18Watt.com

Hello guys, sorry to post with this account: I've found it on the internet because I couldn't signup for months.
I even wrote to Graydon but he said I have to wait the registrations to be opened again.

I've lurked for six months, and while the building experience was great (it's my first build) and the help I got from this forum was unvaluable, I think still something is not quite right with the tone.
It sounds very similar to this: http://guitargear.org/2009/10/26/el84-f ... that-tone/
Well he loved that tone but it don't think is quite right judging from sample sounds and videos I've seen posted here.

What I've done until now (thanks to all the informations I found on this site, I knew nothing about tubes 6 months ago!):

-Increased grid stopper resistor to 220k (helped a bit with the fizz)
-Installed a zobel network (10ohm/2.2uf, helped a bit)
-Decreased leak resistor to 220k (the volume knob became more useful)

I tried the bigger cathod cap with bigger bias resistor, and even stock resistor w/o cathod cap. Didn't help a lot with fizzness and stole some life from the amp, so I putted stock values back.

Now, not that the amp sounds bad (it's a lot better that that sound sample above), but still if I rise the volume over, say, 12 o clock, I get the fizz.

Another funny thing is that I never had the chance to really push it for hours, until recently that I got a palmer speaker sim: today without changing anything I felt a big improvement. Is there such thing as tube break-in?

Anyway, I'm stuck at this point: I don't even know if the problem comes from the preamp or not. Is there also a bit of muddines in the sound.
I've read one person had good result decreasing V1 plate value to 68k.

Could it solve my problem? Is there anything else I should do?

For the record these are the biulding pics: http://s1184.photobucket.com/albums/z33 ... r%20PP-18/

Any help really appreciated!!

Hmmmm, I'm not really sure if this is a "bug or a feature"....

I don't even know if the problem comes from the preamp or not. Is there also a bit of muddines in the sound.
I've read one person had good result decreasing V1 plate value to 68k.

It categorically can't come from the preamp, because this design (basically a Lite II) has barely even got a preamp. Just one gain stage. So it has to lie in the PI and/or power amp. That 68k resistor idea is a joke. All it does is the same as turning down the volume control a little. so anyone who thinks they've actually solved anything that way is delusional.

BTW in Vox type amps, you get a fizz at around the breakup point, due to the PI being cold biased, with a 1.2k cathode resistor. The Marshall 18W's 820 ohm seems to fix that issue pretty well, although you can in fact go a bit lower, down to around 700 ohms, and have the PI running very clean.

One fix in your case might be to try a pair of 7189s, which are beam tetrodes in an EL84 compatible format. They have a much better damping factor than pentodes, such as EL84s, and may give you what you're looking for. Just be aware that 7189As have a slightly different pin configuration from regular 7189s and EL84s. You could use 7189As with a small change to the socket wiring. Or OTOH you could even rewire the power tube socket connections to use 6CM6 tubes, instead of EL84s. 6CM6s are basically a 9-pin version of a 6V6, rated for 12W dissipation, and they sound somewhat in between an EL84 and a 6V6. If you do a search you will be able to find the few resistor value changes also required.

Try a Conjunctive filter instead of the Zobel. Put a 10K 5W in series with a .001uf 1200V cap across the OT primary. Put your grid stoppers back to 8.2K to get your highs back. You can increase the .001uf cap if it doesn't remove enough fizz. You could also try a 47pf-120pf cap across the PI plates like the big amps to help remove a little fizz.

Thanks guys,
I'll try with the simpler mods first (PI cathode, conjuntive filter) and report back.
If still I'm not satisfied I'll change the power section with different tubes as suggested.
I think in that case I'm going to go with 6v6. If not, I'm afraid the changes will be subtle and not worth the effort.

One question: do you think it's better to restore those mods I made before rewiring the power section with different tubes?

It depends on the mods. As for the 6V6s, like I said 6CM6s *are* 6V6s, just in a 9-pin bottle, so you don't need to punch bigge holes and go to octal sockets. And also let me re-iterate 7189s *aren't* EL84s, even though they are plug-compatible, and don't require any change in socket wiring. They are tetrodes and therefore unlikely to suffer from any fizziness within the power amp.

Ok clear enough, really helpful as always.
I'll try the PI cathode resistor and then the congjunctive filter while I look for some 7189 on ebay.
There is an auction now for four NOS 6p14p-ev, 12€ each. Are they the right kind? http://cgi.ebay.it/4-valvole-6p14p-ev-e ... 382wt_1139

The 6p14p-ev is a military version of the EL84, it is more rugged and can handle more voltage. Same pin out as an EL84.

So is it a 7189 equivalent or I'm buying another set of EL84 (more rugged)?
Because reading on Wikipedia:
Other equivalent tubes are the 7189, an extended-ratings version of the tube for industrial applications and the 6P14P produced in the USSR by the Reflektor plant, which is a direct equivalent of EL84/6BQ5. A slightly modified version of the 6P14P is currently being manufactured in Russia for Sovtek. An extended-ratings version of the 6P14P is also available - 6P14P-EV and is known among US guitar players as "EL84M" or the "Russian military EL84". While not necessarily a true "military version" of the tube (in fact it is more comparable to the 7189), 6P14P-EV are known for their low noise and durability.

it seems that they are exactly what ZP suggested as 7189.

Here is the datasheet of the 6P14P's (all versions except 6P14P-K) but it's in russian:
http://www.drtube.com/datasheets/6p14p.gif
According to it they can dissipate up to 14W but they are pentodes and ZP suggested using beam tetrodes in your case.
6P14P-EV has more working hours compared to 6P14P and is also more resistant to external influences (vibrations and shocks).

