Senin, 22 November 2010

DIY Audio pot pourri 2 : Woody film capacitor

Long time ago I came across a Dutch web site explaining electrolyte capacitor modification. Recently a fellow (senior) DIYer Mr. Paul also have done similar mod with paper in oil capacitor, DHT Rob reports improvements on replacing aluminum can with wood filled with beeswax. Also mixed results from Pak Paul from what I've heard.

A longtime DIYer friend said that you'll get sound improvement just by lightly sanding the caps. Nothing but subjective result analysis though, but I've found a freeware just for that, detecting difference of the same music signals one as a reference- the other may be through a different cable - more about this in my next post.

I've got 4 Solen MKP capacitors belong to a friend ( Hi pak Johny..) so I'll give a shot. I've made mahogany tubes, soft wood cover. I've broken in the caps before for  very long time on the bench. So here are they. The tools are small PCB drill for the caps leads through the soft wood cover and a tweezer. This soft wood is press fitted to the mahogany tube.

I still have some space left in the tube, so the Solens are wrapped with the tissue and pressed in to the tube. The input and output leads are marked, they do make sound difference how you connect them even the Solens are non polar.

Wrap like this, plenty of excess on both ends, they can be removed easily.

Excess tissue paper removed with tweezer


There you go!. A pair of Woody Solen :). Notice that the leads are not centered. Good idea if you marked them before drilling


I did a comparison in my Sony CDP -M75 - looks awesome!. There are difference in sound - not much but the Woody tend to sounds smoother. Definitely better than the original CDP ouput (with Elna Duorex 2 electrolytes).

 Here my  friend Aphinx using the Woody for input coupling on his D class amp. Originally his input coupling was the Russian MBM 1uF we found the woody sounds better. Below the Woody tested against Ampohm MKP, he likes the Ampohm better . 


So there are differences in sounds when you do this kind of modification, whether the sound is better or worst is another matter. Any explanation on this phenomenon? Let me guess, first the skin effect - greatly influenced by surrounding material or dielectric- nowadays 'we' prefer natural materials - hence the the sanding, beeswax, natural fibre like wood, cotton, silk and paper. Silk? I think one electrolyte maker used it on Elna Silmic...

Secondly vibration - with tight clamping the caps are protected from it. Lynn Olson told stories about 'microphony' effect of vibrated capacitor - it wasn't pretty picture - he said . I believe them because I've heard the mod my self. My Spica TC-50 crossover is placed outboard , guess what? the caps vibrates like crazy there. So, at the very least we should tie, damped and fixed our caps very well. Good luck !

Senin, 01 November 2010

DIY Audio pot pourri 1 : Spica TC50 crossover tweak

During several months I've carried out many small audio tweaks. Not full blown projects though but I think it's maybe quite interesting sharing them :).

First is my Spica crossover tweak, continuing effort IMO still necessary to replace those old ECI lytics. This time I like to tackle the high frequency section. On September 2002 the Spica designer John Bau was interviewed by  Enjoy the and asked what to do to improve his Spica, one of his answer is to change all the electrolytic caps to Film.

So here are the pics, some caps, made some 'risers' for easy soldering and quick caps change.

first on the board are Wima MKC 4.7uF, Green Nichicon Muse's 10+22uF, Black Nichicon BP-S 4.7uF. Left them for 3 days about 12 hours on play, sound? me not like :). So instead of waiting for another 3 days I just pick - place - listen - change. An advantage of working with broken - in and old caps...

The Wima MKC can stay, it sounds well for coupling capacitor on my CDP-M75, so next on the board is 0.1uF K42y Paper in oil bypassed the BP-S, still not liked it, then I went to my garage pick some big 22uF Solen MKP - the last pair I have, this in parallel with poor small 10uF green Nichicon Muse, sit and!..the bass is like I never heard out of this Spica..I wonder though..I'm working here in the high freq region and I get that bass?..well the Solen and the Wima stays!.

