Re: Thanks for loading the pic Peter
This is what happens when some mods (and members) are not aware how the new censorship system is working 😉
Jim Read said:
This is what happens when some mods (and members) are not aware how the new censorship system is working 😉
Re: Thanks for loading the pic Peter
Koinichiwa,
No, the Circuit SPECIFIED BY ME contains 1uF, not 0.1uF. I will agree that this may make no difference. But the specified capacitor was a Siemens/Epcos stacked film Mylar (1uF/100V) which was specified because of it's 9nH Residual inductance.
So, you build what you liked, not what was specified.
The PSU is not over the top, it is out of MANUFACTURERS ABSOLUTE LIMIT SPECIFICATION for the LM3875. And no, the effect would not be motorboating and grdual shutdown, look up how the protection works please. It will simply throttle the current to what it "thinks" is safe. The result will be a distortion not unlike clipping. If you have access to a Variac simply turn down the primary voltage for your amp from 240V to around 170V, this should get you into the safe and reliable operating area for the Amplifier chip.
Thermal shutdown is another issue but is such that the chip needs to cool significantly before re-enable.
Lastly, the schematic shows two Bridges and seperate routing of the outputs from these bridges and is at least implied (had you read my original comments explicit) to not be shared between channels. Due to the small value local PSU Cap's significant charge currents flow between bridge and capacitor, in your way of wiring up they will invariably end up flowing at least in part through the nominal "signal ground".
Even if the result is not audible hum the rather substantial higher frequency content may be (due the significant inductances) tracked into the amplifier input and intermodulate with the signal.
And can I fo down to circuit city or dixons and buy one for that? So the price as such is not in the picture. When new the Sony cost (inflation adjusted) a good deal of money and was a quite nice amp.
No, most of the criticsm (at least from me) is about FUNDAMENTAL FUNCTIONAL features you omited and about using the Chip in a way that exceeds a critcal "maximal safe" rating and will cause the "SPIKE" protection to operate when signal is applied to the amplifier.
Excuse ME!? A current clipping portection circuit operating would not substantially alter the sound!? I may be ill.
Given your implementation I am surprised it even makes noise, instead of sending up smoke signals. A credit to Nat Semis Engineering, not yours.
Simply measure the Amplifier supply rails. Anything above +/-42V is past the manufacturers ratings. And you have WAY more.
I suggest that you go back to listening to storebought gear, you are not able to follw a simple a circuit and simple instructions on parts sufficiently to expect any form of success.
Sayonara
Koinichiwa,
Jim Read said:
No, the Circuit SPECIFIED BY ME contains 1uF, not 0.1uF. I will agree that this may make no difference. But the specified capacitor was a Siemens/Epcos stacked film Mylar (1uF/100V) which was specified because of it's 9nH Residual inductance.
So, you build what you liked, not what was specified.
Jim Read said:
The PSU is not over the top, it is out of MANUFACTURERS ABSOLUTE LIMIT SPECIFICATION for the LM3875. And no, the effect would not be motorboating and grdual shutdown, look up how the protection works please. It will simply throttle the current to what it "thinks" is safe. The result will be a distortion not unlike clipping. If you have access to a Variac simply turn down the primary voltage for your amp from 240V to around 170V, this should get you into the safe and reliable operating area for the Amplifier chip.
Thermal shutdown is another issue but is such that the chip needs to cool significantly before re-enable.
Lastly, the schematic shows two Bridges and seperate routing of the outputs from these bridges and is at least implied (had you read my original comments explicit) to not be shared between channels. Due to the small value local PSU Cap's significant charge currents flow between bridge and capacitor, in your way of wiring up they will invariably end up flowing at least in part through the nominal "signal ground".
Even if the result is not audible hum the rather substantial higher frequency content may be (due the significant inductances) tracked into the amplifier input and intermodulate with the signal.
Jim Read said:
And can I fo down to circuit city or dixons and buy one for that? So the price as such is not in the picture. When new the Sony cost (inflation adjusted) a good deal of money and was a quite nice amp.
