Worst part of building DIY is the casing especially when you do it like me. I
started with the concept of how big I wanted it to be , and the package of an
external HD seemed just about right.
Along the 2 years of building it , plans change , and it was difficult to get it all
in that small box. After more than a month's worth of afternoons struggling with soft ,
very easily scratched , and bended aluminium 1100 ( all I could get)
and asymmetric heads of M2 bolts , I finished a skintight
case over the electronics.
Building "skintight" is quite different than mounting PCB's in a bigger case !
I did it all wrong and it took much longer and was much harder .
Heatsinks to the case for 2 x 35 mA at about 40 V , is near 3W to dissipate idling
by the LM3x7's , OPA's and the LME49600's (and the transformer) , and in a
28-29 degr room ( yes celsius) with no fan on it , it gets warmer than expected , guessing
beween 40 to 50 degr when used with 32 ohm HP's . With a fan near , it stays at ambient.
How does it sound ? Hard to say , with laptop output and 2 cheap (rather bad)
HP's . No oscillation with oscillation sniffer , but offset not stable , it dances
around 200uV and 450uV for the other , without signal. I can't remember whether
my test amp was stable after trimming it with a passive servo.
Could it be this is residue of keeping the much faster LME49600 in check with
the "slower" OPA828 ?
I'm not worried , anyhing under 1 mV or even 0,5 mV offset is great.
Only one thing leaves me puzzled : when I play a sine frequency sweep from
around 3 kHz to 14 kHz , that I made in Audacity , at some points I hear some
very faint secondairy oscillation which I don't hear when listening directly from the
laptop output. Weird . Is it the LM3x7 + denoiser that can't follow ? Could even
be the laptop output that can't take the difference between a 32 ohm HP and a 1k
HPamp input , I don't know. I did not compensate and it is high bandwidth LME.
Another big disadvantage of builing compact with skintight casing , it is very
service unfriendly and swapping the amp with the testamp I made before , is not
possible. I'll live with it.
When I started , it was never meant to be DIY high end , just to have a HPamp that
could take my 300 ohm senn's .
I'm happy with it . The price of components is only about 60-70 $ , the time spent
on it (on and off) , was a bit much , making the AL case , far too much.
Just some different (better ) M2 bolts and maybe paint , and it is finished and
finally can start on electronics again , the fun stuff , cmos logic .
started with the concept of how big I wanted it to be , and the package of an
external HD seemed just about right.
Along the 2 years of building it , plans change , and it was difficult to get it all
in that small box. After more than a month's worth of afternoons struggling with soft ,
very easily scratched , and bended aluminium 1100 ( all I could get)
and asymmetric heads of M2 bolts , I finished a skintight
case over the electronics.
Building "skintight" is quite different than mounting PCB's in a bigger case !
I did it all wrong and it took much longer and was much harder .
Heatsinks to the case for 2 x 35 mA at about 40 V , is near 3W to dissipate idling
by the LM3x7's , OPA's and the LME49600's (and the transformer) , and in a
28-29 degr room ( yes celsius) with no fan on it , it gets warmer than expected , guessing
beween 40 to 50 degr when used with 32 ohm HP's . With a fan near , it stays at ambient.
How does it sound ? Hard to say , with laptop output and 2 cheap (rather bad)
HP's . No oscillation with oscillation sniffer , but offset not stable , it dances
around 200uV and 450uV for the other , without signal. I can't remember whether
my test amp was stable after trimming it with a passive servo.
Could it be this is residue of keeping the much faster LME49600 in check with
the "slower" OPA828 ?
I'm not worried , anyhing under 1 mV or even 0,5 mV offset is great.
Only one thing leaves me puzzled : when I play a sine frequency sweep from
around 3 kHz to 14 kHz , that I made in Audacity , at some points I hear some
very faint secondairy oscillation which I don't hear when listening directly from the
laptop output. Weird . Is it the LM3x7 + denoiser that can't follow ? Could even
be the laptop output that can't take the difference between a 32 ohm HP and a 1k
HPamp input , I don't know. I did not compensate and it is high bandwidth LME.
Another big disadvantage of builing compact with skintight casing , it is very
service unfriendly and swapping the amp with the testamp I made before , is not
possible. I'll live with it.
When I started , it was never meant to be DIY high end , just to have a HPamp that
could take my 300 ohm senn's .
I'm happy with it . The price of components is only about 60-70 $ , the time spent
on it (on and off) , was a bit much , making the AL case , far too much.
Just some different (better ) M2 bolts and maybe paint , and it is finished and
finally can start on electronics again , the fun stuff , cmos logic .
Attachments
around 3 kHz to 14 kHz , that I made in Audacity , at some points I hear some
very faint secondairy oscillation which I don't hear when listening directly from the
laptop output. Weird .
Try to isolate the problem to track its origin: first thing to do is to plug your HP directly into the laptop.
Level will be different, but if the artefacts are still audible, it means the problem is not in your amplifier.
If you suspect the supply, use batteries for a test (or a good lab supply)..... etc etc
^ Yes at first I didn't hear it straight from the laptop output because the level was too low , but artefacts are there and differ for each headphone.
I transfered the sweep files to the mp3 player , and there were no artifacts with or without the amp. So good news , it is not the HPamp !
Relieved !
I transfered the sweep files to the mp3 player , and there were no artifacts with or without the amp. So good news , it is not the HPamp !
Relieved !
^ Thanks . I already use them now with 32 ohm HP's , but I wondered about the coil of the speakers giving voltages like when you cut the power to a relay coil. I forgot what's that called and whether they could be so powerful that the LME can't handle it or it goes to the powerlines via protection diodes.
^ Yes but then the supply has to absorb it , which it barely can't or through the diodes over the LM317 to the big cap's.
I haven't seen that done in power amps . Mosfets on the output have an internal diode though.
Is counter EMF a worry when only playing at low levels like maximum 6 V in 30 ohm is 0,2 A or 1,2 W ?
I haven't seen that done in power amps . Mosfets on the output have an internal diode though.
Is counter EMF a worry when only playing at low levels like maximum 6 V in 30 ohm is 0,2 A or 1,2 W ?
No. It's not a worry. I have used an lme49600 headphone based amp for several years without any issues. I also have 2 LPUHP based amplifier which uses 8 LME 49600's in parallel to drive loud speakers directly, again no issues. Any amplifier has some or a lot of damping factor. The damping factor controls the movement of the speaker cones as the music plays and helps reduce the EMF generated by the speaker. Damping factor is like a shock absorber on a car.
There are diodes in bipolar transistors too.
There are diodes in bipolar transistors too.