Now the job is much easier, that what I've thought and I was wrong. I got better bass & high, same and maybe better staging, but some tracks still give sibilance. Swapping the BP-S to Evox MMK did not help, same with 'old' Yellow Rubycon.

The last 4.7uF film cap I have is a pair of el-cheapo yellow Audiophiler..yuck.., nonetheless I've soldered them and ...its sounds ok, no, they sounds quite well indeed..and  totally acceptable to me, much better than the Evox.. So the Audiophiler stays! After 2 months listening I tend to let this setup freezed and change the 10uF Muse to MKP and maybe another Wima MKC replacing the Audiophiler, I miss the high notes timbre of acoustic piano sometime. 

My lessons here, tweaking the sound for Mid is very difficult, many times proven cheap parts perform well in combination with other. Thanks again for reading guys..

In the pipeline : 
-Part 2 :  film cap gets woody jacket..
-Tubelover 6N1P tube buffer in depth..

Jumat, 08 Oktober 2010

Electric Fan, MMC holder

Now, some design cases recently I've observed as always - at home. A memory card holder and an electric fan. Nowadays you can barely escape from plastics, everything we use almost certain contains plastics - one form or another. Let us begin with another well executed design first.

A plastic holder made from PP of polypropylene. Very nice, compact and functional. This one comes with my digital camera. It's function is to hold a pair of memory cards, MMC. Lets try to see how such a design took place. 

First is surely the design requirements to function. It needs to hold MMC's, must be transparent or a bit see through so that we can see what cards inside, light, very cheap, not easily breaks, the card must be secure enough - not falling down by itself, easy to use etc.  These are just wordings, no design concept or design sketching whatsoever. 

Now comes the Task. Task comes after function. Tasks are whatever you need in the design in order your design to function properly, still no real design yet at this stage :).

To cut the story short, take one of the design requirement, lets say " easy to use", now let see the 'design' below. What features coming from the 'task' of the holder supporting this requirement?. 

There are several, as this 'easy to use' requirement is the most vague and hard to implement. We can see small recesses on both side of  the holder. Curved recesses on the sides are for your, say left hand thumb and middle finger. Also on the top where you can pull the card inside using your thumb and index eh..?. Last thing you want is to be able to insert a card and can not pull it out....

What else?  yup, easy to use often means easy fabrication as well. Hence the two parts, mate both half, press're done. Why not one part? well it can but the metal mold will not run very long. Lots of small thin 'cores'. Easy to break with high molten plastic pressure and you production will be ruined.
See small slit and a bulb on the right sides? It is an integrated spring, it task is to push the card to the left wall. This creates enough friction, preventing the card from falling down..another smart design.

Such a simple thing like this are full of hidden smarties, which greatly contributed by systematic thinking in terms of Function and Task. Most of innovations are made by thinking hard this way, rather than 'eureka' thing..., smart!.

Now, a saying that says A Chain is as Strong as the Weakest Link is really well said. Look at this beautiful small electric fan. Same brand as my problem rice cooker. Their products have the bad habit of early broken out!.

All the parts, plenty of them laboriously made. The styling is not bad either. The fan cost less that $15, last less than 5 month just because of ..weakest link!. see below.

The weakest link in this case is the transformer wire. It gets hot, even there is thermal protector- the wires are broken somewhere in the winding. So all the labor, effort and resources just wasted. Really? not much though.

Buying this fan keeps the factory workers get their salaries. The seller get profit and can take care of their family. What about I'm throwing this fan in garbage bin then?
Hmmm..there comes the most environmentally friendly job in the world...the Recycling Man!. Yes they live from our garbage, they will sort and pick all the plastics and metals for further recycling. They get money paid for that. To me, the Recycling Man is my Super Man!.

