Side Project: A lower-power power-amp

I'm somewhat obsessed with saving power - just ask my wife who calls me the "Heat Miser". 

So, the idea building power hungry, Class A, Single Ended tube amps seems a bit contrary to my lifestyle.  Well maybe.  My prototype 6B4G Triode based amp draws just over 40 Watts... not too bad and I'm still satisfied with that design.  But while waiting the postman to deliver a few extra supplies for that one, I prototyped a simple Single Ended amp based on a 6V6 pentode.  I chose that tube because it has a rather low heater draw for the power output it produces (2.8 Watt to heat it for up to 4 Watts of output).  I used a 6SN7 for a driver.  It's renowned for it's low distortion characteristics.  After a couple hours crunching numbers to bias the tubes, another hour to scrounge parts, another hour to build it, and 10 minutes to find one missing ground, I had a working channel.  It sounded pretty good, so I copied it to a second channel.


Now it sounded really good, especially on semi-acoustic music with nice vocals (Neko Case, Fleet Foxes, Yes, CSN, Jim White, Iron & Wine, etc.). The bass end was great.  After first spark up around 8pm, I literally listened to various tracks all evening - quite a break-in.  It was somewhat unforgiving on bad or old recordings and didn't move me so much on rock tracks, but overall it was very listenable.  Even before measuring distortion, I knew it had more 2nd order distortion than my Triode amp.  Sure nuf, it measures 6.8% THD @1kHz (-23dB 2nd order, -46 dB 3rd order).  Not super, but there's still some room for tweakin.

Here's the kicker - It draws only 22 Watts while putting out about 1.5 Watts per channel - plenty for living room levels on efficient speakers (90+ dB/m).  So for the same carbon footprint of my wife's 5 minute daily hair drying (1200 Watts), I could listen to music for 4.5 hours!  I'll let the speaker draft and tube heat dry my (orange spikey) hair.

More wood fun

My original plan was to use a pair of dual plastic binding posts for speaker connections.  I drilled appropriate holes in the panel for them, then decided that I'd rather use nicer (single) gold posts.  All the ones I found online required smaller diameter holes than the ones that I already drilled.  They also all seem to have cheesy plastic isolation washers.  To replace those and make up for the large holes in my panel, I decided to make my own insulation blocks out of scrap Ipe.  A simple block would float on the panel surface, so I needed a block with built in shoulder washers. Perfect job for a milling machine...

I started with blocks about 7mm thick and milled off about 2mm of thickness except where I needed the 2 donut shaped shoulders.  The workpiece is mounted in a vice and moved under the milling bit using X/Y screwdrive knobs.  Milling a circular pattern with X/Y knobs reminds me of drawing circles on an Etch-a-sketch.  I guess it's good for the hemispheres of my brain to work together once in a while ;-)

I finished the day by doing the final chassis sanding.  First, I used a pad sander with 100 grit on all surfaces.  Some joints also needed some leveling with a hand block to match their connecting piece.  I then rounded all corners with a long strip of glue backed 100 and 150 grit stuck to a work board.

I finished all surfaces with 220 grit.  Finally, I can clean up all that nasty to breathe Ipe dust.  Love the wood, hate the dust.

Knowing the drill

I've spent my last few evenings marking and drilling the aluminum panel.

Tools: Caliper and rule for measuring, square and scribe for marking, center punch and hammer for making center divots for each hole.

Once marked, it's time to drill...

Center Dill #4; Stepped Unibit; Large Unibit, Carbide tipped Hole Saw

All holes (nearly 100 of them), regardless of size, are started small and are opened larger with progressively larger bits.  All are started with a Center Drill.  I have a #2 and a #4.  For small holes or ones which need very precise placement, I use the #2.  The #4 is great for #4 screws, venting holes, etc.  Larger holes then get the small Unibit.  Stepped Unibits are an amazing time saver.  There's no need to swap out bits in the chuck, just keep going deeper until the step is the right diameter.  If you just graze the hole with the next biggest step, it even deburrs the hole.  For the tube socket and output transformer holes, I use a 35mm Carbide tipped Hole Saw.  It takes a lot of torque to cut a hole that size.  My drillpress really bogged down, but with a bit-o-lube and unclogging the cutters every 100 turns or so, I could create a nice hole in about 10 minutes. 

Now for the irregular hole needed for the AC power connector:

Marking with template; Corners Drilled; More holes; Final hole after filing

Then a final scrub with steel wool and a green pot scrubber.

Final Panel

Woodworking can test my tolerance

I spent 3+ hours obsessing yesterday over mere hundredths of an inch.  My dovetailing jig is cool, but when few hundredth's of an inch make a significant difference in the tightness of a joint, I think Rockler could have done a better job of making their jig more precise.

