seb:diy:audio

Headphone Amplifiers .01

2006-08-01T14:27:00.000+02:00

Here are a few pics and infos about the headphone amps i've built. It's where it all got started, my DIY bug. Proves damn well anybody can do stuff. I've no experience in electronics. Well, before the first project i didn't. Theory or practice-wise. Zero. Nada. Zilch.

What's needed is some documenting and time. And a lot of patience. Loads and loads of it.

Anyways, so there i was, looking for something suitable to amplify the headdies of my iPod, a pair of Sony's they were. HeadWize is the place to be when looking for headphone DIY. Everybody's nice too. The CMoy seemed the best of all worlds : accessible for newbies in electronics, and good result for minimal cost. So i printed the parts list and went to the local "el cheapo electronics" store. I was going to mail-order better parts for better amps if this one turned out good. A proto-type of sorts it was going to be. The first one, i decided to play it sure ; i used the tutorial done by Tangent, and followed it *exactly* as written. It worked. Immediately. And gave results, too. Wow. I was amazed. Here i had this miserable looking result done in a day by myself with miserable low-cost parts, and it sang. What's more, having had done it, and by studying the schema, i now knew the signal path, and started to understand (little by little) how it worked. This meant i could do another design and not be too lost between the resistors.

I spare you the visuals of my first attempt (don't have any pics).

I was fascinated by the clean looking designs of Sijosae, and decided to use one of his as a basis for my second CMoy. So i drew the plans and was ready to go. This is how it turned out (click on it for a bigger one) :

Nice and clean. Here's the underside :

But i found it to lack some bass (i had used 470uF caps), and wished that there was a way to adjust the gain without desoldering and resoldering again everytime. I started to get into socketing everything. This would allow for swapping the caps and the gain resistance whenever it pleased me and save me a lot of headaches. Later on i modified the second CMoy and socketed everything, you'll see a pic of it in the end. In the meantime, here's the final result housed :

Once i had gotten the CMoy how i wanted it, i wished to go a little further. I decided for A47. This time i went directly for the Sijosae design, with a few modifications for my needs. Everything socketed. Here's the result :

And here it is housed :

And the sound in all that ? Very nice, full and round, with deep & punchy basses. The sound is more "airy", less uni-dimensional. But i must say that since i burned my Sony's and bought a pair of Koss PortaPro's, i find less benefit from these amps. The difference is not so enormous anymore. It's still there, sure...it just isn't as obvious as it was with the Sony's. It isn't "Wow" anymore, it's more like "hmm...ah...yes...yes indeed".

We'll see with the next one :)

Here's the last pic with them two in comparaison, and on this one you can see the modified CMoy how it is today, socketed and with the 1000uF caps and all. Loads of fun, these little projects...

Merci DIY :)

Modding the T-Amp

2006-07-30T10:45:00.000+02:00

Intro & disclaimer :

Ok, so this is my take on the ever-growing movement on modding the wonderful T-Amp.
I've not invented anything, nor have i come up with a better method. I'm not even fully understanding the electronics involved ; i just followed the instructions floating around on the web.

So why do yet another page on these mods ? Simply because i didn't find it easy for myself to find a concrete case of building. No photos. A few schemas. A lot of information, a lot of different pages, threads, posts, opinions, parts, notes...you get the picture. So i wanted to show my project as a "real-world" example of how it got done. With pictures, my plan, the parts. Nothing more. All this in the hope of helping someone who, like myself some time ago, is looking for informations on how to mod his/her T-Amp, the beginners way.

