18 watt fizz: please help me diagnose this?

18watt-specific Tech Talk - Building, Fixing, Parts, Mods...

Moderators: CurtissRobin, colossal, zaphod_phil, Daviedawg, Graydon

Post Reply
CurtissRobin
Senior Amp Tech
Senior Amp Tech
Posts: 1169
Joined: Sat 08/09/08 2:00 am
Location: Columbia River Gorge, Oregon Side

Re: 18 watt fizz: please help me diagnose this?

Post by CurtissRobin »

No "issues," just a bit of voicing change.

KennyO
0 x

User avatar
gtrplayr1976
Unrated
Posts: 13
Joined: Mon 06/17/13 10:20 am
Location: WV

Re: 18 watt fizz: please help me diagnose this?

Post by gtrplayr1976 »

I think I'm going to change it to the Ceriatone schematic ,or use the Marshall schematic.
0 x

User avatar
500 Four
Unrated
Posts: 1
Joined: Mon 05/26/14 11:10 pm

Re: 18 watt fizz: please help me diagnose this?

Post by 500 Four »

Old thread here but I figured I'd chime in with my experience.

I built an 18 watt according to the Lite iiB schematic/layout last summer, I used a 14 gauge aluminum cake pan for a chassis (which worked great), Classictone PT and an OT from an old Baldwin organ that was run w/a push pull pair of 6BQ5s. It was my second build and I was very pleased with it, dead quiet at idle, nice overdriven tones from about 1/4 of the way up and onward. Then I tried to use it on some recordings...ugh.

I don't know if you could characterize it as 'fizz' or 'buzz' or crossover distortion or blocking distortion or what, but listening back I could hear this kind of shrill, unpleasant high end frequency accompanying the sound. I successfully mixed it out by rolling down a lot of the high frequencies in 'post', but I was prompted to scrutinize the sound in person to see if it was some fault of my recording technique. Sure enough, playing in a room with the amp a couple weeks later it was absolutely coming from the amp. A lot more tolerable in person but there nonetheless and getting in the way of the amp achieving what I would call 'good tone'. I went through a few tries at modding it out but none seemed to work and I wrote it off as a bad amp and carried on.

Then, last week I was working on a sort of cross between a 36 watt and single channel AC30 inspired amp that I intended to run 6GK6s in. Only have the one quad of 6GK6 I decided to re-wire the 18 watt to accommodate 6GK6s in order to find out definitively if the tubes were faulty. I re-wired the output tube sockets, but now with a couple more builds under my belt I was able to see other parts of the circuit that I could change. If I screwed it up or it didn't work, what the hell, right?

So I changed two things on the amp, 1) I swapped out the 8.2K grid resistors on the output tubes for 33K carbon films, and 2) I changed the cap that connects the plates of V1 to the volume control from .0047 to .01
Now I don't know if it was a different characteristic that the 6GK6s possess which EL84s don't, or if something I did made a difference but not only did I find out the 6GK6 tubes were healthy but holy crap, the 18 watt now sounded strong, able, glorious, confident and just good! I couldn't believe how much better it sounded. The high end harshness was gone, almost all the way up the dial. Topped out the volume did get a bit mushy but that was to be expected and is of no real concern since max volume is achieved much earlier on the dial.

The 30 (36?) watter got switched to EL84s since I had more to go around and I didn't want to change whatever magic happened in the 18 watt. I can't wait to record with it again :)

Anyhow, just wanted to share what worked for me. Death to all fizz, buzz and unwanted high end harshness!
0 x

ilcaccillo
Unrated
Posts: 9
Joined: Sat 12/24/16 9:42 pm

Re: 18 watt fizz: please help me diagnose this?

Post by ilcaccillo »

Bry wrote:Has anyone tried the 2 x 3k3R's + 2 x .005 caps like in the Watkins amp?
http://www.18watt.com/forums/viewtopic.php?f=25&t=24434" onclick="window.open(this.href);return false;
Isn't this a cross between a conjuctive filter and a snubber? (looks like)
It seems like 2 conjunctive filters from each plat to CT of the output transformer, I guess the peak voltage on each filter will be halved.

