The Cathode Follower Mod

or, what does this mod actually do?

Almost everyone who has looked into the heart of a Blues Junior has noticed that extra half of a 12AX7 that’s just sitting there, doing nothing. There are lots of things it could be doing… a second input, a line input, another gain stage…. Or take a page from the venerable Fender Bassman, and use it as a cathode follower. The cathode follower is a non-gain stage that separates the preamplifier from the tone stack.

Why would you want to do that? The tone stack, with multiple paths to ground through capacitors and tone control potentiometers, places a rather heavy load on the previous preamp stage. The preamp stage is high impedance and does not drive a load all that well. The cathode follower has very low input capacitance and very high input impedance, so it is a very light load for the preamp and a better match impedance-to-impedance than preamp-into-tone stack. On the other side, it has very low output impedance and is a powerful current source, so it can drive the tone stack effortlessly.

So why wouldn’t you want to do this?

The biggest reason is that it doesn’t sound all that different. The theoretical bark is worse than the actual bite, and despite all the impedance mismatches, lots of signal gets through to the tone stack. So you have to question whether it’s worth the effort. On the other hand, nobody ever accused a Bassman or a Marshall (which copied the circuit) of sounding bad.

Mark Huss did this mod on the green board Blues Junior years ago. He has instructions and some sound clips on his site:

At a customer’s request, I recently re-engineered and adapted the mod to the cream board. It’s not all that difficult. Here’s how, below. I’ve also re-engineered Huss’s mod for the rev. C and D cream board. Photos are at the end of this page.


This is the circuit we’re trying to create. I cut-and-pasted another tube section into the Blues Junior schematic. It’s laid out just like the schematic for the Bassman, and it’s technically a DC-coupled cathode follower. The half-tube connected by the colored lines is the unused half of V2, designated V2A. The only additional component, other than the wires, is the cathode resistor. Although it’s unmarked on the schematic, it’s 100K, just like the plate resistor of V1A. This may seem strange if you’re used to cathode resistors in the 1K-2K range, but the high-potential balance between the plate and cathode is essential to the cathode follower’s operation.

The voltage drop across the cathode resistor raises the cathode to a high voltage that gets modulated by whatever signal appears on the grid. Unlike amplification stages, the cathode follower does not change the phase of the signal. What goes in is what comes out.

V2B, by the way, is the tone stack recovery amplifier.

(Note: If you’ve changed the plate resistor to some other value, as with the high-voltage preamp, the new cathode resistor needs to be the same value.)


This is the stock circuit board. The first step is to clip both ends of this jumper, close to the board.

This jumper is very convenient for our purposes; it’s the path from the output of V1A to the tone stack. We’ll feed the cathode follower from one end and feed the output of the cathode follower to the tone stack with the other.

It’s essential to clip components when desoldering them on printed circuits because the leads are typically bent onto the trace on the back of the board. Trying to pry or wiggle the entire part loose while keeping the solder molten is an excellent way to tear the solder pad from the circuit board.


I’ve removed one of the stubs from the jumper and am about to remove the other. A solder sucker is the best way to do this because you don’t have to linger on the joint and it cleans out the stub and the excess solder.


This trace brings filtered high voltage from the “X” point of the power supply to the plates of V2 and V1, but only half of V2 is supplied–no sense in sending power to an unused portion of the tube. I’ve marked the best place to tap into the supply.


I used a #60 drill in a Dremel tool fitted with an inexpensive accessory chuck. The chucks are readily available by mail order and in home centers and hardware stores.

If you drill at low speed (I use a variable-speed foot controller), you have excellent control over the drill. If you start at high speed, it could skitter off over the surface of the circuit board.


We also need a place for the new 100K cathode resistor. I created a pair of holes, one that connects to the new wire that will go to the V2A cathode, one that connects to the diamond-patterned ground area on the circuit board.

Scrape away the green coating so you can make a good solder connection. A 1/2 watt resistor is a perfect fit between these two points.


