help me understand this Dukane PA

[alange5]

hello all.  I've come to this forum seeking help understanding the signal path in a Dukane 1A385 PA.  I'm in the early stages of understanding how tube amps work.  I've built a few amps from layouts over the years.  My next project is a (maybe) 5E3 conversion from an old PA.  Before I start modifying it, I'd like to better understand how it functions in its current "stock" state.



Here's a link to a low-res schematic (slightly higher quality version added as an attachment):
http://elektrotanya.com/PREVIEWS/63463243/23432455/egyeb/dukane_1a385_sch.pdf_1.png


5E3 schematic:
https://robrobinette.com/images/Guitar/5E3P_Build/5E3_Schematic_Clean.gif


Some questions pertaining to the Dukane:


I'm a bit confused by the 2nd 12AX7.  Why is the plate of one triode connected to the cathode of the other?


What's the role of the 12AU7?  It looks like the 2nd half is the phase inverter... is the 1st half another gain stage?


Regarding the tone control, it looks like it's placed after the 2nd 12AX7.  The 5E3 has it after the 1st.  What's the difference?


If indeed I decide to rewire rather than just re-voice, my first inclination is to completely bypass the 2nd 12AX7 and wire according to the 5E3 schematic. Thoughts?


Any help / insight / correction is welcomed and very much appreciated

[PRR]

> Why is the plate of one triode connected to the cathode of the other?

Read-up on "cascode".

Cascodes are quite rare in audio.

> tone control, ...the 2nd ...after the 1st.  What's the difference?

Dart-toss. Difference between a 3-in PA mixer and a basically 1-in guitar amp. If the system is distortionless, it does not matter where you tweak the frequency response.

> phase inverter... is the 1st half another gain stage?

Perhaps phase splitter is a better name. Or "Cathodyne". Where you find this configuration, you "always" find a gain-stage before it.

> my first inclination

I would adapt the MIC input to a guitar and see if it rocks-out.

[alange5]

Thanks for the info. I think my first step will be to play with the mic input, as you suggested.  I plan on using this amp as a learning platform.  I may in fact eventually do a full 5E3 conversion, just for the experience.  For the time being, I'll add a guitar input and play around with cap values in the tone control.


I'm reading up on cascodes a bit.  If I'm understanding correctly, the first triode basically "re-amplifies itself" using it's second triode?  I guess the high-gain cascode explains the low-gain 12AU7 for the last gain stage.... 

[Paul1453]

I would adapt the MIC input to a guitar and see if it rocks-out.

This is an excellent suggestion.   

Often this will produce better results than modifying what you have into a different design.   

Just don't ask me how I know that, please.   

[HotBluePlates]

I'm reading up on cascodes a bit.  If I'm understanding correctly, the first triode basically "re-amplifies itself" using it's second triode?  I guess the high-gain cascode explains the low-gain 12AU7 for the last gain stage....

In simplest terms: the Cascode is essentially strapping 2 triodes to act as a pentode. Gain for the composite pentode might be much higher than a 12AX7 stage, but it's also much lower than 2x 12AX7 stages in cascade (on after the other). Since it might be simpler (fewer parts) to just use an actual pentode, a designer would need to have some very good reasons to use it. Hence, it's just a curiosity or "because-I-can" in guitar amps.

I guess the high-gain cascode explains the low-gain 12AU7 for the last gain stage....

Amps get designed from output to input, so that's how you should analyze them.

The 6V6 output stage is shown to have 17v from cathode to ground (bias voltage), so maximum design input to the 6V6's was probably ~16v peak. The split-load or cathodyne inverter at the 2nd half of the 12AU7 offers no voltage gain but does have good balance if the output tubes aren't drawing grid current. No-gain there means the split-load needs ~16v peak of input to drive the 6V6's fully.

The 12AU7 gain stage before the split load probably offers a gain around 10-11, so about 1.5v peak would drive the 6V6's fully. But negative feedback is also returned to the cathode of this stage, and it doesn't have a full-size cathode bypass cap, so actual gain from this stage is more like 4-5 and so ~3v peak at this stage's grid will drive the 6V6's to full output.

You want decent sensitivity to your Aux input, but maybe you think a pentode is too noisy or prone to microphonics. So you opt for a cascode to get pentode gain and hope it answers some other issues. A 12AX7 in cascode (not the best tube choice in general for this circuit) might exhibit a gain for the circuit of ~100-120, so input sensitivity for the Aux input is something like 25-30mV for full output from the 6V6's. And to allow for a wide range of Aux levels (which could be well above 30mV), a Volume control sits between the jack and the cascode to prevent the signal from distorting the cascode.