Anyone can chime in on this?
0 x

ampmadscientist
Unrated
Posts: 3
Joined: Tue 05/26/20 11:44 am

Re: 18 watt fizz: please help me diagnose this?

Post by ampmadscientist »

I'm going to add something to this:
The fizz is caused by cold bias of the output tubes.
Since all output tubes are not the same..
The bias should be adjustable. And to get the cleanest decay of notes, the bias should be adjusted to 70%.
I see from the schematics that many of these amps have no adjustable bias.
That was a design error. The bias must be made adjustable, an easy mod to the amp.

What's happening technically:
The output tube is trying to shut off, during the decay of the note.
Because the bias is set too cold.
This is what creates the fizzy sound in the decay of the note.

On a scope:
You will see crossover distortion. This is the fizz sound.
And if you had a scope, you could recognize these problems quickly.

So you can go on doing all kinds of tweaks and mods...BUT:
you will never solve this until the bias current of the output tubes is set to match whatever type tubes you are using.
And as I said:
Not all tubes are the same. There is very wide variations among EL84 output tubes.

If you need more help, go ahead and ask it's OK. Have fun w/ it.
AMS
0 x

User avatar
JMPGuitars
Super Duper Admin
Super Duper Admin
Posts: 3954
Joined: Tue 09/18/12 8:00 pm
Location: South Central, MA
Contact:

Re: 18 watt fizz: please help me diagnose this?

Post by JMPGuitars »

ampmadscientist wrote:
Tue 05/26/20 11:54 am
I'm going to add something to this:
The fizz is caused by cold bias of the output tubes.
Since all output tubes are not the same..
The bias should be adjustable. And to get the cleanest decay of notes, the bias should be adjusted to 70%.
I see from the schematics that many of these amps have no adjustable bias.
That was a design error. The bias must be made adjustable, an easy mod to the amp.
It's not a design error. It's cathode biased. You should read more about cathode biasing.

You would also benefit from studying this circuit more. It's significantly more common for these amps to be biased too hot rather than too cold.

Crossover distortion can cause fizz, but that's not the only reason it happens. A lot of the time these tweaks mentioned above can be compensating for OT issues, among other things.
1 x
'I installed a skylight in my apartment yesterday... The people who live above me are furious.' - Steven Wright

Modern Ground Schemes
Soldering Technique
B+ Voltage Reduction
Amplifier Tools & Parts Info


Web Design: DolceVittoria.com
Guitars / Amps / Effects: JMPGuitars.com
(anti)Social: Facebook · Instagram
Items for Sale

User avatar
Bieworm
Verbose Moderator
Verbose Moderator
Posts: 2287
Joined: Mon 02/10/20 8:24 am
Location: Belgium

Re: 18 watt fizz: please help me diagnose this?

Post by Bieworm »

ampmadscientist wrote:
Tue 05/26/20 11:54 am
I'm going to add something to this:
The fizz is caused by cold bias of the output tubes.
Since all output tubes are not the same..
The bias should be adjustable. And to get the cleanest decay of notes, the bias should be adjusted to 70%.
I see from the schematics that many of these amps have no adjustable bias.
That was a design error. The bias must be made adjustable, an easy mod to the amp.

What's happening technically:
The output tube is trying to shut off, during the decay of the note.
Because the bias is set too cold.
This is what creates the fizzy sound in the decay of the note.

On a scope:
You will see crossover distortion. This is the fizz sound.
And if you had a scope, you could recognize these problems quickly.

So you can go on doing all kinds of tweaks and mods...BUT:
you will never solve this until the bias current of the output tubes is set to match whatever type tubes you are using.
And as I said:
Not all tubes are the same. There is very wide variations among EL84 output tubes.

If you need more help, go ahead and ask it's OK. Have fun w/ it.
AMS
When experimenting with bias settings on fender amps I have concluded cold bias to result in thin and sterile, not fizz. Thanks to my friends here at 18watt I learned that doing the PR mod with the zeners and diodes the fizz is eliminated to the level that a little bit of fizz on top adds a lot of charm and complexity. The PR mod is only like 1$ on components and some fun on top. 😄
0 x
"THIS should be played at high volume..preferably in a residential area"

User avatar
zaphod_phil
Builder, Admin
Builder, Admin
Posts: 15208
Joined: Wed 03/19/03 2:00 am
Location: YYZ

Re: 18 watt fizz: please help me diagnose this?

