Fender Tube Amp Bypass Caps

Tim Greene · Post by **Tim Greene** » 9 Jan 2009 10:44 am

There are 7 25uf/25v cathode bypass caps in my dual showman reverb and they look orginial which dates them around 30 years old.These need to be replaced with a 25uf/50v and my question is? Are all of the brands available(sprauge atom etc)the same or are there differences in them? Is there one you have used and would recommend over the others?Thanks Tim

Donny Hinson · Post by **Donny Hinson** » 9 Jan 2009 11:02 am

These aren't in the signal path, so I wouldn't consider them critical audio parts. Most available brands should be quite acceptable.

Mark MacKenzie · Post by **Mark MacKenzie** » 9 Jan 2009 2:56 pm

Now that's a good question and I don't have the answer. But, don't cathode bypass caps actually pass certain freqs to ground and even though the audio doesn't pass through to the power amp, doesn't it shape the frequencies of the signal? Won't different capacitance values change the tone? Or is this only in a cathode bias configuration?

Confused again......

Brad Sarno · Post by **Brad Sarno** » 9 Jan 2009 3:31 pm

The cathode bypass cap essentially shorts out the cathode to ground at all frequencies in the audio band, but still blocks DC so the cathode resistor can maintain a DC bias voltage at the cathode relative to the grid. Essentially, the cathode bypass cap allows the tube to generate it's full gain potential since without it, some gain is lost by having the 1.5k bias resistor there. You can run a tube without that cap, but you lose something like 8dB or so of gain.

So that cap absolutely should be considered part of the audio path even though it's not what you may consider directly in the path. It's completely part of the tube's amplification of the audio signal. About 8dB or so of the tube's gain is generated by AC passing thru that cap.

Different cathode bypass cap's have different tonalities. Many of us messed with trying tantalum cap's there, and instantly noticed a very clear yet crispy treble. For some reason the cathode bypass cap has a particularly strong effect on the tone. Many like to go for the lowest ESR cap they can find. Since the values tend to be large, film cap's are typically not an option, so we're left with various electrolytic choices. Any aluminum type will work fine, but if you're tweaking for a special mojo, there are some that do sound better.

The Sprage Atom is the most popular in boutique amps. I don't think that cap is really all that special except for its ruggedness and maybe it's ability to withstand heat. I don't hear much difference between those and your typical run of the mill generic electrolytic type. But there are some special audio electrolytics that do sound especially nice there. The Black Gates were cool there, but they don't make them anymore. The Elna Cerafine is sort of like a Black Gate. They sound pretty hi-fi with a clear top end. My current favorite is the Elna Silmic II that uses a hemp/silk dielectric. It has a distinctly warm and gentle upper mid and treble with a large bass feel. That's all I use lately. They're cheap too.

In a Fender amp, 22uF/25v or 22uF/50v is standard. Alembic chose a 50uF/50v in their FB2 to help extend the bass response and move any low end phase shift way, way down below the audio band. I use 22uF/25v all the time. The Silmic II does break in over time, mabye after 50 hours or so. They're brighter at first, but mellow out in a nice way. They're smooth on top, but not at all dull. They are quite detailed and refined compared to the Sprague or generic types. I love 'em.

Brad

Donny Hinson · Post by **Donny Hinson** » 9 Jan 2009 4:12 pm

...even though the audio doesn't pass through to the power amp, doesn't it shape the frequencies of the signal? Won't different capacitance values change the tone?

The short answers are "Yes" and "Yes". You can alter the frequency response and gain by changing the values, but Tim dodn't say he wanted to change the tone. He said he was just looking to replace old caps. Replacing them with the same value and type will retain the original sound and tone.

However, if you want to change the tone and dynamics, you can, as Brad said, do that by changing the values (or eliminating them entirely). The only "upgrade" that I know of for a standard electrolytic caps would be to use polypropylene caps, but they're often physically too big to go into many amps.

Ken Fox · Post by **Ken Fox** » 10 Jan 2009 6:20 am

The voltage above the cathode bypass cap is typically around 1.8 volts DC. That is the typical positive voltage needed to set the 12AX7 into a class "A" operation mode, keeping the grid negative with respect to the cathode.

The 25 volt cap is quite OK there. I use the Mallory 50 volt only because they are similar in size to the old caps in the Fender amps.

Newer electrolytic caps are much smaller in physical size when compared to the ones made 30 years ago.

Typically the older caps go much higher in value with age. I see these 25uf caps as high as 80uf when tested in circuit with my ESR (equivalent series resistance) tester.

The higher that value the more bass energy gets amplified, as these act as a high pass filter in the circuit (shorting the signal around the 1500 ohm cathode resistor) A new set of caps will tighten up the amp a lot if these have aged and drifted.

The Fender Vibrasonic of the 90's deleted the cathode bypass cap on the first gain stage of the steel channel to reduce gain. When you see one of these caps fail open the amp will have lower than usual gain. Typically I hardly ever see one open, most have drifter very high in value.

Mark MacKenzie · Post by **Mark MacKenzie** » 10 Jan 2009 7:53 am

Thanks, guys! Great info!

Brad Sarno · Post by **Brad Sarno** » 10 Jan 2009 9:32 am

Excellent!

Thanks Ken.

Brad

Larry Robinson · Post by **Larry Robinson** » 10 Jan 2009 10:19 am

The purpose of the cathode bypass cap is to keep the voltage drop accross the cathode resistor constant and provide the bias for the tube. (Pos voltaged at the cathode makes the control grid neg with respect to cathode) I don't agree the capacitors are for tone. Without the caps, bias on the tubes would vary a the audio rate which was cause a decrease in gain. Faithful signal reproduction is what is desired.

