I'm waiting for a few 1.8uF, for trying out at C13. These are PETP film/foil, and the tube crowd seems to like them.
Even if they're not optimal at C13, lower values like 220nF could still be candidates for C4 - we'll see.
Hmm - It's a reasonable idea, but the cable + track length between C1, C2 and the main PSU filter caps on the PSU board shouldn't get too long, or it's going to be a big unshielded transmitting loop antenna. It's possible to twist/braid the cabling with ground, of course.
You should also note that the V1.3/1.4 boards already isolate the rectifier and filter cap current loops, from the power amp rails and ground returns, with only 3 star-termination points common between the two circuits - +Vs at the filter cap +ve, -Vs at the other filter cap -ve, and the PGND star ground point. It's already about as optimal as it gets, though some slight improvements are still possible to reduce the radiating loop area in the bridge-rectifier/filter cap loop. I didn't bother to minimize that loop because it's mostly only 50/60Hz and harmonics that circulate there, and the large metal can caps C3/C8 already act as shielding between those loops and the rest of the board. There are also several wide power and ground traces between PGND and C1, C2 which act as a shields.
Another idea (which I think I mentioned a couple of years ago) is to put the +/- 12V (or 14V) regulators on separate daughter-cards with the same pinouts as 3-pin regulators. This allows improved shunt regulators to be developed and tested independently.
Edit: Oops, I just noticed that you were also referring to the LM318 regulators being moved to separate board(s).
Last edited:
I'm not so sure about it, the FE ground plane is split around the smoothing caps so the rectifiers are well insulated from the rest...
I think it's a bad idea unless you realize a MUCH better regulator for the LM318.
Thanks to it's low impedance and low inductance (wide, short traces on a grouindplane) the FE regulator doen't need decoupling caps on LM318 rails... with great results (soundwise)
Similar impressions, not a bad cap, though.
Do,
I haven't tried them, but they sound like an interesting cap. Given your comments, maybe you should try a small bypass cap with them. That should brighten things up and improve the upper resolution.
Of course, what often happens is the two caps don't sound coherent and smooth, but you might get lucky. I would try anything that you have lying around that is in the neighborhood of 1 nF (say 330 pF to .01 uF). My personal experience is that very bright bypass caps like teflon or polystyrene may not sound as smooth with a cap that is dark to start with. A small polypropylene might be worth a try.
Jac
Actually I have tried them only on my preamplifier as coupling caps and connected it on the Fetzilla and this is where it sounded like this. I connected the same preamp on my ML-2 Clone and it is completely different, I actually quite enjoy them... Now I'm just finishing up my My_Ref FE and will test with it and see how it behaves. Could have just been a bad impedance match of my preamp and amplifier?
I did put in a bypass cap (MKP1837 0.1uF). The whole setup has about 24 hours burn in so far.
Thanks
Do
I did put in a bypass cap (MKP1837 0.1uF). The whole setup has about 24 hours burn in so far.
Thanks
Do
That's generally a good choice for a bypass, though I generally use smaller values in the range of 4.7nF to 10nF, ERO KP1830.
Vaguely related: For a very compact coupling-cap upgrade on consumer-grade equipment (which often use 1..10 uF/50V electrolytics for the purpose), a 5mm-pitch Wima MKS2XL in the range of 1 .. 3.3uF, bypassed on the underside of the board by a Panasonic ECHU SMD PPS film cap in the range of 47 .. 100 nF, gives a huge improvement to the definition of the mids and highs.
It's possible that an ECHU bypass will also work well with the K73-16 - I'll give it a shot after it arrives.
Is this evidence supporting my hypothesis that the bypass must be low inductance to allow it to pass the HF more easily than the big cap can with the long lead outs?
What is the difference between the BC639G and The BC639-16? To me it looks like it is just the factory packaging, One being Bulk the other being ammo pack.
The reason I ask is because it appears that I only ordered the BC639-16's when I placed my order with Mouser. Are the two interchangeable?
