Showing posts with label power supply. Show all posts
Showing posts with label power supply. Show all posts

Monday, February 25, 2013

Rebuilding a pair of Heathkit W-2 tube amplifiers

 The Heathkit W-2M amplifier is a fairly rare beast compared to the more later versions of the Williamson line of amplifiers.  It is also perhaps one of the earliest incarnations of a commercial ultralinear circuit.  The earlier predecessor, the W-1, used 807s connected as triodes, with pretty much the same power supply and driver circuitry.  Of note, it's would be an easy project to convert the W-2 to match the W-1, since the layout and even the output transformers are so similar.

Anyway, for my W-2 rebuild I decided to stick with a retro-modern ethos: keep the same circuit but improve the parts and biasing method.  Since I consider myself a bit of an audio historian, I prefer to hear the amplifier as the designer intended.  If someone wants to tackle an improved Williamson design, I would strongly suggest the Chimera Labs modification which will also reduce distortion, provide cleaner power, and improve square wave performance. 

(right click and open to see more detail)

Power Supply Chassis: The stock power supply is rather anemic with barely enough capacitance to reproduce transients and extended output levels.  In my case, I replaced all of the electrolytics, also increasing the values except for the first bucket since it is attached to the 5V4 rectifier.  I also added bypass resistors to normalize the voltage between the two series capacitors.  Parts required per unit: (2) 22uF/350V capacitors, (2) 150uF/450V capacitors, (4) 470k/2W resistors, capacitor mounting hardware, and nuts & bolts.

(right click and open to see more detail)

Signal Chassis: Once again, new electrolytics, also increasing the values to improve power supply stability.  Potentially leaky paper coupling capacitors were replaced by new film units.  The input capacitor was removed and replaced by a small value (1K) carbon composition grid stopper resistor.  Aged resistors should also be checked to see if they are within specification.  This is especially important for the ones requiring matched pairs.  Parts required per unit: (2) .047uF/400V capacitors, (2) .22uF/600V capacitors, (2) 1K carbon composition resistors, (1) 47uF/450V capacitor with clamp and nuts & bolts, and (1) 22uF / 450V capacitor with two lug terminal strip for mounting.

 (right click and open to see more detail)

Bias Supply: The 1/4" phono test jacks were replaced by easier to use banana jacks and 10 ohm resistors which now allow the output tubes to be balanced by using a multimeter: set the meter to the lowest voltage range and adjust the 100-ohm potentiometer until a reading of zero is found.  This biasing method is pretty much straight from the Heathkit W-5 schematic and is much easier than trying to juggle current meters with the clunky phono jacks.  Parts required per unit: (2) 22uF / 150V capacitor, (2) 10 ohm, .5W resistors, (2) banana female jacks with washers and nut, and one terminal strip.

Miscellaneous: The stock RCA jack can easily be replaced with a modern gold-plated unit.  On the other hand, the single speaker screw terminal strip is a slightly more difficult proposition.  At this time I don't have an easy solution that doesn't require a drill press and new hardware.  For now I suggest using speaker wire with small "vintage" spades, or bare solid-core wire.


Listening results: The stock pair of Heathkit W-2s, even with ancient electrolytic capacitors, were still quite pleasant.  They sounded a little muddled and lacking in power, but still possessing a musical charm that many modern units have a hard time matching.  The excellent Peerless output iron certainly helps.  However, this simple upgrade really helped, making the bass firmer and the dynamics better; all while keeping the golden midrange and easy going treble detail.  With new parts, the W-2s really excel at musicality and now rate among my favorite vintage amplifiers.  At least in my listening situation, I would still give the nod to the Eico HF-60s, but that's due to the higher power and the overall sound of the EL34, which is a tube I prefer over the 6L6GC family.  But for a different listener and a different setup with more efficient speakers, it really would be a matter of personal preference.

