Noise problems - Silverface Deluxe Reverb

[David Vinson]

Have some new parts; bias level suggestions?

I have new JJ 6V6s, two CV4024 to replace the 12AT7s, and am waiting for some Tung-Sol 12AX7s for the gain stages; and am putting 1ohm resistors in for bias adjustment.

Any suggestions for what level to set the bias on this DR?

Thx

[David Vinson]

Voltage w/ 6V6's

Hi Jag,

Here's the latest measurements (previous 6V6 tubes installed):


Pin……To..……....V8………..V7

3………Plate……. +464…… +463

4………Grid2.….. +465.... +465

5………Grid 1….. -43.6..… -44.3

[David Vinson]

Post-Bias voltages

I added the 1-ohm resistors from cathode to ground, and found the bias at ~16milliamps. So I re-biased to get the following voltages (using the old tubes - haven't put the JJ's in yet)

Pin……..To..…….....V8………....V7

8.........Bias....... .024...... .030

3………. Plate…… +440……. +440

4……… Grid 2….. +442……. +444

5……… Grid 1….. -35.1……. -35.8

Do these values look OK? Am I about to fry anything? ;-) Any suggestions for what range can I run the JJ 6V6's? Thanks

[David Vinson]

Tremolo Ticking gone

While fooling around with all this, I disconnected the Vibrato foot switch and the ticking went away. I don't use Vibrato, so unhooking the jack is easier than trying to re-dress the wiring.

[MWJB]

Keep the 470ohm resistors at that value. If they have drifted it's less than 10% and they may have been that far out when installed.

I've yet to see a pre '80 Fender that actually runs the voltages quoted on a schematic! They're are usually higher than the schematic.

Out of interest what rectifier do you have fitted, it should be a 5U4G, not a GZ34/5AR4 or solid state? Nevertheless, as Jag says, JJ 6V6s will cope happily with the voltages present, as you bias hotter the plate voltage will drop, 21-22mA should be fine at 470v on pin 3.

60 cycle hum - check the condition of the 100ohm resistors mounted on the pilot light assembly (brown/black/brown). Sometimes a tube will fail and burn out one of these resistors, but the dead resistor just sits there as if nothing is wrong. Also check the heater voltage (VAC from one lug of the pilot to the other), use clips & connect up with the amp unplugged, this may help determine if there is a PT problem pushing up voltages (but at this stage there is nothing pointing to this so don't panic).

[Jag]

+1 what MWJB said.

With 440v on the plates, I think you're biased a little hot for the tubes you have in there. I agree with MWJB, bias so V7 is about 22ma at the most. Remember, those old tubes are not really designed for those voltages.

The voltages will be different with the JJ's compared to the current tubes and changing the bias will also affect the voltages so measure often.

You still could try a 5Y3 in there, that will lower the voltages a little. GZ34/5AR4 would raise the voltages as MWJB said.

[Jag]

MWJB, do those screen voltages look okay to you in relation to the plates? That's one of my weak areas of understanding. I understand the reason that the screen should be less but, many of the amps I've worked on had screen voltages that were equal to or a little higher than the plates. When should that be a concern?

[MWJB]

Hi Jag,

Ideally, screen voltage should be less than plate (different folks have different preferences as to by exactly how much), but in reality, as plates draw lots of current and screens don't, the typical Fender choke (or 1K CT to screen supply dropping resistor in Champs/Princetons) doesn't always drop enough voltage for this to be the case. Notice how, as the plate current has gone up, the discrepancy between plate & screen voltage has widened.

Many amps have trundled on for years/decades with screen voltages at the same, or a slightly higher potential than the plates and if you didn't measure it, you wouldn't know you had a "problem".

Nothing to stop you experimenting with 1K 5W screen grid resistors (I often use these as I like to keep options open regarding alternative tubes) but some folks report a drop in output.

[Jag]

Yes, I usually go to 1k or 1k5 if I'm going to be using alternate tubes. To be honest, if I'm sticking to 6V6 or 6L6. I actually like the sound with the screens about equal to the plate.

KOC says the amp will be more touch sensitive and open if you go 2k2. I find the opposite. With 470 the amp seems more dynamic to me.

[David Vinson]

Post-bias results

Hi Guys, Thanks for your input MWJB.

I’m using a 5U4. The heater voltage is 6.54VAC. I’ll check the pilot resistors next time I power off.

Sorry I’m slow at this, but what’s the relationship between “pushing the tubes” and bias/plate voltage? Here’s my understanding so far: As I increase the bias voltage at the cathode, this decreases the negative voltage at Grid 1, which allows a higher “idling” current, and this decreases the plate voltage (?)

