Project files: The Bas***ds

What is it?
These are the files for my revised version of the “The Bastard” preamp and RIAA design (see blog posts here and here)

How big are the boards?
- Line stage: 2.15″ x 3.90″ (app. 55 x 99 mm)
- RIAA: 3.0″ x 3.875″ (app. 76 x 98 mm)

What is the status of the boards?
The boards are both in version 1.1. Version 1.0 has been prototyped and version 1.1 incorporates some minor tweaks and fixes to the layouts.

Does it use any special/expensive/hard-to-find parts?
Not really. The 6J5 tubes should be possible to find on ebay and similar places. There is also a 12J5 version which, apart from needing a 12VDC filament voltage, should be identical to the 6J5. As for other substitutes, please search online. The caps in the line stage and the RIAA are Mundorf standard MKP types (the white ones).

Anything else I need to know?
Not really, no. The boards take a +37VDC (or thereabouts) main power supply and draw very little power – you can easily use one of the regulated PSUs I have posted here on the site. The Line Stage additionally needs a +6.3VDC filament supply. Current draw depends on the type of tube chosen, but you can use one of the regulated PSU designs as well as long as you use a 6 or 7VAC transformer.

Downloads:
Download design files here

Related information:
Note: Always read the “intro post” for additional important information about my designs.

Link to the original design/build article (in Danish) is here (20MB scanned pdf).

Project files: New gainclones

What is it?
These are the board files for the gainclone amp and PCBs posted here. They (sort of) supercede the original files from this post – at least the PSU.

How big are the boards?

  • Amp: 2.4″x1.5″ (app. 61×38 mm)
  • PSU: 3.925″x1.95″ (app. 100×50 mm)
  • Bridge: 2.55″x1.6″ (app. 65×41 mm)

What is the status of the boards?
The boards are in version 1.0 or higher. They have all been built and tested.

Does it use any special/expensive/hard-to-find parts?
Not really. You can do what I did and splash out on 0.1% resistors and audiograde caps, but it definitely isn’t necessary.

Anything else I need to know?

  • The LED on the amp board is connected rail-to-rail which means that in a typical application it will drop around 70V, so it should be rated 1W.
  • If you use film caps thicker than 3mm for the bridge board, you may have problems mousing the diodes completely straight (as I had on the one I built)

Downloads:
Download design files here

Related information:
Note: Always read the “intro post” for additional important information about my designs.

You can find additional information about the LM3875 gainclone in the LM3875 build guide over at audiosector.com.

 

New gainclones…

At the start of the year I promised that I was working on some new gainclones – and here they are :)

Actually, it is only really one new amp (and it is “only” a mono-version of the one I’ve already published. The amp design is faithful to the original non-inverting GC circuit that is still used by Audiosector. Along with it is a tweaked version of the original PSU design I did and a new rectifier bridge PCB if you want to stick to the “original” gainclone concept and only have relatively small capacitors in the PSU.

My test-implementation here is a little bit “over the top” as I have used 0.1% Vishay RN55 resistors and Nichicon KZ and KG “Gold Tune” audio-grade caps. Not sure if it makes a difference, but it does look nice I think (and that matters to some people – me included :D). I have also assembled the diode-bridge and I plan to (at some point) make up another set of amp boards and try the minimalist implementation with only 1000-2000uF of capacitance on each PSU rail and see if I can hear a difference.

The assembled set of boards will be used as a dual-mono amp with two 150VA transformers (on order) and the boards are mounted to 25x50x100mm pieces of aluminium for heat sinking (we’ll see if that is enough, but it should be).

The project files will be published soon – I just need to make a few final tweaks to them first :)

Past projects: 6-channel Tripath amp

Well, without time to finish new projects, it’s time to (once again) go digging in the “back catalogue” instead :D

Here is my first DIY multichannel amplifier. This is based on three Tripath boards from Sure Electronics that were “flavour of the month” some years ago. The boards are based on the TK2050 chipset with the output drivers in bridge mode (so up to around 100W output per channel according to the specs). The boards are stock except that I removed the clamping diodes on the input (which protects the ICs from excessive input voltages) and replaced the (small and horrible) ceramic input caps with (slightly better) Nichicon Muse ES bipolar electrolytics. The boards shown here are the first version that was passively cooled, all future versions I have seen are cooled by small fans. The heat sinks get fairly hot (around 50-55C) when the boards are running.

