Friday, July 11, 2014

Headphone Amplifier - Part 2 - And Bad Mistakes, I Made a Few

Previously, I discussed that I started building a headphone amplifier.

I "finished" building it in about a week or so.  I took my time, and didn't spend a lot of time at once.
[ A few in progress pictures are at the top of the post.  Top picture has the hardware in place the capacitors on the input side, audio input wiring, and power rails connected to their source.  The other two show the box taped up for drilling. ]

"Finished" is a poor choice of term, it was assembled.  But it really wasn't finished, testing needed to be done. When powered up for the first time it lit up the tubes.  I took that as a good sign.  As I would learn later, they were WAY too brightly lit up.  Measuring voltages showed me it wasn't correct.

I plugged some cheap earbuds in anyway.  This caused a high pitch whine of protest to emanate from the power supply.  So I shut it down.

The power supply was so pissed at me at this point it stopped working until the next morning.

The capacitors in the output stage were hot.  Even I knew this wasn't correct!

That led me to discover the first problem.  I put the wire from pin 3 of my mosfet on the wrong side of the output caps.  I had misremembered a photograph of a finished build.  Lesson learned, pictures maybe worth a thousand words, but schematics are more meaningful.  Especially when you don't recall the picture correctly.

When I noticed that I pulled the capacitors I ruined out of both channels and alligator clipped the left channel back together with a new capacitor.   No sense in making things permanent until I had it correct this time.  Voltages were measured.  Voltages were still incorrect.

Plugging in cheap earbuds at this time didn't create any squeals of protest (one problem solved). Measured the output rail again.  Still wrong.  Put audio through it.  Still wrong.  No sound.

This led me to start investigating this problem.

Through some trial, error, and conversations I realized I used my soldering iron way too hot for the mosfets.  Aha!  I learned how to check mosfets, and sure enough they were broken.  I put a new mosfet in the left channel and checked it prior to power up.  The mosfet worked.


The circuit didn't.  I was getting 30 volts on pin 3 of the mosfet/pin 4 of the tube.  I should only have had 12-13.

For giggles I tried to put audio through it.  It didn't work.

This led to a bit of frustration.

It also led to an accidental and very poor decision --  Somehow I ended up putting the power supply's full 48 volts on one of my tube's heaters.  It lit up in a beautifully bright golden light, and went dim.  Poop.

Dead tubes aren't much fun.

I'm not sure how I actually hit upon the solution, but while staring at the schematic one morning I simply realized I was feeding pin 1 of the mosfet from the wrong point in the circuit.  I fed it from the wrong side of a resistor.  When I made the change, magic, the voltage was correct at the output stage.

I put audio on it.  It worked. I was partially shocked, and I partially felt like a German soccer fan watching them kick Brazil's butt in the semi-final.

Fast forward a bit and I soldered it in place and redid the right channel too.

So, I got it working.

At times I found this a bit frustrating, but I learned a lot, and this is a lot cheaper than a semester of electronics school.

Things learned:
How to really read a schematic
Soldering is getting better
How to test a mosfet
How to use various functions of a multimeter
Don't trust memories of pictures

Tuition paid:
5 broken mosfets ( some were ruined while building before discovering the iron issue )
2 broken capacitors
1 smoked 12AU7 tube
~30 hours reading, gathering parts, building, thinking, troubleshooting, and correcting. [ I think I could build one in an afternoon now... and it would look more tidy. ]

Curse words invented:
0 [ But I may have used some conventional ones in creative ways.   I can't remember. ]

I've listened to it for a couple of days now and have a few general impressions of the amplifier. For headphones it is the best thing I have.  Bass is tight and big, even out of my headphones which are known for "anemic" bass response.  Mids and highs sound good to me.  I guess I'm not too much of an audiophile, I have no terms to wax poetic about the sound, but it is good.  I don't have any ground buzz which seems to be a common problem if you read some of the build threads on do-it-yourself forums.  This is probably because I used a ground plane, and not about anything skillful I did.

I also appreciate having a physical volume knob while listening from my computer now ( sound board is ASUS Sonar STX and I've got the amp connected through the RCA outs on it ).

This experience also cemented an idea I've always had.  Most audiophiles are full of crap.  If you ever go researching speakers or amplifiers like I did a while back you will encounter all manners people who, if their words are true, possess super-human hearing.   But, I think they should all build their own amplifier at some point.  It SHOULD dispel any of their magic faerie-dust notions regarding audio once and for all.

No comments: