18 watt fizz: please help me diagnose this?

[zaphod_phil]

Under these conditions (hot enough signal from pi for instance) is when the coupling caps charge from the grid current flow causing the unwanted effect of colder bias and unwanted cutoff.

That IS the mechanism that the Paul Ruby fix is addressing. BUT what Phil described is what would be improved with fixed bias (or a bigger cathode cap). Two different mechanisms that cause the same temporary cooling off of bias.

But I WAS talking about Paul Ruby's buzz issue. It's actually far more than just a general fizziness on top of the guitar's tone. Like I said in a previous post, when I first heard it I thought I had a loose tube retainer or other hardware vibrating inside the amp.

They are 22k carbon comps.

I meant the whole amp, not just the grid stoppers. They do make an amp more hissy.

[TomOlsen1]

@Katopan: Hmmm. I'm using JJ's. I'll take a look at some other valves if the suggested fixes don't work.

@Zaphod_Phil: Is the fizz/buzz/mechanical rattle the same as the one heard in my clips? The reason I ask is maybe we need to try compile a collection clips to provide a guide to what the different causes of buzz etc can be. For example, I've heard some clips that are clearly a grounding issue, others a dry joint, but maybe we need a definitive Paul Ruby style clip. Maybe these could help people identify their problem faster?

[guitarmike2107]

Hey tom

After listening to your clips I am wondering if it is a phase inverter problem as it starts to distort after the power valves in an 18watter, which may be why you dont have it so much on the lower settings?

have you tried the Paul RUby mod?

you really need your guitar and amp on the scope..

Mike

[guitarmike2107]

what I was also thinking was a boss Blues driver does the same thing on higher drive settings, one of the reasons why I avoid mine.. sorry not that that helps much

[nyazzip]

i like the idea of specified guitar amp fault audio clips.....!! i can certainly submit case points for: microphonic preamp/power tubes, bad grounding, speaker types/enclosures, and awful playing!
with the advent and availability of the internet imagine the knowledge and progress that might be, in years to come: 30 years of amp technician experience could be had in 1 weeks time... not to mention all the other actually important stuff

[TomOlsen1]

Hi Mike

I haven't done the Paul Ruby mod yet, and I'll have the amp on a scope this weekend, so hoping to get things resolved shortly. When we connected it to the frequency generator we were using a reasonable signal level. We're going to measure the output from my HFS bridge pickup and hit the amp with the same level from the generator. Then we shall hopefully see exactly where the undesired distortion is happening.

Hahaha, yeah the BD-2 does that grainy/spluttery thing which I don't care for either. But one of my fellow session players makes his sound like heaven in to a clean hotrod deville. Go figure.

[zaphod_phil]

@Zaphod_Phil: Is the fizz/buzz/mechanical rattle the same as the one heard in my clips?

Unfortunately my computer's sound is out of action, until I get round to fixing it.

Hopefully, before too long we will have a forum specifically for people to post their amp clips - whether good or bad.

[katopan]

After listening to your clips I am wondering if it is a phase inverter problem as it starts to distort after the power valves in an 18watter

I'm wondering the same. Normally in an 18 Watt the PI only starts to distort so far after the power stage is saturated it basically doesn't contribute to the sound, and the PI never really gets to full on clipping. But if bias or something else was wrong here and it was clipping earlier you'd hear it as a fizzy sound. Anyway, the scope testing you're planning should show up what's happening.

[JdJ]

So I picked up a scope for cheap money this week and did some further investigation. Although I am a scope rookie, it looks to me like the PI is distorting in my 18 watt. I fed the amp a 1k signal at aprox the same level that my humbucking guitar puts out on 10 with thew amp volume at around 70%.

Trem Input (V3) Plates


PI Grids


PI Plates


Output at 100%



It'll be interesting to see if this is the case with Tom's as well...

Ideas?

Josh

[zaphod_phil]

Yes, the 18W PI does distort when you push it. But you tend not to hear it much because the power stage is distorting a lot more.

You can try a 750 ohm cathode resistor in the PI if you want it to distort less.

[katopan]

Here's a doc I posted a while ago with my waveforms in detail.

http://www.18watt.com/modules.php?name= ... +waveforms

You're right - the PI ones look odd. Particularly I don't like the general shape, and the ringing at the top peak will not be adding anything nice to the sound.

Now you've got a CRO you'll be able to really see what's going on. But you've still got to interpret it all. I tested at a lower signal level than you but you're still getting a clean output (as you should) on the preamp. The rest should be the same. I'll have to have a bit more of a think on what could be the difference.