The brown wire will go to V2A’s cathode; the yellow wire (green on the schematic above) goes to the grid.


But first I had to remove the ground connections between pins 1, 2, and 3 of V2. If a tube section is unused, good design practice is to short all of the elements to ground so there is no chance of parasitic oscillation or other electronic weirdness.

Always test with an ohmmeter to make sure there is no hidden thread of copper connecting the pins.

I’ve desoldered pin 1 here. There’s a nice pocket right next to the socket pin. When I desolder the other two, I’ll be able to poke the ends of the wires down along the pins and solder them securely.


All finished! The new 100K cathode resistor is under my thumb; sorry for not having a clearer view. The red wire carries the plate voltage, the yellow wire connects the plate signal from V1A to V2A, and the brown wire carries the signal from the cathode to the tone stack.

You can see how the cathode follower essentially replaced the jumper wire, with the addition of plate power and the 100K bias resistor.

Rev C and D green board cathode follower:


On the rev C and D green board, the layout isn’t nearly as convenient as that of the cream board. The first step is to tap into the trace from the plate of V1A to the grid of V2A, our cathode follower. Huss did this on the tube board; so did I. The yellow wire runs from pin 1 of V1A to pin 2 of V2A. Note the cut traces between pins 1, 2, and 3 of V2A. including the ground.


The trace where the red wire goes into the circuit board is power to the preamp. We tap into it and send it to pin 1, the plate of V2A. The trace just above it is ground. One end of the cathode resistor goes into the ground trace. The other end is in an open area of the board. holes for the brown wire and the black wire shown in the next picture are drilled next to the lower end of the cathode resistor and all three are connected together.


The black wire sends the signal developed on the cathode into the Blues Junior tone stack. Note that the trace between the black wire and the plate side of R10 is cut. Thus V1A receives power from R10, but the signal from V1A never reaches the tone stack. Instead, the yellow wire (above) carries it to the grid of V2A to drive the cathode follower.

One final thought:

The 12AX7 in V2 may not be the best choice for a cathode follower, since the 12AX7 is designed to be a high-gain voltage amplifier. It’s not designed to supply significant amounts of current. It also has high internal impedance, which is part of the problem we’re trying to solve in the preamp/tone stack interface.

The cathode follower, however, is a current amplifier, not a voltage amplifier. Other tubes in the 12A*7 family are better suited to deliver current, most notably the 12AT7 and the current-driving champ, the 12AU7. They have lower internal impedance and are designed to deliver significant amounts of power. Putting a 12AU7 or even a 12AT7 into V2, however, would reduce the gain of the tone stack recovery stage, exactly like turning your master volume down halfway or more. That’s a bad idea.

There’s one tube, however, that is ideally suited for V2 in the Blues Junior: the 12DW7, also called an ECC832. It’s a hybrid tube; one half is that of a 12AX7 and the other is a 12AU7. It’s specifically designed for tasks that require an amplification stage and a driver stage. Fortunately, the triode on pins 6, 7 and 8 is the high-gain 12AX7 side, a perfect match for V2B. The section connecting to pins 1, 2 and 3 is the 12AU7 side, perfect for a cathode follower.

What’s the Difference?

So does the cathode follower actually sound any different? My customer was looking for more touch-responsiveness and feels that cathode follower tone stacks deliver this. I couldn’t say for sure, but I felt that the follower gives the amp an additional measure of clarity, especially in the bass. I haven’t done an analysis of the tone stack’s impedance, but it stands to reason that the modded tone stack, with the increased bass and mids, loads the preamp more on the low end. As a result, you get unequal amounts of distortion on the low strings vs. the high strings. Remove the load on the preamp, and everything evens out. This is not an easy mod to switch on and off, so I couldn’t do A/B comparisons. But I’ve heard enough Blues Juniors that the tones are pretty much engraved into my brain. This mod slightly increases clean headroom; the amp takes pedals better.