But, and understand this well, this is *not* a tutorial. I will not explain the basics of electronics, nor the PCB of T-Amp, nor the soldering procedures, nothing. I just show my project in different phases, with everything noted on the photos. I'm not going to explain *why* i wanted to go where i did, just *how* i got there and what it looks like when done. I recommend you read the basics about modding the T-Amp on the pages i give links to just below, and then come back here to see how i did it, you'll find out that you suddenly understand it all. My mods mainly concern the "Version 3 - STEALTH" mod @ Michael Mardis site. Concerning the power stocking caps, visit and read the thread (and others once there) @ DIYAudio.com. With the information contained in the links & everything you'll see here, you should be fully equipped to jump in and do one on your own. You should be...but as in with everything in life, you could screw up the whole thing as well. And therefore i'm obliged to say that you'll do any of this at your own risk, and i certainly can't be held responsible for any damage generated, should it be on yourself, on your T-Amp, your loved one, dog, house or neighbors ;)

Oh yes, before going any further ; this is definitively *not* the most sophisticated mod procedure, nor is it very advanced (you could go a lot further yet), not necessarily even a very pretty one, but it covers the most important requirements : replacing the most critical parts (for the sound) with better equivalents (the capacitors), which adds up to bettering the overall sound quality, and enhancing the bass.

Some url's to start with :

- the famous 6Moons review of the T-Amp :
http://www.sixmoons.com/audioreviews/sonicimpact/t.html

- different mods explained @ Michael Mardis site :
http://www.michael.mardis.com/sonic/inputmods.htm

- going deeper into the anatomy of the T-Amp :
http://www.diyaudio.com/forums/showthread.php?s=&threadid=50870

Background :

This project started out when i discovered the modern-day-audio-wonder T-Amp on the web. Having had done some headphone amplifiers for starters, i wanted to do one for my living-room as well. I decided to make it a global project, and do speakers as well (more on those in my next post). So i bought a few on eBay, and started to gather information while waiting for them to arrive.

The goals :

The most popular mods remain accessible and concern all the replacement of the main power cap (C10 on the T-Amp pcb), and from there go into further lengths. I decided i wanted to replace all the most important components in the audio circuit, likely to affect the sound the most ; the main power cap (C10), the input caps (C3 & C4), and the output caps (between the speaker terminals). Also, i was going to replace the stock potentiometer with an ALPS one. And add a few power-stocking caps for more reserve.

Of the other components, i wanted to replace the input for mini-jack with something more of quality (i use the T-Amp only with my iPod), add RCA input, and leave one spare input for whatever might come up later on. The speaker terminals had to be replaced as well.
I thought about installing a 12V battery with a chaging circuit, but decided to leave that for later on, and instead re-used the battery case that comes with the T-Amp (is cheaper and keeps it more compact for the moment)...

To finish, everything had to be socketed (versus soldered on-board) so as to allow for experimenting if ever i should wish to try and change a cap here or there. To keep everything organized, i decided to go with an extra board that would house everything once finished. Sure, it adds to signal length, and probably isn't to everybodys liking, but i'm sure it doesn't make *that much* difference, not to my uneducated ears anyway...

Result :

Here's what i got after some never-ending & nerve-wrecking evenings of hard labor. Click on the image to see a larger one :

Getting dirty :

Wow, are we getting to start yet ? Well here's a starting point : an unmodded T-Amp pcb on the left, and the modded one in the middle (the top-side) and right (the bottom-side). The red boxes contain the modified areas. Click on the image to get a larger one :

So whats been modified exactly ? On the top-side :

- the resistor R01 is gone
- the cap C4 is gone and bridged
- the cap C10 is socketed
- left & right output have been socketed
- the power input (from batteries) has been socketed
- the led has been socketed
- the whole block coming from the pot has been socketed

On the bottom-side :

- the resistor R02 is gone
- the cap C3 is gone and bridged

Now go and do it :)

----

Once the T-Amp pcb is prepared, here's my initial plan i used to get everything in place on the extra bread-board. NOTE : this is a top-view. Click on the image to get the big one :

Look confused ? Non-conventional ? It probably is. I've never drawn a circuit in my life. Hah.
Try this to go with it. Click again for a larger one :

Ok, so everything is measured to give a nice size without cutting up the original pcb. It fits snuggly in the middle. You'll notice that i soldered a cannibalized "holder" that'll keep the T-Amp board in place even if you'll rock with it in your pocket (not that i could imagine any reason for that). Economized me having to drill for bolts and nuts too. What you can't make out from the photo (if ever that should be the case), you'll find it on the drawn schema. It's pretty self-explanatory, really.