James
The reason I ask is because it appears that I only ordered the BC639-16's when I placed my order with Mouser. Are the two interchangeable?
James
That's reasonably intuitive - lower series L is usually beneficial. However, for a small cap, a bit of lead inductance doesn't hurt as much as it does for a much larger-valued cap, because sqrt(LC) is much smaller to begin with, ensuring that the resonant frequency is well above the audio band.
In this particular application, they are the beta multipliers in the shunt regulator. There's also one bc639 that drives the speaker-protection relay.
The substitutes for the bc639/640 are harder to find - 2sb647 and 2sd667 come to mind, as well as several other Japanese TO126 chroma transistors.
In a pinch, you can use BC337-16 and BC327-16 in the shunt regulator, but be aware that they have lower Vceo ratings, and that the pinout probably differs from the BCE pinout of the BC639/640.
Andrew,
linuxguru can answer your question from his experience. From my limited experience, I think the inductive effect is small or hard to prove. After your earlier comments, I tried a bypass cap on input DC blocking cap in two positions, directly on the board in a direct line between the input and output. This should be very low inductance as it is separated from the ground plane by only the thickness of the board. I also tried it with additional 30 mm leads which left a loop above the board of about 20 mm. I couldn't tell any sound difference. That said, I think your idea is correct and putting bypass cap and return together for minimum inductance is a very good practice. I still don't understand what happens inductively when using a large diameter capacitor. I think the current averages through the center of the cap and that leaves a loop, but I don't know of any evidence that if affect the sound.
Anyway, I plan on using off board caps and twisting the bypass and return together as a way to use your theory.
Jac
No, they're different.
-16 versions are guaranteed for higher Hfe (100 minimum)
In the regulator it's an important parameter, the higher the Hfe the lower the output impedance.
You should be able to use the -16 version on the protection circuitry without problems.
Here is my progress so far.
SMD stuff went well for only my second project with it. The first being a PUP Dac kit.
Can not wait to finish up the other one and having sound flow from them. Thanks for the great project Dario!
SMD stuff went well for only my second project with it. The first being a PUP Dac kit.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
Can not wait to finish up the other one and having sound flow from them. Thanks for the great project Dario!
Last edited:
OMG!!!
This amp is totally fantastic!! Kind of remind me of a transistor Class A amplifier but with the spaciality of a Single Ended Tube amplifier. Really impressive!!!
I find that in transparency, my Goldmund wins but there's really nothing to hate of that My_Ref FE. It is in my list of preferred amps. Imaging is where I find it extremely pleasing, the soundstage is really deep. In Miles Davis - Kind of Blue on So What track, you can really feel the distances of the instrument, the double-bass is in the back of the scene and you can really feel the room. I find this is the harder part to reproduce. Many amplifiers will have a wide soundstage but the depth is not always well defined this is where this amplifier shines. There's also a lot of air between the instrument and separation is very good. Heavy passages don't come all compressed. Also the reproduction of timbres is right on!
As you can tell, I'm really excited about this amplifier.
Ciao!
Do
This amp is totally fantastic!! Kind of remind me of a transistor Class A amplifier but with the spaciality of a Single Ended Tube amplifier. Really impressive!!!
I find that in transparency, my Goldmund wins but there's really nothing to hate of that My_Ref FE. It is in my list of preferred amps. Imaging is where I find it extremely pleasing, the soundstage is really deep. In Miles Davis - Kind of Blue on So What track, you can really feel the distances of the instrument, the double-bass is in the back of the scene and you can really feel the room. I find this is the harder part to reproduce. Many amplifiers will have a wide soundstage but the depth is not always well defined this is where this amplifier shines. There's also a lot of air between the instrument and separation is very good. Heavy passages don't come all compressed. Also the reproduction of timbres is right on!
As you can tell, I'm really excited about this amplifier.
Ciao!
Do