Swapping the 12AX7 for a 12DW7 (actually a JJ ECC83 and a JJ ECC832) did make a small additional improvement, but it may just have been the small, but audible difference between the two 12AX7 stages in V2B, the tone stack recovery stage. Then again, the internal capacitance of the 12AU7 stage and the 12AX7 stage is different enough that it could also affect tone.

My conclusion: If you make the effort to do this mod, you should try a 12DW7 or ECC832 to see if you get additional benefit.


  1. 12DW7?! All right! Old Ampeg tubes are NOT dead after all! I’m actually going to have to get one now. Now I know why my old SVT head that had been modded from 12DW7’s to all 12AX7’s in the front end, ostensibly to make it “easier” to get tubes in a pre-internet world in those years of wandering through the desert tube-wise, after most of the USA & UK tube factories shut down, but before overseas factories got their act together, sounded a little more harsh than others. I have known for years about the gain stage differences in 12A*7 tubes, but didn’t know about the current characteristics. Did I say I’m now going to have to get one?

  2. Update: the 12DW7 came in. Along with the tube, I got a 47pf capacitor, and modded C2 to a pair of jumpers to a DPDT mini switch installed next to the input jack so I can have a switchable “bright and brighter” switch with a choice of the stock 100 pf or a total 147 pf.

    With this configuration, and the preamp tubes going 5751, 12DW7, 12AT7, and this bright switch in, I am getting a rich, silky, clean sound with just the right sparkle that is almost identical to a 1978 Fender Silverfaced Vibrolux I bought brand new and had to sell a few years ago to make a house payment when I was between jobs.

    • Note: Scott has a green board and the capacitor in question is the bright cap on the volume control that keeps trebles alive when you play at low volume. The effect of this pot diminishes as you turn up the volume. The equivalent pot on the cream board is C3.

  3. Without making this mod. Would you still suggest the 12DWY or go with the 12AT7 in the V2 position? I have a non-mod Blues Junior and was looking to change V1 to 12AY7, V2 & V3 to 12AT7. Your thoughts and suggestions please. Thank you.

    • The What About Tubes page has more info on general tube substitution. The 12AT7 is a driver tube, not a preamp. Use a 12AY7 instead to preserve frequency response and tone. If you want darker tone, turn down the treble. The 12AT7 can be used in V3 to reduce drive to the output tubes, but a 12AY7 would work as well, without dulling the highs. There is absolutely no point in using a 12DW7 in V2 without doing the cathode follower mod.

  4. Bill, another possible use for the unused section of V2 would be for valve reverb. In that situation, the 12AX7 would be fine. This is a mod that is incorporated into Torres BBQ Blue mods. I’ve just got my hands on the hand wired version of the kit so I’ll post back how valve reverb sounds. I’m not a big fan of the stock reverb sound – anything past 3 and I feel I should be wearing shorts and a Hawaiian shirt!

    • You realize, of course, that the Torres BBQ Blue isn’t a modded Blues Junior in any sense of the word. He guts the amp and installs a new circuit board. So you have a different amp in a Blues Junior box.

      I’ve seen others use the unused half for reverb recovery, but you’re still left with solid state drive. If you think the stock reverb comes on too strong, you can install the audio-taper pot, which extends the low end greatly. Also, the basic mods deepen the tone of the amp, which has the byproduct of improving reverb tone, too. Finally, the Ruby reverb tank has longer delay and deeper, slushier reverb. It works well in combination with the audio-taper control.

  5. Yes the Torres kit won’t be a genuine BJR for sure! A comparison of his schematic with the factory schematic shows a much simplified preamp circuit, though the reverb appears to be valve driven as well as valve recovered- output from V2A into a reverb transformer through the pan then on to V2B. The audio taper pot is a definite upgrade! I’ll check the Ruby Tube tank also.

    I’m doing the build myself and planning to incorporate several of the mods you have developed. My plan is to substitute 6V6s in the output stage and hopefully achieve more clean headroom. I was planning a 6V6 kit build when I read your “A Very Special Blues Junior” and was inspired (or deluded). By building a hand wired board (and with careful disassembly) I can always restore the BJR to its present state if the results are truly awful!