What do you get? Next time you buy new product please don't be surprised. 80% of that product was indeed once your garbage :))
And what did I get? just a lousy transformer bobbin with 30ish gauge enameled copper wire from which I planned to make something like this see below and several hours of Audiophile-foolishness fun :). Thanks to broken fan and Bud Purvine and his DIYAudio postings on ground side electrons..what ever that means..

This is the first time I can relates design stuff with DIY audio thing. Thanks for reading my foolishness.

Sabtu, 18 September 2010

TDA1541A , more than I can handle!

Actually I've planned to write about yet another good design bad design thing, this in waiting my mood to finish-tweaking my Spica TC50.
During visit to my audio part supplier friend just before a long Eid el Fitr holiday,  I in all a sudden being blessed by a chance of trying out a bunch of the legendary TDA1541A DAC chips from Philips, not one or three pieces, but 4 tubes with about 8 to 13 chips each!. plus a pair of TDA1541 (non "A")..More than I can I've postponed the Spica tweak and design thing...

My friend is just one out of many victims around the world of TDA1541A frenzy that still on going even today!. It's the Fake or Genuine TDA1541A debacle that started out around 2004. This chip is so high in demand that some people just make  fake ones and sells them. Just read this thread about it in My friend have tested the chip randomly some 6 years earlier, none working he said..

Being a TDA1541 fanatic (see my Sony CDP950 tweak stories starting here ) I took the challenge of testing all 4 tubes, may be, just maybe I can get one or two working chip. As a positive thinking guy I assume there is always some good thing out of even bad guy!. I even told my friend if I can find a good chip I will let him know. So here is the chips, my story & the pics..enjoy!

The tubes, yes in the 90's and early 2000's Philips made the chips in Taiwan, many says even with better production process than in Eindhoven Holland. The highest grade, the TDA1541A/S2 (so called 'double crown') is made in Taiwan.

Testing all the chips here in my salvaged Non Oversampled Sony CDP-M75 with Sennheiser PX100 headphone. This player was totally wrecked and non functional. I've got it from a good friend (thanks mas Yuda!). I planned to take out the DAC only for spare and piggy bagged to DAC in my Lampized Sony CDP950 for higher current output. It turned out that I have the luck to bring it to life :) so here we go..

After testing the chip halfway? Guess what?.. there are 5 chips or so that completely working! But beware!, and I know from the DIYAudio thread later on, there are some chip that can blow your CDP!. Here are the 'test result 'modes' :

  1. the CDP just doesn't want to turn on, no display at all when powered up - the chip felt cold and is busted
  2. as above but the chip felt warm- busted as well
  3. the CDP and the display turns on, the disk spins and worked!....yippee!
  4. the CDP turns on, the display also, the disk spins, chip is very hot and I can hear some music-heavily distorted and interrupted often..then I heard loud DC signal through my daughter's  PX100..frantically  switching off the player.......what... was.... THAT??..
Tried to turn on the player no luck...sigh...yep..sometime you gains..and other time you just loose..within a fraction of a second! This 'mode 4' is by far very dangerous. What I did was repeatedly switching the player on and off during this mode 4, a big No-No, learned later.

After checking the M75 for supply voltages on DAC pins, I found that pin 15 reads should be -14.9 or so... Luckily the other day I got schematic from Mr. Jeffry, thanks Pak! them, testing some components & yes one transistor is surely went busted.

Late in the evening and no electronic shop open for sure, I just browsed my old PCB's junk collection, looking for equivalent replacement, yup, found a better equivalent, 2SB734 in place of the fried 2SA1015. Fit the B734..aah..yes..the player worked again, so I continue testing, until my left hand thumb felt numb putting and removing those 28 legged little monsters...this time I'm very careful, no more switching in on and off again. No display- no disk spin -the chip is simply busted, and yes I recalled another 3 or so chips with 'mode 4'..