Thanks everyone for the thorough information. Time to seek better on ebay I guess...

Uhm, wait. Every 7189 datasheet I've found calls them "pentodes". However I understand that internally are beam tetrodes.
So the first question remain: anyone knows if the 6P14P-EV is the right one for this application? (better dumping factor, less fizzness).

Uhm, wait. Every 7189 datasheet I've found calls them "pentodes".

For some reason, you will find that American manufacturers often called their beam tetrodes "power pentodes" or "beam pentodes" or suchlike. Another triumph of marketing over reality. Same as calling Vox AC30s Class A amps. The tube manufacturers' logic was that the beam plates could be thought of as an additional electrode, then making 5 electrodes, and hence could be called "pentodes" - even though they still aren't pentodes in the normal sense of the word. On the whole RCA and GE wanted to avoid making any power pentodes because they would then have to pay royalties to the patent holder, Philips.

So the first question remain: anyone knows if the 6P14P-EV is the right one for this application? (better dumping factor, less fizzness).

I was under the impression that 6P14P-EV is an EL84 clone, in which case it's a true pentode. But I could be wrong...

I've found an honestly priced RCA 7189, but it's just *one*... consider that I'm in Italy so I'm trying to buy in Europe to avoid custom fees and infinite shipping time : )

According to this http://www.drzamps.com/amp/z_wreck/ the amp is advertised as 4 -7189 (EL84) Cathode Biased. But in the article it says that Dr.Z acquired a massive amount of NOS Russian 6P14P-EV for the project.
Considering that I saw them advertised as power pentodes on another site, they may be what I'm looking for.

*EDIT* bought 2 6cm6 : ) I've read those enthusiastic comments of that user few years ago, couldn't resist : ) They're cheap!!

I'm off on vacation for a week, I'll report back when I put them in!

For all my amps I build for friends, I actually stopped using EL84s and stick to 6CM6s now. I picked up a bunch of NOS ones on ebay a few years ago for about two bucks a piece. They are a helluva lot "smoother" (best word I can use). And remember you need to tweak a few resistors to get them to work correctly.

Great to know!
Is it not enough to tweak the cathode resistor to 220 ohm (as I read)? Are there are other tweaks? Beyond rewiring the socket of course.

Probably 270 ohms would be a safer value to use at the cathodes. Then 22k at the PI tail (instead of 47k or 56k) and 470 ohms on the PI cathodes, instead of 820 ohms. That's it.

"....(in fact it is more comparable to the 7189), 6P14P-EV are known for their low noise and durability.

it seems that they are exactly what ZP suggested as 7189.

No, it says they're "comparable" to 7189s, due to their ruggedness and higher voltage spec. We don't necessarily know if they have a pentode or a tetrode inside the glass. And don't forget the USSR never bothered about paying royalties to anyone. So they wouldn't care whose patents they infringed.

zaphod_phil wrote:Probably 270 ohms would be a safer value to use at the cathodes. Then 22k at the PI tail (instead of 47k or 56k) and 470 ohms on the PI cathodes, instead of 820 ohms. That's it.

Perfect!!

We don't necessarily know if they have a pentode or a tetrode inside the glass. And don't forget the USSR never bothered about paying royalties to anyone. So they wouldn't care whose patents they infringed.

Yeah I know... they don't have any regulation for cracked software and spam and who knows what else : )

I think the best choice will be to buy some and see.

I'll restore the lite to the standard values, then I'll apply those mods you indicated. I should report back in a couple of weeks!!

Since I'm trying to learn:
-270 ohms cathode on the PA is to bias the tubes at lower dissipation (?) and why is that? Because 6cm6 have a lesser plate to cathode max voltage than el84?
-470 ohms on the PI cathode is to bias PI hotter hence to have a warmer overdrive at PI?
-22k at the PI tail is for push less signal into the PI? And why I want that?

Thanks for the kind advice!!

I also got the fizziness which annoyed me a bit. I installed a Zobel filter (can't remember the values now but got them off a thread here somewher) and it cleared that fizziness right up. I took it out the other day to see the difference and put it back in 10 minutes later. Simple mod, just a resistor and capacitor in series between the speaker wires.

-Ewald

Since I'm trying to learn:
-270 ohms cathode on the PA is to bias the tubes at lower dissipation (?) and why is that? Because 6cm6 have a lesser plate to cathode max voltage than el84?
-470 ohms on the PI cathode is to bias PI hotter hence to have a warmer overdrive at PI?
-22k at the PI tail is for push less signal into the PI? And why I want that?

- 270 ohms is for a 6V6, which has a 12W dissipation limit - ie a 6CM6. 6V6s and EL84s bias differently, so you need a bigger cathode resistor. You will still need to measure the dissipation when the amps running, and adjust this value if necessary.
- 470 ohms increases PI gain to help compensate for the lower gain of 6V6s or 6CM6s
- 20k PI tail resistor increase the voltage swing of the drive from the PI to the power valve grids, which also helps compensate for the lower gain of 6V6s and 6CM6s

TomOlsen1 wrote:I also got the fizziness which annoyed me a bit. I installed a Zobel filter (can't remember the values now but got them off a thread here somewher) and it cleared that fizziness right up. I took it out the other day to see the difference and put it back in 10 minutes later. Simple mod, just a resistor and capacitor in series between the speaker wires.

-Ewald

Hi,

Would it be possible for you to take some photographs or to jot down a rough layout of the parts and where they go? I have one of Barry's 18 w kits and have some fizziness but unfortunately I'm not that technically minded so the schematics for Zobel filter that are floating around don't make much sense to me