Help improve pp-18 tone

[TomOlsen1]

Its a 10 ohm 5 watt resistor and a 2.2uF 600V(polyester, non-polarised) capacitor. I connected the resistor and cap in series between the 8 ohm speaker tap and the ground for the output jacks.

-Ewald

[toasteeee]

So at the speaker jacks I solder the cap and resistor from the 8 ohm wire to the ground point? If so then do I have to do the same at my 16 ohm jack? Apologies again for my appalling lack of technical knowledge....

[TomOlsen1]

I'm not that well versed in this stuff either. But yes, mine looks like this


8 Ohm Hot -> resistor -> cap -> speaker ground


Seeing I only use the 8ohm jack (I only have a 8 ohm cab) I put it there. I don't know if there's some magical reason it will function on the other taps too (hey, maybe it does?!) Maybe someone else can help here.


-Ewald

[zaphod_phil]

You leave it connected to the same 8 ohm tap, regardless of which secondary tap the speaker is connected to. It still does the job it's meant to, whatever tap the speaker is on, as it's more about the impedance that the EL84s see reflected onto the OT primary. Having said that, there's nothing to stop you trying out how it sounds on other tappings as well.

[toasteeee]

Thanks ZP! My next (dumb) question is where can I get a 2.2UF 600v cap? I'm based in the UK and a quick search of ebay has thrown up: http://www.ebay.co.uk/itm/160441333244? ... 4277wt_982 but I can't find an equivalent cap. Where would be the best site to order from?

[rock_mumbles]

For a Zobel network which is installed on the secondary (which is a low voltage speaker tap) it doesn't need to be a 600V rated cap, a 100V would more than enough.

Most people use caps designed for speaker crossovers for Zobel networks, the low budget choice would be a nonpolar electrolytic. Film crossover caps would be the more expensive choice.

You cannot use a standard electrolytic cap.

[Gary]

I remember being surprised that Maplins had some non-polarised in stock. You could try there.

[asd]

- 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.

But isn't that the same wattage dissipation of an EL84 or am I missing something?

I finally rewired the socket with the (Sylvania) 6cm6s: I didn't believe the hype but I have to say that they sound *really* smooth. I'm enthusiast!! They took away what I didn't like very much of this amp (maybe that fizzing edge?) and give back the liquid smoothness that I was looking for.

Stock grid stoppers and leaks and zobel network removed. Sounds great!!
(I must add I have PI+PA vvr on this amp and I splitted V1 cathodes)


My cathode-to-plate voltage is 310v --> 33.1 mA plate current --> 10.25W dissipation

Is it a bit on the low side? Could I decrease the cathode resistor? Or maybe something else?
What concerns me at the moment is that according to this datasheet http://www.shinjo.info/frank/sheets/137/6/6CM6.pdf it should be 285v at grid 2 while I have 312...
Any suggestion?

[Chris72]

Search ebay uk for 2.2uf polyester capacitor, I found a few sellers with both 630v and 100v.

[asd]

- 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.

But isn't that the same wattage dissipation of an EL84 or am I missing something?

I finally rewired the socket with the (Sylvania) 6cm6s: I didn't believe the hype but I have to say that they sound *really* smooth. I'm enthusiast!! They took away what I didn't like very much of this amp (maybe that fizzing edge?) and give back the liquid smoothness that I was looking for.

Stock grid stoppers and leaks and zobel network removed. Sounds great!!
(I must add I have PI+PA vvr on this amp and I splitted V1 cathodes)


My cathode-to-plate voltage is 310v --> 33.1 mA plate current --> 10.25W dissipation

Is it a bit on the low side? Could I decrease the cathode resistor? Or maybe something else?
What concerns me at the moment is that according to this datasheet http://www.shinjo.info/frank/sheets/137/6/6CM6.pdf it should be 285v at grid 2 while I have 312...
Any suggestion?

Guys anyone? Should I bias to 12watt and reduce G2 voltage to 285?

Thanks : )

[toasteeee]

Sorry to hijack the thread again but I installed the Zobel filter the other day:

http://i1213.photobucket.com/albums/cc4 ... AG0164.jpg

It's worked really well and has got rid of most of the unwanted fizz. Thank's everyone for the advice and also a big thank you to Michael from Modulus for sending the parts.

[asd]

Sorry to hijack the thread again but I installed the Zobel filter the other day:

No worries, glad it worked for you.

Back to my original thread:

I ended up restoring the original tubes (el84).

