Hello all,
JLH 1969 vs LM3886
I did this test for 8 days I wanted to be sure that my opinions were right. I was also concerned not to use the daft metaphors one sees on, for instance six loons.
I used the setup you can see here;
Input = One channel from my CD/DVD player through to a switch with two metres of wire
Output = left loudspeaker from the JLH 1969, the right one from the LM3886
By doing this I am able to switch from one amplifier to another instantly and can use the return on the CD/DVD remote to repeat a passage in seconds.
My first reaction was there's not that much difference. After half an hour or so I realised that the JLH sounded a little brighter. This was with Mahler's 1st symphony. The following day I tried some old Jazz, Bessie Smith and Clarence Williams's acoustic recordings from the early 1920's and Lady Day's from 1937, both sounding a bit clearer on the JLH.
I'd listened to these recordings many times and realised that I was listening to what I expected to hear i.e. the 3886 and 3876 before it and not what the JLH was giving me. The JLH still sounded a bit brighter though.
Next up was Norah Jones and 'What am I to you' this time I listened to her and realised that her voice was more up front with the JLH, this is where the switch came into it's own, being able to swap amps instantly identifies ones thoughts like nothing else.
The piano version of the Enigma Variations was next and one particularly loud and fast variation sounded quite brighter on the JLH but a little dull from the 3886. Oh dear I thought there must be something wrong with my speaker on the right.
I removed the drive units and the input lugs from both, resoldered them and undid and re-tightened the cables, I then swapped the speakers left for right and right for left.
Back to the variations and the same result, it wasn't my speakers at all! I thought now I'm on to something here. On another forum there is a member who calls himself Poultrygeist he mentioned shimmer in a post with reference to Cymbals. My goodness! A bright light in ones brain, that awoke a recent memory, I'd never heard of the term as aural only as pictorial and thanked him.
Straight back to Norah Jones and yes there it was, a barely heard and rather dull Cymbal from the 3886 and a clear Shimmer from the JLH. What's going on here I thought and back to the Variations, there it was again that instant brightness as I listened carefully to the JLH.
It then dawned on me what was going on (forgive me I'm just a slow old cheapskate git) it was the ADSR the JLH gives me the spike between A and D the 3886 doesn't;
After that I tried Glenn Gould's Goldberg's I could hear him better on the JLH, maybe not such a good thing! Then Julian Breams BBC series Guitarra I could hear his breathing and his jacket rustling and his fingers sliding up the fret (not just the squeaks) at first I didn't believe it and played them over and over again on a couple of days to make sure.
Back to Mahler and the 1st symphony by this time I was used to the brighter sound and listened for the bass drums in the background, a dull rumble on the 3886 I'd never taken much notice of, on the JLH a clearer more concise sound, by this time I wasn't surprised at all everything the 3886 did was a little better on the JLH.
I'd saved Tangerine Dream for a few days, this is the new trio with their wall of full frequency sound recorded at high volume (into the red as I found out in Audacity), deep bass to high treble and everything between. Clarity is what I got from the JLH and a slight confusion from the 3886.
I tried other recordings of course I won't bore you with the names the ones above were the occasions when something dawned on me and was repeated with lots of other music.
The last eight days have been rather strange I did have to look for differences they were not immediately apparent. The 3886 is a fine amplifier no doubt about that, the JLH is slightly better.
What will I be listening to now for the foreseeable future, the JLH 🙂
Cheers -J
JLH 1969 vs LM3886
I did this test for 8 days I wanted to be sure that my opinions were right. I was also concerned not to use the daft metaphors one sees on, for instance six loons.
I used the setup you can see here;
An externally hosted image should be here but it was not working when we last tested it.
Input = One channel from my CD/DVD player through to a switch with two metres of wire
Output = left loudspeaker from the JLH 1969, the right one from the LM3886
By doing this I am able to switch from one amplifier to another instantly and can use the return on the CD/DVD remote to repeat a passage in seconds.
My first reaction was there's not that much difference. After half an hour or so I realised that the JLH sounded a little brighter. This was with Mahler's 1st symphony. The following day I tried some old Jazz, Bessie Smith and Clarence Williams's acoustic recordings from the early 1920's and Lady Day's from 1937, both sounding a bit clearer on the JLH.
I'd listened to these recordings many times and realised that I was listening to what I expected to hear i.e. the 3886 and 3876 before it and not what the JLH was giving me. The JLH still sounded a bit brighter though.