Oh, the only permanently soldered component is the 10000uF Elna power Cap, it just had legs far too thick to socket.

----

Once the extra board was finished, it was time to assembly. Click for a larger image :

That's about all there is to it. What remains is to wire up everything (go back and look at the first photo), and box it up. Mine is still in the open, the speakers wired up directly in the output sockets. Oh, and just so that there's no confusion, the switch goes where it's indicated on my drawing, in case you noticed it was hooked elsewhere on the photo (i forgot it while testing some stuff). In addition, you need to jumper together the V & + sockets that came previously from the pot. If you use external power stocking caps, like i do, and put the switch between the V & +, you'll be letting power permanently into the caps, even when the amp is switched off. This is why you need to put the switch between the power source and the board.
Also, there's no ground input on the ALPS pot, this is why you need a ground jumper, as seen on the last photo. Should you forget it, you'll be sure to hear it...

Parts :

Here's what i bought and how much i paid for them here in France @ selectronic.fr:

- 2 x Black Gate Rubycon BG-C 2.2 uF (input caps, currently in use) -- 2 x 2.50 euros
- 2 x Elna RFS 4.7 uF (input caps, to test and see between them and the Black Gate ones) -- 2 x 0.60 euros
- 2 x Black Gate PK 0.1 uF (output caps, currently in use) -- 2 x 0.60 euros
- 2 x Black Gate PK 0.22 uF (output caps, to test and see between them and the 0.1 uF ones) -- 2 x 0.80 euros
- 1 x Elna Silmic II RFS 470 uF (power cap, replaces the C10) -- 8.80 euros
- 1 x Elna 10000 uF (power cap, on external board) -- 10.80 euros

Here's what i had already and that got used :

- ALPS RK027 potentiometer (replaces the original one)
- 2 x Elna 4700 uF (power caps, on external board)
- Mini toggle switch (for turning the power on/off)
- "bread-board" type pcb board

Conclusion :

There, hope that yours is turning out as much pain and fun as mine did. However, it's while listening to it that the fun really begins ; it sure is some alien stuff, so much wonderful sound from such an ugly and fragile thing, not to mention cheap to top it all. I'm using it to feed my DIY Cornu Spiral Copy Horns (see my other post), and what can i say..? It's a killer combination of some seriously unconventional (unrational, even) thinking. Wonderful.

Merci DIY :)

Cornu Spiral Copy Horn

2006-02-08T07:26:00.000+01:00

Intro :

This is a speaker project i did in parallel with the T-Amp. As all the best projects, in the beginning, it was never meant to be...

I was surfing the full range speaker section @ diyAudio.com in search for a project to go with the modded T-Amp i was doing (see my other entry here). I kinda decided from the beginning that i wanted to go with a single driver design, but couldn't find "the ideal project". There were some very nice designs around that i was interested in, but the budget was an issue ; indeed, i couldn't resolve myself to pay 200+ euros for a pair of drivers, not for the first speakers i was going to do. I ended up posting my question on the forum, and got a lot of helpful replys (the folks over there are very nice). I even decided on the project i wanted to do in the end : the Cyburg's Needles, with VISATON FRS8 drivers instead of the original TangBand ones. Everything was fine in the universe. Back then.
This all was before i saw some photos of the Cornu Spiral Horns...

