3stage + CF options

[Smokebreak]

Who knows which switch type the first switch is?
Which diode type is applied in parallel to the 1 Megohm potentiometer?

I've been through a bunch of scenarios with this internal jumper configuration switch/term strip(?)and I'm still confounded. Seems 1 connection is always getting grounded. I will say that the two input switches are missing a jumper from bottom left, to top right, otherwise the added stage is floating

[Smokebreak]

3 gain stage + CF preamps are tricky to get right. There are a few basic reasons why.

The 3rd gain stage + CF pair clips in a really unique, great sounding way. See here: http://www.valvewizard.co.uk/dccf.html for more info. In the standard arrangement, it will soft clip the negative side of the waveform seen on the grid of the 3rd stage.

Often, the 2nd stage (which may or may not be switchable, like it is shown in the emplexador schematic) is cold biased. This means that it will clip on cutoff rather than grid conduction - clipping the negative half of the waveform seen on its grid.

The problem here is that in addition to clipping one side of the waveform, the 2nd stage also inverts the signal. This means that the 2nd stage and 3rd stage + CF clip opposite sides of the waveform. Too-symmetrical clipping cancels out even order harmonics.

My favorite arrangement biases the switchable 2nd stage hot (around 730 ohms for the cathode resistor). This means that the 2nd stage will clip via grid conduction instead of cutoff (be sure to add a 100k-ish grid resistor to the 2nd stage), clipping the positive half of the waveform seen on its grid. After inversion, this already-clipped side of the waveform becomes the negative side of the waveform seen by the 3rd stage's grid, so the same side of the waveform actually gets clipped twice. In other words, MEGA even order harmonics. You will need a divider in between the 2nd and 3rd stages to do a significant amount of attenuation, which you should tune by ear.

Using a pre-PI MV and running the power section clean can get you a dumble-ish lead tone. Running the pre-PI MV high and then using a post-PI MV brings the PI clipping into the mix. The combination of the highly asymmetrically clipping preamp plus the symmetrically clipping PI gives you the super fat and crunchy cranked marshall sound.

By the way, this is exactly the same strategy used in the trainwreck express, except the express does it slightly differently. The 2nd stage in the express is really nothing more than a recovery stage to make up for the loss from the tone stack. The third stage is super cold biased, which gives the asymmetrical clipping. The power section is pretty much straight up 50 watt plexi.

See my post here which includes spectrum analysis plots justifying why an asymmetrically clipping preamp plus a symmetrically clipping power stage sounds awesome:

https://tubeamparchive.com/viewtopic.ph ... ht=#271577

This is great stuff. Thanks a bunch.

[stiltamp]

That is simple a pair of transistors with the emitters aimed at each other.

I think the switch is a simple DPST my eyes are worse than that file though.

Thanks!
So there are eight different input wirings possible.
I going to detail each wiring omit the switches getting the circuits across.

[stiltamp]

Who knows which switch type the first switch is?
Which diode type is applied in parallel to the 1 Megohm potentiometer?

I've been through a bunch of scenarios with this internal jumper configuration switch/term strip(?)and I'm still confounded. Seems 1 connection is always getting grounded. I will say that the two input switches are missing a jumper from bottom left, to top right, otherwise the added stage is floating

Thanks!

[gktamps]

I normally make the second or third stage switchable after having been through the same questions. If you are dividing equally the resonse is different from a 68k split vs 470k but equal division is still same voltage. 50:50.. so.

Are you talking about the 470k dividers in the first image below?

I dink with the mixers, but generally split each stage 470k/470k except the added stage gets mixed short on the signal side depending on how much signal I am looking for. I have run the ground sides as high as 1M and varied signal legs for response.

Sorry; again, are we talking about these 470k dividers? So, then, ground R could be up to 1M...

I have found it's better to use dividers for everything but that first stage then use the CF master and the normal master like on an 800. You are better to put the control of such a thing after the first stage or never hope for a clean sound.

As far as the CF and simple on the same dual pot, yeah you can make it all switchable.. BUT if you turn either up full they act as if they are bypassed for the most part. I never max the CF master. I always leave at least 5-8% rotation of headroom up there. Just a touch.

