Problem with this fizzy sound

[diagrammatiks]

not optimum. sensical.

the slo makes perfect sense.

chuck and steve why don't you guys plot the lines and I'll explain my changes.

[sickman82]

No it doesn't. You'll hear the same trailing fizz with MV cranked in the clip I posted above, just in case you missed it.

Thatīs interesting, Sickman and I were experimenting the same fizz issue, but it became less noticeable with the MV up, not even close to your clip with the MV dimed, in fact I get that kind of fizz with the volume quite low. Maybe it has something to do with using a load box instead of a speaker.

This is where it gets a bit confusing. I have scoured the internet for any information I can find on this topic. If I remember correctly it was a thread on The Gear Page, and some NT owners claimed they didn't have the problem at all. Some claimed it went away with the MV turned past 1/2 o'clock. Others said that the problem was there wherever the MV was, like yourself, Deltones.

I remember reading another thread, this time on the Vox forum I think. A guy over there said that he complained to Vox, they accepted it back for a repair but it came back exactly the same. So, he got on to Vox again, they then sent him a brand new replacement which he claimed didn't have the problem.

So, although Silvio and I have the same issue.. Deltones doesn't, and from spending quite some time searching for as much information as I can on the internet regarding this issue with the NT, it is fairly clear that the problem is inconsistent - providing the people on the forums are reliable.

That's a lot of changes you propose. Wouldn't even be the same amp. I'm pretty sure the guys (and gals) who buy the Vox Night Train did listen before buying. Some things other than the basic tone do come into evindence post purchase. That's why we're trying to squash the trailing ratty fizz without drastic alteration.

Well said. When I bought the NT, in November 2010, I was trying it out at full tilt.. my local guitar store has a sound proof booth for testing out amps. I wasn't even considering bedroom levels, and even if I was, the chances are that what would be considered bedroom levels in a guitar store would be too loud at home anyway. So like Chuck says, problems like this only arise when you get the damn thing home. Or in my case, for a large part of me owning the NT, I was only using it live and at rehearsal.. until my band broke up in autumn.

Also, a metal guitar could have very high output and drive the amp into areas that the designer never explored (they probably test their creations with a standard issue Strat and Les Paul)

I play a standard Strat, a standard Les Paul, and a standard Tele.. all pickups are vintage style, I think the hottest is in my Tele at about 9k. There isn't a metal bone in my body! I play blues mainly, and a bit of rock. I see what you mean though, and it is a question I posed to Vox when I started getting problems with pedal interaction with the NT. I asked whether the NT was designed to take pedals, as with all my previous amps I have never experienced this issue. Their response was along the lines of 'of course we design the amp with pedals in mind, it must be your pedals that are the problem'. I don't think so. I couldn't even be bothered replying to the guy.

It's most noticeable in simple tube circuits that try to make a lot of gain in a few stages..." VERY TRUE!

Yeah, I'll second that. It is also an obvious factor when considering pedal interaction. I like to use fuzz, for Pink Floyd type stuff.. and I also play a lot of Hendrix. In my past experience with tube amps, it is desirable to have a little break up from the amp when using fuzz, or any other boost/overdrive pedal for that matter. If you set the NT to anything but the cleanest of cleans, it sounds like utter crap. Woe betide if you try and set the pedal up past unity gain. This again, is something I only learned after taking the damn thing out of the shop.

You're not going to get a Twin tone out of one even if you got Keith Richards to play it.

With all due respect, you're completely missing the point. The tone of the amp is great, I love it and I realise it isn't a Fender Twin. But there is an issue with it.. that much should be obvious. I'm not sure how you cannot hear that, especially with Deltones sound sample. I would think that with better speakers, it would be more obvious. I listen to everything at the computer through a set of Grado headphones, top notch, and it is really bloody obvious! Again, all due respect.. I'm not trying to pick a fight here.

I guess what I'm trying to say, in my non-professional experience, is that I've run into this problem a good deal and the only way I ever seem to be able to fix it is by stepping back, taking a breather (and maybe a drink or two...) and revisiting the amp - *all of the amp, not just the wiring or a single resistor or capacitor value* - in its entirety.

Very true. I'm sure that what you said about reducing gain in each stage would be very beneficial to both the fizz issue, and the issue with pedals. To be honest, a reduction in gain would be a welcome thing for me.

