Krisfr - So what? What does that have to do with anything?
1.) Metal MOSFETS sound better.
2.) Metal MOSFETS can take more abuse, thermally, etc.
3.) Metal MOSFETS have two mounting holes. Two bolts hold a transistor better than one. Metal case MOSFETS can be torqued down harder without the possibility of cracking the body. A nicely torqued down metal case transfers heat better from the MOSFET to the heatsink. Plastic transistors only have a single (offset) mounting hole. If you over torque (even just a tad too much) the body warps and one end actually lifts off of the heatsink. Plastic MOSFETS are fragile and can crack.
And the list goes on.
1.) Metal MOSFETS sound better.
2.) Metal MOSFETS can take more abuse, thermally, etc.
3.) Metal MOSFETS have two mounting holes. Two bolts hold a transistor better than one. Metal case MOSFETS can be torqued down harder without the possibility of cracking the body. A nicely torqued down metal case transfers heat better from the MOSFET to the heatsink. Plastic transistors only have a single (offset) mounting hole. If you over torque (even just a tad too much) the body warps and one end actually lifts off of the heatsink. Plastic MOSFETS are fragile and can crack.
And the list goes on.
Last edited:
Just a comment. Doesn't goldmund also use plastic To247 devices on the output, of their newer amps?
I guess the plastic packaged devices cant be too bad or the goldmund couldn't be the best sounding amplifier in the world.
I guess the plastic packaged devices cant be too bad or the goldmund couldn't be the best sounding amplifier in the world.
Last edited:
Melon Head - They sure do!! And that is one of the many, many reasons why their current amplifiers don't sound as good as the original Mimesis series.
Who said their latest amplifiers are the best sounding in the world?
Out of all the current amplifiers today, they still sound the best in the world, but out of any production date amplifiers, the Mimesis series is king.
Who said their latest amplifiers are the best sounding in the world?
Out of all the current amplifiers today, they still sound the best in the world, but out of any production date amplifiers, the Mimesis series is king.
Reminds me of this guy who cuts open capacitors and puts them in wooden cases with beeswax to seal it.
And they sound sooooooo much better of course. 🙄
The page (In Dutch but the pictures tell the tale) || DHTRob - Projecten: Het modificeren van een electrlytische condensator (1)... ||
I thought you said the Goldmund was the best sounding amp in the world. When you first started out you didn't seem too fussed about using plastic output devices until you found you could get the metal can devices.
I won't disagree that the packaging can have an effect on the sound, but I think you might be making a mountain out of a mole hill here. Unless everyone has state of the art everywhere else in their audio chain, I think most people won't hear much difference.
Last edited:
Yes, but I always referred to the Mimesis series, always.
The project wasn't in my hands. I would have built the clone either way, with either style of output MOSFETs.
But if this is to be the best of the best, then certainly metal MOSFETs should be used!!
The project wasn't in my hands. I would have built the clone either way, with either style of output MOSFETs.
But if this is to be the best of the best, then certainly metal MOSFETs should be used!!
hmmm....I for one have used Golmund..quite allot for demonstration purposes...It's good.. but I must say not all is superior..... mids and highs are soft and silky.. but the buttom...lacks the grip an untter control that is clearly present in other designs..in that aspect they sound like what they are mossfet amplifiers...So stating that it's the best of the best....is like saynig butter is better than oranges...
It's very very good HiFi....
But i guess that not the point of this thread.. to me it's taking a respected and good design...trying to understand it and see if it can be improved with all the skills that we have here..and when doing noncommercial design we don't have to cut corners......and evaluate every euro or dollar spent..
It's very very good HiFi....
But i guess that not the point of this thread.. to me it's taking a respected and good design...trying to understand it and see if it can be improved with all the skills that we have here..and when doing noncommercial design we don't have to cut corners......and evaluate every euro or dollar spent..
Last edited:
Indeed, but it s true that metal cases are far better.
My 2X100W mosfet amp using three pairs of 2SJ48/2SK133 per channel
did resist to incredibly bad treatments even it has only a rudimentary
protection that act when temperature is about 90°C,
and this for 15 years.
