Problem with this fizzy sound

[Silvio55]

Hello! I have a vox nigh train 15, and have found this fizz as note decay when you play with the gain set at the point of breakup, (letīs say between 11 oīclock and 1 oīclock in bright mode) itīs not there when you play clean and when overdriven it becames a part of the sound so itīs not that noticeable. Many people have reported this problem on various forums all over the internet, but I canīt find a solution to it, I think one guy said it went away with a good retube but many others have tried that and the problem was still there. This is a sound clip of the issue, taken from another forum

MP3 Player SoundClick

As another guy said it sounds "like a dying transitor radio". Itīs really noticeable when you put down the master volume, and tends to desappear with the master set louder, but still there.
I have noted that the cathodyne PI in this amp is exactly the same as the 5E3, reading the Valve Wizzard site about cathodyne PIīs he states that the problem with this type of PI are not having a grid stopper, and power tubes should have bigger (100K ohms) grid stoppers. Does it has something to do with it or should I be looking for something else?

Thanks!


Silvio

[sickman82]

Hi Silvio, glad to see somebody else with the same issue as me. I have been on a quest for a number of weeks to get to the bottom of this problem. I have done one or two tweaks already, but none have made a difference to the fizzy decay we are experiencing.

One forum member over at FSB suggested I lower the value of the coupling caps. Off the top of my head, some of those coupling caps are rated at .1uf.. Which is apparently quite high and could contribute to 'crossover distortion'. I was told that I should try to reduce them to .022uf. I don't have any caps of that value rated at the correct voltage, so in going to have to wait until I can try that mod. I should mention that a side effect of this mod is a reduction in low end response, something that is already lacking with this amp.. So I'll be interested to see the results and whether it reduces the bass too much.

As a side note, I have installed a grid stopper on V4.. This has helped a great deal when using high gain pedals and the Night Train. If like me, you love fuzz, this amp isn't great for that sort of thing stock. There is a 1M resistor going to pin 2 of V4.. Solder your grid stopper from the upper lead of the resistor, directly to pin 2.

I'd be interested to see if anybody with tube amp knowledge thinks that the Night Train would benefit from the Paul Ruby Buzz Fix.. Apparently the output section of the NT is basically an 18 watter. Would anyone care to chime in on this one? Does anybody have some suggestions as to why this amp produces that fizzy decay, and how can we tame it? Thanks in advance!

EDIT: I should have mentioned, the grid stopper should be somewhere between 470k and 1M. I settled with 1M after some experimenting.

[pdf64]

Yes C10 seems asking for trouble at 0.1uF. As it's feeding a bootstrapped input, 4n7 would be plenty.
Prowess Amplifiers - Vox - Schematics - Night Train Updated
Re Paul Ruby zener mod, Silvio55's clips were done with the master vol low; the mod is on the power tube grids and will only do anything when the power tubes are getting overdriven.
Agreed that a grid stopper on V4 pin2 should be beneficial; a 10k-100k grid stopper on pin 7 may help also.
To me, the clip doesn't seem too bad, but IME, cathodyne phase splitters don't overdrive a nicely as an LTP.
And as it's ac coupled and has a gain of 1, there no benefit (that I can see) from the post phase splitter dual track master; a regular 1M master between the sections of V4 may sound better.
Pete.

[sickman82]

Hi Pete, thanks for your reply. I did try a grid resistor on pin 7 as well, starting with 120k down to 33k.. Overall gain and volume was reduced far too much though. I might try again, but experiment with low values.. working up from 1k5 or so.

I'll take a look at C10 this week, hopefully. Incidentally, I upped the value of the first stage cathode bypass cap to 22uf for a better bass response. It sounds great when clean, but tends to get a bit flabby with gain. In an attempt to lower the gain from the bypass cap increase, I removed R25 and lowered it to 1k. After some pretty unexpected results, it turns out that the value of R25 was actually 1M. Is there any possible explanation as to why the value is so different in the amp, compared to the schematic that is available online? An update perhaps? Maybe the schematic is incorrect? Everything else that I have measured has been correct in respect to the schematic though, it's just R25.. 1M is a big value to have as a cathode resistor isn't it? It is followed by a 2k5 to ground though. It has puzzled me for a while.. Hope you, or someone could shed some light?