A Mic could require an even higher sensitivity, so we have additional 12AX7 gain stages for each Mic input (gain about 55 for sensitivity of ~0.5mV) with level control again to keep things from distorting later stages.

Or at least, that's how the PA Designer looked at it.

[DummyLoad]

you could use the aux input for clean playing - note: i am assuming that the aux input is the one that feeds cascode (v4b).


looks like it'll rock as is. replace the power supply filters and should be good to go for another 50 years.


you could try a 5751 or 12AY7 in either v4 hole: also, try 12AX7 in V3.


split pins 3 & 8 on V4a (first 12AX7) and use plexi (marshall 1959 ckt) values there. 


use A_B_Y switch to inputs.


--pete

[DummyLoad]

attached a schematic redraw for this PA. cool circuit.


--pete

[PRR]

> a schematic redraw for this PA

Much easier to think-at. Thanks.

> 12AX7 gain stages for each Mic input (gain about 55 for sensitivity of ~0.5mV) with level control again

Note that 2 Mic and 1 Aux (not shown on DL's re-draw) feed a mix network. The gain from any one mix input to mix output is around 1/3. So gain from MIC in to cascode grid is nearer 16; from AUX about 100mV.

This gives around 2mV sensitivity at MIC. That's pretty darn hot. (As DL says, the AUX input is a real option, tho lo-gain for headbanger work.)

The 30mV sensitivity at gain-pot recovery (cascode) is rather low. We like 50mV-100mV. Coming all the way up from 30mV suggests that pot-down idle hiss will be an issue. Agree that the cascode is better than a pentode here, for this topology, though many PA amps did do pentode here. An old PA amp might tolerate a high hiss because were often large and speakers less sensitive than a hot guitar speaker.

I have real curiosity about the cascode grid bias divider. For resistor, tube, tube in series, we would think maybe 1/3rd of B+ at the upper grid. Here we apparently have 3/4! The squinty voltage-chart says, I think, 152V?? Taking B+ here as 250V-300V, that's about 2/3rd. (Unless they drive the upper triode into grid current, which would be odd.)

One thing about the cascode: Plate resistance is about infinite so output impedance is essentially the plate resistor, 100K. This gives the treble-bleed network a known source to work from.

Guitar does not need all this gain. I might experiment with making the cascode a simple stage. The way they did it, just jumper pins 1 and 3 (shorts the top triode). Might have to disconnect top triode grid so the 0.1u does not load the new plate node. Snipping the cathode bypass cap too gets gain down by factor of 4. Mix node level now 100mV-150mV, MIC sensitivity now closer to 10mV.

It is indeed a fine foundation for 6 or 8 different classic Fenders and Gibsons. However some of these old PA amps have their own sound. Maybe bland. Lessening the NFB (increase the 6.8K) may jazz them up. And the EQ options were limited, but that promotes self-honesty in playing. There are many other amps with 3 or 13 EQ knobs and I don't think this chassis really begs for featureitis.

[HotBluePlates]

> 12AX7 gain stages for each Mic input (gain about 55 for sensitivity of ~0.5mV) with level control again

Note that 2 Mic and 1 Aux (not shown on DL's re-draw) feed a mix network. The gain from any one mix input to mix output is around 1/3. So gain from MIC in to cascode grid is nearer 16; from AUX about 100mV.

This gives around 2mV sensitivity at MIC. That's pretty darn hot. (As DL says, the AUX input is a real option, tho lo-gain for headbanger work.)

Thanks for pointing these out. I mentally lumped the mix networks in with the volume controls, though they are added loss which reduces input jack sensitivity.

[DummyLoad]

fixed a couple of designations, etc.. added the aux input ckt.

added suggestions for e-cap replacement/rework.

--pete

[jjasilli]

Found this posted on the web

! No longer available

[alange5]

Guitar does not need all this gain. I might experiment with making the cascode a simple stage. The way they did it, just jumper pins 1 and 3 (shorts the top triode). Might have to disconnect top triode grid so the 0.1u does not load the new plate node. Snipping the cathode bypass cap too gets gain down by factor of 4. Mix node level now 100mV-150mV, MIC sensitivity now closer to 10mV.

