First Watt Amplifier & PSU Build

Very enjoyable thread. Thank you!

 

That is an intriguing suggestion! I may look into how much it could cost to have a sheet cut, although the tricky part might be cutting a suitable ventilation grill.

I think I considered an acrylic faceplate for a previous project, and recall shelving it due to concerns about plexiglass and static electricity buildup. I can't quite remember.

Thank you for the idea, and the kind words.

 

That was a project that I had intended to do to cover the custom crossovers on top of the Altec's.

I got as far as researching clear acrylic boxes. The main reason was to keep them in view but without getting covered with dust.

I realized that there would have to be some slits cut in the sides of the box and I was wondering how I might do that?

I am wondering if it might be better to custom order boxes where the builder can cut some slits in them?

 

I haven't worked with acrylic, but I wonder if it can be drilled without cracking. Instead of slits you could have 2-3 rows of drilled holes maybe. How warm do crossovers run?

What you're describing makes me think of a miniature version of the custom turntable covers some members have. Is that semi-accurate?

 

Sure, acrilic can be drilled without cracking. You would want to support it on the inside.

I use custom crossovers from ALK Engineering that were designed to be upgrades for Heritage Klipsch speakers. Since my Altec's have a 511B horn, I can use the Klipsch crossover that is used on the La Scala.

They don't get hot at all, but placing them in an unventilated box would hold in the heat.

While slots would look nice, so would a pattern of drilled holes.

That is exactly accurate. Go to the ALK web site and check the crossovers out. There are different styles.

 

Very much enjoyed reading this thread! Nice piece of work.

Maybe too late, but if you want to ensure that nuts don’t come loose over time you usually use locktite (the right version!) for that.

As for the acrylic sheet, you might have vents laser-cut. Also leaves very nicely finished edges.

 

Yeah I hoped something like laser cutting was an option. In the end though I already have a perfectly good aluminium cover, and the money may be better spent on more parts - which is what I've done. Aleph J parts are incoming.

The plan was to finish the power supply then drop the boards from my F6 into the chassis, but I feel emboldened, so I think I'll continue on and build the Aleph J as the rest of the project.

 

I love this thread. I just like following along and I’m so intrigued and excited to do some project like this myself.

 

Something tells me you may be crazy enough Steve.

 

The Aleph J

I've mentioned earlier there is little in my knowledge basket when it comes to amplifiers, but hopefully I can afford anyone constructing the Aleph J in the future a reasonable view of what they may encounter along the way.

Apart from one of the standoffs needing a bit of forceful coaching, the holes I drilled in the heat sinks were adequately spaced, and since I had the transistors available I test-mounted them along with the boards.

Unlike the F6 and other diyAudio boards, the Aleph J boards are not mirrored; e.g. the resistor on the far left of the PCB in the below photo points toward the front of the case for the left channel, but the rear for the right channel. The two boards are exactly the same.



That isn't a problem, but it does mean the wiring on one side won't be completely duplicated on the other, which messes with the symmetry a bit.

Other points of interest regarding the wiring is a) the very wide spacing between the three voltage inputs that connect to the power supply and b) the positioning of all the wiring connections along the top edge of the PCB. It may be a challenge to satisfactorily route the signal input and output cables around the span of the power carrying ones.



Above is a wire connection test from the power supply to the amp board which illustrates the problem. It may be a better solution to run each wire independently from the terminal block rather than have them gathered together like this. I haven't tried yet.

Once the signal wires are added it's gonna get a bit cluttered. With the power supply lines approaching from the side it may be okay as long as the signal wires are routed in from the top & bottom.



While the heat sinks were attached to the chassis I did a test fitting of the top panel and it was significantly out. Instead of being square, the case was a slight parallelogram shape and I had to loosen a lot of the bolts, and smack the back side to shunt it into the correct place position.

 

Have been slowly working through populating the Aleph J PCBs.

These boards do contain component variables or substitutions in some areas which may make it slightly confusing for someone wanting to build a standard configuration. It was for me. The info is available in the Bill Of Materials and the Aleph J Build Guide, but hopefully I can help by showing the main points below. And if I've made a mistake, maybe someone will tell me.

Firstly, there were five jumpers to be added.



It seems the four jumpers in the middle of the board (J1, J2, R6 & R30) are rarely used unless there's a problem with oscillation. Over on the left R27 is jumpered as it will be superseded by a potentiometer to be placed in R27A above it.