Post by zaphod_phil »

500 Four wrote:
Thu 01/15/15 11:45 pm
Old thread here but I figured I'd chime in with my experience.
I don't know if you could characterize it as 'fizz' or 'buzz' or crossover distortion or blocking distortion or what, but listening back I could hear this kind of shrill, unpleasant high end frequency accompanying the sound.
So I changed two things on the amp, 1) I swapped out the 8.2K grid resistors on the output tubes for 33K carbon films, and 2) I changed the cap that connects the plates of V1 to the volume control from .0047 to .01
Now I don't know if it was a different characteristic that the 6GK6s possess which EL84s don't, or if something I did made a difference but not only did I find out the 6GK6 tubes were healthy but holy crap, the 18 watt now sounded strong, able, glorious, confident and just good! I couldn't believe how much better it sounded. The high end harshness was gone, almost all the way up the dial. Topped out the volume did get a bit mushy, but that was to be expected and is of no real concern since max volume is achieved much earlier on the dial.
Anyhow, just wanted to share what worked for me. Death to all fizz, buzz and unwanted high end harshness!
I'll put my money that it was swapping the 8.2K grid resistors on the output tubes for 33K which fixed the fizz issue. Changing the cap that connects the plates of V1 to the volume control, from .0047 to .01uF will have improved the amp's low end response. You could also have used a 0.022uF there, for yet more bass response.
2 x
Nature abhors a clean tube amp

User avatar
Bieworm
Verbose Moderator
Verbose Moderator
Posts: 2287
Joined: Mon 02/10/20 8:24 am
Location: Belgium

Re: 18 watt fizz: please help me diagnose this?

Post by Bieworm »

zaphod_phil wrote:
Tue 05/26/20 12:55 pm
500 Four wrote:
Thu 01/15/15 11:45 pm
Old thread here but I figured I'd chime in with my experience.
I don't know if you could characterize it as 'fizz' or 'buzz' or crossover distortion or blocking distortion or what, but listening back I could hear this kind of shrill, unpleasant high end frequency accompanying the sound.
So I changed two things on the amp, 1) I swapped out the 8.2K grid resistors on the output tubes for 33K carbon films, and 2) I changed the cap that connects the plates of V1 to the volume control from .0047 to .01
Now I don't know if it was a different characteristic that the 6GK6s possess which EL84s don't, or if something I did made a difference but not only did I find out the 6GK6 tubes were healthy but holy crap, the 18 watt now sounded strong, able, glorious, confident and just good! I couldn't believe how much better it sounded. The high end harshness was gone, almost all the way up the dial. Topped out the volume did get a bit mushy, but that was to be expected and is of no real concern since max volume is achieved much earlier on the dial.
Anyhow, just wanted to share what worked for me. Death to all fizz, buzz and unwanted high end harshness!
I'll put my money that it was swapping the 8.2K grid resistors on the output tubes for 33K which fixed the fizz issue. Changing the cap that connects the plates of V1 to the volume control, from .0047 to .01uF will have improved the amp's low end response. You could also have used a 0.022uF there, for yet more bass response.
Or maybe a .1uf.. you'd get fuzz😃😃😃
0 x
"THIS should be played at high volume..preferably in a residential area"

ampmadscientist
Unrated
Posts: 3
Joined: Tue 05/26/20 11:44 am

Re: 18 watt fizz: please help me diagnose this?

Post by ampmadscientist »

JMPGuitars wrote:
Tue 05/26/20 12:19 pm
ampmadscientist wrote:
Tue 05/26/20 11:54 am
I'm going to add something to this:
The fizz is caused by cold bias of the output tubes.
Since all output tubes are not the same..
The bias should be adjustable. And to get the cleanest decay of notes, the bias should be adjusted to 70%.
I see from the schematics that many of these amps have no adjustable bias.
That was a design error. The bias must be made adjustable, an easy mod to the amp.
It's not a design error. It's cathode biased. You should read more about cathode biasing.

You would also benefit from studying this circuit more. It's significantly more common for these amps to be biased too hot rather than too cold.