I'll be able to play around with it next week.  I may very well make the cascode a simple stage.  I suppose I could also bypass it completely, yes?


To all contributors, I really appreciate all the knowledge that has been dropped thus far.  I'll report back when I have the time to open it up.  Feel free to keep chiming in

[HotBluePlates]

I'll be able to play around with it next week.  I may very well make the cascode a simple stage.  I suppose I could also bypass it completely, yes?

You could, but I think everyone advocates you trying it as-is; as you see from the video jjasilli linked, it probably just needs you to install 1/4" jacks in place of whatever else is there (I don't know if the 1/4" jacks shown in the video replace a different kind of connector).

[jjasilli]

FWIW" my vintage Bogen PA's sport the old Amphenol screw-on mic connectors for the mic channels, and RCA plugs for AUX (line level) input.


Also, note that your trannies seem to be the hi-grade leaf blade type, in contrast to the open OT windings (even if hidden by bell covers) typical to guitar amps.  Inter-leaf OT's usually pass a broader frequency spectrum, though full audio spectrum had to wait until about 1962.   The older trannies typically covered 100 - 10,000 Hz.   Interleaf trannies may add more weight than needed for a guitar amp. 


Also, you may have a really good  tube PA amp if you have use for one; stock except for modern input jacks.  Or a decent mono "hi-fi" amp, which may be calling out for a sibling for stereo.  Great for quality bookshelf speakers (plus maybe:  a powered subwoofer; and maybe supertweeters to optimize hi-end above 7000 Hz).

[PRR]

> Interleaf trannies may add more weight than needed for a guitar amp. 

No. It just costs more to wind, and who is going to hear it?

PA OTs were only a slim step up from guitar OTs. Maybe 3 interleavings instead of 2. (You can't run AB hard without fairly good coupling between push and pull winding.) This in contrast to some DIY Williamson windings with 8 or 12 leaves.

FWIW, the early Sunns used Dynaco hi-fi iron with ample interleaving. No guitarist noticed the extended high end. Sunn could do it because Dynaco's winder friend was churning out so many that it became cheaper (in Sunn-size lots) than any other large stock OT.

[jjasilli]

Sorry, nomenclature snafu.  Didn't mean the windings, but rather the multiple metal laminations.

[dmp]

I have this same amp that I modded to use for guitar a few years ago. It rocks. I basically tweaked it more like a Tweed Super with a Baxandell Tone stack (but kept the cathode biased 6v6 output). The knobs on the front are vol1 vol2 bass treble. I switched the middle cascode stage to a cathode follower like the Super to drive the tone stack.
You don't need to do much at all to make it awesome. The output transformer sounds great.
I can post more detail if you have any questions, but this is basically what I did:
- put in new filter caps on power rail and cathode bypass.  I put in terminal strips and put caps inside the box, and left the old metal can in place.
- put two 1/4" input jacks in the back in place of the mic jacks. If I recall correctly they fit the same size holes. Got rid of the aux.
- added a switch on the back between the jacks that jumpers the two channels together (dpdt, so the tip of the 1/4" are connected and the switch to ground on the jacks are defeated) This is nice because there isn't room to put second jacks on each channel for a patch jumper, ala the jtm45. Once you hear it with both channels jumped, you'll want to run it this way all the time.
- The tiny chief caps are nice - you wouldn't want to take them out most likely

I've been playing a farfisa through it lately. It's a great amp all around.

[alange5]

I have this same amp that I modded to use for guitar a few years ago. It rocks. I basically tweaked it more like a Tweed Super with a Baxandell Tone stack (but kept the cathode biased 6v6 output). The knobs on the front are vol1 vol2 bass treble. I switched the middle cascode stage to a cathode follower like the Super to drive the tone stack.
You don't need to do much at all to make it awesome. The output transformer sounds great.
I can post more detail if you have any questions, but this is basically what I did:
- put in new filter caps on power rail and cathode bypass.  I put in terminal strips and put caps inside the box, and left the old metal can in place.
- put two 1/4" input jacks in the back in place of the mic jacks. If I recall correctly they fit the same size holes. Got rid of the aux.
- added a switch on the back between the jacks that jumpers the two channels together (dpdt, so the tip of the 1/4" are connected and the switch to ground on the jacks are defeated) This is nice because there isn't room to put second jacks on each channel for a patch jumper, ala the jtm45. Once you hear it with both channels jumped, you'll want to run it this way all the time.
- The tiny chief caps are nice - you wouldn't want to take them out most likely

I've been playing a farfisa through it lately. It's a great amp all around.