Next up was Norah Jones and 'What am I to you' this time I listened to her and realised that her voice was more up front with the JLH, this is where the switch came into it's own, being able to swap amps instantly identifies ones thoughts like nothing else.
The piano version of the Enigma Variations was next and one particularly loud and fast variation sounded quite brighter on the JLH but a little dull from the 3886. Oh dear I thought there must be something wrong with my speaker on the right.
I removed the drive units and the input lugs from both, resoldered them and undid and re-tightened the cables, I then swapped the speakers left for right and right for left.
Back to the variations and the same result, it wasn't my speakers at all! I thought now I'm on to something here. On another forum there is a member who calls himself Poultrygeist he mentioned shimmer in a post with reference to Cymbals. My goodness! A bright light in ones brain, that awoke a recent memory, I'd never heard of the term as aural only as pictorial and thanked him.
Straight back to Norah Jones and yes there it was, a barely heard and rather dull Cymbal from the 3886 and a clear Shimmer from the JLH. What's going on here I thought and back to the Variations, there it was again that instant brightness as I listened carefully to the JLH.
It then dawned on me what was going on (forgive me I'm just a slow old cheapskate git) it was the ADSR the JLH gives me the spike between A and D the 3886 doesn't;
An externally hosted image should be here but it was not working when we last tested it.
After that I tried Glenn Gould's Goldberg's I could hear him better on the JLH, maybe not such a good thing! Then Julian Breams BBC series Guitarra I could hear his breathing and his jacket rustling and his fingers sliding up the fret (not just the squeaks) at first I didn't believe it and played them over and over again on a couple of days to make sure.
Back to Mahler and the 1st symphony by this time I was used to the brighter sound and listened for the bass drums in the background, a dull rumble on the 3886 I'd never taken much notice of, on the JLH a clearer more concise sound, by this time I wasn't surprised at all everything the 3886 did was a little better on the JLH.
I'd saved Tangerine Dream for a few days, this is the new trio with their wall of full frequency sound recorded at high volume (into the red as I found out in Audacity), deep bass to high treble and everything between. Clarity is what I got from the JLH and a slight confusion from the 3886.
I tried other recordings of course I won't bore you with the names the ones above were the occasions when something dawned on me and was repeated with lots of other music.
The last eight days have been rather strange I did have to look for differences they were not immediately apparent. The 3886 is a fine amplifier no doubt about that, the JLH is slightly better.
What will I be listening to now for the foreseeable future, the JLH 🙂
Cheers -J
Perhaps the magic is not JLH-ness , perhaps the magic is Class-A-ness . Perhaps _every_ Class-A amp beats the 3886 . . . . ?
Almost exactly my playlist including 78s. The good of class A is layers and detail. Many are not really intended to be heard. The Beatles thought this. Class AB and D can jump dynamically for transients which can sound more impressive.
Julian Bream is said to live just south of Shaftesbury.
I am building a valve amplifier as I have plenty parts and they won't build themselves. My friend Jeanine has a 10 watt EL84 design I love. I heard a similar Heath amplifier. I will try a cathode coupled idea like long tail pair. 3 devices 0.7% distortion.
Julian Bream is said to live just south of Shaftesbury.
I am building a valve amplifier as I have plenty parts and they won't build themselves. My friend Jeanine has a 10 watt EL84 design I love. I heard a similar Heath amplifier. I will try a cathode coupled idea like long tail pair. 3 devices 0.7% distortion.
I think there is more to testing and comparing amplifier performance than might seem obvious or necessary. Your method, using a SMPS power supply and basic PCB assembly in a loose wire lash-up, has none of the the usual trappings of a shielded, grounded case or a suitable preamp which ensures the input is matched well enough to the amplifier. Otherwise, it will probably perform differently when fully built as recommended.
Impedance mismatches affect bandwidth, hence tonality, stereo effects and most often the high frequencies which define the sound source and direction, instrument types and some vocal qualities etc. If you arrange and complete the whole amplifier as JLH recommended so long ago, you might find things are different. Perhaps likewise for your LM3886 build. Compare when you complete the whole deal with more normal, linear power supplies for each amplifier. You may be just as pleased either way this time.
I'm not suggesting you shouldn't be happy with the results so far (success is always sweet) but let's look at how far you are from a full working audio amplifier before making comparisons. I think many would agree that it's very hard to find such obvious flaws as you found, in the balance of treble from almost any modern chipamp build done right.