The general idea is that once you have the jig set up for a given width and thickness of blank, you should be able to clamp a new set of blanks in, route away and Bingo... perfect joint.  One problem is that the fence that sets the depth of the cut is attached to the upper clamp, but the clamp has a substantial amount of play.  So the fence can shift up to about 0.05" between clampings.  It took me a while to figure out that even though two passes had identical adjustments, things were shifting around every time I reclamped.  Ok, so I'll remeasure and readjust the fence depth every time.

The next problem is alignment/squaring of the blanks.  The main magic of a dovetailing jig like this is that you route both blanks at the same time.  There's a little alignment stop which is supposed to keep the two blanks aligned relative to each other and the jig.  My first beef is that it is free to float way out of square to the jig.  Ok, I'll square it each time I adjust it (3 screws).  But wait, it's made of fricken no-so-hard plastic and it flexes!  The upper part of the stop has 2 screws so squaring it pretty much keeps the upper blank square.  The lower portion has only one screw.  Trying to square it is like trying to benchpress with one arm.  Without a second tie point the stop rotates around the single screw and flexes based on the position of the upper.  That would be just fine if it was metal which doesn't flex, but as is, the lower blank is almost never square if you simply rely on the stop to keep it square.  In the end what should take about 1 minute of set up per joint takes about 10 and still may result in a bad (loose, tight or misaligned) joint.

To make up for looseness of tightness, the height of the router bit can be adjusted.  I found that the difference between loose and tight could be as little as 0.015".  You girls take note.  Yeah right, like any girls would read this blog.

Anyway, enough bitchin'.  After about 16 test joints and wasting about 10+ inches of length of my good blanks, I made the final cuts and I'm pretty satisfied with results.  It holds together without glue and only one of the 4 joints needs some filing to get the pins and tails to align.

The next step is to saw a kerf (slot) just below the top of the frame for sliding in the aluminum panel.

Enclosure layout design and mockups

Enclosure layout design, despised by many builders, is actually something I've always enjoyed.  Probably because I started making enclosures for my projects with my Dad when I was about 6 years old.  From cardboard with holes cut with scissors to drilled aluminum.  Those projects usually had just a few switches and lights, buzzers and bells, but the basic concepts are the same.  The main work is drilling the holes, but planning where to put the holes is definitely an art.  It involves ergonomic evaluation to make the device usable, but part placement must consider the functional constraints of the innards.

In an amp, that generally means making sure that components which connect to each other are close to each other and things that could induce noise to each other are far from each other.  Since most components connect to multiple other components, various placement permutations result in the need to consider oodles (technical term) of tradeoffs simultaneously.  Unlike semiconductor block/module placement and routing (which I did professionally for 10 years), the amp layout problem is solvable in polynomial time.  In fact, the problem space is so relatively small that even the meager human brain can figure out a set of component placements which are nearly ideal.

For me, noise is the number one consideration.  Electromagnetic induction is used to our advantage in many components of an amp.  Power transformers, output transformers and chokes all work because of electromagnetic induction.  However, that phenomena is also the leading cause for noise in an amp.  Any conductor (wire, resistor, capacitor, transformer, etc.) is susceptible to induced current from another conductor. To be induced, that current must be oscillating and thus has a frequency which can be heard if it or it's harmonics are within the audible range.  Most hum heard in an amp powered in the US will be either 60Hz like the incoming AC line voltage, or 120Hz (2 x 60Hz).  Either is certainly audible and annoying.

It's near impossible to eliminate the source of electromagnetic induction, but we can shield sensitive components from it, or keep sensitive components far from problematic sources.  Fortunately, induction has an inverse square relationship to distance.  So a small bit of distance between components can have a dramatic effect in the amount of induced current.  Thanks Physics!

Since transformers are the #1 source of EMF in an amp, I feel it's most important to place those first starting with the power tranny.  In this amp, I decided centralize all the power components in the center rear of the amp (see below).  Then the audio portions of the system can exist "far away" on either side. I did allow one twisted pair carrying AC to run along the side, to the front to the power switch and lamp.  This was the single case where I let ergonomics and aesthetics trump performance.

My other layout goals were:

• Arrange all inductors (transformers and choke) for minimal electromagnetic interaction (eg. Orthogonally oriented cores, spaced as far as possible or on opposite sides of a grounded shield)
• Keep AC lines > 2cm from DC rails or signal lines/components (AC lines include power entry/fuse/switch/lamp and tube heater lines)
• Keep AC lines as short as possible (power switch and lamp exempted)
• Arrange signal path components so the path is roughly linear without recrossing (ie. start from front center, out to the power tube, back to the output tranny, then back to the speaker terminals).
• Employ a "Star" grounding scheme with the lug very close to the amp center and close to the power channel
• Optimize heat dissipation.  Keep heat generating components (transformers, power resistors and tubes) away from each other and away from heat sensitive capacitors.