The PSU is a Mean Well USP-225-24. This is a little under-powered for five/six channels if driven at high levels, but I have a small room and even driving the relatively difficult 4 ohm Elac speakers I never had any issues with clipping. I actually have a spare USP-350 PSU that I have planned to put in instead for a bit more headroom. However, since I don’t use the amp at the moment, for now that is just one more point on a very long to-do list :D

I was originally not going to show this until after it’s been rebuilt with the more powerful PSU, but as I am not getting to that anytime soon here are some pictures anyway :)

Work in progress….

It’s been a while since I have posted here, but real life is still getting in the way of my build time. At the moment this means I am mostly starting up new projects, because for some strange reason I can always find time to start new projects – even when time to finish the old ones is nowhere in sight :D

I have been shopping a little at diyinhk.com recently and bought several DAC boards and power supplies, so one a few DACs are in the cards for sure :) Also, one or two gainclones might make an appearance as well at some point. One reason for this is that I rediscovered some pictures of the Audiosector “Patek” amp a couple of weeks ago, so guess who has been scouring ebay etc. for reasonably-priced copper bars lately? ;)

Lastly, the Borbely and the Le Monstre front/rear panels have arrived and I am pretty happy with those, so now I am just waiting for my transformer order and a free weekend and then at least that should be sorted out.

However, I have to say that even with not much new going on there are still plenty of people that find their way here. The blog was at 15k page views in the beginning of January and it is now over 21k views total – I can’t complain about that, so thanks a lot for stopping by :D

More monsters…

As mentioned in my previous post on the Hiraga “Le Monstre” design, I wanted to build my own version of this classic amplifier. I was pretty happy with my original board layout and it seemed to be working fine as well, but there was still some room for improvement so I could justify a version 1.1.

I received the PCBs of the new amp board layout around a week ago and put a couple together along with a couple of extra PSUs for a complete monoblock pair. The front and rear panels have been ordered and the transformer order should be ready to go within the next week or so, but here are some quick and dirty “work-in-progress” shots of one of the monoblocks :-)

The JFET-matcher

While building the Borbely amp from the previous post I had to measure and match a number of JFETs, both for the input circuit and for the current sources. I have done that before using a test setup based on a solderless bredboard and I used that here as well, but afterwards I started thinking if there was an easier way because I always have to rebuild the setup and it can be a little bit fiddly to use I think.

The result is this little PCB. It has screw clamps for input and output connections which means you can connect the leads from a multimeter directly – no crocodile clips and resistor leads required :D. There are connections for the power source and for volt-, amp- and ohmmeters (if you are measuring current sources with a resistor and not the raw Idss of the JFET). In addition, there are a few other features which I thought might be handy:

  • Silkscreen on the back of the boards to keep board connections to hand.
  • A jumper (marked “Idss”) to bypass the resistor/trimpot and measure Idss.
  • 2×3 pattern for a trimpot to allow connections of pots with both inline and staggered pins (soldered or via socket strips)
  • Connections for a switch to control whether the ohmmeter is connected or not. This means that you can connect the ohmmeter permanently to the board using the screwclamps. Then measuring is as simple as connecting the JFET, adjusting the trimpot to read the right current and then remove the JFET and pressing the switch to get the corresponding resistance reading – should help if measuring a larger batch of FETs.

The board measures app. 50x41mm and if there’s any interest I can easily share the eagle files. I also have a few spare boards (as I don’t really need 10 of these :D) so if  you just want a single board send me a message.