Now, i'm a sucker for beauty. Really. To the point of being able to put aside rational thinking when a mirage of magnificence presents itself before me. To the point of forgetting all the "reasonable", "practical" arguments that go against that beauty. To the point of being possessed, obsessed even, by that beauty. To be able to convince myself that the obtention of that beauty could be within my reach, if only i try hard enough. No matter if i have no idea how, nor the skills needed. And that's exactly what happened here. Daniel, i blame you for all this ;)
Boy, once i saw the Cornu Spiral Horns, i knew i was in trouble...

Exit Cyburg's Needles, exit all rational thought, i had no idea how to get there, but i knew i wanted to construct a pair of Cornu Spiral clones. In a way, i was in love. To each his own.

I surfed around, trying to find more information about them. I was sure that there were clones built already, i just couldn't be the only one in my condition ? Surprisingly enough, the informations were rare. No documented DIY project. But a few projects nevertheless (see below for links). None in "two layer" design, all "simple layer" (two squares with the spirals rolled inside). None in all-wood. I found it strange. What's more, the design seemed to work, even with the modified projects. That's when i decided i wanted to go all the way, two layers, all-wood, the works. So i tried to track down every bit of information available on the net. I found out in a comment @ the Cornu site, that for budget drivers the MONACOR SP-50X will work. Even better then. I downloaded the photos, the drawings, made plans, took notes. It started to shape up. At times i wondered how in heaven i was going to be able to pull this off, but i knew i just had to try. And that's how it all began...

...and this is how it came to an end. Click on the photo for a bigger one :

Cornu DIY :

Before i start, i might add that all this information is available on the net. Well buried, certainly, but freely available. What i couldn't find, i had to invent by myself. I think that the right idea is there, but have no idea of how dependant the design is of the details, the exact curves, the millimeter stuff and such. I have never had the chance to listen to the real deal. But, once again, i think that the right idea is there. Therefore, i wouldn't dare call this a Cornu clone, but rather a tribute to Cornu. A copy, if you will. Hence "Cornu Spiral Copy Horn"...

Some links to the few rare DIY projects i found about :

- Poiram's Cornu project -- http://www.freewebs.com/ulzog/poiramsspiralhorns.htm
- micro-io project -- http://www.micro-io.dk/speaker.htm

Wow. Two ? Well that's all i could find.

Here's the Official Site : http://www.cornu.de/en/minispir.html

The Specs :

Reportedly, the best drivers that’ll work for Cornu, are = FOSTEX FE108EZ (the most expensive), Lowther DX4 (middle-prized), and the MONACOR SP-50X (the cheapest). Alternatively, i found this info in a post from Daniel (the creator of the Cornu) on his message board :
"The first step in the upgrade path is to replace the SP50X drive unit with Lowther, AER, REPS, Fertin or whatever fullrange driver you prefer. It should have fs=36Hz and Qt=0.25 or values close to that."

So pick a driver that suits your taste & budget, my choice was based on economics. I'll upgrade them one of these days.
The rest of the specs (in my project) are :

Top & bottom plates (the "cover" and the "back") = 70cm x 70cm x 10mm
Middle plate (where the cuts are made) = 74cm x 74cm x 10mm
The flow guides :
- lower-left & upper right (longer) = 224,5cm x 14cm x 3mm
- upper-left & lower right (shorter) = 159cm x 14cm x 3mm
The cover plate has a central opening (read : a hole) for the speaker --> 11,5cm for the MONACOR SP-50X.

So, in my project, this and the rest necessary gave me a total of :

- 4 square boards of 70cm x 70cm x 10mm
- 2 square boards of 74cm x 74cm x 10mm
- 4 flow quides of 224,5cm x 14cm x 3mm
- 4 flow guides of 159 x 14cm x 3mm
- 2 Monacor SP-50X full range drivers
- 2 speaker terminals
- 4 meters of speaker cable
- glue
- wood lacquer

So why the middle plate is 4 cm bigger than the cover & bottom ? For the cuts for the flow guides that'll go flush with the two other plates. I started out by imagining only 2 cm more (72 x 72 cm), but in the end decided to go with 4 cm. It's a whole lot easier to work with the cuts when you have 2 cm of secured space from the borders. Also, i find that the aesthetics call for this as well. Today i wouldn't imagine them otherwise...