It now lives as my 24/7 amp, downgraded to 2X10W...
Surely that with plastic case trannies, it wouldn t be there..
That is exactly why once this is designed and built and Stable, it can be expanded to incorporate 4 5 6 7 8 pairs to accomadate what ever.
TO EACH HIS OWN....😉😀😉
Chocolate Vanilla Strawberry
I Like each one on different days...😛
Great, but unfortunately, this will not be significantly better
than the simplistic original , if not no better at all...
So far, in my tries, although it s possible to reduce THD,
the improvements i got make me think that to relegate the ancestor
as a ripe for replacement, this is simply not the adequate topology,
or not its good implementation.
That said, i m interested if one can bring a real breakthrough rather
than a collection of a patchwork of subcircuits that bring nothing
either in performances or reliability..
I would believe the simulator as opposed to you . Even as it is just simulated the A/B FFT comparison of the goldmund and my MUCH improved version speakes for itself. EVEN if the simulator was "off" , it can't be that far "off".
(attachment 1) is the original , it's crude design shows itself in H2/H3 distortion , but it does have pretty low higher order harmonics ... in other words it shouldn't sound to bad. BUT , it is NOT the worlds best amplifier ... not by a long shot.
(Attachment 2) shows mine , which is CLOSE TO KEENS.. it has 20-30db !!! LESS H2/3 and practically no upper harmonics.
To interpret this into the real world is simple . one would have a sound (flavor , tone, character) and the other would be as a perfect magnifying glass adding nothing ... you would JUST hear your source and the noise of your input differential.
To all the disbelievers , such a massive reduction in distortion (in simulation) would set the two circuits apart in a living room and in the real world circuit (on a scope or distortion analyzer) .The differences would be substantial either way. One might like the distortion of a particular amp , just as the tube lovers like thier warm second harmonic (H2) good for them ... I like nothing.
More udder *@** , I have built 4 or 5 versions of this topology , and with my modular setup , try out 5 more variants within the hour. ALL TALK , NO WALK !!
PS. notice how the spectra of the 2 amps is so similar , the beta enhancement and the better devices (ksa/c's) are the big factors in the performance differences. Amp #2 has 120v/us slew and is quite stable into any capacitive load. I make my mouser order tonight.. 🙂
OS
Attachments
Last edited:
I agree , this topology is not the best sounding or the highest performer. Carlos's (DX) amp sounds better , I am going to reference this amp to it...sonically. Fully complimentary topologies will have higher slew , just by design. Each has thier own sound , pick a "flavor". 😀
OS
All these hitachi "spinoffs" . There is more than one way to skin a cat. look at all the different designs out there and ask yourself why would this one be "the best sounding amplifier in the world".
OS
Quite a long time that i abandonned this kind of topology,
as whatever i tried, it remained far inferior to a full complementary
topology in matter of linearity, while the latter was fast as well
and had better stability, both having Hitachi s laterals as OPS.
Last edited:
AS a software engineer, whose worked with C, assembly language and even raw machine code (poking bytes into memory in the good old days of ZX Spectrums and Commodore Amigas) I can tell you that you're absolutely wrong. Older processors like Z80, 8080, 6502 etc did not have hardware multiply or divide, but x86, Motorola 68000, and several others certainly do have opcodes for it. This is all done by the ALU. The x86 is a hybrid architecture so the i386 opcodes get converted into a RISC-like microcode that is executed, but this is transparent to the user.
And if this is even a suggestion that computation by simulation is somehow inferior to manual labour.... well that is just laughable.
1) Utter tosh. The same die is often used. There is just no basis for this remark.
2) Agreed... but thermal coupling of plastic packages can be improved significantly, see next answer...
3) You can also mount plastic pack transistors using a bar to apply force into the middle of the package, with multiple holes between devices. This gets many of the advantages of TO-3 without the hassle of drilling through holes for leads, having the TO-3 case exposed outside, etc.