Gainclone vs Tripath T Amps Vs Nelson Pass F5

Hiten

Well-Known Member
Joined
Oct 17, 2008
Messages
3,222
Points
113
Location
Kalyan
Hi guys, help me with this.

Sound quality wise and technically which is better amongst these three...
Gainclone / Tripath T Amps / Nelson Pass F5
I know SQ is subjective matter and all three may be different (Class) design and may have different circuit design philosophy, but all three have got good reviews on various audiophile websites so comparing them will be interesting.

Thanks !
 
I'll let you know in a couple of months - am in the process of building one of each :eek:hyeah:

I've compared a ultra cheap TA2024 based amp running on a laptop charger power supply to a Miniwatt and I liked the tripath more (comparison based on a pair of Fostex DIY speakers). I've also tried a TA2022. It doesnt sound as good as its low power cousin but does deal with tougher speakers better. I'm in the process of putting together a TA2020 base tripath amp using a better board and power supply (parts ordered off ebay, awaiting delivery).
 
Hi Jai
I'll let you know in a couple of months - am in the process of building one of each
I can wait. Pl. post your reviews here.
I'm in the process of putting together a TA2020 base tripath amp using a better board and power supply (parts ordered off ebay, awaiting delivery).
On ebay A valve based preamp and Gainclone kit combo was also interesting. will look if still available.
(comparison based on a pair of Fostex DIY speakers).
How much did this DIY speakers cost ? (speaker+cabinet etc.)
Thanks
 
Last edited:
There is no comparison between the 3 of them..THe pass labs is just in a different league compared to the gainclone and likewise comparing the Tripath T amp to a well made gainclone.

If you can build the F5..go for it.
A gainclone no matter how good it is will lack authority in the lower octaves..no comment on the T amps ;)
 
OK Sachi
So your vote goes to F5. :)

Reason for putting T amp was it (Trend Audio's TA-10) was reviewed on 6 moons audio review and was seriously considered to be good.
6moons audio reviews: Trends Audio TA-10
But I dont like digital manipulation in amplification anyway. Some remarks written about "Audiophiles" are good to know.

Though I am non technical person I find minimalist simple approach to quality sound amplification interesting.
Thanks for your views on F5.
 
OK Sachi
So your vote goes to F5. :)

Reason for putting T amp was it (Trend Audio's TA-10) was reviewed on 6 moons audio review and was seriously considered to be good.
6moons audio reviews: Trends Audio TA-10
But I dont like digital manipulation in amplification anyway. Some remarks written about "Audiophiles" are good to know.

Though I am non technical person I find minimalist simple approach to quality sound amplification interesting.
Thanks for your views on F5.

I don't want to be overly critical about the Trends Audio, but i consider it to be low end budget-fi.
Regarding simplicity of amplifcation, the hypex Class D modules deserve a serious look. Phenomenal sounding amplifiers. I haven't heard an ICE powered amp yet but have heard the Hypex one in the Veloce (custom tube input buffer i think) and is another brand which escapes my mind.

Of course you could always DIY it yourself, they come as fully built and tested units. Lots of support on diyaudio.com

Advatages include, build is just assembling different boards (amp, psu) heatsink issues are largely taken care of thanks to high efficiency.
The F5 on the other hand..will double up as a nice room heater :p
 
the hypex Class D modules deserve a serious look. Phenomenal sounding amplifiers.
Hmm, one more option to choose from.
But first will try to get more info on F5
thanks
 
My F5 should be up next week- it's my winter amplifier, and a Class D of some kind (either a TDA8950 bridged, or a TA3020) by January or February - that will basically hold up the duties for summer.

Meantime, the gainclone, depending on how you build it, can be a pretty serious piece of kit. There are too many variations to be able to test all of them, but I find that with some nice Holcos and Blackgates, direct-coupled inputs, and lots (80,000uF) of capacitance, they kick some serious high-end butt.

Their biggest issue is that they run out of steam with inefficient speakers and at very high volumes, but for normal everyday listening they can't be bettered by mid-fi equipment at similar prices.