Crossover distortion can cause fizz, but that's not the only reason it happens. A lot of the time these tweaks mentioned above can be compensating for OT issues, among other things.
There are 2 types bias in 18 watt amplifiers:

1. is the cathode bias method, this is class A.
The current is 90-100% of the rated plate dissipation.
The bias is adjusted by changing the cathode resistor.

2. Grid bias: this is class AB
The current is typically 70% of the rated plate dissipation.
the bias is adjusted by B minus voltage on the grids of the power tubes.

I know a little bit about it. I have worked on tube amps for 45 years.

But no matter which method, Class A or Class AB, the bias is adjustable.
0 x

User avatar
JMPGuitars
Super Duper Admin
Super Duper Admin
Posts: 3954
Joined: Tue 09/18/12 8:00 pm
Location: South Central, MA
Contact:

Re: 18 watt fizz: please help me diagnose this?

Post by JMPGuitars »

ampmadscientist wrote:
Tue 05/26/20 5:19 pm
There are 2 types bias in 18 watt amplifiers:

1. is the cathode bias method, this is class A.
The current is 90-100% of the rated plate dissipation.
The bias is adjusted by changing the cathode resistor.

2. Grid bias: this is class AB
The current is typically 70% of the rated plate dissipation.
the bias is adjusted by B minus voltage on the grids of the power tubes.

I know a little bit about it. I have worked on tube amps for 45 years.

But no matter which method, Class A or Class AB, the bias is adjustable.
Cathode bias does not make it class A. Nor does grid bias make an amp AB. It sounds like you've been holding on to misconceptions for a long time.

The 18 watters discussed on this site are almost all cathode biased. They're pushed towards class A, but that doesn't make them class A.

You sincerely need to do some reading about cathode biasing so you can understand how it works. Yes, bias can be adjusted, but once set correctly for a cathode biased amp circuit, it isn't that likely that it will need any adjusting unless you're changing to drastically different tubes.

Again, nearly none of these amps are biased cold. If you look at the circuits, and the voltages on this site, you will see most people are pushing 85% to 100% dissipation (and yes, those at or near 100% should bias them better).
1 x
'I installed a skylight in my apartment yesterday... The people who live above me are furious.' - Steven Wright

Modern Ground Schemes
Soldering Technique
B+ Voltage Reduction
Amplifier Tools & Parts Info


Web Design: DolceVittoria.com
Guitars / Amps / Effects: JMPGuitars.com
(anti)Social: Facebook · Instagram
Items for Sale

User avatar
zaphod_phil
Builder, Admin
Builder, Admin
Posts: 15208
Joined: Wed 03/19/03 2:00 am
Location: YYZ

Re: 18 watt fizz: please help me diagnose this?

Post by zaphod_phil »

Bieworm wrote:
Tue 05/26/20 1:10 pm
Or maybe a .1uf.. you'd get fuzz😃😃😃
Maybe more like flubb :lol:
2 x
Nature abhors a clean tube amp

User avatar
zaphod_phil
Builder, Admin
Builder, Admin
Posts: 15208
Joined: Wed 03/19/03 2:00 am
Location: YYZ

Re: 18 watt fizz: please help me diagnose this?

Post by zaphod_phil »

JMPGuitars wrote:
Tue 05/26/20 5:44 pm
ampmadscientist wrote:
Tue 05/26/20 5:19 pm
There are 2 types bias in 18 watt amplifiers:

1. is the cathode bias method, this is class A.
The current is 90-100% of the rated plate dissipation.
The bias is adjusted by changing the cathode resistor.

2. Grid bias: this is class AB
The current is typically 70% of the rated plate dissipation.
the bias is adjusted by B minus voltage on the grids of the power tubes.

I know a little bit about it. I have worked on tube amps for 45 years.

But no matter which method, Class A or Class AB, the bias is adjustable.
Cathode bias does not make it class A. Nor does grid bias make an amp AB. It sounds like you've been holding on to misconceptions for a long time.
JMPGuitars wrote:
Tue 05/26/20 5:44 pm
The 18 watters discussed on this site are almost all cathode biased.
Actually about 99% of them are, apart from some experimental circuits. Likewise for Vox AC30s, AC15s, Matchless amps and Fender Tweed Deluxes. They're also all Class AB, as a pair of EL84s in Class A can't provide any more than 9W total. Only Class AB operation enables them to output 15W to 18W per pair.