I decided to use Google sketchup for doing my layout.  It's a trip constructing in 3-D and took me a few nights to become productive.  It's an awesome tool though and one of the coolest things is that there's oodles of models for most everything you can think of online.  I found switches, jacks, tubes, etc.  The things I was left to construct myself were the James output transformers, power supply caps, power transformer cover, and circuit boards.  Here's what I came up with for the top view:

The output tubes don't quite look like 6B4G's, but good enough for government work...

... and after lots of jockeying trade-offs, the underside view:

... now "eyeballing" with real (top and bottom) parts:

Time to start planning the final build

With most of my design work complete and measurement and listening tests of my prototype showing nice results, it's time to start planning the final build.  That means final part selection, wood frame prototyping and layout planning.  I've made probably 15 orders in the last 6 months from 6 different US vendors, 1 Chinese, and the rest on Ebay.  Lots of part swapping means my parts bins are brimming with parts to be used on future prototypes and projects.  My current prototype uses all of the components of my most resent design revision except for the power switch, jacks, etc. and it still uses carbon resistors which will be swapped with metal film.

For the wood frame, I've decided to use some Ipe (aka Brazilian Walnut or Iron Wood) boards which my coworker had left over from his very nice deck (Thanks Jeff!).  It's super dense, fairly unforgiving to work and it creates a nasty dust which I shouldn't have been breathing, but it's fricken beautiful wood.  The boards were roughly 6" wide by 3/4" thick.  I had a 2.5' board and a 1.5' board.  I ripped them (roughly) in half and had 4 2.5" strips.  Perfect!  I even justified a new carbide tipped ripping blade which did an amazing clean cut.

I wanted to finally try my hand at some nice half blind dove tail joints.  I've had a jig to do those for years.  I spent a many hours one weekend learning how to set it up and use it.  There's about a dozen adjustments and it's imprecise enough that the only way to really get a tight joint is by iterative trial an error.  Keeping everything aligned and square is the key.  Also logging measurements for each trial.  I had some hemlock blanks which had very similar dimensions to my new Ipe blanks.  I used those to hone my dovetailing skills and made a few small frames.  Here's a "half size" frame about 7" x 5" with a kerf (slot) for the aluminum top panel to slide into:

Driver Design and Fun with Distortion

My original *mono* prototypes had the 6B4G output tube biased using "Grounded Cathode biasing" or "Auto Biasing".  They included a 6SL7 driver tube configured in SRPP (Series Regulated Push-Pull).  The driver preformed pretty well: decent gain and distortion, but because SRPP has the cathode of one triode at half the HT voltage (125v in my case), I had to drive it's heater from a supply separate from the heaters of the output tubes or else I'd exceed the maximum heater-to-cathode voltage of the driver tube.  In the end, my power transformer had 2 heater windings, so I probably could have made it work by using one for the power tubes and one for the driver tubes, but I had already set off into the land of fixed bias for my output tubes with a simple 6SL7 common cathode driver (see my previous post).  This reduced my tube count to three (2 output tubes plus the 6SL7 with one triode to driver each channel). 

My first test with that configuration in October resulted in a much better overall sound.  The details were nice, gain was lower than expected but the distortion profile was generally more pleasing (~4.5% THD@1kHz, -27dB 2nd order, -35dB 3rd - see red spectrum graph below).

Still not outstanding, but better.  However, that was just one channel.  Unfortunately, duplicating the circuit to the second channel resulted in almost double the distortion, but more gain.  WTF!  Both channels were using identical components.  The only difference was the cathode bypass capacitor on the driver triode.  I had used a junk box generic 47uF cap on the original channel (the one with the good distortion) and a decent Nichicon VX series 47uF on the second channel.  Sure enough, swapping the caps pretty much swapped their distortion and gain profiles. So the crappy generic cap had some magic mojo for reducing distortion...

Well, kindof.  Turns out, it's value measured by my multimeter was about 0.02uF.  For a bypass cap, that's essentially not bypassing much other than very high frequencies.  It wasn't a magic cap, it was a blown cap which wasn't far from no cap at all.   Many designers believe that with a proper design, a bypass cap shouldn't really be necessary.  It increases gain, but also increases distortion.  It's used most everywhere because we generally don't want to "throw away" gain.  So to not include a bypass cap means that the lower gain has to be designed into the whole amp.  Fortunately, the 6SL7 driver I'm using has a high transconductance and yields about 25x gain even without a bypass cap.  Thus with an input signal of 1V RMS, I could expect about 25V RMS (~ 35V peak to peak) on the grid of the output tube which is just below the maximum for the operating point that I selected for my output tube.  Almost like the 6SL7 was born to drive the 6B4G!

Look Ma, no driver bypass caps!