729 boards
Jim,
Had a look in a 82 and copied the layout schematic for the 729 section. Andys totally right; the 82 implementation of the time align buffer filter is almost identical to the one he posted with just a couple of component value changes.
Seems to me the guts of an 82 is really little different to a 72 (as far as the gain and time align filter 'boards' go)- tape buffer (I assume it is such) has puzzled me but haven't investigated deeply) just with much better layout. Most of the additional circuitry appears to revolve around remote control.
My impression is that the difference in sound between 72 and 82 boils down to putting everything on 1 motherboard (avoids those crappy clip in boards) allowing better star earthing, implementing star Positive rail, better actual 'board' layout, and tweaking of the actual component values of the 'boards'. Higher tolerance Component selection could be a factor but then only Naim would know that.
Any comments anyone?
PS- will post the 82 board layouts later.
Ced
Jim,
Had a look in a 82 and copied the layout schematic for the 729 section. Andys totally right; the 82 implementation of the time align buffer filter is almost identical to the one he posted with just a couple of component value changes.
Seems to me the guts of an 82 is really little different to a 72 (as far as the gain and time align filter 'boards' go)- tape buffer (I assume it is such) has puzzled me but haven't investigated deeply) just with much better layout. Most of the additional circuitry appears to revolve around remote control.
My impression is that the difference in sound between 72 and 82 boils down to putting everything on 1 motherboard (avoids those crappy clip in boards) allowing better star earthing, implementing star Positive rail, better actual 'board' layout, and tweaking of the actual component values of the 'boards'. Higher tolerance Component selection could be a factor but then only Naim would know that.
Any comments anyone?
PS- will post the 82 board layouts later.
Ced
Ced
Great post, can I confirm that the 82 does use star wiring for the +ve rail?
Any internal regualtion?
I've long puzzled that the lower-end pre's have loads of shared +ve current paths - my engineer hat says that's crap and shouldn't be done, but my Naim faithful hat tells me they usually know what they are doing.
I'm strongly suspecting that the improvements through the range, in many areas, are just simple, one-off engineering improvments, that cost so little, it's shocking. Difficult to criticise as a business model, and hierarchies have to be maintained, I suppose.
I've not had time to investigate fully though.
Can't wait for the layouts - get snapping!
A.
Great post, can I confirm that the 82 does use star wiring for the +ve rail?
Any internal regualtion?
I've long puzzled that the lower-end pre's have loads of shared +ve current paths - my engineer hat says that's crap and shouldn't be done, but my Naim faithful hat tells me they usually know what they are doing.
I'm strongly suspecting that the improvements through the range, in many areas, are just simple, one-off engineering improvments, that cost so little, it's shocking. Difficult to criticise as a business model, and hierarchies have to be maintained, I suppose.
I've not had time to investigate fully though.
Can't wait for the layouts - get snapping!
A.
82
Andy,
can indeed confirm, +ve rails are star earthed (hardwired) from the din sockets to each part of the motherboard. With the 2 hicap arrangement, 1 hicap feeds the gain 'boards', other hicap rest of 'boards' (by 'boards' I mean the part of the motherboard that would be a pull out board in a 72)
Makes you wonder what could be achieved with a little rewiring of a 32.5 or 72.
No internal regulation other than 2 LM317Ts which are for the digital control sections only. Don't affect the audio circuits.
schematics follow shortly.
ced
Andy,
can indeed confirm, +ve rails are star earthed (hardwired) from the din sockets to each part of the motherboard. With the 2 hicap arrangement, 1 hicap feeds the gain 'boards', other hicap rest of 'boards' (by 'boards' I mean the part of the motherboard that would be a pull out board in a 72)
Makes you wonder what could be achieved with a little rewiring of a 32.5 or 72.
No internal regulation other than 2 LM317Ts which are for the digital control sections only. Don't affect the audio circuits.
schematics follow shortly.
ced
schematics
are here.
They main differences I've noticed are a)a little 5.6pf ceramic cap bypassing one of the 384 transistors on one of the boards, some of the polystyrene caps are down to 100pf from 470 etc. Some of the resistors have changed value. Note the 47uf input filter caps (SLCE coi series if I remember) seem to be important as there's some minor +ve rail star tracing from them.
On the mother board in the 82 the gain and filter boards are laid out 2 each side of the main star earth so its really only a couple of inches to the earth point for each board.
The board layouts are accurate to the 82 ones though haven't double checked the gain board yet, and there was one bizzare input or output on it that I could not for the life of me understand the reason for; but will post it in a couple of days and also if there's any amends.
Cedric
are here.
They main differences I've noticed are a)a little 5.6pf ceramic cap bypassing one of the 384 transistors on one of the boards, some of the polystyrene caps are down to 100pf from 470 etc. Some of the resistors have changed value. Note the 47uf input filter caps (SLCE coi series if I remember) seem to be important as there's some minor +ve rail star tracing from them.
