Trainwreck Hiss

[Firestorm]

We've been over this ad infinitum, but I want to start an "experimental thread."

There's no reason an amp should hiss this much. So what can be done (besides sorting through different $100 NOS tubes)?

Increasing plate resistor power ratings can only reduce one type of resistor noise (but it's a start). Dropping resistors in the power supply can be very noisy -- who out there as found the quietest 9K1 resistors to use?

Does anyone else think the 1M pots in the tone stack add noise? I ran The Duncan tone stack tool with Express values versus Fender values and found only slight differences. Anybody try 250K pots instead of 1M? How did it sound?

I know I'm grasping at straws here, but it would be very cool to design an amp that sounded like a Trainwreck, but didn't hiss (as much).

[bluegate]

I got mine tamed.
Originally there is one shielded cable from input to the grid V1.
Mine has a lot more, I replaced all wires from the board to the lugs of the pots (except those connected to ground) with shielded wire, and from the PI to the PPIMV.

[j-po]

We have one unused triode, right. How about doubling up the first stage with two parallel triode sections and also increasing the load twofold - tone stack with 50k slope resistor, 1n, 47n, 47n caps, 100k, 100k, 10k pots and a 500k volume pot.

Any amp with the master volume on full or no master volume will hiss though.

[Jana]

If you're going to experiment, maybe try and push it and see how far you can go. Try wiring the first stage as a cathode follower, that will drop the impedance much more than just paralleling the tube. Then, change the tone stack to 10k slope, .0047uf for treble cap, .22uf for mid and bass caps, 100k for bass, treble and volume pots and 5k for mid (2.5k is the scaled value).

[skyboltone]

Mine don't hiss. I think it's 99% lead dress. Sorry.

[Firestorm]

Sky

Seriously, NO HISS? You must be only one ...

[geetarpicker]

They all seem to hiss some. However once I did a comparison in the studio between a '68 Marshall plexi 100 superlead and my original '89 Express. I set the Marshall on 10, with a somewhat standard EQ setting. Then I set my Express for a similar gain and tone which was mostly everything on 1/2. With the amps set for a similar tone and gain the Marshall was actually NOISIER than the Express. Yes the Express was noisier when cranked further, but alot gainier. I think the Express is actually not that noisy IF you consider how much gain it has. Most folks nowdays are used to kicking off gain pedals are/or switching channels and such when they are done with the high gain and then they of course hear a lower noise floor. With the Wreck you are simply cranking the thing into high gain territory and leaving it there all the time. I don't think there really is any way around the noise, though in a clone proper lead dress seems to make a big difference. In my original Express and 2 clones I built the only shielded cable is from the input jack to V1. I think sometimes shielded cable can actually hurt the tone and even add noise issues in high gain tube circuits. I've seen some amps where the shielded cable loaded down the high end so much that it sounded bad, and had less noise simply because it had less high end. In most of my videos I did edit out the noise at the start and endings, however in the middle of the track the noise is there but barely a problem. Another good example is my "New Mexico" Youtube video where I set my amp for a fairly high gain (all knobs on 1/2, full bright switch) and used my Gretsch. I then rolled the guitar down to about 3 for a clean sound, but the amp is still cranked. Even with the guitar backed way down the noise isn't that bad. I only gated it off in the beginning before I hit the first note, but then it's left on.

http://www.youtube.com/watch?v=-U36Uxwi ... annel_page

[jjman]

Jana helped me understand the some of the hiss could come from stage 2 even if there is no hiss when the volume is on zero. One check is to turn the B,M, and T controls to min then set the volume to 50%. If you have no hiss at this point, only stage 1 would need to be targeted for possible hiss reduction. Any hiss you do hear at this point is NOT from the 1st stage. This is because the BMT at min is blocking the 1st stage buy grounding it’s output. Now, turn up the B,M and T and note that the INCREASE in hiss is from stage 1.

I was able to verify that my 3watt MF plate resistor on stage 1 is pretty quiet and that the CF one on stage 2 is not a significant source of hiss in mine. I would like less hiss but I’ve done all I can think of on that quest. If I wasn’t using the unused stage elsewhere I would try the parallel idea since lowering the volume pot’s value would address the root cause of the stage 2 hiss Jana identified in the other thread.

[skyboltone]

Sky

Seriously, NO HISS? You must be only one ...

Ok, exceedingly smallish, tiny, non-annoying hiss. But remember, without my aids in I'm down 70db at 2K. With them I'm in pretty good shape. Dartanion's amps don't hiss either so I know I'm not alone. I would be willing to bet that the other Halls build TW that don't hiss either. I sometimes use little strips of sticky aluminum flashing tape, to tape down NFB, OT primary, second stage grid. I also make sure to keep the leads back to the board from the bass pot parallel from start to finish. I will sometimes slip a 1" peice of non shrunk shrink tube over them to keep them together.