Cool!  I noticed my amp is actually the model 1A385B - the schematic is identical.  Only difference is mine has a screw-on terminal strip for the speaker outputs.  I got around to tinkering with it a little bit.  I replaced one of the mic terminals with a 1/4" jack.  It was indeed a perfect fit.  I also added a 3-prong cord.  Totally stock, the amp is a monster.  It's got some serious crunch - not exactly in a good way, but encouraging to know it's loud and mean and operates quietly.  The plan is to slowly start replacing components headed toward a 5E3, starting with the preamp.  I'm going to completely remove the aux input and most likely bypass one of the triodes in the cathode stage.


Question:  Can I use one of the existing inputs as a 1/4" speaker OUT?  Will having the output so close to the input introduce noise or other problems?

[HotBluePlates]

Question:  Can I use one of the existing inputs as a 1/4" speaker OUT?  Will having the output so close to the input introduce noise or other problems?

It might turn out okay, but I recommend going to great lengths to avoid putting a circuit output near a circuit input. Look at any amp you have or any layout of a classic amp: the speaker jacks are usually on the opposite side of the chassis away from the input jacks. Close-spacing invites oscillation.

It should be a simple matter to get a couple of spade terminals at a home improvement store or your local Radio Shack (if it's still open). Get a speaker cable, unsolder one 1/4" plug, and crimp/solder those spade terminals on to the end of the cable. Now you have a speaker cable ready to go for your Dukane. If you only have speakers in a cabinet with their own hard-wired cable with a 1/4" plug, then look to add a 1/4" jack to your Dukane next to the existing barrier strip (and away from the input jacks).

[alange5]

I jumpered pins 1 and 3 on v2 (bypassing one triode in the cascode) and disconnected the .1 cap feeding the now-unused grid.  Gain and crunch are much more manageable - way less aggressive-sounding.  I then placed the tone control after v1 and wired it like a 5E3.  I must have messed up... the amp got downright nasty.  It was breaking up at 3 on the volume knob.  All distortion and no clean.  I wasn't sure what to do with the 470k resistors in between V1 plates and V2 grid.  I experimented with them in and out of the circuit - similar results.  Tried clipping the v2 cathode bypass cap - not much difference.


I felt a bit in over my head, so I rewired it back to stock.  The V2 cathode bypass cap is still clipped.  I swapped in a 12AX7 in place of the 12AU7 and the amp sounds really nice.  Any danger in having a 12AX7 in V3?  What can I do to the stock tone control to make it more guitar-friendly?  As it sits, the amp sounds beautiful.  It stays crystal clean until about 5 on the volume knob, then starts to break up gently.  Full-on overdrive around 7-8.  But the stock tone control is barely audible.  Slightly more treble clockwise, slightly less counter-clockwise.  Any suggestions?  I'd prefer a 1-knob tone control to a T-B configuration

[DummyLoad]

> I swapped in a 12AX7 in place of the 12AU7 and the amp sounds really nice.

see reply #5. 

next, try lower gain tube in V1 such as 12AY7 or 5751. you could even give 12AU7 in V1 a go as well, however, it'll probably sound "dark" and "lifeless".

--pete

[alange5]

> I swapped in a 12AX7 in place of the 12AU7 and the amp sounds really nice.

see reply #5. 

next, try lower gain tube in V1 such as 12A&7 or 5751. you could even give 12AU7 in V1 a go as well, however, it'll probably sound "dark" and "lifeless".

--pete

awesome, thanks!  I'm definitely keeping the 12AX7 in V3.  Once I get the tone control sorted, I'll be a happy camper.

[alange5]

I've had the amp for a couple weeks and I've been able to tweak it a bit.  Some changes I've made:

• removed the aux channel
• added 1/4" jacks to mic 1 and 2
• removed cathode bypass cap on V2
• bypassed one triode on V2
• replaced slightly out-of-spec 470K resistors from pins 5-8 on the 6V6's
• installed terminals and new filter caps/resistors in the power supply
• swapped V3 for a 12AT7
• wired the 2 volume controls and 1 tone control like a 5E3

Tubes are:

• V1, V2 - Sylvania 12AX7 (came installed in the amp - sound great)
• V3 - GT 12AT7 (tried various X7's and U7's - the 12AT7 sounded best to my ears)
• V4, V5 - JJ 6V6
• V6 - NOS RCA 5Y3

The amp sounds beautiful.  I guess "tweedy" is a good way to describe it... Not a ton of headroom but the breakup is really nice sounding.  It's clean up to about 4 on the volume knob.  4-7 is a really nice "edge-of-breakup" blues sound, and 7-10 is full on crunch.  Currently playing it through a console radio cab retro-fitted with a Weber Chicago 12 (ceramic).