JMPGuitars wrote:
Tue 05/26/20 5:44 pm
You sincerely need to do some reading about cathode biasing so you can understand how it works. Yes, bias can be adjusted, but once set correctly for a cathode biased amp circuit, it isn't that likely that it will need any adjusting unless you're changing to drastically different tubes.

Again, nearly none of these amps are biased cold. If you look at the circuits, and the voltages on this site, you will see most people are pushing 85% to 100% dissipation (and yes, those at or near 100% should bias them better).
Cathode bias is sometimes referred to as "auto bias", as it self adjusts. Basically, as the tube's current flow increases, the voltage across the cathode resistor also rises, biasing the tube colder, which then reduces current through the tube. So then the only "adjustment" needed is to chose a cathode resistor value to put the tube approximately in the right bias range, and then the auto-bias will take care of the fine adjustment.

I also recommend this article to clear up questions on Class A vs Class AB amplifiers. http://aikenamps.com/index.php/the-last-word-on-class-a
2 x
Nature abhors a clean tube amp

User avatar
PaulRuby
The MAN
Posts: 342
Joined: Mon 03/08/04 2:00 am
Location: Folsom, CA
Contact:

Re: 18 watt fizz: please help me diagnose this?

Post by PaulRuby »

The fizz has nothing to do with design of the bias. It was very clear when studied with a scope. When the EL84's were over-driven, grid voltage swings up above cathode voltage and the grid current very quickly charges the grid coupling cap which forces the bias to go cold and cross-over distortion results. The resulting kinks in the output current hit the OT and the OT reacts with kick-back. This kick back shows up as large voltage spikes on the EL84 plates and in the output signal at those kinks. Adding zeners to clamp the grid voltage eliminates the cross over distortion. I documented this long ago.

http://paulamps.com/18watterbuzz.html

What I was never able to figure out is why some amps have the fizz sound and some don't. They all have the blocking distortion happening.
2 x

User avatar
JMPGuitars
Super Duper Admin
Super Duper Admin
Posts: 3954
Joined: Tue 09/18/12 8:00 pm
Location: South Central, MA
Contact:

Re: 18 watt fizz: please help me diagnose this?

Post by JMPGuitars »

PaulRuby wrote:
Fri 08/21/20 8:25 pm
What I was never able to figure out is why some amps have the fizz sound and some don't. They all have the blocking distortion happening.
Hey Paul!

I assume it has to do with inconsistencies in the manufacturing process from one transformer to the next (even among the same lines). I could be wrong, obviously it's just a guess, but I figure it has to do with human error in the manufacturing process.

Thanks,
Josh
0 x
'I installed a skylight in my apartment yesterday... The people who live above me are furious.' - Steven Wright

Modern Ground Schemes
Soldering Technique
B+ Voltage Reduction
Amplifier Tools & Parts Info


Web Design: DolceVittoria.com
Guitars / Amps / Effects: JMPGuitars.com
(anti)Social: Facebook · Instagram
Items for Sale

User avatar
PaulRuby
The MAN
Posts: 342
Joined: Mon 03/08/04 2:00 am
Location: Folsom, CA
Contact:

Re: 18 watt fizz: please help me diagnose this?

Post by PaulRuby »

I tried swapping OTs between a good and bad amp. The problem did not follow the OT. Also left the OT unbolted and tried moving it around a bit as I wondered about the magnetic spike coupling elsewhere. That didn't work either. Very frustrating. I really wish I could solve that because I still consider the zener trick a band-aid for something I can't figure out. The last hypothesis I had was ground paths for the current spikes, but old repurposed amps mostly tied right to the chassis willy nilly, so is not an easy experiment if that's the bad one.
2 x

User avatar
JMPGuitars
Super Duper Admin
Super Duper Admin
Posts: 3954
Joined: Tue 09/18/12 8:00 pm
Location: South Central, MA
Contact:

Re: 18 watt fizz: please help me diagnose this?

