Derailment Build

[jarrodthebobo]

Wouldn't you use the negative feedback loop to minimise distortion and tighten bass response.

Yes you might be right. Although I never really understood what a tighter bass response really means when you play.

I always took it to mean less "flub"

[Mark]

I hear the bass response still being present but the amp has better control over the response of the speaker and it can't ring on beyond the note.

[Riffmonster]

Thanks Mark and Jarrod. I will let you know I will implement the NFB switch, however, I am quite busy these days.

[jarrodthebobo]

Thanks Mark and Jarrod. I will let you know I will implement the NFB switch, however, I am quite busy these days.

From my own experiences building a Single ended Marshall plexi/jcm800 thing, NFB is CRUCIAL to the overall sound of the amp as a whole. Id 100% give it a go and throw in a presence knob as well. I've found presence to be pretty important to the overall Trainwreck sound too.

[Riffmonster]

Thanks Mark and Jarrod. I will let you know I will implement the NFB switch, however, I am quite busy these days.

From my own experiences building a Single ended Marshall plexi/jcm800 thing, NFB is CRUCIAL to the overall sound of the amp as a whole. Id 100% give it a go and throw in a presence knob as well. I've found presence to be pretty important to the overall Trainwreck sound too.

So, which values would you use? 5k pot with a 0.1 uf cap like in the Express? And how about the resistor coming from the 8 ohm tab? It is 47k in the express. Colossal said earlier the original had a 100k though.
Thanks.

[Colossal]

Colossal said earlier the original had a 100k though.

Express is 100k NFB off the 8R tap into 5k/0.1uF Presence.

[Riffmonster]

Colossal said earlier the original had a 100k though.

Express is 100k NFB off the 8R tap into 5k/0.1uF Presence.

Right! My bad, I was looking at version A0.

[Mark]

I'm wondering how you would implement negative feedback and a presence control without messing with the cold clipping stage?

I'm thinking the negative feedback would have to go to a stage after the cold clipper that simulates the PI stage. The Tweed PI does have a certain tonal character.

[jarrodthebobo]

I'm wondering how you would implement negative feedback and a presence control without messing with the cold clipping stage?

I'm thinking the negative feedback would have to go to a stage after the cold clipper that simulates the PI stage. The Tweed PI does have a certain tonal character.

Exactly; in both builds of mine where I've built a single ended amp with negative feedback, I've implemented an extra triode stage to be a fake PI of sorts. I believe that if you use a 100k plate resistor and a 2.2k cathode resistor, unbypassed with a cap, you end up with close to the same gain you'd get from a PI.

[Mark]

I'm wondering how you would implement negative feedback and a presence control without messing with the cold clipping stage?

I'm thinking the negative feedback would have to go to a stage after the cold clipper that simulates the PI stage. The Tweed PI does have a certain tonal character.

Exactly; in both builds of mine where I've built a single ended amp with negative feedback, I've implemented an extra triode stage to be a fake PI of sorts. I believe that if you use a 100k plate resistor and a 2.2k cathode resistor, unbypassed with a cap, you end up with close to the same gain you'd get from a PI.

So you're hanging the 100K feedback resistor onto the 2.2K cathode resistor, but how are you implementing a presence control?

I had thought of using a 1K cathode resistor, stick with with 10K tail resistor andthe 5K presence pot on the end of the 10K tail resistor.

[Riffmonster]

Happy new year guys!
Very interesting thoughts from both of you. I have indeed one triode unused of one 12ax7, so I could implement that fake PI. But wouldn’t it exceed the overall gain? The amp has a lot of gain already.

Cheers

[jarrodthebobo]

I'm wondering how you would implement negative feedback and a presence control without messing with the cold clipping stage?

I'm thinking the negative feedback would have to go to a stage after the cold clipper that simulates the PI stage. The Tweed PI does have a certain tonal character.

Exactly; in both builds of mine where I've built a single ended amp with negative feedback, I've implemented an extra triode stage to be a fake PI of sorts. I believe that if you use a 100k plate resistor and a 2.2k cathode resistor, unbypassed with a cap, you end up with close to the same gain you'd get from a PI.

So you're hanging the 100K feedback resistor onto the 2.2K cathode resistor, but how are you implementing a presence control?

I had thought of using a 1K cathode resistor, stick with with 10K tail resistor andthe 5K presence pot on the end of the 10K tail resistor.

You run the presence control like you would in any amp; it's just where you put it that's different. The 100k feedback resistor goes to the cathode of the triode, then you tap the presence control off of the cathode as well

[jarrodthebobo]

Happy new year guys!
Very interesting thoughts from both of you. I have indeed one triode unused of one 12ax7, so I could implement that fake PI. But wouldn’t it exceed the overall gain? The amp has a lot of gain already.

Cheers

If you're pretty reserved in how much gain you give that extra triode, you should be able to get close to the same amount of gain a PI gives. If my calculations are correct; a PI typically has a gain of around 23, and a normal gainstage with a 2.2k cathode resistor, unbypassed, and with a plate resistor of 100k, has a gain of around 25; so close to the same ballpark. Although the clipping characteristics will be slightly different, it'll be better than building this without the fake PI, imho.

[Riffmonster]

Happy new year guys!
Very interesting thoughts from both of you. I have indeed one triode unused of one 12ax7, so I could implement that fake PI. But wouldn’t it exceed the overall gain? The amp has a lot of gain already.

Cheers

If you're pretty reserved in how much gain you give that extra triode, you should be able to get close to the same amount of gain a PI gives. If my calculations are correct; a PI typically has a gain of around 23, and a normal gainstage with a 2.2k cathode resistor, unbypassed, and with a plate resistor of 100k, has a gain of around 25; so close to the same ballpark. Although the clipping characteristics will be slightly different, it'll be better than building this without the fake PI, imho.

I think I got now what you mean. So I would still leave one triode unused of the both 12AX7?
Still, I am super happy how the amp sounds now. Especially for home use with an attenuator. Sounds better than my 18 watters and another Trainwreck with ECC99 power stage.
I might build it again with an EL34 and a FX Loop. Then I could consider the NFB. For now I leave it untouched, I think. Thanks a lot for all the input!

[Riffmonster]

I still haven't implemented the NFB, but I had the chance to a quick recording with an impulse response. Just give it a listen of you are interested!
Thanks