Tubes: With the rated 4A filament supply, it is possible for the power transformer to handle an output tube with a little more heft than the 6L6GC.  Since I'm not one to trust vintage iron too far, I probably wouldn't go beyond a KT66 type.  I would be curious to see how a set of Gold Lion re-issue KT66s would stack up against the vintage Tungsol 5881s.  Another option for the cheapskate would be to rewire the output sockets to use the 6AR6 which is still cheap and plentiful.

With two 6SN7s handling gain and phase-splitting duties, there is a vast world of tube-rolling opportunities here.  For now, I'm sticking with GE 6SN7GTAs since they are easy to get and sound great for the dollar. 

The 5V4 rectifier can be replaced with the 5AR4 which will slightly bump up the voltage and offer some of that classic Mullard sound.  However, with the high price of good vintage versions, one may want to stick with the more common 5V4 with all the various manufacturers and versions.

Conclusion: Bringing an old tube amplifier back to life and ready for another twenty (or more) years of service is always a feeling of accomplishment.  The Heathkits aren't the easiest to rebuild, not with some of the strange capacitor can sizes and small interior chassis depth, but the time and work was worth the effort.  These are wonderful amplifiers with that vintage smoothness but still modern enough for electronic music and pop.  Highly recommended.

Tuesday, February 12, 2013

New Project: a SE 6CB5A amplifier

After the completion of my 1625 "Command" amplifier, I began to cast my eye around for another project.  This time I wanted to yet again tackle a triode amplifier, perhaps doing a push-pull 2A3/6B4G type.  However, the pesky cheapskate that lives inside my head just couldn't spend that sort of coin on a set of output tubes.  While searching around, I read a thread on diyaudio that lead me to this blog post on Vinylsavor.  Check it out for the signal schematic.

The 6CB5A is a 1950s television sweep tube that wasn't designed specifically for audio, but the curve traces in triode mode are quite good.  This cheap tube can also produce 6-8Ws in single-ended, which is pretty close to the mojo of a 300B.  I also liked the idea of using the 6N7 driver, thought I opted to not implement the expensive interstage transformer, and instead went with RC-coupling.  Hey, this is a budget build after all.

Output and power iron will be from Edcor - I really liked the quality of sound I got from them in the 1625 amplifier.  Chokes will be from Triad and the rest of the parts will be standard bog stuff you can get from Mouser, Allied, and Digikey.  I've also opted for some current meters to monitor the health of the output tubes.


Since a single 6CB5A takes a mighty 2.5A @ 6.3V to light up, I had to purchase a separate filament transformer to power the 6N7 tubes.  Other than that, this is a pretty standard power supply with a fake dual-mono separation for the two channels.

Output iron has been ordered.  Everything but the top plate metalwork has already arrived, so I will provide another post when all the parts have come in and everything has been wired together.  I'm quite curious to hear what this amplifier sounds like since the build/design approach is much different than the 1625 amplifier.  The 6CB5A amplifier will have no negative feedback, a pure triode amplification chain, and minimal parts.  In my experience this should lead to a big sound with a slightly underdamped bass.  Stay tuned to find out!

Thursday, February 7, 2013

The Most Important Amplifier In the World: The Dynaco 70



 
The Dynaco 70 is the most important tube amplifier ever made.  Sure, some may quibble at that characterization, but I can’t think of any other amplifier – at least within North America – that has turned on more audiophiles to the possibilities of tube amplification.  Like countless others, it was the amplifier that turned me on to an alternative way of listening to music. 

Back in the late 1950s, the stereo wars were heating up.  Many audiophiles had invested in the expensive monoblocks or stereo amplifiers from Marantz, McIntosh, and Fisher.  The smaller players – Heath, Dynaco, EICO, and Pilot were nipping on the heels of their more upscale brethren.   David Hafler made the right move at the right time by introducing a $99 stereo wonder – the Dynaco 70.

Sure, this little budget amplifier couldn’t compete with the big boys at a performance level, but it was such a bargain that no one seemed to mind.  The sales numbers certainly back this up.  Yes, the power transformer was woefully underrated for supplying the current to four EL34s, and yes, the 7199 tube is not an ideal driver/phase-splitter, but whatever failings the Dynaco 70 has is ameliorated by the warm forgiving sound that throws a big wide soundstage.  This is an amplifier made for playing music, and if paired with good tubes, the performance is damn well good enough to have made me a tube convert for life.