Thanks again.

[Silvio55]

Yes, the Paul Ruby zener mod is for another sort of fizz in the power amp section (sounds like a fuzz with a phase shifter) when you crank the MV, and is another issue with this amp, I also hear that, but it seems to be pretty common with other amps (specially with the 18 watter), I didīt made that mod yet, but at least I know the cure.
The "fizz" I mention is another thing, itīs like a nasty raspy sound, more noticeable when you play with MV lower, and the preamp and/or PI starts to breakup, I didīt find a solution to it nowhere, but other night train users had reported earlier.
Sickman82, that schematic is wrong!!! be careful, R25 is not a cathode resistor!! itīs actually R36 in the amp layout (I will search and post the correct schematic later ), about the coupling caps, yeah, this amps is somewhat laking of low end, so it makes it more problematic...
I was plannig to add a grid stopper in pin 2 of V4, solder it in the upper face of the PCB between the 1M resistor and pin 2, but you have to cut the strip of the PCB that links the resistor with the pin, donīt you?
Pdf64 Do you think itīs a problem with the MV? nothing to do with the PI or just the PI canīt be tweeked to sound good?

Edit: here are the correct schematics thanks to Blinker from the Telecaster Forum

http://s590.photobucket.com/albums/s...x_Night_Train/

[Chuck H]

I'm not so sure about that schem.

[Silvio55]

I'm not so sure about that schem.

As far as I checked it was OK, R25 is V3B (named V3A in that schem, because itīs V3 pin 2) grid resistor, and itīs indeed 1M in the PCB, the other schem is generally good but has at least that error.

[sickman82]

To me, the clip doesn't seem too bad

It really is quite bad when you're trying to play some clean or very mildly overdriven chords/notes and you get this fizzyness accompanying it.

Yes, the Paul Ruby zener mod is for another sort of fizz in the power amp section

Ahh, ok.. I was just trying to find possible solutions, I didn't know whether it would help with this problem or not. I am experiencing the exact same problem as you Silvio.. MV low, gain around 11-1 o'clock, but that depends on the guitar.. I have a Les Paul and a Tele, both with pretty hot pickups, so when I use those guitars the range is more like 9-1 o'clock. But, with my Strat then it is 11-1 o'clock.

Sickman82, that schematic is wrong

I had a feeling it was wrong. I'll take a look in my amp this morning (UK time here), and see if I can follow the circuit through. Chuck H seems to think the one you posted is incorrect? Is that right Chuck, or were you referring to the schematic over at Prowess Amps that Pete posted?

I was plannig to add a grid stopper in pin 2 of V4, solder it in the upper face of the PCB between the 1M resistor and pin 2, but you have to cut the strip of the PCB that links the resistor with the pin, donīt you?

No, at least to my knowledge you don't need to do that.

So, all we have to go on so far is a reduction of C10 (which matches on both schematics). How much do you think that would influence bass response? Pete, it would be interesting to hear more on your theory behind the regular 1M master too.

Cheers!

EDIT: After tracing through the circuit a little this morning, the schematic that Silvio posted seems correct. The R25 mod (which I got from a forum member at Vintage Amps) now makes sense.. by reducing the value, it sends more signal to ground before it reaches the second stage which would go some way to help with blocking distortion (if that is where it is occurring). I'm going to try a lower value grid resistor on pin 7 of V4 now, I'll report back soon.

[pdf64]

'I did try a grid resistor on pin 7 as well, starting with 120k down to 33k.. Overall gain and volume was reduced far too much though'
A grid stopper won't have that effect on a 12AX7 common cathode stage; above 100k it might roll the brightness off a little. Could you check exactly how you did this?
Regards the arrangement of a cathodyne phase splitter followed by post phase splitter dual track MV, when I've tried this the cathodyne didn't seem to overdrive as nicely as an LTP or paraphase.
To my mind, the significant benefit that a post phase splitter dual track MV has over a regular pre phase splitter type is the additional gain and excellent overdrive character of an LTP; it seems to overdrive in a very similar way to the power tubes.
Whereas a cathodyne doesn't have any voltage gain and it doesn't sound great when overdriven (maybe Merlin's mods would help with that aspect)?
So the conclusion I drew from the above is that the Vox NT might benefit tonally from putting a regular 1M MV between V4B and V4A.
Actually implementing that on a PCB amp won't be easy (an additional coupling cap will be necessary etc).

