LAME 3.97b2 -V 2/FLAC --> Rio Karma/Chaintech AV710
--> Shure E2G
Total spent on playback equipment (not counting DIY equipment or player): $370
Total spent on DIY tools and materials ~$70
__________________________________________________
oh, and... I'M A TRAP!
holy goly... that's indeed the fattest smoothing stage i've seen for headamps... ...but where are the regulators?
very nice built, banfi! could you please elaborate on your p2p-technique: why did you solder each and every point of your wires to the board?
__________________
E-MU 0404 ~ Fid-Audio 0404 Breakout Adapter ~ Fid-Audio Cantus IC ~ WNA HA mkI ~ DT880
holy goly... that's indeed the fattest smoothing stage i've seen for headamps... ...but where are the regulators?
very nice built, banfi! could you please elaborate on your p2p-technique: why did you solder each and every point of your wires to the board?
The PSU's capacitors are not really for smoothing, rather for energy storage to bypass any power fluctuation common in domestic power networks. The headamp has 4700uF puffers and stabilzers quiet enough to power MC headamps (for vinyl playback I mean) I doubt that any ripple rejection is gained by using that many capacitor in the external PSU. On the other hand you can now disconnet and reconnect the mains cable without hearing it Anyway, the stablizers are on the same board as the headamp, the four smaller heatsinks are for them. (The regulators are also non-feedback discrete designes.)
As for the p2p...Untill recently I did work with pcbs, but I had to realize that for prototypes the protoboards are much more suitable. The first headamps I built this way were done in a traditinal way, I used the legs of the components to make connections. Many times even the components themselfs were at the downside of the panel. While this certainly worked for smiple circuits this can be rather messy with more complicated projects. I ended up using more and more wires which were crossing other tracks and I did not liked this. On the other hand, if you use the leg of a part, so that you bend it and solder it to an other part it is not too easy to remove that part if you want to. (With prototypes it happens all the time). So, instead this I figured that I put every component in as they were in a real PCB, so I cut down their legs. Then I use the legs as connections, but this time I build up the track like if it were a real PCB. In this way I could do a much cleaner board with less cross wires and the parts can be easily changed. The following picture illustrates the headamp + regulator board having one side done in the old way and the other in the new. I think the picture speaks for itself. (Note, that the side done in the new way is more complete having the output stage also almost completely connected)
transformer with netfilter separately encased. 2 of tangent's TREADs as regulators with discrete rectifying diodes, offboard smoothing stage and complementary darlington stages for each rail:
__________________
E-MU 0404 ~ Fid-Audio 0404 Breakout Adapter ~ Fid-Audio Cantus IC ~ WNA HA mkI ~ DT880
The PSU's capacitors are not really for smoothing, rather for energy storage to bypass any power fluctuation common in domestic power networks. The headamp has 4700uF puffers and stabilzers quiet enough to power MC headamps (for vinyl playback I mean) I doubt that any ripple rejection is gained by using that many capacitor in the external PSU. On the other hand you can now disconnet and reconnect the mains cable without hearing it Anyway, the stablizers are on the same board as the headamp, the four smaller heatsinks are for them. (The regulators are also non-feedback discrete designes.)
ah... i see... interesting concept, but most likely lots of stress for the transformer. do you use regular fuses on the primary and/or secondary side?
As for the p2p...Untill recently I did work with pcbs, but I had to realize that for prototypes the protoboards are much more suitable. The first headamps I built this way were done in a traditinal way, I used the legs of the components to make connections. Many times even the components themselfs were at the downside of the panel. While this certainly worked for smiple circuits this can be rather messy with more complicated projects. I ended up using more and more wires which were crossing other tracks and I did not liked this. On the other hand, if you use the leg of a part, so that you bend it and solder it to an other part it is not too easy to remove that part if you want to. (With prototypes it happens all the time). So, instead this I figured that I put every component in as they were in a real PCB, so I cut down their legs. Then I use the legs as connections, but this time I build up the track like if it were a real PCB. In this way I could do a much cleaner board with less cross wires and the parts can be easily changed. The following picture illustrates the headamp + regulator board having one side done in the old way and the other in the new. I think the picture speaks for itself. (Note, that the side done in the new way is more complete having the output stage also almost completely connected)
i understand exactly what you mean and switched myself short time ago from simply connecting parts to pre-wiring some sort of traces before i solder in the main components. but i guess, it'd be better to reduce the solderpoints to the necessary minimum for optimal current-flow... well... just a guess.