I recall this experience of the JLH sounding bright when I built it in 1976. This original design has no low-pass filtering of the input so it will amplify above 20kHz.
The modular pre-amplifier design to match this included a steep low-pass active filter - turnover at 13.8kHz down to -14dB at 20kHz. The article describing this was published in Wireless World issue of July 1969.
By 1982 in Hood's own words re the foregoing, the "types of circuit block were merely a large improvement of a rather off-colour 1950 d.i.y. hotchpotch."
The 1996 version of his Class A amplifier has a simple RC low pass filter - a series resistor of 4k7 and a capacitor to earth of 330 pF.
This presents the difficulty of modifying the 1969 pcb layout. It is possible to allow for this by including the RC elements elsewhere in a line stage having a gain of 2-3 times.
Member
Joined 2009
Paid Member
if I were comparing, with an SMPS, I’d want an RC. And same goes when you consider all the appliances operating in your house that likely were different back in ‘69.....
Hello Mark,
I was thinking the same, maybe I'm just at the bottom end of it 🙂
Hello Nigel,
JB knocks spots of anyone else, recently I found Tiffany Poon and her Chopin, blimey!
Hello Ian,
Many thanks for your explanation good of you to take so much trouble to do it. I do feel somewhat embarassed saying this I am quite happy with what I've got, I'm a cheapskate I build the simplest stuff I can find. I'm quite satisfied that my JLH 1969 is better than my 3886 and I look forward to years of happy listening. I also make model railways and likewise a 50 year old 3 semiconductor circuit is better than the commercial offerings, for what I like doing, shunting at slow speeds and never stalling. I take them to MR shows and people from other layouts are alway amazed when they have a 'go'.
Cheers all
I was thinking the same, maybe I'm just at the bottom end of it 🙂
Hello Nigel,
JB knocks spots of anyone else, recently I found Tiffany Poon and her Chopin, blimey!
Hello Ian,
Many thanks for your explanation good of you to take so much trouble to do it. I do feel somewhat embarassed saying this I am quite happy with what I've got, I'm a cheapskate I build the simplest stuff I can find. I'm quite satisfied that my JLH 1969 is better than my 3886 and I look forward to years of happy listening. I also make model railways and likewise a 50 year old 3 semiconductor circuit is better than the commercial offerings, for what I like doing, shunting at slow speeds and never stalling. I take them to MR shows and people from other layouts are alway amazed when they have a 'go'.
Cheers all
It's very likely a class A amplifier can be influenced by the power supply. The valve brigade often like choke smoothing. With.high voltage and low current that is a reasonable option although parts are harder to find.
Valve designs are delightfully simple on the basics. Valves tend to sidestep the potential problems like bias and current. If only the JLH was so easy and unfussy. I would say RH34 is the JLH of valves or Baby Huey. JLH had a Williamson amplifier. A 1947 design from a 19 year old engineer. The same man is said to have helped Quad with the electrostatic speakers and 2/22 amplifiers. JLH tried to clone that sound. The Williamson output transformer is still available via a few companies and is considered how a transformer should be made. Really good valve amplifiers have an analytical sound that is bright and fast. Less good valve amplifiers sound soft and pleasing. Most valve lovers let their designs drift towards the soft. A true Williamson which is rare are not soft.
My son bought a tiny Class D amplifier that didn't work Took me a hour to find the problem. I have been running it with a DAB radio and JPW Minims It's a remarkably nice sound. I had been powering it from the computer USB. It did current limit a bit. So I tried one of the very cheap 2.1 amp USB supplies. What a transformation . Reluctant to clip and much more bass colour. I have heard very expensive hi fi that wouldn't be as pleasing. I am not a fan of DAB. I use it because it works where signal is poor on FM. Even the speakers that are great for the price but not actually great might be great in this combination. The message is that most things no matter how humble can be made to work I should try a linear supply. Good class D is very good. It has no audio band crossover distortion unlike class AB that can. The new designs don't need output chokes.
Valve designs are delightfully simple on the basics. Valves tend to sidestep the potential problems like bias and current. If only the JLH was so easy and unfussy. I would say RH34 is the JLH of valves or Baby Huey. JLH had a Williamson amplifier. A 1947 design from a 19 year old engineer. The same man is said to have helped Quad with the electrostatic speakers and 2/22 amplifiers. JLH tried to clone that sound. The Williamson output transformer is still available via a few companies and is considered how a transformer should be made. Really good valve amplifiers have an analytical sound that is bright and fast. Less good valve amplifiers sound soft and pleasing. Most valve lovers let their designs drift towards the soft. A true Williamson which is rare are not soft.