Jim Read said:
No, most of the criticsm (at least from me) is about FUNDAMENTAL FUNCTIONAL features you omited and about using the Chip in a way that exceeds a critcal "maximal safe" rating and will cause the "SPIKE" protection to operate when signal is applied to the amplifier.
Jim Read said:
Excuse ME!? A current clipping portection circuit operating would not substantially alter the sound!? I may be ill.
Jim Read said:
Given your implementation I am surprised it even makes noise, instead of sending up smoke signals. A credit to Nat Semis Engineering, not yours.
Simply measure the Amplifier supply rails. Anything above +/-42V is past the manufacturers ratings. And you have WAY more.
I suggest that you go back to listening to storebought gear, you are not able to follw a simple a circuit and simple instructions on parts sufficiently to expect any form of success.
Sayonara
Re: Thanks for loading the pic Peter
1) I did go, again, to the chip ampforum and I did find the 1 uF cap (in red) I thought would be there. Not the 0.1 uF cap you claim is there, Where did YOU look?
2) Over the top would probably be more OK. Exceeding the limits greatly would probably not. As mentioned is it surprising that is works at all.
5) You have changed most things in the original diagram and all the changes most likely can cause great changes in the sound. On top of that the importance of the feedback resistor and how it is mounted has be covered in quite a few places. You have a near perfect implementation of how it should not be done.
And as you seem not to know, the person signing off with sayonara has the username Kuei Yang Wang but his real name is "Thorsten" and I would doubt that very few people know better than him what his own circuit looks like. Could be me making a mistake though...
I suggest building it properly according to diagram and recommendations using the chip in safe configuration within limits and THEN, and only then, judging whether it is any good at all. I for certain would not trust myself as much as to discard all the extremely competent people here who have built it and are impressed with what the amp can do. I would most likely seriously consider having made a mistake or several if it was that obviously bad when I built it. The way things are set up there is no way you can make it sound good IMHO. Have you not realised this yet?
Jim Read said:
1) I did go, again, to the chip ampforum and I did find the 1 uF cap (in red) I thought would be there. Not the 0.1 uF cap you claim is there, Where did YOU look?
2) Over the top would probably be more OK. Exceeding the limits greatly would probably not. As mentioned is it surprising that is works at all.
5) You have changed most things in the original diagram and all the changes most likely can cause great changes in the sound. On top of that the importance of the feedback resistor and how it is mounted has be covered in quite a few places. You have a near perfect implementation of how it should not be done.
And as you seem not to know, the person signing off with sayonara has the username Kuei Yang Wang but his real name is "Thorsten" and I would doubt that very few people know better than him what his own circuit looks like. Could be me making a mistake though...
I suggest building it properly according to diagram and recommendations using the chip in safe configuration within limits and THEN, and only then, judging whether it is any good at all. I for certain would not trust myself as much as to discard all the extremely competent people here who have built it and are impressed with what the amp can do. I would most likely seriously consider having made a mistake or several if it was that obviously bad when I built it. The way things are set up there is no way you can make it sound good IMHO. Have you not realised this yet?
Jim Read said:
Hi,
output resistor on your picture look like 1ohm (brown,black,gold,gold). Is it true?If is, this can be problem with thin bass. Put simple wire instead this resistor.IMHO
And, maybe more caps in PS..
Anyway, I was build few GCs, sounds average, nothing special.IMHO
regards
Jim, can you post a picture of your Sony amp (outside and inside)so that I know which amp we are talking about.
Regarding your GC, I think you have problems with your ground arrangements.
Also if you have a variac, try reducing your rail voltages to the recommended voltages.
Eric.
Regarding your GC, I think you have problems with your ground arrangements.
Also if you have a variac, try reducing your rail voltages to the recommended voltages.
Eric.
a smoothed input connection
While your waiting resolve your gainclone issues, try this simple experiment -
Take your Sony amp and feed it AC routed through that filtered, industrial IEC connection. Let us know how it sounds with the same 'improvement' you've used to evaluate you gainclone.