The Hypex are very, very reputed and I was hoping to try them but they're too rich for my blood. There's not much to DIY - they come fully assembled and you just need to provide a power supply, case and case hardware to fire one up. I've been looking closely at the modules offered by Connex, which seem to be much better value for slightly lower performance.

Can you provide a link where more information about Connex can be had?
 
Hi Cranky.
My F5 should be up next week- it's my winter amplifier,
Pl. give us review and pic of F5 when ready.
Meantime, the gainclone, depending on how you build it, can be a pretty serious piece of kit. There are too many variations to be able to test all of them, but I find that with some nice Holcos and Blackgates, direct-coupled inputs, and lots (80,000uF) of capacitance, they kick some serious high-end butt.
That's good to know, so with quality parts we can improve performance.

Hi Sachi
A gainclone no matter how good it is will lack authority in the lower octaves
Can't this be overcome with modification or some other way ? Is this problem noticeable to average listener ?
 
Last edited:
It depends on the topology of the Gainclone. A plain-vanilla single-chip chipamp like the LM3886 and TDA7294 are capable of serious audio and sub-Rs.1000 prices for a populated gainclone board.

The next step up could be a buffered gainclone with a high-end opamp buffer, and this can reach high-end audiophile territory depending on the board design and layout. I've just built a buffered gainclone (details later) and it already sounds better than almost all of my Technics power amps.

The next step above this is a nested or multi-loop gainclone; these are hard to design and even harder to layout and build. However, a well-built nested gainclone with maybe an LM3875 or LM3886 as the power stage and a high-end op-amp as the front-end, can beat $10k discrete designs. I have a nested gainclone in the early stages of simulation and design. How good is it? Very good - THD20 numbers are likely to be around 0.0003% at full power.
 
There's currently a group-buy that's been initiated at DIYaudio for a premium version of the best-sounding high-end nested gainclone designed so far - the Mauro Penasa MyRef Rev C. Both bare PCBs and full kits (boards + components) are being offered; anybody who is interested can use the information at this thread to sign up:

MyRef_C with Ultimate BOM - diyAudio

My SPICE simulations with a similar topology (nested Howland Current-Pump with global voltage-series feedback), but less expensive op-amps (NE5532 and LM1875), show that it is very hard to beat this topology on any metric - distortion, harmonics/sonics, stability into difficult loads, etc. My design will happily drive a 1000 uF capacitor at the speaker terminals without oscillating.

I'd highly recommend the Mauro Penasa ultimate version group-buy that's being offered now.
 
Thanks for the post, I think a couple of my friends might be interested. I wonder if Cranky had seen this post, I'd love to hear what he thinks of the Mauro Penasa vis-a-vis the F5...
 
I've never heard the nested GC which Mauro uses. It is pretty heavily dependent on the opamp used, as the gainclone runs in slave mode - so super-fast opamps are a problem without compensation. That may be a worthwhile experiment at some point, I am on to my last 2 LM3886 chips from my Digikey reserves, but those are emergency replacements for the two amps I have with active chips in them. Right now I'm busy building out a Class D amplifier, should have comparisons with a gainclone and the F5 pretty soon.

Between a (regular) LM3886/4766/4780 amp and the F5 clone, there is really no comparison. The F5 beats out all of them handily, offering better bass control, a airy and extended top end, and extremely detailed midrange. The GCs are very, very good, but once you build a good discrete you really can't touch it with a good chip. And my GC is pretty well pimped - Holcos and Black Gates throughout, with a huge capacitance reserve for better bass depth. It was already better than most $1000 amplifiers I had heard, but the F5 outclasses it very handily. The GCs have a slightly 'tubey' midrange, for a more mellow and sweetened presentation.

Once the Class D gets built (I guess will be ready by Wed/Thursday) I will have a chance to compare with the F5 clone - the only problem is it takes a long time to warm up and reach operating temperature, so it will rely heavily on aural memory, not very reliable at the best of times.