I wouldn't recommend cutting the flow guides length-wise before doing the middle plates, as the exact length can change a few cm, depending on the cuts made in the middle plates. I prepared mine in straight cuts of 14cm wide, and then, in the end, cut them in length as i needed.

Imitating beauty :

Ok, so i had my square plates cut (all 6 of them), it was time to do some drawing. For the two cover plates, it couldn't be simpler : locate the center, and draw a circle with a diameter of 11,5 cm. This applies to Monacor SP-50X. You'll need to change it of course if different drivers are used. This was going to house the drivers. As for the middle plates...well, this is where the fun part begins. Kind of.

I had prepared a dummy plate in cardboard on which i had drawn and cut the following schema, a reproduction from the photos seen of the entrails of the Cornu. Click on it to get a bigger one :

My idea was to reproduce the drawing on placing the dummy plate on top of the wooden one, and trace with a pencil within the cuts. I'll save you the trouble by saying it didn't work. The cardboard loses all the rigidity once all the cuts are made, and thus can not be trusted when tracing, it moves in every direction. Needless to say the results were *very* approximate. I had no choice : i reproduced the grid from the schema on the wooden plate, and traced the flow guides, like when i was a kid and had to rescale a drawing. It takes time and is rather tedious, but i kept telling myself it was all for the best possible result. And it did give a perfect result. Some more patience, and repeated procedure on the second plate. It was well worth the effort.

There, all the plates were drawn. Now i just needed to cut them. "Just" ? Man, that was a lot of work. It was indeed so much work that i remember telling myself that had i known the amount of work necessary, i'm not sure i would've picked it up as my first project. There's a lot of trails to be cut, i can tell you that. More than 15 meters, to be exact. And i did some of them more than once also. Why ? Because i didn't cut them wide enough from the beginning. Imagine that they have to be *at least* the double the thickness of the flow guides. In my case the guides were 3 mm thick, so at least 6 mm for the cuts, then. Count more for the bends. The wood stresses and gives an outward pression when bent. This means that if the cuts are individually exactly what is needed, when you insert the first guide and mount further towards the center of the spiral, each turn adds more pression, and since the plate now ressembles a swiss cheece (with all the trails), this pression is distributed everywhere on it. Clearly put : what was an exact fit before, is nearly closed for access now.

Here's a happy end result. Click on it for a bigger one :

In practice, i'd guess i cut them finally to about 1 cm wide everywhere. This gave me the freedom to mount "rather easily" the flow guides, but raised another problem : i had to fill those gaps later on, and larger they are...well, you get the picture. If i did it today, i'd go somewhere between, say, 6-8 mm.

So i cut the trails. Did i mention that it was *a lot of work* ?

When all the trails were made, it was time to do some measuring and cut the flow guides into correct lengths. They did measure up to what i indicated in the specs. The shorter ones were a no-brainer ; they bent from start to finish in one piece. The longer ones however, i knew i had to do the center of the spirals in bits & pieces. So what i did concretely, was the following :
i cut the shorter flow guides first, then mounted them. This allowed for an easier way to work on the longer ones. I cut the starting pieces roughly the same length than the two shorter guides, then mounted them. This left me with the center of the spiral. I simply went and cut whatever was needed in length for the tight bends and lines. Piece by piece. And cut a small bias on each piece, so as to fit it perfectly with the one before, and so on...

Here are some photos with everything "dry-mounted" for testing. You'll notice on some of them that all the pieces are numbered. I did that from the beginning so as to not lose the track of things, as i was going to glue them in two sessions. Click on the photos for bigger ones.