I've no doubt that the filtered inlet is within 'spec' for your gainclone use, but I've always found such improvements to be subjectively quite deleterious in poweramps......
sorta like comparing apples to manacled oranges 😱
While your waiting resolve your gainclone issues, try this simple experiment -
Take your Sony amp and feed it AC routed through that filtered, industrial IEC connection. Let us know how it sounds with the same 'improvement' you've used to evaluate you gainclone.
I've no doubt that the filtered inlet is within 'spec' for your gainclone use, but I've always found such improvements to be subjectively quite deleterious in poweramps......
sorta like comparing apples to manacled oranges 😱
Jim there are lot of possible little faults that your GC doesn´t sound as good as it should be;
but before hunting after those I´d sort out the most fundamental: the transformer.
As this is not just a recommendation or joke from National Semiconductor you´re exceeding the absolute maximum ratings!
The datasheet doesn´t actually tell in detail what happens but an overvoltage protection will somehow limit the signal.
And so you´ll lose dynamics, bass and everything you were looking for.
Nice would be if you had a variac to hand to lower the voltage.
But as a DIYer you probably have another transformer to hand.
I got good results even with an underrated transformer(VA´s).
But really bad ones with an "overrated" voltagewise.
Jens - Isn´t really wondering that the Sony sounds much better
but before hunting after those I´d sort out the most fundamental: the transformer.
As this is not just a recommendation or joke from National Semiconductor you´re exceeding the absolute maximum ratings!
The datasheet doesn´t actually tell in detail what happens but an overvoltage protection will somehow limit the signal.
And so you´ll lose dynamics, bass and everything you were looking for.
Nice would be if you had a variac to hand to lower the voltage.
But as a DIYer you probably have another transformer to hand.
I got good results even with an underrated transformer(VA´s).
But really bad ones with an "overrated" voltagewise.
Jens - Isn´t really wondering that the Sony sounds much better
While I am no fan of the ginclone's sound, I do know that you are in no way doing it justice. Go back and rebuild the psu with less voltage, 18-24VAC would be a much better choice. That alone should show you that it can easilly out do any sony you compare it to.
I'm a little confused. According to the app notes from National Semiconductor, my read is that you can use up to 70V rails for LM3875 safely before the SPIKE protection circuit kicks in. I hope that I am interpreting this correctly. According to the posts, the maximum should be 24V???
Could someone please clarify this.
Thanks,
Vic
Could someone please clarify this.
Thanks,
Vic
Vic,
What the datasheet is specifying is differential voltage which is really +35VDC and -35VDC at the rails. The reason is that you can use this amp with a single voltage supply with one rail at ground and the other up to 70VDC before SPIKE enables. The 24VDC is just what we have found to sound good. Nothing says you can't experiment with the upper limits of the manufacturer's spec.
What the datasheet is specifying is differential voltage which is really +35VDC and -35VDC at the rails. The reason is that you can use this amp with a single voltage supply with one rail at ground and the other up to 70VDC before SPIKE enables. The 24VDC is just what we have found to sound good. Nothing says you can't experiment with the upper limits of the manufacturer's spec.
Exactly, and many people (like me 🙂 ) use the Gainclone at 24 or 25 Vac. This results in 34-36 Vdc and approx. 50 W output power...
Fedde
Fedde
Hi Vic,
PS range from manual is
"Wide supply voltage range: |V+| + |V−| = 20V to 84V"
That means absolute value. Guys here talks about symmetrical (relative) voltage and recommended is +-24V for optimal results.(what is 2*24=48V absolute)
regards
PS range from manual is
"Wide supply voltage range: |V+| + |V−| = 20V to 84V"
That means absolute value. Guys here talks about symmetrical (relative) voltage and recommended is +-24V for optimal results.(what is 2*24=48V absolute)
regards
Koinichiwa
According to the Datasheet the maximum safe Rail Voltages |V+| + |V-| equals 84V. This means that absolute maximum rail voltages when playing music are +42V & -42V.
No, the maximum recommend transformer voltage is 24V+24V AC. When running with no load or very little load the transformer will actually give a certain percentage more voltage than rated. In case of a 160V 24V+24V Torriod I would recommend the regulation is specified as 9%. This means that with low load (Amplifier playing a quiet passage or no signal) the Voltage from the transformer is not 24V but 24V +9% or 26.16V.