My son bought a tiny Class D amplifier that didn't work Took me a hour to find the problem. I have been running it with a DAB radio and JPW Minims It's a remarkably nice sound. I had been powering it from the computer USB. It did current limit a bit. So I tried one of the very cheap 2.1 amp USB supplies. What a transformation . Reluctant to clip and much more bass colour. I have heard very expensive hi fi that wouldn't be as pleasing. I am not a fan of DAB. I use it because it works where signal is poor on FM. Even the speakers that are great for the price but not actually great might be great in this combination. The message is that most things no matter how humble can be made to work I should try a linear supply. Good class D is very good. It has no audio band crossover distortion unlike class AB that can. The new designs don't need output chokes.
Hello Nigel.
Thanks for the info good of you, one can see it's a matter of taste. Your right about the fast my 3886 could not catch that shimmer or the bright spike as the hammers hit the strings. I used to call at a place in Long Eaton where they made 1,000's of piano parts, fascinating to see how complex the linkages are. I may have mentioned Tiffany Poon, she's got a Steinway on loan in NY and the video on YT that she did after it was delivered is quite revealing, a very bright instrument.
I also have some Chopin played on an 1840's Pleyel also very bright, they were made loud to stop people chatting at soirée's. If anyone can get near a piano give middle C a good bash, ditto some Cymbals with rivets around the perimeter.
Hello KP,
Thanks for the questions, the 3886 has no input or output capacitor and I matched the volumes by ear. One often reads that output capacitors degrade the sound, I think not 🙂
Cheers both
Thanks for the info good of you, one can see it's a matter of taste. Your right about the fast my 3886 could not catch that shimmer or the bright spike as the hammers hit the strings. I used to call at a place in Long Eaton where they made 1,000's of piano parts, fascinating to see how complex the linkages are. I may have mentioned Tiffany Poon, she's got a Steinway on loan in NY and the video on YT that she did after it was delivered is quite revealing, a very bright instrument.
I also have some Chopin played on an 1840's Pleyel also very bright, they were made loud to stop people chatting at soirée's. If anyone can get near a piano give middle C a good bash, ditto some Cymbals with rivets around the perimeter.
Hello KP,
Thanks for the questions, the 3886 has no input or output capacitor and I matched the volumes by ear. One often reads that output capacitors degrade the sound, I think not 🙂
Cheers both
AB comparisons are difficult. The brain remembers so can hear things as identical when only superficially so. I recorded some cassette tapes on a Nakamichi Dragon. As each tape can be calibrated superficially TDK D was good enough. That was untill the next day and the recording whilst tonally excellent wasn't excellent overall. Mostly boring. Maxell MX was a cheap metal that was a top quality cassette tape. It truly was excellent in everyway.
You're right there Nigel and I'd say they were near impossible without some random selection test apparatus which effectively works like the AB-BOX presented here at DIYAudio last year and similar foolproof designs too. Unless they are truly blind tests, logically, we're limited to only confirming the bleeding obvious differences because we all have our expectation and sentimental biases that we add as an unavoidable part of tests when we have even a hint of the actual devices we're comparing.
I've periodically done bench, living room, auditorium and stage listening tests for audio gear over more than 50 years and you may say I must have cloth ears or be tone deaf by now if I can't hear differences that others claim. However, I do still play music and listen to an occasional live classical concert when possible - finally free of hi-fi fanatic concerns about stupid microdetails or quibbles about recording techniques, stereo sound stages and so on.
Problems also arise from using our favourite test music tracks and room conditions, certain speakers etc. that will variously play well with some types of input source and program but not with others. We'll no longer be testing the amplifier but our preferred combinations and likely just wasting time in such an endless tail-chasing pursuit.
The differences are a bit like the climate with and without aircraft. People say the skies are more blue and the air quality nicer. I started to question that as to be honest I wasn't so sure it was obvious. Nearly everyone I see says look at the sky. Being a friend I go along with it. All I can say is Dorset looks like a Constable painting right now so maybe
With music I can be more certain. Boring or not boring. Harsh or not harsh. Bright, neutral or dull. A zero loop feedback valve amplifier my brother made for me was near perfect. It had 1% THD 14 watts and rolled off at 7.5 kHz due to a transfomer mistake. You would never guess the later. It had a very true to life sound despitie not being a post 1947 idea of hi fi. It was reasonably bright !! Bright and harsh are not the same. Bright usually is good it not too exagerated.