So which values represent a “hotter” operating range? When there is a higher cathode current, lower plate voltage, and less negative grid 1 voltage, are we pushing the tube harder? Or as we lower the cathode current, increase plate voltage, and make grid 1 go more negative, is this pushing the tubes harder? Or am I trying to peel an onion with a greased eel? ;-)

And, is there a range that the JJ’s like to see in the SF DR? I gather it is now safe to put the JJ’s in now?

Below are results for two different bias levels, using the previous 6V6 tubes installed. (I want to tweak this before I put in the JJ’s and start swapping the 12’s in the preamps out).

The settings at the higher bias voltage (~.030) sound much better to my ears – the amp is louder, and it doesn’t break up until 5-6 (breakup occurs ~2-4 on the lower bias voltage). So with a bias of .030, will I burn through the JJ's too soon?


Pin……To..……....V8……...…..V7

8………Bias….… .014…….. .022

3………Plate….… +454.….… +452

4………Grid2…… +453…..… +454

5………Grid 1….. -40.6…..… -40.6


--and—


8………Bias…….. .024…..…. .031

3………Plate……. +440…..… +440

4………Grid2.….. +442...... +444

5………Grid 1….. -35.3….... -35.8

Thanks for staying with this!

[Jag]

As you decrease the negative voltage to the input grids the plate current increases causing a smaller voltage at the plate. In other words adjusting bias from -40.6v to -35.8v, as in your samples, the current draw increases and the plate voltage drops. So your understanding is correct.

In a fixed bias amp, a tube is being "pushed" (in this context I am not referring to push-pull operation) when it's idle dissipation is significantly higher than 70% of its max rating. Current times plate voltage, with all controls at zero and no signal input, will give you the idle dissipation.

Max dissipation for a single (vintage spec) 6V6 = 12w
70% of max = 12 * .7
= 8.4w

In your first set of measurements you have a plate voltage for V7 of 452 and a current draw of 22ma (.022a). The idle plate dissipation for that exmple is

452 * 0.022 = 9.944w (a little high or "pushed")

While this may be a little hot, as long as the tubes are not redplating when you play, it's probably okay.

In your second set of measurements you have a plate voltage for V7 of 440 and a current draw of 31ma (.031a). The idle plate dissipation for that exmple is

440 * 0.031 = 13.6w!!! (1.6w over max spec!)

The amp might sound really good with the high current, but the tubes won't live long and you are risking an internal short in a tube which could take out your power transformer ($$$$).

At this point, you would be risking less to have the JJ's in there. They are a far tougher tube. The max plate dissipation for the JJ 6V6 is 14w and they can handle 500v plate and 450v screen.

Also, it's fairly obvious that your current tubes are not very closely matched. I'd fire them out of there.

BTW, I always do my calculations based on the tube running at the highest levels (V7 in your samples).

[Jag]

BTW, setting the bias with one set of tubes and then installing a new set does not work. Any change necessitates a double check of the values aat the least.

[David Vinson]

JJ's in

Installed the JJs; reset bias to the following:

8………Bias........ .022…..…. .023

3………Plate....... +445…..… +445

Watts............... 9.8........ 10.2

% Dissipate....... 70......... 73

Based on Bias and plate voltage, this pair is better matched :-D
Thx for the calcs, Jag.

The sound is good: a bit louder before breakup, and generally cleaner. Woo hoo!

Next I'll start by replacing the PI w/ a Mullard 4024...

[Jag]

Yup, that would be considered matched.

Good job, I bet it sounds a lot better now.

[David Vinson]

Mo' betta now - summary of solutions

Jag & MWJB, thanks for all your help. It sounds WAY better than before.

I'll wait until I get the 12AX7's replaced before I make the next move. We nailed nearly all of the issues on my original list:

A) There's a constant "ticking" or "popping" noise:
Unplugged the Vibrato foot switch since I don't use it; another workaround is to re-route wire dress per Fender Service Bulletin #9, found at
http://www.harpamps.com/micKfender/V...vc-Bltn-9.html

B) The amp just doesn't have that Fender "sparkle" - tone seems a little muffled, or muddy:
Re-tubing is helping. While inserting the Bias resistors, I removed the 1200pF "tone-sucker" disc caps. Sometimes this isn't removable due to wire dress, here's instructions on re-routing for that problem:
http://www.geocities.com/rdiubaldi/blackface.htm

C) There is some constant pink noise along with the ticking:
I'll address this after all tubes are replaced.

D) Is a little bit of 60-cycle hum normal?
Same as (C)

E) one of the 6V6 power tubes is glowing brighter/glowing red:
Fixed here in this thread.

You guys Rock!