12sq7->50L6 SE...request for comments

[Zipslack]

I decided to try to do something with an old (40's or 50's) phonograph that somebody gave me.  It has a 35z5 12SQ7 50L6 typical cheap amp in it with no power transformer.  I decided it might be useful as a low-power bedroom/practice/recording amp.  I am currently using a 2:1 transformer from another junk donor with a voltage-doubler and just using the 12SQ7 and 50L6 as series-string using the 63VAC before the doubler.  The one gain stage works okay for me - looking for mostly clean to slight overdrive that I can use with pedals for more gain/drive.  The basic prototype build seems to work okay, but there is a slight "buzz" that I need to look into...it may be due to my "spaghetti-wired" prototype and proliferation of flourescent lights, or I may need a small cap across the output transformer...don't know yet until I box it up properly.

Disclaimer:  I used absolutely no curves, charts, or calculators for this...strictly piecemeal and experiment.  I assume that there are places that can be revised for better performance.

Links to inspiration projects:  Sonic MA-1 http://kirtland.com/guitar1/guitar1/MA1.html
Alfred Morgan's 3 Tube Audio Amplifier http://www.bignick.net/images/RadioPix/MorganAmp.pdf
Gibson EH-100 http://www.prowessamplifiers.com/schematics/Gibson/eh100_2.pdf

Okay, here is my circuit as built (using all junkbox stuff, which explains some of the funky values):

[sluckey]

Adding another RC stage between D2 and node A may help. Certainly won't hurt.

[Zipslack]

Thanks for the feedback.  It doesn't sound like 60 or 120Hz buzz...more like a fuzzy oscillation/ringing tone.  I may try the extra RC stage just to see what happens, but I think things will be better once I package everything.  Considering the environment (open wiring, flourescents, computer/LCD monitor, etc), I would have been surprised if I didn't have issues.  Just for giggles, here's a pic of my "butt-ugly" prototype/testing area....

[printer2]

You are using the same winding to supply the heaters as the voltage doubler, right (on the buzz part).

[PRR]

> a slight "buzz"

SE amps need good B+ filtering, even on the final plate.

You obviously are not trying to set any loudness-records with this build.

The change below will clean-up buzz in the final with negligible drop of power output.

[Zipslack]

Thanks for the advice. Turns out the buzz was coming from the speaker...shouldn't leave speakers laying around where students can touch them.  I may go ahead and do the PS revision and see if that helps the overall noise.  Also actually did a quick curve plot and decided to go with 100K and 1.2K on the preamp.

Yeah, I'm going for wife-friendly volume and easy portability. I even have a shell and speaker from a 50's portable radio to put it in. One day I need to build the 6au6 Third-Watter.  I always have my Blues Jr. and '73 Twin Reverb Deluxe if I want to rattle the windows.

[PRR]

> did a quick curve plot and decided to go with 100K and 1.2K on the preamp.

Experimentation is needed.

100K+1.5V is well-loved.... in >250V work.

That sets plate rather close to B+. My gut says to lower the cathode resistor, get more current, get plate more centered in the available voltage swing.

However look at what the Old Men suggested. Tung-Sol 12SQ7 data, page 2, has suggested values. We start at Ebb=90V, the closest to your 110V (we may round-up/down a bit toward the 180V values). In this case the load Rc is fixed: the 500K pot on grid of the power tube. This is a pentode and it's only guitar so we needn't worry about capacitive loading. The nearest chart point is 510K.

We need gain but we also need maximum output swing, since the 50L6 needs considerable voltage *and* you are throwing-away swing with the volume pot (if not full-up).

The best seems to be R1 (plate resistor) at 240K. This is somewhat more than the >7V possible with the best 100K loading. The suggested cathode resistor for 240K load is a stunning 8K! (Maybe 7K if Ebb is 104V.)

So I don't know. Might be worth a 10K pot and a long dial-in session, comparing gain (which is already low) against overload in heavy strumming.

For a buck more, 12AX7 gives two such triodes, ample-ample gain, enough to cover some NFB or tone-control. I agree that a bedroom amp may not warrant the effort.

[Zipslack]

Thanks for the insight.  My electronics education came well after tubes disappeared, so I'm still learning as I go.  Based on what you are saying and what I see on the chart, it would seem that I could squeeze a little more gain (43 as opposed to 40) by using a 1 Meg volume pot with a 500K plate resistor and 15K cathode resistance - am I reading this right?