I have listened where possible to with or without capacitor coupling. I would say the difference is not vast and bass can be tuned with a output capacitor. Less bass can sound better. Port loaded speakers perhaps need this to be found. Oversized also is a tuning idea. 4700 uF is OK. It is a wonderful protection against DC. Bryan amplifiers made a point of that. It was the 1961 Tobey Dinsdale circuit sold by Bryan.
When AB testing is done a different part of the brain is used. It is useless for this task. It is the learing side. Stress causes us to revert to learning. AB is learning. Do different wines really taste different ? People who think they know don't on a blind test. Not even red from white and these are ones who think they can. Stress of being judged means they can't do it.
Some people are very forceful and say all amplifiers sound the same once a hi fi standard it met. This to me is about them being right and you being wrong. Bob Carver got to the truth of that. I think Bob got it right. One amplifer can be forced to sound like another using a Null principle. Bob said his competitors amplifiers sounded better, he could clone them although not copy their circuits. He wasn't liked for this.
The Carver Challenge | Stereophile.com
With music I can be more certain. Boring or not boring. Harsh or not harsh. Bright, neutral or dull. A zero loop feedback valve amplifier my brother made for me was near perfect. It had 1% THD 14 watts and rolled off at 7.5 kHz due to a transfomer mistake. You would never guess the later. It had a very true to life sound despitie not being a post 1947 idea of hi fi. It was reasonably bright !! Bright and harsh are not the same. Bright usually is good it not too exagerated.
I have listened where possible to with or without capacitor coupling. I would say the difference is not vast and bass can be tuned with a output capacitor. Less bass can sound better. Port loaded speakers perhaps need this to be found. Oversized also is a tuning idea. 4700 uF is OK. It is a wonderful protection against DC. Bryan amplifiers made a point of that. It was the 1961 Tobey Dinsdale circuit sold by Bryan.
When AB testing is done a different part of the brain is used. It is useless for this task. It is the learing side. Stress causes us to revert to learning. AB is learning. Do different wines really taste different ? People who think they know don't on a blind test. Not even red from white and these are ones who think they can. Stress of being judged means they can't do it.
Some people are very forceful and say all amplifiers sound the same once a hi fi standard it met. This to me is about them being right and you being wrong. Bob Carver got to the truth of that. I think Bob got it right. One amplifer can be forced to sound like another using a Null principle. Bob said his competitors amplifiers sounded better, he could clone them although not copy their circuits. He wasn't liked for this.
The Carver Challenge | Stereophile.com
To explain Null. The conjecture is that things like phase shift and bandwidth are making an amplifier sound the way it does. If two amplifiers are fed the same signal the sound difference between them can for example be shown as the output difference between the positive output terminal. If by manipulation the difference is zero Bob Carver thought that the sound should be the same. If you read the Carver Challenge he almost lost the challenge due to unforeseen weeekesses in the method and the challenge. What Bob started to prove is that most theories of sound reproduction are subtly wrong. Indeed very low distortion amplifiers can mimic other amplifiers. The unsuspected factor is the ears of the designer of the good amplifier cones into it. When you buy a great sounding amplifier you are buying the designers music ear. In nearly all other musical devices the designer is credited with the genius. In hi fi the designer is explained away as he who had the luck to choose a good design. As a golf professional said, the more he practiced the luckier he got.
Friends say that the hi fi press chooses to ignore this as it didn't suit their beliefs. What it says to me is if you have musical ears or whatever tune your amplifier. Sometimes an oscilloscope is required to avoid bad choices. Supposed bad choices can sound better. Often this is due to very out of date theory that probably is finding problems that in real music and real loudspeakers don't exist. I doubt that the JLH has this problem.
A good example of tuning is crossover distortion. Get rid of it. The JLH can not have any. Now to challenge Bob Carver . Can his class AB be made to sound exactly like the JLH. I am sure over two days of listening once cloned it can. Over a year I doubt it. This shows the impossibility of any subjective side by side test. It's really like a glass of wine and how you find it.
The NAD 3020 is said to have sold more than any other. It is remarkably unusual in some of the design. Tons of low cost parts. The conclusion is that it's designer had good ears as to it's success. Proton who stopped making it when NAD got their own factory made an upgraded NAD. It wasn't very good! Bland and slightly harsh.