I bought my first Dynaco 70 at the tender age of nineteen.  I was heading off to college and needed a stereo system.  At the time I didn’t know anything about tube amplifiers but came enamored with them after hearing a McIntosh 240 over at a friend’s house.  I loved the black transformer and chrome chassis look and lusted after the glowing tubes.  When he asked if I wanted a tube amp, I readily agreed.  However, I will admit that I had to hide my disappointment when he brought out the brown caged Dynaco 70.  It certainly couldn’t compare to the McIntosh in the looks department.  But little did I know how this little tube amplifier would change my life.

The amplifier started stock with Siemens EL34s, a Mullard 5AR4, and RCA 7199s.  After a few weeks of that, the same friend suggested running an outboard tubed regulator to supply the front end.  With the shortage of 7199 tubes – this was before the Internet, mind you – I converted the PCB, through the use of a RCA manual and my first soldering iron, to use 6GH8As.  After that, it was modified to run the EL34s in triode.  After that, it was sadly sold to another friend who began his own audio journey with tubes.  At that point I was running modified Scott or EICO integrated amps, a small step before my DIY journey.
A few years ago, I exchanged emails with the fellow who bought the Dynaco 70.  He still had it!  We did a swap and I had my very first amplifier back in my possession.  The original circuit board was long gone, the tube sockets were shot, and the can cap had been replaced by a giant oil power supply capacitor.  This old amplifier was due for a rebuild.

In order to stay true with the spirit of the Dynaco house sound, my restoration was fairly limited in scope.  The input circuitry was replaced with a PCB from Triode Electronics that utilizes a pair of EF86 tubes and a single 12AU7.  The power supply can-cap was replaced with a SDS circuit board that fitted underneath one of the output transformers.  A bit of wire and solder, some new output tube sockets, and the amplifier was ready to sing again.  

For tube selection, I went with some budget Russian valves – EH EL34s, Sovtek EF86s and a 5AR4.  The sole American tube was a late production Phillips 12AU7.  The sound, as to be expected, was a little on the harsh side.  This was due to the “solid-state in a bottle sound” that afflicts many of the lower-end Russian tubes.  So it was time for a spate of rolling.  In the end, the amplifier finished with Matsushita EL34s, Dutch EF86s, a Hitachi 5AR4, and a 1950s-era RCA blackplate 12AU7.  This change made for a much smoother and cohesive amplifier that was quite enjoyable to listen to.

But, I’m not the sort of chap who wears rose-colored glasses.  The Dynaco, even with an improved power supply and new input circuit, suffers from several flaws that stop it from being a truly great amplifier.  
First of all, the power transformer is helpless underrated in the current department which will cause the voltage to sag at higher output powers.  Another issue is the use of a single 5AR4 rectifier to handle all four output tubes.  Sure, it’s within the specification of that tube, but still it is hitting the top of the envelope.  Of course these two problems can be rectified (ha! – editor) by an aftermarket power transformer and a different rectifier or even the use of solid-state diodes (which have their own issues).

However, even with those changes, I believe the fatal flaw of the Dynaco 70 rests with the A-470 output transformers.  Like many things Dynaco – especially on their budget gear – this output transformer was built to a price point.  At higher output levels, the amplifier sounds like the core is saturating, giving that pleasant “tube-like” compression.  It sounds like the images are on springs, causing the soundstage to shrink in size as large dynamic swings occur.  Why do I pin the blame on the output transformers?  I’ve heard the same effect on the Dynaco Mark IV, the monoblock version of the Dynaco 70 that have a much beefier supply that isn’t shared between two channels.

So yes, the Dynaco 70 – and all the new kit versions now available – is still a great introduction to the world of tubes, but it isn’t the end game.  There are much better - and admittedly more expensive – amplifiers that really lift the veil over the music and provide the sort of output that an audiophile wants to hear.