my new project-Acoustic Valve Amplifier

[bigugly]

I bought a piezo contact pick-up and a sound hole pick-up for my acoustic to see which one I liked most and I like them both so I decided to design an amp so I can blend the two. Any suggestions would be welcome.

James

[tubenit]

The size of your posted schematic made it difficult to see all the topology in one glance. So I resized it for you. However, you labled the components with such small print that it is difficult to see.

Hopefully, this will allow some forum members to post some comments about the topology and give you some suggestions.  I wish I could comment, but your design is beyond my understanding & knowledge.

With respect, Tubenit

[tubenit]

I recorded this with the posted amp design here. I played my acoustic thru this with the band mic'd into the sound system for months and it worked out quite well.  I used a Boss digital delay in the passive effects loop. Got quite a few compliments on the acoustic tone.

This is with a K&K western pickup. With the band, I had the volume on about 3 & the master volume at about 7.

http://www.soundclick.com/player/single_player.cfm?songid=10590844&q=hi&newref=1

Note that Larry Carlton plays his acoustic thru a Dumble amp. My guess is he may be using the FET input?

Kevin O'Connor suggests using a larger input resistor like a 10M instead of 1M. I think you could make that switchable and use the amp for both acoustic and electric.  I probably would build an acoustic amp that's about 25-35 watts using 6L6's.

Just offering another idea. Only comment that I have is that I am not sure the single ended idea will give you enough power?  I probably would want the "blend" in the guitar controls or preamp instead of in the guitar amp. That way you can plug into anything.

With respect, Tubenit

[tubeswell]

@bigugly - seeing as how your designs a SE amp for acoustic guitar, why not have GFB (instead of the piddly bit of local NFB at V3b)?

[bigugly]

Sorry 'bout the pic being so big. I broke it up into three parts so I hope it helps a bit. I'll also reference Merlin's first book, in bold, where it pertains to the circuit as most everything I used in designing this amp came right out of those pages.

So here's the pre-amp. J1 is the high Z input. It's roughly 12M due to the boot-strapping of the self-biased cathode follower of V1B.(pg.120) This feeds into the grid of V2A. J2 is the normal input. X1 is a female XLR which connects to a 1:10 mic input transformer which is connected to the switching contact of J2 so I can use a mic instead of the sound hole pup on my gtr. V1A is configured as a regular common cathode gain stage. This then feeds the grid of V2B.

R8, R9 and RV1 form an attenuating network which allows me to blend the two signals by shunting more of one signal to ground in proportion to the other. V2 is a common anode mixer with dissimilar gain.(pg.105) The gain of V2A is more than V2B to make up for the additional gain in the preceding circuit of V1A. A typical FMV tonestack with a master volume is fed from the plate of V2.

Next up is V3A which is another common cathode stage. This is the drive stage of the FX loop. The cathode of V3A feeds into J3 through RV6 which is the FX send. J4 is the FX return. V3B is a variable gain virtual earth mixer.(pg.272) RV7 sets the "wet to dry" mix level when the loop is set in "parallel" mode. It acts as a level control when the loop is set to "series" mode. This then goes into the grid of V4 which is a 6550 in SE, which should be good for around 20 watts.

tubenit, thanks for resizing my pic. I didn't realize how big it was. Great playing btw. My chops are rather sad in comparison. I thought about building a little battery powered FET pre-amp to put inside the gtr but my skills with solid state stuff isn't so good. Besides this amp is just for jamming in my bedroom. I don't play in a band.

tubeswell, I usually don't include NFB in my SE output stages but I have on occasion grounded the OP tube cathode resistor/cap combination through the OP trans secondary which adds some neg feedback into the power tube itself. I think I'll wait until I get to the listening test phase before I make a decision. I'll design the turret board so I can easily add NFB around the OP section of V3B/V4.

[tubenit]

If you're not playing in a band, then I think the SE design is reasonable.  I have limited understanding about your design, but to the degree I understand it, ........... it sounds reasonable and do-able.  Plus you're referencing some solid design info from Merlin.

I'd drawn up a layout for the next step & then post that for comments.  I like designs that are outside the box, so I hope you go for it.

Only comment I'd have at this point is to experiment with different 12A_7 tubes after you build it. I built an external FX-Reverb unit and found I liked lower gain 12A_7 tubes for the FX then the 12AX7.

With respect, Tubenit

[bigugly]

If you're not playing in a band, then I think the SE design is reasonable.  I have limited understanding about your design, but to the degree I understand it, ........... it sounds reasonable and do-able.  Plus you're referencing some solid design info from Merlin.

I'd drawn up a layout for the next step & then post that for comments.  I like designs that are outside the box, so I hope you go for it.