Once all the pieces of the spiral puzzle were fabricated, it was time to glue ! Glue dude glue ! But just before that, i had to make sure that the plate was *exactly* in the middle of the flow guides. The advanced hi-tech method used was a plastic meter and loads of time ; and around we go, again and again. Pushing and pulling. I removed the bits & pieces from the center, and glued only the four long flow guides. This is better done on a plane surface, as it helps to make sure the guides are leveled. That way the cover and the bottom should fit with no problem. I filled the bigger gaps in the trails with left-over wood, and applied a generous coating of glue on the base of the guides. Wait and let dry. Once dry, turn over and start again. Wait and let dry.

The rest (the center) was glued piece by piece, on my sunny balcony and with a bottle of cold beer always available (i rather found that the tequila-flavored Desperados© did the job perfectly). Once everything was dry, i filled all the cracks and holes that were left, and sanded the whole lot. Dude, did i mention already that that was all some huge amount of work. At that moment i remember thinking that had i knew how much work it required, i would have probably not chosen to do it as my first project.

Once everything was nice and smooth, i lacquered everything. Two layers, yes sir. A lot of lacquer it took. The thing with the Cornu's is that they just look like a square plate of 70x70 cm, when in reality there is an enormous surface of wood, everything counted. Let's see : there's 15 meters of flow guides. 2 sides. That makes it 30 meters. Add almost 8 square meters with the plates (with the two sides). Apply two layers. You can feel the glue and lacquer running...

You can see a nice colour taking on after two layers :

So a lot of lacquer later, i prepared the cables from the center to the edge. Making the holes through all the flow guides once they were in place was not an easy task, but i fail to imagine how it could be done otherwise. It all needs to be glued before making holes, otherwise who knows where the holes will end up to...

Once the cables were done, what remained were the cover and bottom plates. Since i was not lucky (or pro enough) in my choice of wood, they were slightly bent in the middle. This meant that i had to find a way to apply enough weight during the collage so as to force the contact everywhere. I resolved this by placing a small bookshelf filled with heavy stuff on top of it. I decided to apply the glue on the board of the guides, and then turn it upside down, to glue it against the plate on the floor instead of the other way around. In my sense this saved me the hassle of glue running down the guide's walls. On the downside, the structure started to get heavy already, and to handle it carefully upside-down placing it flush from each side proved to be somewhat...delicate. You don't have much time before the glue starts to take, but aside a few details, it worked out very nice. Once dry, turn around and repeat.

And then, the last morning, after yet another night under my bookshelf, there it was : this object of my desire that i had learned to accept as something that would never be finished (in a way it had almost been reassuring in the same way that it had been frustrating). But now, there it was, slightly coloured and reflecting the increasing daylight that filled the room. The last of them two. It...was...finished. Here's the last picture before the Monacors went in :

It's Alive !

So how do they sound ? Well, my first reaction was that of sheer amazement at the amount of bass they produced. And i mean *a lot*. So much so, that i felt it ate the middle frequency. Some low stuff indeed. Then i realized that their placement is something of critical. No chance to leave them on the floor. So i elevated them. Muchos better. But they are some heavy beasts, and i haven't yet had the time to fix a secured system for them to hang from, so for the moment they are on some stuff. But elevated. Also, i'm using a modded T-Amp to drive them (see my other post), which means some new big capacitors that'll need some time to settle down. All in all, the sound is starting to change already (after some 20 hours of listening), and the bass is becoming more controlled. The mids are still lacking in some type of music, but there are others where everything is *spot on*. Mainly electronic. And when it's perfect, it's heaven for my ears. The overall sound changes as soon as i move one or the other speaker more than 1 meter, so i'll have to experiment. Then there are the drivers that i'll change one day. All this leaves a lot to wait for, but in general, i'm enchanted. And the sheer beauty of them leaves me contemplating each time i look at their direction.

This is definitively a project that i would have *never* started had i know the amount of work involved, but today, i am so glad i didn't have a clue about it.

Bottom line : i love them.

Merci DIY :)