placebo. I should have checked bias before adjusting, but I might actually get along with this amp now. I really do wish my VT-22 was about 80lbs lighter so I could carry it to and from practice. Nothing is going to sound as good as that to me. This little GVT52 will have to suffice for now.

[HotBluePlates]

... The story continued that the drive channel just sounded gross and I could not EQ that grossness out. ...

The Drive channel is EQ'd to be very bright.

   - 1st Stage (V1-B) uses C34 that is 1/100th the size of C38 for the Clean channel.

   - 4th Stage (V2-A) uses a cathode bypass cap (C81) that probably strips out all lows & mids.

You could probably alter those parts to get a little more Beef.

... So I plugged directly into the FX loop return. Sounded fine. ...

Sure it did, because that is after all the preamp shenanigans.

I know "tubes are better" but you'd probably be happier adding a pedal to the Clean channel to get your "distortion" sound.  And there's a ba-jillion of them to choose from.

[jacobarber]

... The story continued that the drive channel just sounded gross and I could not EQ that grossness out. ...

The Drive channel is EQ'd to be very bright.

   - 1st Stage (V1-B) uses C34 that is 1/100th the size of C38 for the Clean channel.

   - 4th Stage (V2-A) uses a cathode bypass cap (C81) that probably strips out all lows & mids.

You could probably alter those parts to get a little more Beef.

... So I plugged directly into the FX loop return. Sounded fine. ...

Sure it did, because that is after all the preamp shenanigans.

I know "tubes are better" but you'd probably be happier adding a pedal to the Clean channel to get your "distortion" sound.  And there's a ba=jillion of them to choose from.

I dug in further and completely see what you're saying about that drive channel EQ. From what I can see, that .001uF cap coupling the V1-B plate to the channel 2 gain control is doing a similar thing to the Ampeg V4 ultra-hi switch set to boost. That control boosts presence frequencies. In the GVT though, there is nothing to control them at all. They're just on 10 all the damn time!

I went ahead and slapped the original Celestion 70/80 back in just to see if there's a big difference between that and the Eminence Cannabis Rex. The difference is there, but not gigantic.

I can still hear the presence frequencies quite strong, so I'm not completely out of the woods. I think if I want to make this amp 100% usable, I'd need to go changing component values in a few spots. But honestly, I need an EQ pedal too, and I think that would probably help fix 99% of this mess.

In the end I'm probably going to sell it and try to find something more friendly to my ears. I don't want to have to fight my gear like this.