[zaphod_phil]

I think you would need to use waveforms at the same input signal level, same frequency and the same volume/tone control settings, to be able to make a meaningful comparison.

[katopan]

Phil - Totally agree.

I'd be interested in knowing what input signal level you used, and what the voltage scale (V/div) is for each of those CRO shots. Also to see shots with the Ruby zeners removed (you still have them in right? - then the balanced nature of your PI plate waveform makes sense). In terms of using the CRO, turning the brightness down a fair bit will give you a thinner trace making it possible to see some detail.

[zaphod_phil]

Also as another interesting amp to compare with, you can take a look at waveforms for this November, built by Romero da Silva. The November is basically a full Marshall Plexi, but with EL84 tubes. The PI in this amp therefore has more gain and output voltage swing than a typical 18W, but it may still prove useful as an additional point of comparison.

http://www.silvatone.bravepages.com/Nov ... hnical.htm

[JdJ]

I'd be interested in knowing what input signal level you used, and what the voltage scale (V/div) is for each of those CRO shots.

The input signal was 1k at 400mV which is what I measured as my guitar peak output both with the scope and with a rms dmm. I'm not sure of the scale - I ended up bouncing around a bit to maximize realestate, but I will be a little more careful today. I'm planning on a 333hz signal as you used katopan for at least a bit more direct comparison. Although I find it very interesting at the end of your PDF where you inject a 1v and 2v signal - I'm amazed that the PI isn't more saw-toothed given what I have seen with my 18w thus far, but I guess I have to attribute a bit to your volume only being on 2.

In terms of using the CRO, turning the brightness down a fair bit will give you a thinner trace making it possible to see some detail.

I did notice that, but in the pictures it comes out a bit fuzzy none the less.

The PI in this amp therefore has more gain and output voltage swing than a typical 18W, but it may still prove useful as an additional point of comparison.

Definitely. That's for that! I'm getting to the bottom of this slowly but surely...

Thanks a ton guys! This has been very informative.

Josh

[JdJ]

Well at least my ears aren't broken...

Testing with a 400mV 333Hz signal injected into the normal jack with various volume levels - zeners removed this time and with the PI cathode resistor dropped to 760ohms (the scope shots of the PI were barely different compared to the 820ohm BTW).

These shots were measured with a speaker connected and the probe on the speaker jack:

Volume at 2 - 10VOLTS/DIV


Volume at 5 - 10VOLTS/DIV


Volume at 7 - 5VOLTS/DIV (had to mess with the trigger a bunch to get it to behave...)


Next I'm to switch the cathode bypass cap from 500uF to 2200uF and retest.

Any ideas what the leading edge crossover spike is all about?!?!!

[guitarmike2107]

JDJ, Since you are new to scopes , Have you calibrated the probe to your scope, it will be in the scopes user manual as how to do so.

also try it with a pure resistive load, which is the norm for this sort if thing

Mike

[JdJ]

Hey Mike-

I haven't calibrated the probe as of yet - I'll do that, although voltage wise it seems to be in tune with my DMM.

As far as the load goes - all of my previous traces (on the last page) were with an 8 ohm resistive load. I am checking with the speaker following along with the Ruby research in the buzz info pdf and that's really when I see the spike as the volume on the amp is increased. I am also seeing it in the waveform as I play and it coincides with the fizzyness.

I have now also tried:

9.1v, 15v and 16v Zeners
8.2k[edit](stock) & 15k grid blockers to the Power Tubes

Still getting the fizz...

[guitarmike2107]

Hey JDJ

Calibrating the probe/setting the compensation will ensure the square waves are shown as proper square waves, has nothing to do with voltage accuracy.. I would do that first and see how much it sorts your signal out.. might not, but at least you then know for sure its the amp and not the Scope/probe causing all those spikes.

Mike

[katopan]

Although I find it very interesting at the end of your PDF where you inject a 1v and 2v signal

With that I was investigating the waveform when it was getting close to and a bit over the cathode bias point and therefore overdriving the first stage a little.

The spike is like what is described in the Paul Ruby doc - inductance spiking up the sharp edge of the crossover distortion. You do seem to have quite a bit of crossover distortion, even at mild overdrive (your vol at 2 pic). Interesting given that you're not running a real cold bias or anything. I'll be interested to see how your next tests go and how things look with a resistive load vs the inductive one. Also just humour me and move things around in the power stage & OT with a chopstick with the spike up on the CRO and see if it changes. I don't think it will but just in case.