I've never heard anyone criticise the sound of the Leach amp. In fact, it seems like everything I've read about it has been positive. I've seen criticisms of some of the general features of the amp, e.g., triple darlingtons, fully complementary, old transistors, etc., but just as many positive comments about the same structural features; i.e., use of the same output transistors in the JLH amp, etc. Looking for positive/negative discussion on the amp schematic, and how it compares with other amps.
I build the Leach some years ago using original PCB's from Mr. Leach.
It was tested for app. 14 days and then scrapped again due to it's weak bass reproduction....
However I find the mid and trebble reproduction very fine, so I would maybe use it again for in an active system, but only for frq's above 100 Hz.
Please note that sound is objective to the individual person, so you may find it as the "right" amp for you....
It was tested for app. 14 days and then scrapped again due to it's weak bass reproduction....
However I find the mid and trebble reproduction very fine, so I would maybe use it again for in an active system, but only for frq's above 100 Hz.
Please note that sound is objective to the individual person, so you may find it as the "right" amp for you....
hi,
i built my first leach amp in 1984, before that, in the mid to late 70's i cloned a poineer sx770 power amp, and other amps in the tiger series of Dan Meyer....
about the leach amp...it is not tiring to listen to at high levels, the amps seem to work effortlessly....mids and highs are crystal clear, bass is decent....the tigersaurus on the other hand provided a lot of puch in the low end, even with just 10kufd in the supply rails...but mids and highs are nowhere near the leach amps in clarity....i'd choose the leach over the tigersaurus anytime...
i built my first leach amp in 1984, before that, in the mid to late 70's i cloned a poineer sx770 power amp, and other amps in the tiger series of Dan Meyer....
about the leach amp...it is not tiring to listen to at high levels, the amps seem to work effortlessly....mids and highs are crystal clear, bass is decent....the tigersaurus on the other hand provided a lot of puch in the low end, even with just 10kufd in the supply rails...but mids and highs are nowhere near the leach amps in clarity....i'd choose the leach over the tigersaurus anytime...
I can only get one channel working on my Leach amp. I don't like the mounting scheme of the transistors outside and all of the wires to the inside. I understand you have to keep the heat out, but a connector seems like a better option, but it may effect sound quality. There is no silkscreen on the board. Not that I could design anything better.
meat said:
I assume you're talking about TO-3 style cases?
This amplifier works fine with the MJ21193/94 and should
work equally well with the MJL plastic case series. Haven't
tried it with the popular Toshiba transistors yet; they might
be fast enough to cause stability problems with the stock
circuit. BrianGT designed a Leach board that has the exact
hole spacing for the plastic case leads.
I'm curious about the perception of lack of deep bass, but I've
never had the chance to listen to an amplifier of similar quality
and power on my system so it's hard to comment. I need to
get out more. :/
It would be fun to compare the Leach amp against ASKA, one
of the OPTIMOS amps and some of the other popular solid
state amp projects, and some really good commercial amps.
I'm sure there's some room for improvement, maybe just
good parts tweaking.
People that don't like the bass on the Leach like ...
MUD.
What you all perceive as 'bass' is power supply mud from un-bypassed large electrolytic capacitors.
If you want your Leach to sound just like the Tigersaurus in the bass, just take out the four pieces of 22µF that bypass the main rail filter caps.
If you want the Leach to sound like the Tigersaurus in the highs, just remove the 0.1µF bypass from the feedback cap and add a Tantalum input coupling cap.
Want your Leach to sound like a Stasis? Move all the feedback to the feedforward point on the pre-drivers (v4.5), even with un-matched outputs the distortion will be less than 0.5% With matching you can get less than 0.05%
MUD.
What you all perceive as 'bass' is power supply mud from un-bypassed large electrolytic capacitors.
If you want your Leach to sound just like the Tigersaurus in the bass, just take out the four pieces of 22µF that bypass the main rail filter caps.
If you want the Leach to sound like the Tigersaurus in the highs, just remove the 0.1µF bypass from the feedback cap and add a Tantalum input coupling cap.
Want your Leach to sound like a Stasis? Move all the feedback to the feedforward point on the pre-drivers (v4.5), even with un-matched outputs the distortion will be less than 0.5% With matching you can get less than 0.05%
I have strong suspection that AKSA will be the one!