Regards the value of C10, the actual input impedance of the cathodyne will be at least 2M, probably more like 5M, due to the bootstrapping of the 1M grid leak resistor.
With C10 at 0.1uF, that gives a -3dB frequency of 0.8Hz (assuming 2M, probably even lower)!
4n7F would provide a -3dB of 17Hz, so you could go even lower in value (eg 1nF) without losing any perceivable bottom end.
Pete.

[Chuck H]

With C10 at 0.1uF, that gives a -3dB frequency of 0.8Hz (assuming 2M, probably even lower)!
4n7F would provide a -3dB of 17Hz, so you could go even lower in value (eg 1nF) without losing any perceivable bottom end.

This is a very important point. Bigger caps won't make for more audible bottom end once the useful knee frequency has been exceeded. In fact, it can have the opposite affect. With the amp working hard and wasting power reproducing frequencies too low to be reproduced by the speaker. Power that could otherwise be used for audible low end. I would definitely reduce that cap value to something like .0022uf

I do think the clips are bad. I've heard this same problem in poorly biased preamps. Since the cathodyne PI is where most of the clipping is happening I might try altering the bias of the PI to see how that may affect the problem.

Also, there are a few things about the design that are not typical of guitar amp circuitry. One thing would be the film caps parallel to the filters in the power supply. This improves the caps decoupling by lowering ESR. Electrolytic caps typically have comparably high impedance at higher frequencies. Now the way I've always looked at it, since guitar amps often have more gain than other amps, there is some interaction with the power supply impressed on the signal path due to decoupling resistance. To put it another way, there are small positive and negative feedback loops that occur due to poor decoupling in the B+ rail that have an affect on the tone. Now...

By "improving" the power supply filter performance with additional film caps across the electrolytics, these feedback loops are reduced. So, perhaps there is some beneficial affect that is being lost by the addition of small film caps across the power supply filters. It's also possible that this particular amp suffered more problems than benefits due to decoupling resistance and the film caps were added to eliminate oscillation. I might try lifting one end of each of those caps to see how it affects the problem.

EDIT: As for the amp lacking low end, this may be a consequence of limitations of the OT. A different speaker would be an easy enough thing to try. I can't say enough about choosing the right speaker for a given amp. I can't physically see the amp, but if the OT is small, upgrading it may help. You may also try increasing the filtering on the first filter node. Not much to gain there but it may make the amp "feel" tighter on the bottom end.

[Silvio55]

I was plannig to add a grid stopper in pin 2 of V4, solder it in the upper face of the PCB between the 1M resistor and pin 2, but you have to cut the strip of the PCB that links the resistor with the pin, donīt you?
No, at least to my knowledge you don't need to do that.

But if you donīt cut the strip, the added resistor gets bypassed, you need all the signal going thru the resistor, or am I wrong?

As for the amp lacking low end, this may be a consequence of limitations of the OT. A different speaker would be an easy enough thing to try. I can't say enough about choosing the right speaker for a given amp. I can't physically see the amp, but if the OT is small, upgrading it may help.

I donīt think the OT is the problem, itīs really a big piece of iron, looks like a 30 watt OT. Iīm using the amp with the 1x12 matching cab, it has a Greenback and itīs open back, maybe with another cab you get better low end.

[sickman82]

A grid stopper won't have that effect on a 12AX7 common cathode stage; above 100k it might roll the brightness off a little. Could you check exactly how you did this?

No, you're right there Pete. After tracing through the circuit earlier today I realised that when I added in the resistor originally, I had soldered the lead of the resistor to the wrong end of C22. I added it in correctly today. I went up in various stages but settled at 100k. It hasn't made any difference to the present issue of fizzy decay.

So the conclusion I drew from the above is that the Vox NT might benefit tonally from putting a regular 1M MV between V4B and V4A.
Actually implementing that on a PCB amp won't be easy (an additional coupling cap will be necessary etc).

This sounds like a great idea, but it is far and away from what I am capable of. Unfortunately.