Once rectified the DC Voltage becomes around 1.414 times the AC voltage with no load, so the DC voltage will be around 26.16 * 1.414 = 36.99V.
So, a 160VA; 24V+24V transformer will give around +/-37V rail voltages when the amplifier is not putting out much power. This is only around 5V below the rated maximum.
If now our local mains voltage is higher than implied by the nominal 230V (in europe) your DC voltage will be higher. In the UK the highest permissable Voltage is 240V+6% or 11% higher than 230V. In this case our 37V will also be 11% higher, so we now end up with 41V.
Hence, with the LM3875 the highest Transformer Voltage that should be used is 24/25V. Using a significantly higher voltage will put you into risk areas where the voltage sepcfications are concerned, due to the issues dioscussed above.
Sayonara
vic said:
According to the Datasheet the maximum safe Rail Voltages |V+| + |V-| equals 84V. This means that absolute maximum rail voltages when playing music are +42V & -42V.
vic said:
No, the maximum recommend transformer voltage is 24V+24V AC. When running with no load or very little load the transformer will actually give a certain percentage more voltage than rated. In case of a 160V 24V+24V Torriod I would recommend the regulation is specified as 9%. This means that with low load (Amplifier playing a quiet passage or no signal) the Voltage from the transformer is not 24V but 24V +9% or 26.16V.
Once rectified the DC Voltage becomes around 1.414 times the AC voltage with no load, so the DC voltage will be around 26.16 * 1.414 = 36.99V.
So, a 160VA; 24V+24V transformer will give around +/-37V rail voltages when the amplifier is not putting out much power. This is only around 5V below the rated maximum.
If now our local mains voltage is higher than implied by the nominal 230V (in europe) your DC voltage will be higher. In the UK the highest permissable Voltage is 240V+6% or 11% higher than 230V. In this case our 37V will also be 11% higher, so we now end up with 41V.
Hence, with the LM3875 the highest Transformer Voltage that should be used is 24/25V. Using a significantly higher voltage will put you into risk areas where the voltage sepcfications are concerned, due to the issues dioscussed above.
Sayonara
Kuei,
The 70VDC Vic was talking is about is from the note in the Operating Ratings section saying SPIKE might introduce distortion above 70VDC if proper heatsinking is not used.
Fedde,
How hot does your heatsinks get when you operate at those voltages?
The 70VDC Vic was talking is about is from the note in the Operating Ratings section saying SPIKE might introduce distortion above 70VDC if proper heatsinking is not used.
Fedde,
How hot does your heatsinks get when you operate at those voltages?
I felt so free to echo this post from the Amp discussion forum:
"My apologies to all who read and answered my previous posts.
I thought I would give the GC another chance and this morning started a listening session that ended about 7.00ish this evening.
By lunchtime I began to realise that the amp was performing better than my dear old Sony.
Highlights of the day have been, Sidney Bechet’s ‘Blue Horizon’ I could feel his huge chest cavity in the sound of that Clarinet. Julian Breams wonderfully expressive playing of ‘Recuerdos de la Alhambra’ I have never heard this so clearly and this tape is my ‘test’ piece. Janet Baker’s singing in Mahler 2, rapturous! Bart van Oort playing Chopin Nocturnes on an 1842 Pleyel pianoforte, this has always sounded on the edge of distortion, not so now, wonderful! Oooh and Martha Argerich with the Chopin Polonaise No6 I could feel her playing it, a piano tour de force that only a great amp and speakers can cope with.
By late afternoon the sound was sublime and after Mahler5, Simon Rattle, CBSO, what a perfect joy, I decided to give my all time No1 favourite a chance, Jacqueline Du Pre, the Elgar Cello Concerto, Barbarolli and the LSO, this usually brings tears to my eyes, they rolled down my cheeks as this truly wonderful performance unfolded before me, I have listened to this hundreds of times over the past 25 yrs or so, but never like this, revelatory!