Just read entire forum and you will see who's the men.
Not only hugh, but some others too are special, some of them do not like to show themselves, " but has also ".
Some of them we love because they know they are special and take care not to say, others we hate them because special but no shy to show us, others we are tollerant because they hardly knows their quality than has no patient enough related treat our poor insignificance mortal. And others, as a computer (sorry, but only chocoholic knows what i am telling about.. and choco is not the person) ... those, make analises without no feeling going together.
I know one, that is the most precise brain in earth, sinthetic, clear and very kind... but you have to be precise with him, if not, a red lamp start to flash and a buzzer start to run:
- No register!!... bad income information!... attention.!... failure error in register!..... chocked in windows register!
Do not misunderstand, this one i like him very much, but have to be precise when you speak, this way:
- I will be happy to see the schematics!
He do not move itself, you have to say:
- Please, can you send me schematics!
He move fast and gladly.
So, more than technologies, we have to understant one each other. In my language we talk inverted related english, the small sentence;
Can I? in my language i ask I can?
So, whats real meaning.... can or cannot..... i can?, is it some kind of joke, ironic words?... there are cultural an languages problems to understand one each other.
My culture, man value, for a big part of my society, is how good he is, how patient and delicated with sons, how much time spend with family related to their own things, the tolerant way he understand the world, and how much he is not "defended" inside... those that can see error only in others and never can see their own defects and problems. Last evaluation your degrée, and money you have. In a society were money is not well divided, this difference, if great, can be a "bad evaluated" person, because has 10 tv sets listening one each time. 5 cars and driven only one each time....... and, others walking.... this way, rich can have bad evaluation.
Indians teached us that we have nothing, we own nothing, we are passing some time in mother earth, ground, sand is our mother, we will return to be dust, ground, earth and some material forming plants and bacteries... how can we be the owner of something?... this feeling make us flexible... if world finishes figthing, i am sure, we will survive (Gloria Gaynor song i suppose).
Never heard AKSA, and i will not, unemployed, can not face those things, also schematic do not exists, and i suppose can be so simple you understand looking into the board 5 seconds!.... the secrets are!.... and i already discovered with kind of secrets are there. It may be the Best!.... because philosophy precise in this imprecise world, where "values" are changing a lot, adapting persons one each others... and this is profound... deep as ocean!
Carlos
Just read entire forum and you will see who's the men.
Not only hugh, but some others too are special, some of them do not like to show themselves, " but has also ".
Some of them we love because they know they are special and take care not to say, others we hate them because special but no shy to show us, others we are tollerant because they hardly knows their quality than has no patient enough related treat our poor insignificance mortal. And others, as a computer (sorry, but only chocoholic knows what i am telling about.. and choco is not the person) ... those, make analises without no feeling going together.
I know one, that is the most precise brain in earth, sinthetic, clear and very kind... but you have to be precise with him, if not, a red lamp start to flash and a buzzer start to run:
- No register!!... bad income information!... attention.!... failure error in register!..... chocked in windows register!
Do not misunderstand, this one i like him very much, but have to be precise when you speak, this way:
- I will be happy to see the schematics!
He do not move itself, you have to say:
- Please, can you send me schematics!
He move fast and gladly.
So, more than technologies, we have to understant one each other. In my language we talk inverted related english, the small sentence;
Can I? in my language i ask I can?
So, whats real meaning.... can or cannot..... i can?, is it some kind of joke, ironic words?... there are cultural an languages problems to understand one each other.
My culture, man value, for a big part of my society, is how good he is, how patient and delicated with sons, how much time spend with family related to their own things, the tolerant way he understand the world, and how much he is not "defended" inside... those that can see error only in others and never can see their own defects and problems. Last evaluation your degrée, and money you have. In a society were money is not well divided, this difference, if great, can be a "bad evaluated" person, because has 10 tv sets listening one each time. 5 cars and driven only one each time....... and, others walking.... this way, rich can have bad evaluation.