Gearhead invented the "build the tonestack outside the chassis trick" to keep that pot of worms from taking off. Just make a template out of 1/16" FR-4 using the chassis front as a guide and then assemble it all of a piece outside the chassis, then stick it in there.

Make it look like the Francesca picture.

[Structo]

Sky

Seriously, NO HISS? You must be only one ...

Ok, exceedingly smallish, tiny, non-annoying hiss. But remember, without my aids in I'm down 70db at 2K. With them I'm in pretty good shape. Dartanion's amps don't hiss either so I know I'm not alone. I would be willing to bet that the other Halls build TW that don't hiss either. I sometimes use little strips of sticky aluminum flashing tape, to tape down NFB, OT primary, second stage grid. I also make sure to keep the leads back to the board from the bass pot parallel from start to finish. I will sometimes slip a 1" peice of non shrunk shrink tube over them to keep them together.

Gearhead invented the "build the tonestack outside the chassis trick" to keep that pot of worms from taking off. Just make a template out of 1/16" FR-4 using the chassis front as a guide and then assemble it all of a piece outside the chassis, then stick it in there.

Make it look like the Francesca picture.

How do you ground the aluminum tape since the adhesive is not conductive?
That gives me an idea of using the Stew Mac copper shielding tape that does have conductive adhesive as well as having the ability to solder to the copper directly as well.

[skyboltone]

I burnish it with a polished rod. My guess is that it presses through the adhesive at some point. Anyway, a check with the VOM on low ohms setting shows it's making contact. But I don't really use it as shield, just hold down. Strips about 1/4" wide by 1" long, perpendicular to the direction of the wire. Just to keep it good and flush to the chassis.

[j-po]

If you're going to experiment, maybe try and push it and see how far you can go. Try wiring the first stage as a cathode follower, that will drop the impedance much more than just paralleling the tube. Then, change the tone stack to 10k slope, .0047uf for treble cap, .22uf for mid and bass caps, 100k for bass, treble and volume pots and 5k for mid (2.5k is the scaled value).

I think the point of paralleling is not only dropping output impedance. The noise (random) from two parallel stages will be fighting each other to some extent while the output signals are in phase, thus better s/n. Of course less resistance in the tone stack to pass through helps also.

[xk49w]

I think the point of paralleling is not only dropping output impedance. The noise (random) from two parallel stages will be fighting each other to some extent while the output signals are in phase, thus better s/n. Of course less resistance in the tone stack to pass through helps also.

Random noise sources add as square-root of sum-of-the-squares fashion.

I'll plunge in. It looks like you should be able to get 3dB improvement by paralleling two tubes where each is run at the same current as the single-tube case. This works if the noise source is primarily shot noise. Shot noise varies with the square root of the current. The parallel tube case will double the current allowing you to halve the plate load resistance while keeping the same stage gain (doubles the effective Gm?). The signal current from the pair is doubled but the shot noise only goes up by sqrt(2), hence 3dB improvement.

I'm sure there are other factors. I'm just a beginner at this tube noise model stuff.

[Rimy]

i just compared my express clone to a mesa stiletto trident.
both set up with the same gain. the express was not even near the noise level coming out of the mesa!

[Tubetwang]

They all seem to hiss some. However once I did a comparison in the studio between a '68 Marshall plexi 100 superlead and my original '89 Express. I set the Marshall on 10, with a somewhat standard EQ setting. Then I set my Express for a similar gain and tone which was mostly everything on 1/2. With the amps set for a similar tone and gain the Marshall was actually NOISIER than the Express. Yes the Express was noisier when cranked further, but alot gainier. I think the Express is actually not that noisy IF you consider how much gain it has. Most folks nowdays are used to kicking off gain pedals are/or switching channels and such when they are done with the high gain and then they of course hear a lower noise floor. With the Wreck you are simply cranking the thing into high gain territory and leaving it there all the time. I don't think there really is any way around the noise, though in a clone proper lead dress seems to make a big difference. In my original Express and 2 clones I built the only shielded cable is from the input jack to V1. I think sometimes shielded cable can actually hurt the tone and even add noise issues in high gain tube circuits. I've seen some amps where the shielded cable loaded down the high end so much that it sounded bad, and had less noise simply because it had less high end. In most of my videos I did edit out the noise at the start and endings, however in the middle of the track the noise is there but barely a problem. Another good example is my "New Mexico" Youtube video where I set my amp for a fairly high gain (all knobs on 1/2, full bright switch) and used my Gretsch. I then rolled the guitar down to about 3 for a clean sound, but the amp is still cranked. Even with the guitar backed way down the noise isn't that bad. I only gated it off in the beginning before I hit the first note, but then it's left on.

http://www.youtube.com/watch?v=-U36Uxwi ... annel_page

Thanks for the comments Glen!
Always interesting!!

Great video + nice playing!!

How do you manage to have such a low volume with the amp set at half and guitar volume at 3?

My Trainwrecks are way too loud at half volume and guitar on 3...

Do you change V1-V2 and V3?

Thanks