I'm very much up for making a project out of this, providing we have a logical path to follow. Silvio, Deltones.. are you up for it? Is everyone else, with the technical know how willing to help us out on the journey?! It would be cool to get a solid solution to the problem, without altering the tone of the amp - yep, I realise that reducing the gain is a fairly big alteration, but hopefully not to the tone of the amp. I never use the 'thick' mode anyway, unless I'm in a Texas Power Blues mood! But I'd prefer to use pedals for that anyway, seeing as I build pedals for a hobby!

On the subject of valves. I warmed them up for a few hours last night and had a jam pretty late on. It sounds great, it obviously needed a revalve.. I have hit it pretty hard over the last year or so. It runs much quieter too. The JJ EL84's are big and bold sounding valves, it gives the amp a nice solid bottom end. I found the stock Sovteks pretty flabby. The EHX ECC83's are ok, I didn't really see much of a difference in the nature of the preamp. I am yet to try out the ECC81 in V4. I will give that a shot tomorrow hopefully.

Hope everyone is well. Cheers!

[diagrammatiks]

there's no way that the amp was designed or optimized for pedals or hotter pickups.

again, the load line for the first stage makes it very apparent.

a hot enough pickup will clip the first stage pretty hard.

[Steve Conner]

not optimum. sensical.

the slo makes perfect sense.

chuck and steve why don't you guys plot the lines and I'll explain my changes.

The SLO makes a lot more sense than you're making. What do you mean by "lines"?

I'd guess load lines, but who ever used those in guitar amp design? They're for hi-fi amps. Guitar amps are designed by ear, the math is just to check whether the thing will blow up.

[diagrammatiks]

I really hope you are being factious.

[Steve Conner]

Well, I've designed loads of amps by my method and none of them fizzed as bad as that. I doubt Chuck draws load lines either.

So no, I'm deadly serious and I still want you to explain what on earth "Class-A Saturation" is and why it's a bad thing.

I was wrong about the single-ended thing though, I got confused with the Lil Night Train that I saw in my local guitar store.

[diagrammatiks]

how do you help people regarding an amp that you don't have in front of you without quantifying it's operating parameters?

[Steve Conner]

Well why don't you explain how you quantified the parameters and what your recommendations are, instead of posting a list of changes to a schematic that might not even be the right one, according to some earlier posts.

I wouldn't have the confidence to go into that much detail unless I did have the amp in front of me, so I tried to offer some more general advice.

Earlier someone mentioned that a grid stopper in some place was helpful. That seems like a simple mod to try.

[Deltones]

I'm very much up for making a project out of this, providing we have a logical path to follow. Silvio, Deltones.. are you up for it?

Well, the help I can bring has unfortunately nothing to do with the technical know-how you guys are displaying with amp circuits. I'm not even a noob in that respect. Where I can contribute is by testing different cab/mic configurations with the Torpedo because I'm not limited in how far I can push the Night Train volume-wise. Unfortunately, since most guys here who have the Night Train play will a real cab (at least, I think), I'm not sure if my tests are going to help much.

As reference, on the clip I posted previously, I used the Torpedo's H&K Triamp cab (Greenback) paired with a M160 ribbon. mic.

[diagrammatiks]

Sure.

Starting from the input.

- There's a ferrite bead, capacitor network there that serves to keep the amp quiet when nothing is plugged into the input. You'll see that it filter/grounds the driver for the PI as well as the input. Nothing to worry about there.

- The actual signal input network is formed by R21/R20/C19/R22. If we ignore R21/C19 for now R20/R22 forms a fairly standard gamma network. R20 sets the input impedance and R22 acts as the grid stopper to v3b. However, the circuit also includes R21 and C19 which are not strictly necessary. R21 and R20 will form a voltage divider that attenuates the input signal by 1%. C19's low value acts as a path to shunt hi RF frequencies. The purpose of R21 and C19 are to aid in keeping super hi RF frequencies from the input. However, they are almost never necessary in a guitar amp. Since R21 will also be affected by the output impedance of the guitar pickup circuit I would remove it.


on the first triode stage.
-D+ is indicated on the schematic as between 240-250. The anode resistor r9 is 220k. That sets one end of the load line at 0vdc/1.1ma and the other end at 245vdc/3.6ma.
The cathode resistor R23 is 1.5k which sets the idle bias of the stage at 1vdc at the cathode and 110vdc at the anode. C9 again rolls of hight trebles.

The problem here is that a 0-240vdc line is a very compressed load line especially with a 220k anode resistor. It only allows, ideally, a 70v swing towards saturation and a 140v swing towards cut-off. The stage is biased very hot already. However, since a tube isn't true rail-to-rail on the positive swing, the swing towards saturation is nominally only about 50v. A 1 volt signal swing at the input will send the tube into saturation. A fender champ in comparison has an input sensitivity of 1.5volts and a vox ac30 top-boost pre has an input sensitivity of almost 2volts.