    Thanks for the many useful tips you have generously given us over the years. I will email you separately about mod kits for the Torrestein!

  6. And if you can’t find 12DW7 or ECC832 anywhere, go to Hammond organ repair shop and ask for 7247. That quite rare tube seems to have many names. Found that out only few days ago. (Could have helped a Ampeg owner on tour earlier with that knowledge, since we had 7247s for Hammonds but no information about 12DW7)

  7. I just did this mod but havent tryed it yet because i ordered some of your other mods and im not gonna close it all up just to open it again, but does the 12DW7 make a large difference, i just bought new preamp tubes but if it does ill get the 12dw7

  8. first question, did the fender ’57 Twin (5E8A) have a cathode follower? i think it does… but not quite sure. can you clarify please?

    second question, the main part of the cathode follower is that the resistor between pin 1 and 2 should be equal to the resistor between pin 3 and ground?

    finally, your wires you used look pretty thick. can I use a 22~24 AWG stranded wire for this modification?

    • Yes, the ’57 Twin used a cathode follower to drive the tweed tone stack. Yes, the resistors should be equal in size for a full, symmetrical waveform. The wires I use are 22 gauge unistrand with 600V insulation. There’s vibration on any freestanding wire in a combo amp, so the heavier insulation helps to damp the vibration. I also reinforce the connection on the main circuit board with a dab of hot melt glue.

      • Hey Bill, thanks for the response.

        I installed the cathode follower this morning, and here’s my outcome:
        with the tone stack mod (.1uF bass cap), the amp sounds very very sharp. It sounds somewhat “metallic”. I guess it’s because the increased bass cap, but the bass knob governs bass AND low mid frequencies. From this, the mid knob governs the higher mids and some treble. Treble knob controls the “chime”. What could I do to get a bit less treble? What do you think of replacing the bass pot with a 1M, and the mids pot with a 50k?

        Thanks Bill

        • Did you do the TwinStack mod? That gives you more separation and independent operation of the tone controls. A 50K mids pot with the TwinStack is sort of a “super mids” control. I do that on harp amps along with the .015uF / .1uF tone stack mods. I haven’t found it as useful on guitar, but if you like more mids, it’s worth trying. I think a 1M pot in the bass will throw the stack out of whack. You can model the changes with the Duncan Tone Stack Calculator.

          I can’t think of any reason that the amp would sound more metallic with the cathode follower.

  9. Why not turn the unused half of the 12AX7 into an extra gain stage? Would it make a difference? Could you use the “farty” fat foot switch to trigger it instead of the farty fat =p


    • With the other mods, the Fat switch is not as farty. Turning the extra triode into a gain stage is a lot of work because you probably don’t want it there all the time. Making it switchable is a hassle. You’re better off getting a Hot Rod Deluxe and saving the effort.

    • Sure–a tube is a tube. You’d have to figure out a way to do bias modulation or opto modulation, though. The bigger problem is where to mount two additional controls.

  10. wouldn’t using the 12AU7 side of the 12DW7 for the cathode follower draw a lot more current, dropping the HT supply? i think leaving V2 a 12AX7, then lowering the cathode load resistor on the cathode follower, along with the plate resistor of the driving stage to maybe 68k/68k would give nice results? i put in a 5751 in V1 on my blues jr. also, it might be nice to adjust the cathode resistor of the driving stage. i’ve been experimenting with this part of my blues jr. putting a cathode load resistor of 47k instead of 100k gave me a little bit smoother sound.

    • The 12AU7 can deliver more current, but the tone stack doesn’t load it all that heavily, so it doesn’t draw down the HT supply.

      47K will give you a different harmonic structure than 100K, but a smaller voltage swing and thus more compression. You also get some loss in volume, so it’s a tradeoff.