As far as the 12ax7, where's the fun in that? Everybody does 12ax7 -- I thought it would be interesting to try to use some tubes that most people don't bother with as well as using up some parts that are just taking up space.  Also, I'm not looking for a gain monster - mostly clean with a little breakup at the top - this almost achieves that now.  I've also considered a 12AU6, but I'm not sure if I have one in my stash (6AU6, but not 12AU6).  I'm even using a switch on the pentode bypass cap for a volume boost - without it, it is talking-volume and clean at the top, with the cap it's almost yelling-range with a little bit of dirt on chords/hard strumming.

Either way, it's a fun little diversion that could be useful for pedal-testing and experimentation. I may need to go back and review my understanding of curve plotting. I used the cutoff/saturation values and plotted for the standard 100K and 220K and calculated based on the -0.5V and -1.0V grid values for experimentation purposes.  Guess I need to spend more time with the charts - saves a lot of time.

[PRR]

> am I reading this right?

That IS what it says.

50L6 specs say max G1 resistance is 0.5Meg, so you are "in violation". A worst-case-leaky tube may throw enough grid-current through 1Meg to raise the grid positive, increasing tube current, melt-down. But grid-R specs are often violated. Apparently 99 of 100 tubes don't leak so bad that a 2X spec resistor does any harm. And you are not making a million of these with only go/no-go testing, you have a meter and not afraid to use it. If you got a soft tube, you'll change it or over-bias it or take Rg back down to spec.

12SQ7 current is down to ~~0.05mA, node impedance is up around 250K. Stray capacitance wants to suck-off your treble, and such high impedance radily accepts stray hum/buzz.

I think if the plate-cap-pot-grid wiring is short and clean it will work.

Can't find 50L6 capacitance, assume 50C5 is similar
50C5 - 0.55uuFd grid-plate
Gain grid-plate = 13
0.55*13= 7pFd Miller cap
30pFd stray wiring
12SQ5 Plate = ~5pFd

~~42pFd loading on node

At 0.05mA: 12Sq7 plate resistance ~~1Meg
1Meg||510K||1Meg = 250K

250K against 42pFd is 15KHz

You can't sell this in the 20CPS-20KC Hi-Fi market; it is entirely fine for AM radio. Many-stage guitar amps with "zinger" tone can't afford a cascade of 15KHz stages because it would shave the 5KHz end of e-guitar, a 1-stage is probably fine.

> curve plotting

Way down in the corner, 50V 0.1mA 0.5Vg1, it's hard for me to squint values. Also hard to believe the junior engineer who plotted a sample tube really sweated all those points. OTOH, I'm not sure the R-C table is entirely well-thought. If nothing else, there should be two "optimums": max gain and max output. They are close, and many makers only quote one.

That's why I might try a 10K pot under 12SQ7.

OTOH, an overall Optimum Design would move to dual triodes and a 6V6. Same number of sockets, same iron-count, better sensitivity, lots more power. But with low output demands and pedalboard in front that is the wrong way to go.

BTW: dump the doubler and run 52V B+, that's a quarter-Watt amp. Use a less-sensitive speaker, you can run it flat-out and talk over it without going hoarse. Input overload may be quite low, pot-down the guitar. Or use a 110V/55V switch only on the OT/50L6 stage, leave the 12SQ7 at the big 105V.

[Zipslack]

I guess I'm going to violate the 50L6 then (aesthetic reasons - I have a 1M pot with a nice taper and the proper length shaft and a matching 3-position rotary switch that will control Off/On/Boost).  I tried the pot with 240K and 500K plate resistance.  I think the best ended up being the 240K - oddly enough the pot ended up almost exactly 8.2K at the "sweet spot" of max gain/volume.  I like the overall volume as it is, and I'm using an old Golden Voice radio speaker (6"), so I think I'll stick with the higher voltage on the 50L6.  Just dropped the interstage cap down to 0.001u to reduce some bass...may reduce the bypass caps, too - it has an abundance of bass/full-body and I want more "chicken-pickin" tone that I can roll-off with the guitar tone.  Otherwise, I think I'll call this thing ready to re-build and mount.  Hopefully, next week I'll get it boxed up and done.  I'll post pics and a soundclip when it's done.  Wish I could find a schematic for the Harmony 303 or Silvertone 1330/1341/1339 amps - supposedly also used 12SQ7 and 50L6...be interesting to see how similar this is, or not.

Thanks for all the help - it's been educational...hopefully, I'll make sense of most of it eventually.  At this point, I'm more of an "intuitive" tinkerer with tubes as opposed to having a "quantitative" understanding.

Now, I still have an extra 50L6 and it turns out I do have a 12au6....maybe a Kay 703 or Harmony 303A derivative next....

[sluckey]

The Harmony 303 schematic is right here in Doug's schematic library.