-The tone stack is unremarkable except for the bypasses switching. Switching out the tone stack puts r24 in parallel with r23 causing the cathode bias resistance to go to 1.2k and biases the tube even hotter, decreasing input sensitivity even more.

An even moderately hot pickup or pedal signal will be sufficient to clip the input stage.

-Stage two.
D+ is again 240-250. The input network with the tone stack shunted consists of r35/vr2/r25. If the gain knob is turned up all the way that makes 150k in series with 1m//1m effectively 500k causing a 30 percent signal attenuation. Otherwise the stage has the same characteristics as stage 1.

That's 2 stages with very low input sensitivity, very low headroom stacked together.

-the cathodyne driver has 100k anode resistor so there's more headroom there.
The cathodyne has 56k load resistors which makes is fairly compressed.

The amp tries to get as much distortion as possible from as few stages as possible without the opportunity for any sort of harmonic cancelation/recovery in the gain stages.

It adds up to a harsh fuzz overtone to the notes. I wouldn't call what I'm hearing in the clips fizz though. That is more the sound that a number of very cold biased stages gets. The night train is exhibiting symptoms of a series of hot biased stages coupled with a slight amount of blocking distortion.

[Chuck H]

On a much less technical and purely theoretical note...

First, it's not buzzy tone that we're trying to fix. It's the assymetrical cracle that accompanies the note decay. It's very audible in the OP.

I've heard this problem in certain amps, and as noted earlier I've never had a clear moment when I could say "Ah ha, this is what causes it". But there are some common characteristics I've noticed. One is low plate voltage in the preamp. So, my theory in purely non technical jargon is as follows...

With low plate voltage in the preamp any "sag" is going to affect a greater % of the voltage available. This creates a greater bias shift. If the tubes self bias characteristics cannot compensate well enough the tube will suffer a temporary shift too far out of any nominal range. I don't figure the high plate load resistance helps with this at all. So now you have a tube with plate volts that are flopping all over the place (relatively speaking, since a higher plate voltage would handle any sag more gracefully). and a high plate resistance. When the voltage sags and the bias condition is, it seems to me, nearest to cut off, and the signal from the instrument is fading, you get the trailing chop into cutoff from input signal warbles in the low end causing the tube to cut off in an assymetrical way. Of course, when the signal is consistent and strong the chopping is more symmetrical and simply part of the clipping characteristics. But as the note decays there are less symmetrical jaunts of input signal that intermittantly drive the tube into cut off. So

Try reducing resistance in the HV rail to the preamp and maybe reduce the resistance of the plate load. If gain is lost then maube heat the bias of the preamp stages a little to both compensate gain and keep the tube further from cut off.

Just my laymans view. Go ahead and tear at it technically but I'd sure check into this if it were my amp.

[diagrammatiks]

yup chuck. that sounds right. That's the changes I recommended.

The power line resistor changes will bring the b+ to 300,305,310.

the cathode biasing resistor changes will recenter the bias at 1.5v to compensate.

[Steve Conner]

Thanks for the in-depth analysis, diagrammatiks!

Even without analysing load lines, a 220k plate resistor with 1.5k cathode resistor is kind of weird. I've seen 100/1.5, 220/2.2, one weird old Selmer with 100/5.6, and of course the cold clipping stages with much bigger cathode resistors. But I've never seen it unbalanced the other way, 220/1.5. I guess that would be a "hot clipping stage" by analogy.

So anyway, if changing the dropper resistors doesn't do it, how about decreasing the plate load resistors to 100k.

Another thing that might be relevant, on old amps with two input jacks, the input network often gave a 6dB attenuation and 136k load impedance. Sometimes I think that sounds better than the modern approach of going straight into the grid. It handles pedals and hot pickups more gracefully.

I wonder if the people suffering the worst fizz live in areas of below nominal mains voltage.

[jmaf]

I would try a 12AU7 for V1. That's a cookbook Fender tone stack normally driven from a cathode follower, trying a lower impedance tube with lower gain as well may help.

[Deltones]

I would try a 12AU7 for V1. That's a cookbook Fender tone stack normally driven from a cathode follower, trying a lower impedance tube with lower gain as well may help.

Ok, this is a complete noob question, but with the amp knobs facing me, what are the V positions for the tubes in the Night Train?