Indians teached us that we have nothing, we own nothing, we are passing some time in mother earth, ground, sand is our mother, we will return to be dust, ground, earth and some material forming plants and bacteries... how can we be the owner of something?... this feeling make us flexible... if world finishes figthing, i am sure, we will survive (Gloria Gaynor song i suppose).
Never heard AKSA, and i will not, unemployed, can not face those things, also schematic do not exists, and i suppose can be so simple you understand looking into the board 5 seconds!.... the secrets are!.... and i already discovered with kind of secrets are there. It may be the Best!.... because philosophy precise in this imprecise world, where "values" are changing a lot, adapting persons one each others... and this is profound... deep as ocean!
Carlos
I built the first Leach amp right after it appeared in Audio magazine. I took it to a friend's house and swapped it for his Dynaco SS amp that he built. I normally don't advocate making evaluations like this, but it stomped the Dynaco. Everyone in the room was amazed at the difference. I subsequently bought the version 4 boards and upgraded the parts to 4.3, but have not yet improved to 4.5. I have not had an opportunity to compare the amp with other amps since my initial build. Has anyone noticed a difference between the 4.3 and 4.5 versions, which added back the diodes in the cascode front end? BTW, the schematics of the amp are here: Leach amp schematic
Hi!
I built up the Leach-Amp and I can say NICE SOUND. Even the bass. The mids and the highs were reproduced very clear. The bass comes very clear and not bumpy. The reason for that might be a big power supply. I used a 500VA transformer with 2x66.000µF reservoir capacitors. You can visit my homepage, were I present some pictures and the layout. Unfortunaltely it in German, but I plan to translate it into english. You will find it under DIY.
All the best
Wopo
http://homepage.univie.ac.at/Wolfgang.Postl/
I built up the Leach-Amp and I can say NICE SOUND. Even the bass. The mids and the highs were reproduced very clear. The bass comes very clear and not bumpy. The reason for that might be a big power supply. I used a 500VA transformer with 2x66.000µF reservoir capacitors. You can visit my homepage, were I present some pictures and the layout. Unfortunaltely it in German, but I plan to translate it into english. You will find it under DIY.
All the best
Wopo
http://homepage.univie.ac.at/Wolfgang.Postl/
I upgraded my old Leach amp from a modified 2.something
to the 4.4 and found it much improved; possible
oscillation on the old boards was suspected anyway.
I found the imaging much improved, a faint hum had
disappeared due to the newer grounding scheme, less
residual noise. I was a very happy camper.
I tossed in a few premium parts, matched resistors
practically down to the quantum level, and rewired
with some teflon wiring. Used MJ21193/94 outputs
as they're more rugged and a bit more linear.
One modification: a 1 uF film cap between the driver
emitters, as per Self and Sloan. Supposed to help
turn off the outputs and improve high frequency
distortion. I set bias for a minimum of crossover
distortion, there's a definite null.
I don't have any pictures of my amplifier; all I can say
it's mostly power supply...
to the 4.4 and found it much improved; possible
oscillation on the old boards was suspected anyway.
I found the imaging much improved, a faint hum had
disappeared due to the newer grounding scheme, less
residual noise. I was a very happy camper.
I tossed in a few premium parts, matched resistors
practically down to the quantum level, and rewired
with some teflon wiring. Used MJ21193/94 outputs
as they're more rugged and a bit more linear.
One modification: a 1 uF film cap between the driver
emitters, as per Self and Sloan. Supposed to help
turn off the outputs and improve high frequency
distortion. I set bias for a minimum of crossover
distortion, there's a definite null.
I don't have any pictures of my amplifier; all I can say
it's mostly power supply...
Damon Hill said:
Don't think this is an issue with the T-circuit output. That's supposed to help turnoff. I don't think Self ever mentioned the triple Darlington, at least that I could find. He mostly discussed regular two-transistor Darlington. Not as familiar with Sloan.
I'm curious about Leach's removal of the zeners in the input cascodes, and later reintroduction. I never heard anything about that history.