  11. Hi Bill,
    Modded Jr user here! A question regarding 12DW7 and cathode follower circuits: You say that with this mod 12DW7 tubes can become a good option for V2. I was wondering if 12DW7s could be also a good option for V2 in the Bassman 59… I was looking for a good 12AX7 for this position, and your article made me wonder. Thanks

    • I read a learned dissertation by Merlin Blencowe on how the values chosen in the Bassman and the 12AX7 get along perfectly. But it’s just a tube, no harm in trying it.

  12. Leaving a tube on with no plate voltage/cathode current whatsoever will wear down the tube alot faster than in normal use. (See: Cathode Poisoning)

    You’d think that they would start making 6C4’s again if they needed one triode (1/2 of a 12AU7). (or, could just use a 6/12AV6 and ground the diode plates)

    Excellent mod! I myself hate having unused tube sections. (Even in AM tube radios where one diode plate in a 12AV6 is left unused)

    • On the stock amp, the unused cathode, plate, and grid elements are grounded together. There’s no voltage present. So the cathode may be hot, and may even poison itself, but if that half of the tube is unused, I guess it doesn’t matter.

  13. Just did this mod. I like it. To my ears, there’s a definite difference. The overall tone balance is definitely brighter, and the bass is much cleaner.

    A 12DW7 in v2 leaves the overall sound a little cleaner. A 12AT7 in v3 makes the amp a little quieter. A 12AY7 in v1 makes it quieter still, and also pushes the breakup to a higher volume setting.

    For jazz and clean fingerstyle strat, the alternate tubes are great. The clean boost mod compensates nicely for the drop in volume. Crunchy strat needs a 12AX7 in v1. The breakup is much more uniform across the frequencies with this mod. Before the bass would tend to breakup before the treble.

  14. Additional observation – I usually run the amp with no fat. This makes the output impedance of the second preamp stage 68k. The output impedance of the 12AU7 half of a 12DW7 in the cathode follower is 425R. According to Duncan’s tone stack calculator, under my typical paramters (TMB all on 12) going from 68k to 425R gives a 6db boost in treble, 4db in mids and 1db in bass. Your results may differ. Turning treble down to about 9 fixes it.

    • i think changing the tone stack’s slope resistor to 68k might make it a bit better. i’ve noticed that after a cathode follower, standard fender tonestack tend to sound thin, and if you look at a 5F6 tone stack, the slope resistor is 56k. we can experiment with that too.

      • Your tone stack sounds thin with the .1µF bass cap and .015 mids cap? Have you changed the coupling caps to .02µF?

        • Actually it turned out I had a bad solder joint in the tone stack. Getting essentially no bass or mids! After fixing that, it sounds just great.

          PS – Bill is correct in his advice not to use a cheap soldering iron…

  15. Hi bill

    I think I have the necessary resistor for this mod, its definitely 100k but is there anyway of telling if it is a 1/2 watt resistor? I received it in a byoc kit for an overdrive, it is blue and about 6mm in length. this may be a stupid question!
    I’ve just ordered the to-20 OT and an input jack, along with coupling caps. any chance you could include one of these 1/2 watt resistors if this isn’t correct? ill happily pay for it, would save me having to pay extra postage!

  16. Do you experience more noise (hiss) with the cathode follower since there is one tube stage more? I’m thinking about this mod, but I’m afraid to get some more noise.

    • I have not heard any additional noise of any kind with the cathode follower. No one who has installed it has mentioned it, either.

  17. Series II owners may want to make sure to add the treble taming cap before doing this mod.

    I added the cathode follower mod to a Series III, and it almost made the amp too bright. I had also done the twinstack mod, and changed the tone caps to .1 bass and .015 mid, but the treble was just ice pick at anything over 9 o’clock on the treble knob. I did, however really like the touch response of the amp and wide open springy bass after doing the cathode follower mod. My solution was to change the slope resitor to 82k, and add a 300pf cap to C9, the treble knob is much more useable as a result.

  18. I’ve got all the mods in place, and I love the sound of my amp. One question did occur to me after I completed this mod.