Next the resistors were soldered to the PCB. There's quite a few, of many differing values, so I worked slowly, methodically through them and hopefully placed them all correctly.



The only notable variable here is R8, which is recommended as a fixed resistor of 1K but can be substituted for a potentiometer if desired. It can be seen in the area labelled LTP Bias. Another alteration that one might consider is the value of the LED resistors, to make the LED indicators brighter or dimmer as needed.

Two trimmers were then to be added, along with the diodes.



Over to the left, in the aforementioned R27A location, a 100K potentiometer will allow the amp bias to be adjusted. It's advantageous for this pot to be set at 68K before installation. The second pot, located in the centre, has a value of 2K and controls the DC offset at the amp output and would best be set at 1K prior to install.

 

I bought my Aleph J already built, from someone who bought a M2. So I was not involved in the build process.

I will note that I kept many of my amps behind the TV, mostly for convience.

I found a photo of the Aleph-J that I had posted on the forum some time ago.


This is the Aleph J.



Here it is sitting on some shelving boards on top of a pair of Rogue M-150 monoblocks that are not currently connected to the Altec Lansing A7's, because the Aleph-J is.

The photo appears to have been taken with the main room florescent overhead lights off and the indirect room LED lighting on.

The main red power LED, as you can see, is rather bright.

Though they are not that noticible during the daytime, the two blue LEDs are rather noticible at night, so the dimming option might be something to consider, as would be a dimmer option on the main power LED.

Although it is not apparent from this photo angle, there is a second blue LED on the PCB board on the right side of the cabinet.

 

I'd just feel lucky that front LED wasn't blue also.

I'm going to be using the same front indicator LED that I did with my preamp. In that case, it turned out to be a little bright but not too bad. Calculating a slightly dimmer setting for the Aleph J, I surprisingly ended up close to the value listed on the schematic anyway.

It'll be interesting to see if it works out as intended, proving or disproving whether I understand what I'm doing in this area.

These are the LEDs I'm using. They have a plastic shell and using the Front Panel Express software I can get a pretty nice fit without any glue. This way they can be removed along with the amp boards, which they will be attached to.

 

You’re doing a very nice (and neat) job on this, keep up the good work!

 

Thanks jrjhoe. Not much longer to go!

 

When I built the Pass ACA I had the choice of red or blue, both of them glaringly bright. I wasn't happy with either choice so went into my parts bin and fished out a pair of old yellow leds. These need more current than the newer ones, so the result was more of a dim glow, evocative of the warm glow of tubes. Very happy with it.

 

Alternatively, you could also just put a resistor in series - of course not very efficient, but with the tiny current draw of leds a perfectly suitable solution to tweak it to the brightness you want.

 

Looking back, I almost wish I'd chosen orange LEDs for that very reason.

This little discussion has me looking into what I'm using a bit closer.

They are red LEDs from the Dialight 558 series which have a forward voltage of 1.7V and a max current of 7mA. The reason I don't have a problem with their brightness may be that they're only 2.3 millicandela which appears to be on the low end of the scale.

Doing the math for the Dialight indicator in my preamp, powered from the remote control power supply of 5V shows I'm right on the max current rating.

5V - 1.7V = 3.3V through a 470 ohm resistor, making the current 3.3V / 470 equaling 0.007A (7mA).

For the Aleph J: The 24V power rail minus the 1.7V drop from the LED equals 22.3V. Divide that by the 4750 ohm value of the LED resistor equals 4.7 mA. Halfway in between the typical & max current rating.

That should produce a satisfactory light level.

 

I think a lot depends on the day or night factor.

They are apparent during the day, which makes this level of brightness a bit high at nighttime.

Without a switch, it becomes a matter of either or.

 

Tbh, I usually just ‘guesstimate’ an initial value and then play around with it. Mostly because it is hard to know what ‘final’ light intensity you’re aiming for.

Maybe obvious, but one of the things with leds can be the lens, so apart from the day/night factor it can also help quite a bit to make sure you’re not staring directly into it.

 

Perhaps the main hindrance when approaching many a First Watt build is the now unobtanium JFETs Nelson Pass may have used in their design.