Are we talking about this master after the tone stack in the second image, or something different ahead of the CF?

Sorry for being dense - you guys know this stuff so well you talk in shorthand. I'm trying to soak it in!

Cheers,

Greg

[Smokebreak]

I'll give this a shot :

1. Kinda...I was asking about the difference between a 68K vs 470K divider, generally, and have learned the frequency response will change, but the same amout of signal will pass. Specifically, the Jose scheme has the added stage 1st, and the pic you posted from the 800 has the added stage 2nd. The divider you posted is between 1st and 2nd, but that can't be in the Jose, bc you'd have a 470k grid resistor in plexi mode. That's why the Jose scheme has 68K split there, in the same spot the 800 has the 470k split. I hope that made sense


2. Yes, for the 800 pic you posted(after "added" stage ). So 470K/1M would be "mixing the signal leg short"

3. The pic you posted is referred to as simple or normal master. The 800 doesn't have a CF master. Check out the Jose scheme that has both.

[Reeltarded]

The division at those mixers is a ratio of 100% of the signal that hits the entrance. If you are hitting the entrance of 1M total with 200v and the mixer is 900k/100k you get 20v. 10%. If you used a pair of 47k instead it would be 50/50, if you used a pair of 470k also 50/50.

The difference is total resistance in the first example is 1M, second is 94k, and third is 940k. That is where you can tinker with the too much hair syndrome of fuzzy defocused bass response by going down in total resistance, or ramp it up by giving it the beezness.

In the first example the stage is set as if there is a 1M volume control just barely cracked, the second and third act more like a 30% taper just beyond the middle.

You can go way higher than 1M. It isn't a limit... if you needed to do that. Been 6M8 on the ground side of a problem mixer before trying to keep it angry but only splitting off maybe 10-15% signal to feed with. Ehh.

Normal values that you see everywhere were arrived at by a process of time. I like the commonly seen variations best too.

If you build an amp with no clean sound you can make it have a clean sound by switchable paralleled signal mixer over that first stage usually. Calm down bitch!! hehe

[gktamps]

OK, thanks Smokebreak. That helps.

Interested to see which way you go with this. Sounds like you're not looking to pound salt into dust with the added gain, just get some flexibility to go from Plexi to a bit wild?

[gktamps]

Miles - posted my response just before seeing your explanation. Thanks for that. Starting to make sense, and the paralleled input mixer to the first stage is interesting.

Mad scientist.

[Reeltarded]

No no, after first stage mixer. Say switch a 470k over the 470k signal, and then add a 220k over the 470k/s with a second position on the switch.

One of my amps sounds like Satan playing with uranium. I cut that mix in half and now he's playing a blonde Bandmaster. Cut that in half and now he's playing the Weather Channel background. All just from that one leg on that first mixer. Whole different amp.

[gktamps]

Aha. Got it now - thanks!

I like it.

[Smokebreak]

OK, thanks Smokebreak. That helps.

Interested to see which way you go with this. Sounds like you're not looking to pound salt into dust with the added gain, just get some flexibility to go from Plexi to a bit wild?

Flexibility for sure, but I'm going for an army of elephants storming a herd of mammoths across a field of pumpkins stuffed with the wildfire

I thought about adding a 4th gain stage, and I've got a SLO type i blindly built from the straight up Avenger scheme, but for me, all these elements we're talking about get waaay more complicated, with their interactions and whatnot. I would have to straight copy something like that again, as they're terribly hard to get right, throwing the right amount of gain down the drain at the right spot. Maybe down the road, as high gain design is top o the pops in my eyes.

At least for this 3 stage one, I can copy from different designs, and actually see what elements I like, and hopefully actually understand them by modding. Plus I just need to build the prerequisite 6V6 plexi

I'm hoping to have something scribbled up today for review..

[Reeltarded]

hahaha you need a trunk to store that amp in.. a trunk.

[Smokebreak]

hahaha you need a trunk to store that amp in.. a trunk.

And Byzantine carved ivory knobs ta boot

[Reeltarded]

I want simple versions of everything made out of natural quartz crystals and platinum with meteorite face plates and transformer cores with quantum suspension for the carbon nanotube light traces holding the components in this dimension.

Working on it.