Regards the value of C10, the actual input impedance of the cathodyne will be at least 2M, probably more like 5M, due to the bootstrapping of the 1M grid leak resistor.
With C10 at 0.1uF, that gives a -3dB frequency of 0.8Hz (assuming 2M, probably even lower)!
4n7F would provide a -3dB of 17Hz, so you could go even lower in value (eg 1nF) without losing any perceivable bottom end.

Great. I'll get this sorted out this week. I just need to order in some caps.

I do think the clips are bad. I've heard this same problem in poorly biased preamps. Since the cathodyne PI is where most of the clipping is happening I might try altering the bias of the PI to see how that may affect the problem.

I'm not sure I would feel confident enough to do this either, unless some poor soul would care to guide me through the process?! It's been an extremely steep learning curve so far, I might as well carry on the way I have started.

By "improving" the power supply filter performance with additional film caps across the electrolytics, these feedback loops are reduced. So, perhaps there is some beneficial affect that is being lost by the addition of small film caps across the power supply filters. It's also possible that this particular amp suffered more problems than benefits due to decoupling resistance and the film caps were added to eliminate oscillation. I might try lifting one end of each of those caps to see how it affects the problem.

I think you might be on to something here. Some of the amp modders out there, who will mod your Night Train, and Tiny Terror, talk about making changes to the power supply.. maybe this is what they are talking about. It would be great to get some gut shots of a modded Night Train and see exactly what is going on in there after they have been in and worked their 'magic'. I'll try this as well.

I donīt think the OT is the problem, itīs really a big piece of iron, looks like a 30 watt OT. Iīm using the amp with the 1x12 matching cab, it has a Greenback and itīs open back, maybe with another cab you get better low end.

Nope, it doesn't seem to be the OT. As Silvio says, it is a really big piece of iron. I use the 1x12 matching cab as well. It sounds really tight with the master cranked, but it lacks that low end growl I like so much about Vox amps.. and was expecting with this amp. With regard to cabs.. I have tried an Orange closed back 2x12 (not sure of speakers) with the Night Train, there was no significant difference with regard to bottom end. If you swap C21 for 10uf, it gives a really nice rounded bottom end.. it isn't too much of a difference, but it does round things out a bit. I tried 22uf in there initially, but it ended up turning to mush as you crank the gain.. IMO 10uf is a good balance between tight bottom, and round bottom.. lol.

But if you donīt cut the strip, the added resistor gets bypassed, you need all the signal going thru the resistor, or am I wrong?

Somebody with better knowledge than me should explain this one.. I am pretty sure you don't need to cut the trace, but let's see what Pete or Chuck has to say about it.

[Jazz P Bass]

Do you have a scope?
How about a signal generator?
I feel that without this equipment you are chasing the problem.
First off: when the amp just starts; to clip is this occurring in the preamp or the power amp section.
You need to narrow it down.

[Silvio55]

Do you have a scope?
How about a signal generator?
I feel that without this equipment you are chasing the problem.
First off: when the amp just starts; to clip is this occurring in the preamp or the power amp section.
You need to narrow it down.

No, I donīt have a scope and donīt have access to one, neither a signal generator.
The clipping is in the preamp/PI section as the problem occurs with the MV down (more noticeable the lower the MV is), so the power amp section is not into clipping.

[Silvio55]

By "improving" the power supply filter performance with additional film caps across the electrolytics, these feedback loops are reduced. So, perhaps there is some beneficial affect that is being lost by the addition of small film caps across the power supply filters. It's also possible that this particular amp suffered more problems than benefits due to decoupling resistance and the film caps were added to eliminate oscillation. I might try lifting one end of each of those caps to see how it affects the problem.
I think you might be on to something here. Some of the amp modders out there, who will mod your Night Train, and Tiny Terror, talk about making changes to the power supply.. maybe this is what they are talking about. It would be great to get some gut shots of a modded Night Train and see exactly what is going on in there after they have been in and worked their 'magic'. I'll try this as well.

That`s interesting, something more worth trying.

But if you donīt cut the strip, the added resistor gets bypassed, you need all the signal going thru the resistor, or am I wrong?
Somebody with better knowledge than me should explain this one.. I am pretty sure you don't need to cut the trace, but let's see what Pete or Chuck has to say about it.

Yes please!! I thought you have to cut the trace