For example, the 2SJ74 units have been discontinued by Toshiba and although there may be still stock around buying blindly will likely only net counterfeits. Thankfully the diyAudio store sells new versions made by Linear Systems (the LSJ74 I'm using here), although stock levels can be limited or intermittent. A bit pricey too, but they do come selected and matched for this particular purpose. DIY builders are lucky to have this source.



The two LSJ74s are mated face-to-face to facilitate thermal equilibrium, if I recall correctly, and are usually cable-tied or heat-shrinked together. Curiously, the schematic does not specify any of the transistor positions by designation, although that information can be found in the Bill of Materials.

For anyone's reference:
Q1A, Q1B - 2SJ74 JFETs
Q2 - ZTX550 BJT
Q3, Q4 - ZTX450 BJTs
Q6, Q7, Q8, Q9 - IRFP240 MOSFETs

 

For a tense couple of minutes there I suddenly thought I'd put some of the transistors in the wrong way around. Turns out I was just going mad. On another note, the last of the capacitor components I needed arrived, so let's chuck 'em in!



In a nod to audiophile excess I used an Audience Auricap XO for the input capacitor at C1. I couldn't find many exotic caps small enough to comfortably fit in this space - only this and the Mundorf EVO series really.

The electroytics are Elna Silmics, which contain silk fiber for the maximisation of the silky sounds.

Deviating from the schematic, I have used a 10pF polystyrene capacitor in position C5 instead of the usual 5pF Silver Mica. This was for two, admittedly shaky reasons. First, a couple of mentions about 10pF producing more stable & better measuring performance, and second, an opinion voiced that Silver Mica doesn't sound very good.

I don't claim the ability to hear any benefit the inclusion of these components may make, but it does sooth the inner audio nerd just knowing they're there.

The other capacitor positions of interest are C6 & C7. These are film bypass caps for the electrolytics, and could be a range of values, but from the builds I've seen 0.1uF is generally used.



I've researched bypass caps before and from what I know they can be beneficial for high frequency filtering. However, it would seem if the correct value is not selected it could actually be a detriment to performance. So, without the knowledge necessary to find the best value, I have decided to leave them off.

It looks like these caps don't exist in the production version of the Aleph J anyway, either on the official schematic or inside the final product.

Now the components were all in place a rigorous cleaning of the flux residue was performed. These were quite difficult boards to clean, with so many parts (many of them quite delicate) to work around.

 

Before attaching the wiring to the boards, I decided to do a mod that I had seen others perform in order to eliminate hum in their builds. It appears that the Aleph J can be a bit sensitive to interference.

As I understand it, this mod decouples the input ground from the main ground so the speaker return does not pollute the incoming signal with noise. The mod itself is relatively benign so I figured there'd be no harm doing it in advance, rather than having to dive back in if there was an issue after construction was completed.



All the ground connections shown in the above image are connected by one trace on the underside of the PCB. The trace will be cut between the input GND and the rest of the numbered GND holes.



With a sharp knife I scored a 1mm section of the trace and lifted it away. I then gave it a quick rubbing with a small file. This would have been much easier to do on an unpopulated board.

I have seen different options for linking the two grounds back together. A resistor and diode network can be used, or just a resistor. I had also considered using a couple of spare CL-60 thermistors, as they are resistive cold, but should become almost non-resistive during a current fault.

In the end I took the advice of another builder and used a 10 ohm, 5 watt, fusible resistor. As its name implies, this resistor type is designed to behave like a fuse under a fault condition.



Ignore the other wiring here for the moment - only the resistor is part of the actual modification. You can see it bridges the gap created by the cutting of the trace behind.

Also attached now are the signal input wires IN+ and IN-. If using single-ended inputs (RCA) like I am, the IN- must also be jumpered to the input GND. The Aleph J can be wired with XLRs for balanced input and in that case the IN- and GND would be wired separately. I'm not sure if the entire circuit is balanced or not however.

 

Here's another angle, with the speaker output cable now soldered in.



The speaker cable is Van Damme Blue and the shielded patch cable used for the input is also from the brand. I did buy a length of their more expensive shotgun style speaker cable as well, but I thought I could see some faint green oxidisation through its clear jacket and decided not to use it.

Shame really. The copper looked really nice next to all the blue inside the chassis.

More wiring - now attaching the LED indicators for the front panel.



And lastly the power wires are soldered in, and the board is complete.



I'm not sure this is the best way to do it. It's gonna be a bit of a fiddle to get this, with all the dangling wires, into the case.