6cm6's were great, but when I got a chance to run them at loud volume, I discovered that they're too much clean to my taste and that el84 were definately more aggressive and overall more interesting.

The only drawback is the dreaded fizz BUT with 220k grid leak resistors it is kept at a minumum.

Now that I could experiment a bit removing and engaging all this mods, I'm enough confident to say that THIS was the only mod that greatly improved the tone (to my liking): I thought it would just change the volume control behaviuor but instead it changed the amp tone and behaviour quite a bit and made it a lot more useful.

On the other hand the zobel network didn't change so much the tone for me: I've read that it's speaker dependant so I guess it's beacause my speakers didn't need it (v30 + g12h30).

My final word is that for crunch tones you can't beat this thing : ) I'm REALLY loving it!!


And... don't hate my but I have another couple of questions:

-I've left the PI cathode + tail resistor unchanged (820 ohms and 22k ohms): I understand this made the PI more imbalanced and more clean with more drive, i.e. the tone is in favour of the PA overdrive.
Any problems to run the PI this way?

-For experiment I changed the valve socket and restored the EL84's but left the cathode resistor at 270ohm. Measured everything it dissipated about 7watts: it sounded terrible (as I understand, is this because it's biased too cold?)

Now back to 150ohm cathode resistor the tone is back and it's great. After measuring the value I'm dissipating 9.95watts. Is it worth to bias it even warmer? At about 12watts?

Thanks, Luca

[asd]

Biased a little hotter (11.2 W / 120 ohms Rk), and put back in the PI stock tail resistor.
So at present the mods are:

- 480 ohms PI bias resistor
- 220k ohms PA grid leak resistors
- 120 ohms PA cathode resistor

I don't think it could sound better : )

[zaphod_phil]

Biased a little hotter (11.2 W / 120 ohms Rk), and put back in the PI stock tail resistor.
So at present the mods are:

- 480 ohms PI bias resistor
- 220k ohms PA grid leak resistors
- 120 ohms PA cathode resistor

I don't think it could sound better : )

If you're running the EL84s that hot, then you should use a 50uF or maybe even 25uF cathode cap, to allow the bias voltage to swing upward freely when the amp is working harder, so that the EL84s safely bias themselves cooler.

If you're using 220k grid reference resistors in the power amp, then really you should go back to the 22k PI tail resistor, otherwise you will be somewhat under-driving the EL84s for an 18W.

Are you really using a 480 ohm boas resistor in the PI, or did you mean 470 ohms? Of course the 10 ohm difference doesn't really matter, but I'm just curious.

[asd]

If you're running the EL84s that hot, then you should use a 50uF or maybe even 25uF cathode cap, to allow the bias voltage to swing upward freely when the amp is working harder, so that the EL84s safely bias themselves cooler.

I do have a 47uF cap as per this schematic http://www.ampmaker.com/images/ak03kit/ak03sc2.jpg
Shouldn't EL84s be biased at 12W?
Is 11.something still too hot?

If you're using 220k grid reference resistors in the power amp, then really you should go back to the 22k PI tail resistor, otherwise you will be somewhat under-driving the EL84s for an 18W.

My rationale was: I want more distortion from the PI and less from the PA because it seems that driving the PA too hard gives that kind of dentist drill sound (on notes' decay) that I'm not really fond of.
So first I tried to restore PI cathode resistor and to keep the tail resistor at 22k: it gave me a lot of PA distortion with very pronunciated dentist drill sound (I think that this is what is referred here as the fizz), and that PI distortion edge was lost.

Then I thought that biasing the PI hotter while restoring the tail resitor (which, if I undestood it correctly, diminishes the voltage swing of the PI output) would give me a bit and better distortion from the PI and less crazy drive from the PA.
The result seems to be exactly what I want. 22k + 470 in the PI sounds really good too, but it distorts more and sooner : )

Isn't decreasing leak resistors but then increasing PI voltage swing a contraddiction?

Are you really using a 480 ohm bias resistor in the PI, or did you mean 470 ohms? Of course the 10 ohm differe nce doesn't really matter, but I'm just curious.

My bad, of course it's 470 : ) I really am a noob so I coudn't remember standard resistor values : )

[zaphod_phil]

Shouldn't EL84s be biased at 12W?
Is 11.something still too hot?

