Building: Amp Boards

Now, on to the amplifier boards! These do the actual work of taking a line-level input signal from your audio source or preamplifier and amplifying it so that it has enough juice to drive your speakers. I'm always surprised how small the actual amplifier components are -- most of the chassis is taken up with the power supply and the heat sinks!  

What does an amplifier do? 

The output from your CD player or preamplifier is low voltage (generally, a maximum of 2 volts) and low current. This signal is fine for sending between a source and a preamplifier. But your speakers, in contrast, require a signal with both higher voltage and a great deal more current in order to produce sound.  The job of the amplifier is essentially to remedy this mis-match, by taking a low-voltage, low-current input signal and creating an identical replica of it, but at a higher voltage and current, suitable for driving speakers. 

The Aleph J manages this feat with two gain stages. First is the input stage, which uses a closely matched pair of Linear Systems LSJ74 or Toshiba 2SJ74 JFET transistors. These transistors increase the voltage. The signal then travels to a pair of MOSFET transistors, which operate as the second, or output stage, increasing the signal voltage and current to a level that will drive speakers. The result is a signal that emerges at the speaker binding posts that is now ready to power your speakers. 

There is, of course, a lot more to be said about exactly how the Aleph J accomplishes this amplification, why there are only two gain stages, what the other transistors are doing, and how transistors generally manage all of this. But we're here to build the amp boards, so I will direct you back to Introduction: What Makes An Aleph An Aleph?, for more details. 

Step-by-Step Build Guide for the Aleph J Amplifier Boards

These instructions cover assembly of the amp boards, but without mounting the big power transistors. We will leave them for a future post, when we mount the amp boards to the heat sinks. 

Step #1: Jumpers and Resistors

Once again following the "small components before large" rule, we will start with five jumpers: 

• R27 (the vertical one right next to C2 -- NOT the square R27a above it) 
• R6
• R30 
• J1  
• J2 

When soldering any components, use enough solder to fill the hole and most of the copper pad. If you've done it correctly, you'll see that a little solder has also migrated to the top side of the board, filling the hole and most of the top pad, as in the photo below. 

I like to use discarded resistor leads for jumpers (you did save them from the PSU build, right? if not, you can skip ahead to solder some resistors, harvest the leads, and come back). You can just solder them in as bare jumpers (easy), but I like to insulate them, if the board position provides the space. My jumpers are red in the photos, except J1 and J2, which are bare. 

To insulate the jumper, I strip a little insulation from the end of some hookup wire (you can use the 20 gauge that we recommended), make sure there are no remaining strands of wire hiding inside (we don't want any unexpected circuit paths should a strand partially slip out and touch something!), then slide the insulation over the jumper lead. If this all seems a bother, then just install the jumper bare.

Continue with soldering the resistors, starting with the small ones, moving up through the big ones, and leaving the potentiometers (R7 and R27) for last (you should set those before you solder them in). 

Because the big, 3 watt resistors can get hot, leave a little space between them and the board surface.

MEASURE EACH RESISTOR BEFORE YOU SOLDER IT, AND DOUBLE CHECK ON THE SCHEMATIC OR BOM TO BE SURE YOU HAVE THE CORRECT VALUE FOR THE POSITION. Be slow and methodical. If you get frustrated or tired or have the urge to rush, stop and come back later. I almost put the wrong part in the wrong place several times, only to catch myself by double-checking (for example, don't trust the part numbers on the Mouser bags--sometimes their label printer will run out of room and truncate an R13 into an R1...). 

Recall the following modification from the BOM: R8 should be a 1k ohm resistor, as shown in the schematic, rather than the potentiometer in the BOM. 

The rest of the resistors and their values are as shown in the BOM and schematic. 

When you get to the potentiometers (R7 and R27), make sure to set them to the correct value before soldering them in. For R7, it should be 1k ohms; for R27, it should be 68k ohms. To measure and adjust their resistance, you'll want to put the negative lead of your multimeter on the lead marked #1 (one of the "ends") and the positive on the middle lead, marked #2 (the "wiper"). Then turn the potentiometer screw until you get the desired reading. 

When positioning the potentiometers, the notch on the housing should reflect the drawing on the board. 

Step #3: Caps

C1 and C5 are straightforward to solder. 

Before soldering C2 and C3, double check that they are installed with the correct polarity. Match the white stripe on the cap (negative, short lead) to the white bar on the board.  

Also be sure you have the polarity correct for C4. Look closely on the board and you will see a polarity "+" sign that is partially obscured. Make sure the positive lead goes in that hole. 