A bit more complicated than usual…

Most of my projects tend to be quite simple designs. Apart from the obvious reasons of making the builds smaller, easier and cheaper, I very much like the thought of being able to make something play music with just a few parts. However, every so often I find myself attracted to designs that are more complex. One of these is this headamp/preamp.

This a variant of a couple of well-known designs by “Mr. FET” Erno Borbely, yet another of the greats of audio design. Sadly, the Borbely webshop closed up a couple of years ago, but many of the designs are still used today. This one is a discrete JFET input/BJT output headphone/preamp coupled with a Sulzer-style series-regulated discrete voltage regulator (phew! ;) ). The boards were purchased from ebay seller al_tsankov in Bulgaria. The boards are decent enough in quality but very importantly, they come accompanied by detailed and clear instructions and BoMs/schematics you can easily read and understand – a welcome change from many of the chinese ebay sellers to be honest…

As with (nearly) all of Borbelys designs, this one incorporates some parts that are quite hard to find these days (Toshiba JFETs). I happen to have a small stock of these, but if you don’t this layout incorporates the option of using BJTs for the input as well (which is also why the board isn’t fully populated in the pictures). The circuit is fairly complex for what it does  (but no doubt, as Einstein said: “as simple as possible, but not simpler” :D) with constant current sources, LEDs to keep noise down and so on. Building it wasn’t too difficult as long as you concentrate on the parts and check everything before fitting. Turning something like this on for the first time is always with a sense of trepidation for me because you know the troubleshooting is going to be a pain if it doesn’t work, but in this case everything was fine the first time :)

As you can see, the boards are assembled and tested and working fine. All that is missing is final case assembly when the front and rear panels arrive from Schaeffer (FPX for the Americans) in a couple of weeks.

EDIT 24th Jan. 2014: I got a message from Alex (the PCB designer) correcting my text a little. The Sulzer regulator is a series-type, not a shunt-type as I had written above – sorry. Alex also asked me to attribute the regulator design, i.e. the modifications to the original Sulzer circuit, to Angel Despotov of Analogdomain which I am of course happy to do :-)

A pair of monsters…

As mentioned in my short Hiraga-post a while ago, I have also been trying my hand at Hiraga’s “Le Monstre” design. The Monstre is a simple 8W class A amplifier that originally ran off a huge 12V battery supply.

The red monstre is built on a set of PCBs from ebay seller jimsaudio. This uses the original transistors (except the output transistors which are newer MJL/MJW-types) but the supply rails are raised to 18V for an expected 15W output. The power supply is a CRC-design also from jimsaudio and I am using 35A bridges and 22mF capacitors (because I already had them).

The green monstre is my own PCB layout. This is built on the original circuit but with never transistors based on the modified schematic by Kristian Kluric and deliberately made to be as compact as possible, just under 5×5 cm. The power supply is a compact CRC-type as well, but because of the lower supply rails the capacitors could be a bit bigger (47mF) even if the board is just 8×10 cm.

The amplifier board seems to work fine as it is, but I haven’t tested it in a way where I could check that it is thermally stable over a longer period of time, only that I am able to zero the output offset and play music through it. I made some small revisions to the amp layout and once those new boards arrive I’ll start building a complete dual-mono amplifier on the new boards. The output transistors in this version are TIP2955/TIP3055 but I will probably try to replace them with faster NJW/MJW-types later on.

Happy New Year!!

Just wanted to wish everyone out there a Happy New Year :-) I have throughly enjoyed posting my projects here and it seems you enjoyed reading as well – the blog has just passed 15k page views total and it has been visited by people from more than 70 countries so far, which I think is quite impressive.

I look forward to continue posting in 2014 as well. In the project pipeline already are more Hiraga-amplifiers, more ICEpower-amps, as well as a discrete FET-based pre/headamp – but of course you never know what else will come up :-)

Have a great new year!

PS: Here is a picture of my box of prototype PCBs (app. 3 kgs worth…). Not everything in it is published yet, so here’s something to look forward to as well ;-)

boxofpcbs-1

Follow

Get every new post delivered to your Inbox.