Finally Murray Perahia playing the Bach Goldberg Variations, well 10 of them, I’ve only listened to these a few times and until now was unable to grasp their musicality, I heard the theme from the first in all of the others, and No6 which had been rather indistinct was clear and focused.
The amp is now, as far as I am concerned the best thing since sliced bread, it is sharp and clear with a dynamic range that easily surpasses my Sony, and that was good, honestly! Orchestral music through the small drive units I use has always sounded a bit confused, not so now. My Wife said that she thought Julian Bream was playing before her, she’s never said that before.
I will make another power supply with a lower voltage transformer, the thick alloy plate I used as heatsinks for the chips got very hot!
So well there we are, such is life, etc.
Again my apologies.
Jim Read"
"My apologies to all who read and answered my previous posts.
I thought I would give the GC another chance and this morning started a listening session that ended about 7.00ish this evening.
By lunchtime I began to realise that the amp was performing better than my dear old Sony.
Highlights of the day have been, Sidney Bechet’s ‘Blue Horizon’ I could feel his huge chest cavity in the sound of that Clarinet. Julian Breams wonderfully expressive playing of ‘Recuerdos de la Alhambra’ I have never heard this so clearly and this tape is my ‘test’ piece. Janet Baker’s singing in Mahler 2, rapturous! Bart van Oort playing Chopin Nocturnes on an 1842 Pleyel pianoforte, this has always sounded on the edge of distortion, not so now, wonderful! Oooh and Martha Argerich with the Chopin Polonaise No6 I could feel her playing it, a piano tour de force that only a great amp and speakers can cope with.
By late afternoon the sound was sublime and after Mahler5, Simon Rattle, CBSO, what a perfect joy, I decided to give my all time No1 favourite a chance, Jacqueline Du Pre, the Elgar Cello Concerto, Barbarolli and the LSO, this usually brings tears to my eyes, they rolled down my cheeks as this truly wonderful performance unfolded before me, I have listened to this hundreds of times over the past 25 yrs or so, but never like this, revelatory!
Finally Murray Perahia playing the Bach Goldberg Variations, well 10 of them, I’ve only listened to these a few times and until now was unable to grasp their musicality, I heard the theme from the first in all of the others, and No6 which had been rather indistinct was clear and focused.
The amp is now, as far as I am concerned the best thing since sliced bread, it is sharp and clear with a dynamic range that easily surpasses my Sony, and that was good, honestly! Orchestral music through the small drive units I use has always sounded a bit confused, not so now. My Wife said that she thought Julian Bream was playing before her, she’s never said that before.
I will make another power supply with a lower voltage transformer, the thick alloy plate I used as heatsinks for the chips got very hot!
So well there we are, such is life, etc.
Again my apologies.
Jim Read"
Philo said:
Well, it depends. Usually the heatsink remains cold. If I play at soft to 'normal' volumes I hardly can feel the opamp gets warm. If I kick it's tail it gets hot. And if I play ***LOUD*** (93 dB/W Tannoys) the chip gets maybe a little too warm. I now use silicon insulation, I will try mica with pasta soon...
(I feel that the thermic resistance is too high)
Fedde
Jim Read said:
.. please try to compare your 1055 with this chip amp: http://www.diyaudio.com/forums/showthread.php?postid=112829#post112829 ..
regards:
ragil.hastomo
Sudden about turn
Hi Jim,
Just so us newbies can understand.......
What is it exactly that you changed in the circuit that had such a dramatic effect changing it from laughably crappy to having visions of Julian Bream infront of your eyes. (given my taste in music I honestly could't tell you which I thought was worst, but horses for courses)
This is as I understand it also still using the overspec power supply as well?
Cheerrs
Paul
Hi Jim,
Just so us newbies can understand.......
What is it exactly that you changed in the circuit that had such a dramatic effect changing it from laughably crappy to having visions of Julian Bream infront of your eyes. (given my taste in music I honestly could't tell you which I thought was worst, but horses for courses)
This is as I understand it also still using the overspec power supply as well?
Cheerrs
Paul
- Status
- Not open for further replies.