__________________
E-MU 0404 ~ Fid-Audio 0404 Breakout Adapter ~ Fid-Audio Cantus IC ~ WNA HA mkI ~ DT880
ah... i see... interesting concept, but most likely lots of stress for the transformer. do you use regular fuses on the primary and/or secondary side?
i understand exactly what you mean and switched myself short time ago from simply connecting parts to pre-wiring some sort of traces before i solder in the main components. but i guess, it'd be better to reduce the solderpoints to the necessary minimum for optimal current-flow... well... just a guess.
I did use fuses during the tests, but not in the version I listen to. That toroid can handle the surge and I don't do power up/downs too often. (Heck, there is no on/off switch )
As for the solder points: you could be right, but could you please define what is the "optional current-flow"? As long there is a stable and good electrical contact between two points with a proper current handling capacity (often a problem with PCB tracks) it is fine for me. Anyway, I did solder down those points because some of the parts moved during assembly, and with them the lines too. After a while a lot of unused joint posts were used and I couldn't move the legs to their posotion without reheating them. To solder all the points was just a little cosmetics
[...]but could you please define what is the "optional current-flow"? [...]
ok... i'll try it... but i meant "optimal", not "optional" current flow: there should be small stray capacitances at each solder point. so most likely current flows without further (small!) degradation if you leave the wire as it is: without additional solder. maybe some educated diyers could explain the matter better...
__________________
E-MU 0404 ~ Fid-Audio 0404 Breakout Adapter ~ Fid-Audio Cantus IC ~ WNA HA mkI ~ DT880
On the other hand nice work with that amp, BrokenEnglish. As far as I can see it the external PSU doesn't has any puffering/filtering and the amp has the puffer/regulator sections as well. Am I right?
I did a few years back a phone preamp with such an arragement and it worked fine until for some reason I crossed the input cable coming from the pick-up with the cable going from the external PSU. I was welcommed instantly with a strong hum. If I crossed the input cable with the mains I heard nothing. (Crossing means that the two cable lies in each other in some angle). The reason was (of course) that the secondary voltage of the PSU is much lower (if we speak about solid state designes) than the mains voltage. If you pull 200mA for the secondaries and you have 22V DC, then you are pulling 20mA from the 220V (such is the ideal transformator with no loss). The 200mA together with the cable going from the secondary did presented a much higher 50Hz energy emittion as the mains cable. Of course it was with a high level MC pick-up (max. 3mV voltage) and the phone amp has a much higher gain then any headphone amp. But since then I either filter the voltage in the external PSU or I don't use a separate case.
ok... i'll try it... but i meant "optimal", not "optional" current flow: there should be small stray capacitances at each solder point. so most likely current flows without further (small!) degradation if you leave the wire as it is: without additional solder. maybe some educated diyers could explain the matter better...
Thanks, I wanted to write optimal of course, not optional. I also heard about high frequency "edges" regarding connection points but I never saw any measurement underlying this theory, nor couldn't I hear any difference when I tried. But could be true...
On the second thought (if it really an existing effect) it could be dangerous, since random capacitances everwhere in the circuit could result to unwanted oscillations in feedback based circuits. It seems plus one point to have NFB designes
On the other hand nice work with that amp, BrokenEnglish. As far as I can see it the external PSU doesn't has any puffering/filtering and the amp has the puffer/regulator sections as well. Am I right?
no... that's just a regulated dual voltage psu. i separated the transformer mainly because my initial goal was to use it for different psus. so far i can't comment on the sound, because (a) i'd need my reference phones (dt880) and (b) those caps are heavily breakin' in (sound changed markedly over the few hours i listened to some music). there's no hum, no audible noise so far, even with my low-impedant cans (dt440) on the wna. small hiss at max. volume with paused source and the dt440, but it is lower than with my other psus.
__________________
E-MU 0404 ~ Fid-Audio 0404 Breakout Adapter ~ Fid-Audio Cantus IC ~ WNA HA mkI ~ DT880
...but where are the regulators?
Anyway, the stablizers are on the same board as the headamp, the four smaller heatsinks are for them. (The regulators are also non-feedback discrete designes.)
Linear Mode