[Zipslack]

The 303A is the more plentiful version that used a 12au6 pentode for the preamp...the original 303 used a 12SQ7.  There are 2 youtube videos of a guy playing one with an SG.  I've searched all over the place for one of these schematics before I started prototyping this.

BTW, here's a pic of the old radio I'm going to use for the housing...I bought it as a "basket-case" (failed repair/rebuild).  Maybe one day I'll do something with those low-voltage tubes....

[sluckey]

I like the radio idea. I just built an Ampeg J12B Jet into an old 1941 RCA Model T-28 radio that was beyond restoration.

[Willabe]

Sluckey that radio Ampeg amp would look very cool on stage and probably freak a few people out to boot.


                 Brad     

[sluckey]

Maybe I can keep this one in the living room. There's a kinda neat story about that old radio, but I'm not gonna hijack this thread. I may build a webpage about it.

[Zipslack]

Very nice build, Sluckey.  

This was a portable (AC/DC) radio...about lunchbox size, 4" speaker, low-voltage tubes, with a leather handle on top.  The back has a full panel with a small door on the bottom corner for the power cable storage.  The radio assembly was mounted on a wood plank that slid into grooves to hold it above the cable storage area.  I'm going to re-use the plank and mount my stuff to it.  It will be uglier than yours, but the back panel will hide the gory detail.  The front viewport where the tuner dial was should line up with the 50L6 so you can see the glow through the window.  The speaker cutout is a little small - I'm actually using a 6" speaker from another radio and will add a speaker-out jack for connecting to a cabinet.  

I already have a donor for my next project - it's a wooden callbox/intercom unit - should make a nice "living room" type of amp.  Possibly a candidate for an old 6BM8 and xformers that I have from an old tape unit...

I guess this will be the finalized schematic, complete with "violated" 50L6, unless somebody sees a major mistake.  Maybe I should call this the Little Violator or Juvenile Delinquent...

[Zipslack]

Okay, I finally got around to moving my prototype mess over to the radio chassis and cleaning things up.  Here's some pictures  (the external chassis shots show a speaker I used for testing, but ended up replacing with a different speaker - the internal shot shows the final speaker and hookup).  Sort of sloppy and redneck-ish, but using only salvage/junkbox parts...total cost $10.00 (that I paid for the non-working radio).  I still need to fashion some kind of barrel sleeves to clean up the front holes, or cut that section out and make a faceplate to sit behind the front panel. I may add a speaker-out jack and I still need to clean the exterior. I made some slight changes to the circuit and experimented with different speakers to find the best fit.  I ended up going back to the 500K pot, no boost switch (pentode bypassed always). 

No soundclip yet, but it sounds pretty decent - volume is plenty loud for bedroom level.  Mostly clean, with a little "hair" at the max volume.  I'm actually really liking what it does with a cheap Squier where it responds well to playing technique.  I'm guessing my prototype spaghetti wiring was causing some problems (interference, oscillation, etc) because the amp is somewhat louder now with slightly more gain than when its guts were spilled all over the table.  Maybe I'll be able to record a very amateurish clip while I'm off next week using a decent guitar and maybe an external speaker for comparison.

[smackoj]

cool lookin' rig zip. love to hear it ROAR!   

[Zipslack]

Not professional playing or recording - just a quick clip to give an idea of what it produces.  I'm using a 1987 G&L ASAT (tele-style, maple neck, MFD pickups) straight into the amp and capturing using an American Audio Pocket Recorder straight to MP3.  I tweaked it in Audacity to get rid of the dead space at the beginning and end.  There is some clipping/distortion in spots where I overloaded the built-in mics, but I didn't feel like re-doing or tweaking things - just my lousy skills on display.  Mainly just noodling and playing with different pickup selections and guitar volume/tone controls.  My 12SQ7 was too microphonic, so I made an adapter and put in a 12AV6 - the sound is mostly the same.


Dropbox link to the MP3 file
https://www.dropbox.com/s/923pqucozdy404r/Radio%20Amp%20Demo.mp3

[smackoj]

Nice! love the jazz influence and the CandW sound. I bet it would sound great for harmonica too....

Jack D     

[Zipslack]

Thanks, my metal/rock days are far behind me (along with a lot of my hair).  I'm mostly into blues and some Southern gospel nowadays, but did a bit of country, 60's boogie, Southern Rock, etc. in my younger days.  I don't even bother to learn stuff now...just noodle around and try to pick out stuff from my (faulty) memory - I never play anything the way it was actually intended or recorded.

I'm not a harp player and I don't know any right off-hand, but I think it would probably do pretty well as a harp amp with a good high-output mic.