Post by JMPGuitars »

PaulRuby wrote:
Sat 08/22/20 10:27 am
I tried swapping OTs between a good and bad amp. The problem did not follow the OT. Also left the OT unbolted and tried moving it around a bit as I wondered about the magnetic spike coupling elsewhere. That didn't work either. Very frustrating. I really wish I could solve that because I still consider the zener trick a band-aid for something I can't figure out. The last hypothesis I had was ground paths for the current spikes, but old repurposed amps mostly tied right to the chassis willy nilly, so is not an easy experiment if that's the bad one.
Weird! That's much stranger than my (wrong) guess. If the problem doesn't follow the set of tubes, or the OT, then it has to be ghosts. 👻👻👻
1 x
'I installed a skylight in my apartment yesterday... The people who live above me are furious.' - Steven Wright

Modern Ground Schemes
Soldering Technique
B+ Voltage Reduction
Amplifier Tools & Parts Info


Web Design: DolceVittoria.com
Guitars / Amps / Effects: JMPGuitars.com
(anti)Social: Facebook · Instagram
Items for Sale

User avatar
crgfrench
Frequent poster
Frequent poster
Posts: 827
Joined: Fri 04/27/18 3:02 am

Re: 18 watt fizz: please help me diagnose this?

Post by crgfrench »

It could be "ghosts in the OT". Let me identify potential ghosts in the OT, with a list of factors that come into play -- The ghosts are parasitic capacitance and leakage inductance. Both of these ghosts can be influenced by the following factors:

1) is the core a strip wound C-type with grain oriented steel or is it a stamped/laminated silicon-iron alloy?

2) is it a single coil shell or a 2 coil core construction? If single is it E&I or T&U layout, and if double is it U&I or L layout?

3) is it sectionalized to reduce leakage inductance, and if so is the sectionalizing vertical, horizontal, or both?

4) are the windings connected in series or parallel?

5) is it interleaved, and if so what is the interleaving pattern (see chart below)?

6) what are the wire gages?

7) are the windings layered or random, and if layered what is the insulation material?

8 ) is the secondary winding(s) paralleled using smaller gage?

9) what is the distribution of the primary winding around the center tap (is it 50% of the windings, 50% of the resistance, or something else)?

10) what specific insulation material is used?

11) what is the bobbin constructed of?

12) does it use a U (conventional traverse) or a Z (flyback) winding style?

13) what is the wire material and wire insulation material (polyurethane etc)?

14) is there an air gap (I know its double ended for PP operation, but there could be one) - and if so what is the gap length?

15) does it have shielding and if so what type?

With respect to #5, leakage inductance decreases with higher levels of interleaving, see the attached chart below. One manufacturer likely uses different interleaving patterns than the next, and perhaps even changes patterns over time within a given firm (e.g., later production runs are improved with higher N2, vs. earlier production runs).

With respect to #7 and #10, even otherwise identical OTs could have different "fizz" characteristics if A) The bobbin was wound using random winding rather than layer winding -- since, by definition, random winding will have more variance sample to sample; and B) the type and thickness of insulation used between the layers of the primary when the layer winding method is used (not random winding). Over time, acid in the paper can corrode the wire insulation. Paper also may not maintain its thickness over time, and is hygroscopic. Over time, "fizz" may develop in trannies that use "vintage" paper insulation between layers, as opposed to those using mylar, polypropylene or kapton. This is because of the increased parasitic capacitance which develops as the dielectric constant of the paper insulation degrades over time.
You do not have the required permissions to view the files attached to this post.
1 x

User avatar
PaulRuby
The MAN
Posts: 342
Joined: Mon 03/08/04 2:00 am
Location: Folsom, CA
Contact:

Re: 18 watt fizz: please help me diagnose this?

Post by PaulRuby »

That is all very cool stuff. But, swapping the OTs between a good and bad amp, the problem stays with the bad amp. The davis 104 amp has the problem, no matter what OT I put in it. Magnetic coupling?? Ground paths??? I don't know.
2 x

User avatar
crgfrench
Frequent poster
Frequent poster
Posts: 827
Joined: Fri 04/27/18 3:02 am

Re: 18 watt fizz: please help me diagnose this?

Post by crgfrench »

Is the fizz different than the hiss from carbon comp resistors?
0 x

Post Reply