I had an input ground loop problem at first. I couldn't really do the fix he suggests, because I have my input jacks on opposite sides of the back panel, rather than close together. I didn't want to ground one jack through the other at that distance. I put in some Jensen input transformers, and all hum dropped out completely. In fact, the amp is almost silent, even with 107db sensitive speakers, until I put my ear up to the tweeter.
i never had any hum problems in any of my amps, i used the star grounding system....
my power traffos have their secondary windings bifilliar wound to exactly locate the center tap....this practice was mentioned by jim bonjiorno in hi GAS Ampzilla article published in AUDIO magazine sometime in 1975.....
my power traffos have their secondary windings bifilliar wound to exactly locate the center tap....this practice was mentioned by jim bonjiorno in hi GAS Ampzilla article published in AUDIO magazine sometime in 1975.....
Attachments
Self just says triples are troublesome and prone to oscillation, so he would not consider using them. Says much the same about MOSFET outputs -- a bit of a head in the sand attitude, or head somewhere else perhaps. Slone follows the same line, but Slone actually does some SPICE modelling to show the potential instability, so I respect that more. Self just handwaves and calls them impractical.pooge said:
Yet we have ancient designs like the JBL T-circuit from 1967 and the Leach that work quite well to this day. With care, triples work just fine.
While I greatly respect the methodology and objective findings Self has presented regarding the front-end design, his output stages are simply too conservative given the outstanding devices we have available today.
It should make a slight improvement in PSRR, but I suspect a resistive divider with small cap filter would work almost as well.pooge said:
Overall, the Leach amp is a superb, well-balanced compromise and much ahead of its time (1976). This amp is about as bulletproof a design as you can find, has been beat up by thousands of students for over almost three decades. No gimmicks, just solid engineering design and great sound.
slowhands said:
As I understand it, "Self's" Complementary Feedback Pair output is also prone to oscillation. If I recall, his published "blameless" amp used EF output.
slowhands said:
I was just curious why he removed them in the first place.
slowhands said:
It seems like he's eliminated oscillation problems with this amp, despite the darlington's.
What version of Self are you referring to?
You might want to refer to the third edition of Self's book, as well as some of his more recent publications; in it his prime candidate for a load invariant amp, and the circuit with which he studies the extended beta devices is a straight EF3 T circuit, with the class A configuration for the drivers.
There are potential issues with stabilizing the T circuit because of the gain and bandwith of the transistors- the parastic capacitive load can make the circuit look like a Colpitts oscillator at very high frequencies. The oscillation, though, is local to the transistor, though it propagates through the rest of the circuit. But it's not a local loop to stabilize, unlike his dalliance/romance with the CFP, which is substantially harder to stabilize over a wide range of loads without an output inductor.
It's possible to make a CFP circuit that is stable with capacitive output loads with no output inductor, but only in my experience by using low GM devices (MOSFETs) as a class A driver, and having very low turn-off impedace for the upper devices (5 ohm driving the distributed gate load). The output stage was pretty brute force, having 8 each P and N channel four chip Magnatek iso modules.
The net output stage transfer function isn't as whimpy this way as a straight MOSFET follower; I'm sure CH would argue it doesn't sound as good, but the client was looking for bottom end control, as well as a smooth, clear mid and highs, and he was quite happy with the results.
Regards,
Jon
You might want to refer to the third edition of Self's book, as well as some of his more recent publications; in it his prime candidate for a load invariant amp, and the circuit with which he studies the extended beta devices is a straight EF3 T circuit, with the class A configuration for the drivers.
There are potential issues with stabilizing the T circuit because of the gain and bandwith of the transistors- the parastic capacitive load can make the circuit look like a Colpitts oscillator at very high frequencies. The oscillation, though, is local to the transistor, though it propagates through the rest of the circuit. But it's not a local loop to stabilize, unlike his dalliance/romance with the CFP, which is substantially harder to stabilize over a wide range of loads without an output inductor.
It's possible to make a CFP circuit that is stable with capacitive output loads with no output inductor, but only in my experience by using low GM devices (MOSFETs) as a class A driver, and having very low turn-off impedace for the upper devices (5 ohm driving the distributed gate load). The output stage was pretty brute force, having 8 each P and N channel four chip Magnatek iso modules.
The net output stage transfer function isn't as whimpy this way as a straight MOSFET follower; I'm sure CH would argue it doesn't sound as good, but the client was looking for bottom end control, as well as a smooth, clear mid and highs, and he was quite happy with the results.
Regards,
Jon
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.