Thanks, appreciate your response!
 
hey cranky, i'd love to know what you think of the tripath chips (especially the TA2020) which I am listening to and quite liking. I am sure they don't compare to an F5, but I wonder what an experienced DIY-er like you thinks of them.

I've never heard the nested GC which Mauro uses. It is pretty heavily dependent on the opamp used, as the gainclone runs in slave mode - so super-fast opamps are a problem without compensation. That may be a worthwhile experiment at some point, I am on to my last 2 LM3886 chips from my Digikey reserves, but those are emergency replacements for the two amps I have with active chips in them. Right now I'm busy building out a Class D amplifier, should have comparisons with a gainclone and the F5 pretty soon.

Between a (regular) LM3886/4766/4780 amp and the F5 clone, there is really no comparison. The F5 beats out all of them handily, offering better bass control, a airy and extended top end, and extremely detailed midrange. The GCs are very, very good, but once you build a good discrete you really can't touch it with a good chip. And my GC is pretty well pimped - Holcos and Black Gates throughout, with a huge capacitance reserve for better bass depth. It was already better than most $1000 amplifiers I had heard, but the F5 outclasses it very handily. The GCs have a slightly 'tubey' midrange, for a more mellow and sweetened presentation.

Once the Class D gets built (I guess will be ready by Wed/Thursday) I will have a chance to compare with the F5 clone - the only problem is it takes a long time to warm up and reach operating temperature, so it will rely heavily on aural memory, not very reliable at the best of times.
 
Cranky, thanks for your comments. Is it worth picking up the F5 bare PCB from EBay that's offered by one of the Hong Kong/Chinese sellers? I'm not terribly keen on Class A because of the efficiency/heat-sinking/bias-sensitivity issues, but I'm willing to make a low-power Class-A just for experimentation, for which my money is on the JLH 2003.

On another note, there is an adequate supply of authentic LM3886TF (isolated package) at S.P.Road, Bangalore at sub-Rs.200 price points. I have a couple lying around, but can get any reasonable quantity off-the-shelf.
 
Right now I'm busy building out a Class D amplifier, should have comparisons with a gainclone and the F5 pretty soon.
That would be great cranky.
------------------------------------------------------------------------------------------------------------------------------
I have high regards for Nelson pass as a person because he encourages DIY and has not commercialised his designs.
But I think it should be commercially available as a kit or ready module after 2 to 3 years of launching a circuit design. Why to restrict pleasure of good hifi sound to DIYrs only.
 
Oh I misunderstood. I thought Pass Designs were for DIYrs only and have to take his permission to build one..

--Thanks
 
Hi,

Although I had read couple of posts,I later somehow missed the recent postings :p !

Anyways I will not comment on what sounds better and such,many have already done the same and there's no reason for me to start the same !!

I just wanted to inform the folks who are interested in Gainclone Amp especially the one mentioned by Linuxguru

There's currently a group-buy that's been initiated at DIYaudio for a premium version of the best-sounding high-end nested gainclone designed so far - the Mauro Penasa MyRef Rev C. Both bare PCBs and full kits (boards + components) are being offered; anybody who is interested can use the information at this thread to sign up:

This very Amp was offered as GB long time back aswell and our forum member Quadtech has successfully built the unit and will be able to tell the whole lot on them as he has some tube gear and single fullrange speakers to test it and compare etc etc etc.............

Here's his build details (pics)

MauroMyRefRevC


Folks interested in reading a bit bout what he says bout the Amp can read it,He posted the details on another forum ages ago !

LM318+LM3886 MyRef amp completed

I myself have recently done a gainclone with NE5532 as Buffer have put it on burn-in mode,but one thing I can say is without much of burnin itself it sounds very good and resolution is also neat ! NOT A TYPICAL CHIP SOUND FOR SURE coz have tried all the chips that there's to try( although not the latest one's from NS) and somehow this one pleases me more !!

Regards.
 
Last edited:
A beautiful, well-constructed speaker with class-leading soundstage, imaging and bass that is fast, deep, and precise.
Back
Top