Only one thing I'm concerned about:  I want to make sure it's biased properly.  I have a bias meter.... With tubes installed, on power-up V6 climbs steadily up to about 37ma and stays there.  V5 abruptly shoots right up to about 40ma before leveling out.  V5 (unless I'm imagining things) is always ever-so-slightly hotter to the touch than V6.  V5 plate voltage is about 305vdc.  V6 is about 298vdc.  Cathode resistor measures 190 ohms (schematic calls for 200 ohms). I tried taking readings and doing the manual math but I'm not sure I'm measuring the cathode voltage properly.... Pin 8 to ground only reads 5v.  I'd appreciate some insight...


Thanks everyone for your help with this amp. 

[DummyLoad]

updated the schematic - have no idea what you used for value of R10 - you could make that a master volume with the AUX input pot, otherwise R10 needs to be "something".


V2 cascode removed - i presume you removed the associated components; V2 bypass removed; V3 updated to 12AT77.


there is no bias adjustment on this amp - it's auto-bias, (AKA, cathode bias) 300V * 34mA is about 10.2W. so Pdiss for class A is fine. it's actually less than that. i can guess it's about 19-20V vor Vk of R19,  but then that' s still less than 10.2Watts, so bias is OK. scrap the JJ, get some RCA or sylvania NOS. the only new production 6V6 tube i'd run are the EH 6V6.   


please see attached schema's. reply with comments if you want me to update it.


--pete

[sluckey]

There's a discrepancy between your current reading with your bias meter and the measured voltage on the cathodes (pin 8). Your bias meter says you have 37mA and 40mA flowing through the tubes and that's believable. That's a total of 77mA for both tubes and that current must also flow through that 200Ω cathode resistor. 77mA through a 200Ω resistor will cause 15.4v to be dropped across that resistor. But you only measure 5V! Something is wrong with that 5v reading. Recheck it. If still only 5V, then replace the cap (C2) that's across that 200Ω resistor (R16).

[HotBluePlates]

What Sluckey said.

However, your highest plate voltage reading (305v) and your highest measured current (40mA) still only totals 12.2w and is likely fine. Which is also exactly what you'd expect of cathode bias...

[alange5]

This morning I swapped in a different 5Y3 and took new readings.  This time the measurements were a little more consistent:


Pins 3: 272vdc
Pins 8: 10.5vdc
Bias meter readings: 37ma for one and about 37.5ma for the other


are those OK numbers?  I might swap out that 25u cap across the resistor... 10.5v still seems low

[HotBluePlates]

Lift one leg of the 6V6 cathode bypass cap. Now what voltage do you get across that resistor?

We're asking because the schematic says 17v.

[alange5]

Lift one leg of the 6V6 cathode bypass cap. Now what voltage do you get across that resistor?

We're asking because the schematic says 17v.

about 15.5v with one leg lifted

[sluckey]

Replace the cap.

[alange5]

OK last two question, I promise!


I replaced the 6V6 cathode bypass cap and all is well.  I added the following features:


Single-input and channel selector switch mod as described here:
https://robrobinette.com/5e3_Modifications.htm#Channel_Selector_Switch


Bright switch on channel 2 as described here:
https://robrobinette.com/5e3_Modifications.htm#Humbucker_Switch




Everything works as it should.  My only issue: volume knob interaction has disappeared when using either channel 1 or 2.  For instance: with the channel selector switch set to channel 1, the channel 2 volume knob only adds noise when turned up.  Is this an inherent drawback in the "either/or" switching design?  Or is something wired wrong?  From the 5E3 mods page: "One issue with this setup is the channel that is not in use is not grounded for silence"


Second question:  With a tele, the last quarter turn of the tone knob is useable.  Anything lower is downright woofy.  With a humbucker, it's even darker, even with the bright switch engaged.  Other than lowering coupling cap values, how can I brighten things up?


Thank you all for your input.  This has been a really fun project.  Pete, I never thanked you for posting those cleaned-up schematics.  So... thank you!! 