C6 and C7 can present a challenge because of the very limited space on the board for them. If you cannot find a part that fits, you can simply omit them and the amp will work fine. They do not appear in the official First Watt Aleph J schematic. So why are they on our board and in the BOM? There seems to be a widespread practice of using a small film capacitor alongside every larger electrolytic capacitor in the audio signal path. Something about electrolytic capacitors inherently performing more poorly than film capacitors at high frequencies, and thus the use of a film cap in parallel compensates to some degree for this shortcoming of the electrolytic cap on its own. (In the case of my build, I mistakenly ordered WIMA caps with the wrong lead spacing, but I managed to bend the leads and fit them into the board anyway. You should probably go with the correct caps from the BOM instead.)  

Step #4: Diodes

The LEDs (LED2, LED3) have a polarity, which is marked by a flat spot on the side of the LED (negative), as well as by differing lead lengths (short is negative). In fact, there are lots of ways to determine LED polarity. Just match the flat side of the LED to the flat side shown on the board. 

According to the schematic, LED2 is supposed to be red and LED3 green. You can use whatever color you like. Blue is the default for commercial Pass Labs/First Watt LEDs, so I included 4 blue LEDs in the Mouser cart, so you should have red, green, and blue to choose from. I ended up using red and green for one channel, red and blue for the other, just because I was short on greens.

NOTE: if you think you might want a front-panel LED, now would be a good time to think about where you might drill that hole in the front panel and solder an appropriate length of twisted-pair extension wire to the amp board, then solder your LED onto that extension wire (don't forget heat shrink tubing or other insulation to cover any exposed wire or leads). For more about front-panel LED options, read ahead here. 

The Zener diode (D1) also has a polarity. The diode has a dark stripe on one side, and that marks the negative lead. Once you locate the stripe, the picture on the board tells you how the part should be oriented. 

Mount the part flush on the board. The photo above is meant to let you clearly see the board illustration. Here's what it looks like installed: 

Step #5: Transistors

Before handling transistors, it's a good idea to ground yourself to eliminate any static electricity. Just touch something that is properly grounded (anything that would give you a static zap on a day you were wearing a wool sweater) (oh, and don't wear anything wool or otherwise static attracting while handling transistors!).

Time to get out your magnifying glass or phone camera on zoom. Why? Because transistors Q2, Q3, and Q4 all look identical, but they are NOT. 

• Q2 = PNP transistor = ZTX 550

• Q3, Q4 = NPN transistor = ZTX 450

You'll want to look very carefully the transistors, where you should be able to discern their part numbers.

Also, you need to be sure that you have the correct pin in the correct hole. The pin arrangement on the ZTX parts are as follows, as you face the flat side of the part: 

The three holes on the board are marked C, B, and E, to correspond with Collector, Base, and Emitter. Make sure you have the correct leads in the correct holes. The flat side should face the CBE labels on the board. (WARNING: if you are using equivalent Fairchild transistors, instead of ZTX, they have a different pinout!)

This brings us to Q1A and Q1B, the precious matched pair Toshiba/Linear Systems JFETs. 

First, be sure you keep the matched pair together (easy if you bought a matched quad, but don't mix and match if you bought two matched pairs separately, as I did). 

Second, the two transistors should be mounted so that their flat faces are held flush against each other. This ensures that they are at the same temperature when operating. You can strap them together with a tiny zip tie, or with a small piece of heat shrink tubing (what I did). But you will want to bend all six leads to accommodate a flush, face-to-face position before you solder. If you solder them in first, it may be impossible to get them into the correct final position, so do your all of your lead adjusting before soldering either part. 

I found it easiest to bend the leads for each transistor separately, adjusting until the two transistor faces were resting flush (or nearly so) when inserted in the board. I then soldered them into position, and finally wrapped them in heat shrink tubing. I also did it the other way around, by wrapping the two transistors together first, and then adjusting the resulting 6 leads until I could insert the bonded pair into the board all 6 leads at once. That proved more fiddly, but it also worked. 

Be sure that none of the legs are touching when you're done! 

Step #6: Inspect and Clean

Inspect your solder joints (ideally with a magnifying glass or phone camera with close-up zoom) to make sure they look clean and show enough solder on both sides of the board. Clean up any yellow rosin residue using isopropyl alcohol and cotton swabs, making sure to clear away any cotton lint strands. (OK, so cleaning isn't strictly vital, but over time, rosin residue is corrosive and can attract dust.)

And, with that, your boards are done! (Well, except for the power MOSFETs, Q5-8. Those will be covered in the next post!)