Friends say that the hi fi press chooses to ignore this as it didn't suit their beliefs. What it says to me is if you have musical ears or whatever tune your amplifier. Sometimes an oscilloscope is required to avoid bad choices. Supposed bad choices can sound better. Often this is due to very out of date theory that probably is finding problems that in real music and real loudspeakers don't exist. I doubt that the JLH has this problem.
A good example of tuning is crossover distortion. Get rid of it. The JLH can not have any. Now to challenge Bob Carver . Can his class AB be made to sound exactly like the JLH. I am sure over two days of listening once cloned it can. Over a year I doubt it. This shows the impossibility of any subjective side by side test. It's really like a glass of wine and how you find it.
The NAD 3020 is said to have sold more than any other. It is remarkably unusual in some of the design. Tons of low cost parts. The conclusion is that it's designer had good ears as to it's success. Proton who stopped making it when NAD got their own factory made an upgraded NAD. It wasn't very good! Bland and slightly harsh.
The main, real world problem that the Carver Challenge conveniently ignores (along with the other "all amplifiers sound the same" tests ) is that it disallows clipping.
Now, most of my listening doesn't involve clipping but any attempt at "concert levels" will rapidly run into clipping with "normal" speakers. (Do the math)
Now, most of my listening doesn't involve clipping but any attempt at "concert levels" will rapidly run into clipping with "normal" speakers. (Do the math)
Often it's the speaker and the listener that clips. My ears have clipped in live opera.
The only objection to the Carver test I can see is it uses another persons work. Some Carver amps have very high maximimum power albeit briefly. They won't clip. They were used in Pro Audio. I don't think any amplifier I heard was better when it clipped. Some resisted until the last moment better than others. The Quad 405 was a bit weird when it did.
Carver really upset people. Unlike me they felt that they had been found out. They were happy with you can't hear something or you can. They didn't like the idea it was more complicated than that.
The only objection to the Carver test I can see is it uses another persons work. Some Carver amps have very high maximimum power albeit briefly. They won't clip. They were used in Pro Audio. I don't think any amplifier I heard was better when it clipped. Some resisted until the last moment better than others. The Quad 405 was a bit weird when it did.
Carver really upset people. Unlike me they felt that they had been found out. They were happy with you can't hear something or you can. They didn't like the idea it was more complicated than that.
Reply to post #6526
I have tried now with 2sc3423 instead of 2sc2240. I actually put them both on 3pdt switch to see if I can notice any difference in A-B comparison. With 2sc3423 the amplifier became more stable and not so sensitive to EMI and ground loops. However, the difference in sound is obvious, 2240 is a clear winner. With 3423 the sound becomes “dead” and sounds like a cheap conventional AB class amplifier.
I do not have an explanation to this. Maybe the 3423 I’ve got are fakes, but they look like quality parts and measure within datasheet specs.
I have also tried to compare 2sc2240 to 2SC2320 which I soldered out of an old Japanese amp. These two sound very close to each other. I could only distinguish slight sibiliance with 2240 while 2320 are a bit more neutral.
Note, neither 2240 nor 2320 become even warm. They remain within 30-40 C even without any heat sink. So I do not think there is a problem with too much current they cant handle.
I have already a 47 pf capacitor in parallel with the feedback resistor. What else can I do to get the amp with 2240 more stable? It sounds really good with them and I want to keep them. Currently the problem is occurring when I plug in cables with form a ground loop between channels and input device.
I have tried now with 2sc3423 instead of 2sc2240. I actually put them both on 3pdt switch to see if I can notice any difference in A-B comparison. With 2sc3423 the amplifier became more stable and not so sensitive to EMI and ground loops. However, the difference in sound is obvious, 2240 is a clear winner. With 3423 the sound becomes “dead” and sounds like a cheap conventional AB class amplifier.
I do not have an explanation to this. Maybe the 3423 I’ve got are fakes, but they look like quality parts and measure within datasheet specs.
I have also tried to compare 2sc2240 to 2SC2320 which I soldered out of an old Japanese amp. These two sound very close to each other. I could only distinguish slight sibiliance with 2240 while 2320 are a bit more neutral.
Note, neither 2240 nor 2320 become even warm. They remain within 30-40 C even without any heat sink. So I do not think there is a problem with too much current they cant handle.
I have already a 47 pf capacitor in parallel with the feedback resistor. What else can I do to get the amp with 2240 more stable? It sounds really good with them and I want to keep them. Currently the problem is occurring when I plug in cables with form a ground loop between channels and input device.