[HotBluePlates]

... My only issue: volume knob interaction has disappeared when using either channel 1 or 2.  For instance: with the channel selector switch set to channel 1, the channel 2 volume knob only adds noise when turned up.  ...

You volume controls are wired "backwards" like a 5E3? Output from 1st tube stage goes to the wiper, and output from pot to next stage comes from the non-grounded outer lug of the pot.


To hear if there's interaction, set one channel to max volume, set the other channel to half-volume. When switching from one to the other, the max'd channel should have a lot of midrange, while the half-channel should have a scooped-sounding midrange.

Second question:  With a tele, the last quarter turn of the tone knob is useable.  Anything lower is downright woofy.  With a humbucker, it's even darker, even with the bright switch engaged.  Other than lowering coupling cap values, how can I brighten things up?

Just drop the coupling caps from 0.1uF to 0.022uF, then tell us if you still need things brighter afterwards.

[alange5]

final update:


A lot of my noise was coming from the 1/2/1&2 switch with the single input jack, and the switchable coupling caps.  Lots of long leads going to and from the switches.  I messed with it 'til I was blue in the face, but ended up going back to two inputs and no switches.  No more noise, and the two volume knobs interact as they should a la 5E3.  I currently have .022 coupling caps off the 1st 12AX7 and .047 for the others.  Normal input is still dark at low volumes but the bright input is sufficiently bright.  My last and final change will be to add two more jacks to make use of the hi/lo input configuration and leave the option for jumpering the inputs.


Pete, I'm using 1M for R10 and ended up back with a 12AX7 in V3.


The amp sounds great.  I have it housed in a big old RCA console radio cabinet.  I replaced the grill cloth, made a baffle for a Weber Chicago 12", and built a sound-reactive LED circuit housed behind the original frequency selector faceplate.  It glows red as you play, and includes a knob to adjust the microphone sensitivity.  Too cool!


Thanks to everyone who contributed.  This has been a really fun build and I've learned a lot.

[sluckey]

We would love to see some pics of that!

[DummyLoad]

We would love to see some pics of that!

+1


--pete

[alange5]

We would love to see some pics of that!

Front:



Here's the light in action:



Still needs knobs.  For the unused hole, I'm either going to install some sort of jewel light, or I'll relocate the LED sensitivity pot (currently mounted on the rear).  Overkill for a PA conversion?  Absolutely.  But I got the cabinet for free and had the speaker lying around, and I see this being a studio fixture or nice living room piece.


One thing I forgot to ask... What should I do with the heater on the unused triode in V2?  Currently, it's still connected.

[HotBluePlates]

... What should I do with the heater on the unused triode in V2?  Currently, it's still connected.

Leave it connected.

There's no such thing as an "unused heater" for this tube, even if you're only using one section.

[DummyLoad]

... What should I do with the heater on the unused triode in V2?  Currently, it's still connected.

Leave it connected.

There's no such thing as an "unused heater" for this tube, even if you're only using one section.

actually there is. remove either pin 4 or 5 out of ckt. IOW, apply 6.3V to pin 4 & 9 only for section 1 -or- 6.3V to pins 5 & 9 for section 2

EDIT: but then again...why?

--pete

[HotBluePlates]

actually there is. remove either pin 4 or 5 out of ckt. IOW, apply 6.3V to pin 4 & 9 only for section 1 -or- 6.3V to pins 5 & 9 for section 2

This is not true for all 12A_7 types from all makers (kind of like the pin 1 issue in some octals claimed to be "internal connection").

For instance, I just picked up a random old-production 12AX7 which had pins 4 & 5 go into a single triode, pin 9 go into the other triode, then a bridging connection from 4/5 to 9 to form the heater center-tap. So that's both "halves" in both triode sections. Then I looked at a different-brand 12AU7, and its heaters were as you described: pin 9 common to both, with pin 4 going only in 1 triode section and pin 5 going only in the other triode section.

Point being he'd need to inspect each 12A_7 before use to verify which way that particular tube's heaters are arranged. To be fair, the tube will still work but I suspect the lower heater power would cause the tube's performance to fall off sooner than it otherwise would (though this might be 10 years of use instead of 15 years).

[angelodp]

I realize this is a very old thread. I have the very same amp and was wondering if the OP might post any gut shots? I am also curious if any experimentation was done with the NFB circuit. I have replaced the Cathode caps and replaced input/ouput jacks. I will be starting up soon to see what the stock sound is.


Best A