Only comment I'd have at this point is to experiment with different 12A_7 tubes after you build it. I built an external FX-Reverb unit and found I liked lower gain 12A_7 tubes for the FX then the 12AX7.

With respect, Tubenit

Yeah, layout is next. I have a pretty crude proto board with a bunch of tag strips so I'll build it there and get it sounding how I want it before I start drilling holes in garolite.

[Leevi]

I add here a topic which was recently handled here.

According to Tubenit the CBS sounds excellent with an acoustic guitar.
Is the FET or transformer at input really necessary? The CBS has not even the 10M input resistor.

http://www.el34world.com/Forum/index.php?topic=11501.0

/leevi

[tubenit]

My son in law's Carolina Blues Special doesn't have a 10M input resistor either.  That particular CBS has a James tone stack, uses 6V6's at pretty high voltages 450+.  It has a very warm full cleaner tone almost more like you'd think of for jazz amp.

With respect, Tubenit

[PRR]

Mixing may NOT be better-- try it with a mixing console first.

Your two pickups will never be in-phase for all frequencies. There may be a "better" and "less good" phasing. If the rig is always the same and you can re-wire one pickup, OK, Otherwise you may need a phase-switch. In your plan, V1B cathode follower can be changed to a cathodyne offering both phases (actually polarity).

Balance control is way asymmetrical. Suggest R8=240K, R9 200-220K.

Similar issue at R20 R21.

Assuming 1Meg balance pot, assuming -6dB loss at 1/4 or 3/4 turn, you want the resistance feeding the pot near 250K, not 10K. Allow 40K for plate resistance-- 240K-250K from low-Z sources, 200K-220K from plate sources.

I'm confused at V3A. Unity gain to effects loop, but gain of 25 to straight path? Seems like unless FX has large gain, straight will overwhelm FX except when turned to the extreme? Ah, R25 against R20 R21 tends to adjust gain. I'd have to try that to be sure it was smooth and ample. I'd be inclined to abandon the 2-triode plate-mixer at V2 (one triode and 2 resistors can mix) and add some gain after the FX return.

J1 J2 J3 J4 appear to be "PhonE" (1/4") not "PhonO" (RCA) jacks.

Stages have "B+1" etc notes but power supply does not. I'd ass-ume, but I see B+1-5 and 7 nodes on the B+ string. If you know what-goes-where, I'm happy. (I'm not sure why 7 nodes; that's a lot.)

The canonical Fender ToneStack compensates the sound of near-naked strings (nasal). A good acoustic "should" need less compensation, though pick-up-ed acoustic will need more than simply miking for recording. Yes, put Bass Treb to zero, turn MID to full, the Fender TS is fairly flat. So go with it, but I wonder if another plan may be more what the acoustic wants. James? (I'd lean to jacks to insert a $69 7-band graphic EQ... you can't build graf EQ for what they sell for.)

I'm not sure what the levels and gains would/should be for this combination (piezo, hole, mike) and I suspect there will be some adjust-on-test.

[G._Hoffman]

Well, the only thing I would comment on is the input impedance seems low for a piezo pickup.  Piezo's output impedance is MUCH higher than a magnetic pickup.  (Yes, PRR, you've explained to me before how that is incomplete - none the less, the effective impedance that you need to design for is very high.)  That, as I understand it, is what the Dumble FET input is all about - you can not effectively get a tube input stage impedance high enough while still being low noise to sound the way you want it to sound (i.e., good).  So, a solid state input stage for the piezo would seem, to me, a very good idea.  You could probably at least start with the Dumble FET circuit.  Otherwise, you are still going to need a preamp before the amp, and at that point, what's the point. 

I haven't looked at the rest of the amp, so I can't comment, but global NFB seems almost a must, to me, for an acoustic amp. 


Gabriel

[PRR]

> the input impedance seems low for a piezo pickup

James alleges: "high Z input. It's roughly 12M due to the boot-strapping "

Using 330K, 12AX7, and two thumbs, I figure like 15Meg... whatever. Ample.

> you can not effectively get a tube input stage impedance high enough while still being low noise

Tube condenser mikes get higher impedance and low hiss. And some are using "common" tubes, not electrometer types.

[G._Hoffman]

> the input impedance seems low for a piezo pickup

James alleges: "high Z input. It's roughly 12M due to the boot-strapping "

Using 330K, 12AX7, and two thumbs, I figure like 15Meg... whatever. Ample.

> you can not effectively get a tube input stage impedance high enough while still being low noise

Tube condenser mikes get higher impedance and low hiss. And some are using "common" tubes, not electrometer types.

Whoops!  I guess I didn't look closely enough.


Gabriel