Ken Fox · Post by **Ken Fox** » 10 Jan 2009 11:44 am

Sorry, that is not correct. The DC voltage is set by the resistor, not the capacitor. Ac capacitor does not pass DC, unless it is dead!!

The capacitor is an AC (signal) bypass proving a low impedance path around the typical 1500 ohm cathode resistor. The 1500 ohm resistor keeps the cathode positive with respect to the grid.

Larry Robinson · Post by **Larry Robinson** » 10 Jan 2009 3:31 pm

I didn't say the capacitor set the voltage. I said the capacitor keeps the voltage drop accross the resistor CONSTANT. If you remove the capacitor, there will be a varying voltage drop accross the cathode resistor whenever a varying signal is present on the control which will cause degeration and lower the gain.. Again, the purpose of the capacitor is to keep the voltage accross the resistor constant. I stand by my analysis and I am not incorrect.

Blake Hawkins · Post by **Blake Hawkins** » 10 Jan 2009 7:14 pm

Ken and Larry,
You are both correct.
My reference book on this is "Radiotron Designers
Handbook" Fourth Edition.
On page 554 under "Cathode Bypassing":

"It is important for the bypass condenser to be sufficiently large to maintain the bias voltage consistant over each cycle - any fluctuation leads to increased distortion and loss of power output.
When operating at low levels it is usually sufficient to ensure that the reactance of the bypass capacitor, at the lowest frequency to be amplified, does not exceed one tenth of the resistance of the cathode bias resistor."

So you can see that the bypass capacitor does
affect frequency response as well as smoothing out the bias voltage.

In "Elements of Radio Servicing" by Marcus and Levy
On page 104 are several paragraphs stating that
the cathode bypass capacitor largely effects low frequency response.

Blake

J Fletcher · Post by **J Fletcher** » 12 Jan 2009 5:45 am

I look at it like this. The cap keeps the DC voltage on the cathode resistor more or less constant, by charging and discharging. This eliminates the AC component that would otherwise ride on top of the DC voltage. The AC component is the signal voltage that is developed by the cathode current.
If these voltage fluctuations were present on the cathode resistor, they would provide negative feedback, thereby decreasing the gain of the circuit.
Kind of depends whether or not you look at it from an AC perspective, or a DC perspective...Jerry

Brad Sarno · Post by **Brad Sarno** » 12 Jan 2009 6:34 am

Jerry,

I like how you put it. That explanation helps to illustrate why a small value cap there would act somewhat as a hi-pass filter, at least hi-passing in the gain range addressed by eliminating the negative feedback. As I recall, that's roughly 8dB or so in a typical 12AX7 application. You can see in many guitar amp circuits that this cap will be very small, often under 1uF, and the result is essentially a treble boost. The Garnet amps have a unique tone control circuit that actually has that cap (small value) routed thru a pot to ground.

I think your description also helped me to better see exactly why the bypass cap imparts its tone character into the audio. If if's constantly working with the signal voltage swings to keep the DC bias stable, then whatever audio performance characteristics the cap has will obviously become part of the signal.

Thanks,
Brad

Jim Sliff · Post by **Jim Sliff** » 12 Jan 2009 6:54 am

I also stay with 25V caps, and usually use 22uf...but unlike some, I only replace those caps if they're bad (and they are rarely bad). They are not nearly as "stressed" as filter caps and, for, example, the blue tubular Fender caps rarely fail; I've heard some amps really fouled by wholesale "cap jobs" with all the bypass caps changed and 715 orange drops dumped just about everywhere else. Does the term "icepick treble" sound familiar.

On the Tele forum I used to tell guys to mail me their blue caps if they were doing a cap job, and that I was working on a sculpture with them...or any other weird reason to get hold of 'em. I usually have a dozen or so of those and the white Mallorys I'll use to "retune" an amp that's ben "improved"....

J Fletcher · Post by **J Fletcher** » 12 Jan 2009 7:13 am

Brad, glad my post was helpful, wish I had your ears!
Sometimes when looking at capacitors in a circuit, I remind myself that what they do is charge and discharge. The bigger they are, the longer it takes to do that. A very simplified view, but essentially that's it...Jerry

Tim Greene · Post by **Tim Greene** » 13 Jan 2009 3:19 pm

Thanks guy for all the great info.I will post some pics of the chassis when I find my camera.The tone and coupling caps that are .1 value are dull brown.There are 3 .047's that are blue.Ceramics are orginial. 2 of the 7 25/25 caps have been changed.The 470ohm 1watt resistors on pins 4 and 6 on the power tubes look somewhat cooked.This amp has a chassis date of 1973 so its around 35 years old.There seems to be a lot of debate as to only fix whats out of spec vs overall maintence.My main concern is the 25/25 cathode caps and the 470ohm resistors along with 3 of the ceramics(250pf and the .022) Tim

Donny Hinson · Post by **Donny Hinson** » 14 Jan 2009 12:53 pm

I'd change out the electrolytics, and check the coupling and tone caps for leakage. (Either a voltage check or IR thermo check to see if they're leaking DC will work for that.) The bottom line, to me (if you're happy with the sound), is to change only what really needs changing. The 470 ohm resistors may need replacement - replace with 2-watters if they've drifted out due to thermal stress.

At any rate, save any internal parts removed so you can give them to any future owner. What originally came in the amp, even if it went bad, may be important for the historians and collectors of the future.