It depends on which approach you're going for, and whether or not your power tubes are expensive NOS. At one end of the scale, you can run the tubes at 80% of max dissipation, with a 1000uF+ cathode cap. At the other end you can run them close to 100% with say a 25uF cathode cap. The latter approach is far more squishy, while the former is stiffer, somewhat more like a bigger Marshall.

My rationale was: I want more distortion from the PI and less from the PA because it seems that driving the PA too hard gives that kind of dentist drill sound (on notes' decay) that I'm not really fond of.

However, PA distortion (when done properly) is richer and infinitely more satisfying, and is what the 18W vibe is all about. Warm-bias and plus the Ruby mod should stop any dentist drill sounds - as well as the resistor values mentioned below .

Isn't decreasing leak resistors but then increasing PI voltage swing a contraddiction?

Not quite. The lower value 220k grid reference resistors, allow the coupling caps on the EL84 grids to discharge more easily. While the 22k PI tail resistor then compensates for the reduced drive. Otherwise, with 220k on the grids and a 47k PI tail, you're not driving the EL84s much, so you've basically got a Vox and not a Marshall. The basic "dentist drill" problem is to do with blocking caused by grid current charging up the grid coupling caps. Apart from the 220k reference resistors, using larger grid stopper resistors on the EL84s, like 22k or even 47k can really help. Incidentally the EL84 design guide says that EL84 grid stoppers should be at least 10k. So it's lind of strange how both Marshall and Vox use lower values than that.

[asd]

Thanks as always for your useful advice. I don't have any NOS tube (yet? ) in this build, just regular JJ EL84s.
The problem with the ruby fix (which I installed once) is that I use a VVR in this amp, so it's rarely being pushed as high as it's needed for the fix to be useful.

Maybe the whole problem is coming from the VVR?
In the sense that, since I scale PI + PA their balancing should remain the same. So basically I'm smashing with a lot of preamp signal a voltage scaled PI.
So could it be that it's the PI that is getting overdriven too much? Could this explain why when I biased the PI hotter (leaving the stock tail resistor) and decreased the PA grid leaks I had an improvement in tone?

I already tried to increase grid stoppers but with no noticeable improvement.

And BTW I get plenty of drive as it is. Maybe it's just that I still couldn't run this thing at loud volume for any decent period of time to get a clue : )

[zaphod_phil]

When you VVR-scale the PI+PA, you're meant to use the amp's Volume control to educe the preamp drive to the PI by a corresponding amount. Don't try to turn up the control to the same level as with the VVR maxed.

You may also want to try a 100k grid stopper on the PI input as well, to control PI grid current and blocking.

It would also be interesting to know if you get the dentist drill effect with the VVR maxed, or only when scaled down.

[asd]

When you VVR-scale the PI+PA, you're meant to use the amp's Volume control to educe the preamp drive to the PI by a corresponding amount. Don't try to turn up the control to the same level as with the VVR maxed.

I'm tipically running VVR at 9 o clock and, with stock values, I obtain dentist drill effect when volume is about 12-1 o clock...

You may also want to try a 100k grid stopper on the PI input as well, to control PI grid current and blocking.

So referring to this schematic http://www.ampmaker.com/images/ak03kit/ak03sc2.jpg
I should add these resistor between R7 and pin2 of V2 and between R8 and pin7?

It would also be interesting to know if you get the dentist drill effect with the VVR maxed, or only when scaled down.

In the past I played it at full volume with stock components, (didn't have a VVR back then) I can remember haveing this effect.

This saturday I will try to play at max to see, but first I have to lower PI tail resistor again...

In any case, I've just ordered a full set of NOS tubes from sgt. overdrive: what a character! : )

It will be 2 sylvania 12ax7 and 2 saratov 7189, so that's in line of what you suggested few days ago Phil.

When they arrive, I want to restore everything at stock values, and see where how I like the sound and if I still getting the drill problem.

[Gary]

the best de-fizzing effect I achieved on my lite 2b with pi+pa vvr, bearing in mind it's operating at family friendly volume, was to disconnect the cap bypassing the shared PA cathode resistor. I believe this approach allows a bit of negative feedback to tame the stage down a touch? I can still get a lot of crunch/distortion ( perhaps too much ) by maxing the volume control, without any annoying fizz.

An easy mod to try - and reverse.

I had previously tried a zobel filter and a large bypass cap without as much success. The zener fix wasn't an option, of course.

I'd be interested to know if operating the stage without a bypass cap is considered detrimental to the amp ( bias voltage swing ). I'm on low vvr voltages 99% of the time.