    For the cathode resistor, you recommend 100k (or 82k if you’ve done the high voltage preamp mod). Is this still appropriate when using the 12DW7/ECC832 instead of the 12AX7?

  19. I did the cathode follower mod and it does take pedals better! I’m glad I did it!!! ,,,Thank’s for all you info on yor site I’m glad I found it…..THe only thing this amp dosen’t have is the reverb pot, the volume pot and the switchcraft input jack and of course the 6v6’s mod …but it dosen’t really need it…SOUND’s AWESOME!!!!!!

  20. I am doing the cathode follower mod and the high-voltage mod. Since the cathode resistor of the cathode follower (V2A) has to be changed from 100K to 82K, should the plate resistor (R16) of the tone stack recovery amplifier (V2B) also be changed from 100K to 82K?

    • No, there is no need to change the plate resistor of V2B. It does not interact with the cathode follower.

    • What brand of tube was it? Don’t use a Tung-Sol tube in V2 with the cathode follower. I think the most reliable are the standard Sovtek and JJ 12AX7s. The small, boxy plate structure is probably better than the long, thin plates.

  21. bill, In order to reduce the number of wires crossing from board to board would it be okay to run the follower’s grid wire (green in your little schematic above) direct from pin 1 of V1-A also on the tube board rather than from the main board?

    • Yes, you can connect right on the tube board. My approach is more “schematic” and uses the two jumper holes, but you can go either way. Since the signal and voltage levels are high feeding the cathode follower, there is no downside to the longer wire–it is not sensitive to interference.

  22. Thank you. In the same context, I checked the full schematic and it appears also possible to connect the 100k cathode resistor from V2’s pin 3 on the tube board (shared with the wire back to the tone stack) to the V2 ribbon’s ground lead which seems otherwise abandoned on the tube board. If there wouldn’t be any significant signal or noise problems with that, it would avoid drilling one of the holes in the main board.

    To take this to extremes, one could avoid drilling the other hole by snagging the high-voltage plate source off the positive lead of C28 (TP29), but that wire would then be pretty long passing over a lot of circuitry; maybe causing problems?

    I agree your “schematic” approach is cleaner, more elegant, in that it puts all the related modifications in a physical proximity that is lost when the pieces are spread around.

    Thanks again.

    • I don’t like soldering components to the back of the board. Drilling a couple of holes is easy and gives a secure mounting and professional appearance.

  23. I noticed that you said you have been using the JJ 12dw7 in your follower circuits. I have read not so good things about the tone of the JJ 12ax7’s. Have you noticed any ill effects on the tone from the ax7 half of the JJ tube?

    • V2 is not that important for tone. It just re-amplifies the output from the tone stack. V1 has a much greater effect on tone. So the JJ 12DW7/ECC832 works fine.

  24. I did get and install the JJ 12dw7 and it made a nice difference. That plus changing from a Mullard 12ax7 to an EH12at7 in the PI has made a nice change to the amp. It added a good deal more punch to the tone, more solid sounding. Can’t wait to get the TO20 installed and eventually go octal. Tons of thanks Bill for sharing info as you do. You ROCK.

    • As an additional benefit I was able to replace V1 with the Mullard I removed from the PI. Also a nice change from the Chinese I had in there. I like the OD of the Chinese but the Mullard smokes it in overall tone quality.

      • I’ve had a number of early failures with the Mullard reissue. If it starts to rattle or oscillate, you’ll surely hear it in the V1 slot.

  25. I’m looking at your photos for the green board, in particular the one showing the black wire carrying signal from V2A into the tone stack (the trace on the plate side of R10). Your caption also says to cut the trace so that the signal from V1A no longer gets into the tone stack. But there it looks like C25 is still connected to V1A. Based on the circuit diagram at the top of the page, that didn’t look right.

    Sorry if I am missing something, but I am about to do this one on a green board and wasn’t sure what do with C25, which is the 250pF cap going into the treble.

    • Sorry, I’m going to be away from my tech docs for a while. And I hate doing the cathode follower on the green board.

Comments are closed.