At the end of this tense testing session I can smile, BIG TIME :)...the TDA1541 non A worked very well, and hmmm let just say ~20% out of 50 TDA1541A chips are working too. My positive thinking and all the effort paid off well...the harvest? here are they..:)

My conclusions are that these chips were : original and working well (mode 3), the original damaged-rejected ones (mode 4) , original totally damaged rejected ones (mode 2) and the rest is just plain fakes (mode 1)

Me and my friend strokes a win-win deal. I think I'm getting closer to my dream, a dedicated much about these and our audiophiles earthly needs...sigh..., thanks for reading my friend..

Minggu, 22 Agustus 2010

Rejuvenating Spica TC50

I'm currently listening to my Spica and also waiting for my fastening break time, why not continue the Spica story?. This legend is known for its 'disappearing' and staging quality. When you listen to it soon you'll  forgot that there are speakers in front of you. Just sounds floating around and in front of you. That's the way I liked it just like my more refined Magnepan SMGa.

My Spica : pre 1988 built !

One box sounds not so loud as the other. So I starts measuring both with my DIY Speaker Workshop Jig II, here is the result, yes the "B' channel just sounds softer.

Now it's confirmed I planned to move the internal crossover to outside the box for better tweaking and checking. I opened both boxes & smells 22 years old odors of this spica. All drivers were in unbelievable good condition, original French Audax including hand markings for each driver to match the crossover. Not to forget, thick genuine sheep wool blankets

So lets see the crossover, as said in the forums, electrolyte capacitors , those yellow things says ECI on the marking. Hell of electrolytes, after more than 20 years !. Inductors are cheap ones with ferrite cores both for high & low frequencies. Resistors are good metal oxyde though.

One elco is not in good condition, white thing comes out so must be leaking then. Here is my old story about electrolytes in loud speaker. Not bad, one out of 10 elco after 20 years!. I just replaced it temporarily with cheap MKP until I have the mood for tweaking.

Leaking capacitor
Now it's the turn of the enclosures - a closed box with no leaks. I have to make sure of no leaks all the time.
I reversed the bolt and nuts for crossover PCB to be mounted out the box on the rear side. Notice white glue and rubber leakage.

Hmm..does not look pretty yet. I'll put the board in a box when all is tweaked right..that's a promise for an old friend. How its measures now? well quite ok I think, both box sounds more balanced now.

Here is the crossover as per today :) waiting for better components, some are almost ready and still in 'burning' process. Nichicon Muses and BP, recycled Philipes- Epcos- Wima MKC , some green and silverish Russian paper in oil and maybe a pair of 30uF big oil can?..who knows.

As for now I felt little to be modified, its sounds already good enough for me, groans-moans and throat sounds of Diana Krall's voice is just awesome also her piano 'kick'.  Maybe I need to remove a little harshness and sibilance a bit for complex music. That's the next tweaks are for. Thanks for visiting.

Kamis, 19 Agustus 2010

There are second chances in my simple life : Spica TC50 speaker

Have you ever thought that you'd have to make choices in the past, well a luxury choices that is. And all a sudden there you are, after say more than 10 good years being confronted again with the second chance right in front of you, dejavu all over again :).

Mine happened long time ago when I was still a near graduation student in the Netherlands. I started looking for a pair of serious loudspeaker to take home and live with it..just like a wife isn't ? except it comes in pair :).

At that time I was considering an unusual speaker, Magnepan SMGa and another pair which I have listened at my well funded friend apartment, Spica TC50 driven by high-endish gears such as Goldmund turn table with Van den Hul Grasshopper cartridge, Aitos tube MC Pre-pre and Aitos tube power. Really an experience not to forget to hear the sound.

Finally I married to the Maggie SMGa, took her back to my country home after graduation. Why SMGa? they are more expensive than the Spica, more than 3 times. The reason was that Maggie in combination with my system at that time, second hands Counterpoint tube pre & hybrid power amp sounds quite good, maybe scoring a decent 6 points out of 10 points of my friends gears.

Fast forward to present time, I started doing DIY audio thing and with several 'hardcore' audio friends planned a visit to a place known as electronic ghetto in Jakarta, a Bronx kind of place with many unexpected audio jewels if you're lucky.

And how lucky was I indeed. In one of the shop I saw a familiar sight...a TC50 be honest I'm bit disappointed to see both grills full of holes and one of them are totally broken into 3 pieces. Hmm..lets see the speaker driver then..looks unfamiliar though, especially the woofer, not the same as my friends. Need to gather more info on the net after asking the price, quite pricey for jewel from the past. Got home and started browsing.

Here is a link on the net.  Yup exactly, the Spica's have stood bravely the cruel tests of the time. With many fanatics. A proof that the TC50 is a legendary loudspeaker. Woofers and tweeters are no longer in production, the designer John Bau just don't want to continue his creation, etc. etc...

To cut the story short, the next day I went back and got myself one of my 'youth' mistakes home to correct, a pair of legendary TC50-Spica, here is the pic. the date was 12th April 2009, just maybe after 10 years, nope, more than 18 years later, can you believe it?

Talk about second chance? I will give this Spica a real second needed too.., one speaker sounds not as loud as the other, the grill have to be made anew..hmmm..sounds yummy for a tweaker like the story will continue for sure..thanks for reading

Kamis, 08 Juli 2010

DIY resistor for DAC current to voltage conversion

June this year is my zero writing month. Don't want to skip July so here it is another DIY audio adventure.
My tweak darling, the Sony CDP950 is almost weekly being changed. I used it for learning both digital & analogue wise. At DIYAudio I came across a strange resistor made by John a.k.a EC Design. It is quite complex to be made- although fully doable in my workshop. Here is my 'dream' Ri/v - the resistor that convert current out of the DAC to voltage that goes to the tube grids. In my case its valued 33ohm.

A beauty isn't? (hopes John wouldn't mind I use his pic.). Why all the troubles? well this DIY resistor promises low capacitance - non inductance - and high wattage. All good for tiny current out DAC.

What I can not get are a coated resistance wires & patience for winding it.  Being a naughty DIYer like me,  I have once hammered several of those white bricks, just to see what's inside :). It turned out that for low R value the R is wire wound. Higher value some kind of carbon film. Hmm..this brings me to an idea..

So, armed only with double tapes and 2 pairs of white bricks 39R/5W here we go :

The tools are 1 hammer, 2 pliers, cello-tape, double tapes (thick & thin semi transparent), scissor or a cutter, a tweezer.teflon tape...thats it.

Take care not to break the wire from the end cap. Tape the wire to the cap & legs for strength

Thick double tape both end stripped & glued to the table and start making zig-zag pattern of the first wire lay up about half of total wire length (my case 28cm ~ 1ft). plan for the double tape length.

Cover the first lay up with thin double tape, next lay up

The first lay up is now 'insulated'. Start now with a second layer, now you know why I need a semi transparent tape :).

Measurement shows a lower R value than the original 39ohm. This happen once you unwind the wire from the ceramic center piece. Not to worry & its closer to my 30ohm target :).

Cut both ends, add another thin double tape, peel wax paper on both sides & start rolling

Two DIY Resistors :)

Rolling & teflon taping. measurement shows quite acceptable 'tolerance'. After rolling the tapes the R value goes higher a bit.

In action :)

The sound? well the 'cocoon' beats my Philips metal film for sure, and about a 'tie' with the pink Takman's, not yet decided though ;). Highs of this DIY Ri/v are quite breath taking...

But...hold your horses, recent discussions on DIYAudio with EC reveals that this DIY Ri/v is not completely non inductance yet. But it can be made non inductance, using a so called Ayrton-Perry_winding and for this I need to insulate the wires using a proper lacquer. Not so hard to do compared with 5m long EC's wire :).

So adventure is to be continued...thanks John..