On the mother board in the 82 the gain and filter boards are laid out 2 each side of the main star earth so its really only a couple of inches to the earth point for each board.
The board layouts are accurate to the 82 ones though haven't double checked the gain board yet, and there was one bizzare input or output on it that I could not for the life of me understand the reason for; but will post it in a couple of days and also if there's any amends.
Cedric
The Wizard of Oz revealled
ALW wrote: "...just simple, one-off engineering improvments, that cost so little, it's shocking. Difficult to criticise as a business model"
Yep. JV was a businessman alright.
Still, gives the rest of us hope as the mystique is blown away and engineering prevails.
BAM
(ps: don't mention this in the Naim forum...
)
ALW wrote: "...just simple, one-off engineering improvments, that cost so little, it's shocking. Difficult to criticise as a business model"
Yep. JV was a businessman alright.
Still, gives the rest of us hope as the mystique is blown away and engineering prevails.
BAM
(ps: don't mention this in the Naim forum...
)Bam
"(ps: don't mention this in the Naim forum... )"
I actually thought last night that posting the 729 schematic had got me banished from there - my posting rights were removed.
All seems to be right today though, so maybe just an admin thing?
While I have your attention, my thoughts are turning to amplifier design at present as my next challenge, and your obvious vastly greater knowledge and experience in this area may be able to help.
Can you recommend some good books on discrete amplifier design, that don't look like someone's phD thesis?
I once bought a book called feedback amplifier principles - I don't think I got much beyond the first few pages
In fact, would you like it? I have a feeling it may be of much more use to you, than it ever will to me (it was a SAMS publication that I picked up cheap at a radio rally).
A.
"(ps: don't mention this in the Naim forum... )"
I actually thought last night that posting the 729 schematic had got me banished from there - my posting rights were removed.
All seems to be right today though, so maybe just an admin thing?
While I have your attention, my thoughts are turning to amplifier design at present as my next challenge, and your obvious vastly greater knowledge and experience in this area may be able to help.
Can you recommend some good books on discrete amplifier design, that don't look like someone's phD thesis?
I once bought a book called feedback amplifier principles - I don't think I got much beyond the first few pages
In fact, would you like it? I have a feeling it may be of much more use to you, than it ever will to me (it was a SAMS publication that I picked up cheap at a radio rally).
A.
Hi Andy,
I've got a book at home called 'analog circuit techniques' or something like that - it's quite good, I'll look up the correct title and author tonight.
Another good place to look is op-amp design notes. Please bear in mind though that some of circuits used are there specifically to address problems of manufacturing on silicon - namely transistors are easier/cheaper to make in a silicon design than resistors or capacitors. Hence you will find current mirrors and sources just about everywhere in op-amps. They are not always the best way to go when trying to build a really good audio amp.
I think the naim approach is fundamentally a good one: namely(sic) aim for a simple design, and spend your time optimising it.
Einstein quote "things should be made as simple as possible, but no simpler". Probably not word for word, but you get the idea.
Luck,
Jim.
I've got a book at home called 'analog circuit techniques' or something like that - it's quite good, I'll look up the correct title and author tonight.
Another good place to look is op-amp design notes. Please bear in mind though that some of circuits used are there specifically to address problems of manufacturing on silicon - namely transistors are easier/cheaper to make in a silicon design than resistors or capacitors. Hence you will find current mirrors and sources just about everywhere in op-amps. They are not always the best way to go when trying to build a really good audio amp.
I think the naim approach is fundamentally a good one: namely(sic) aim for a simple design, and spend your time optimising it.
Einstein quote "things should be made as simple as possible, but no simpler". Probably not word for word, but you get the idea.
Luck,
Jim.
Another question for traderbam / others
Bam,
I've been pondering the 'time-aligned' thing, and the circuit above does have zero GD, but surely constant GD amounts to the same thing, unless I'm missing something.
Since the goal is to retain the timing integrity of the full bandwidth we are trying to pass, surely it matters not whether everything emerges from the filter after several hundred ms, or almost immediately?
Maybe it amounts to the same design challenge, hence the choice of zero GD?
CED the component changes you mention are as a direct result of the filter addition, in order to ensure the preamp circuitry does not impact GD / BW, leaving the filter as the dominant factor.
Andy.
Bam,
I've been pondering the 'time-aligned' thing, and the circuit above does have zero GD, but surely constant GD amounts to the same thing, unless I'm missing something.
Since the goal is to retain the timing integrity of the full bandwidth we are trying to pass, surely it matters not whether everything emerges from the filter after several hundred ms, or almost immediately?
Maybe it amounts to the same design challenge, hence the choice of zero GD?
CED the component changes you mention are as a direct result of the filter addition, in order to ensure the preamp circuitry does not impact GD / BW, leaving the filter as the dominant factor.
Andy.
Andy,
as I understand it, group delay refers to the rate of change of the phase delay with respect to the frequency of the signal.
The total phase delay of the signal through the filter has no bearing on the group delay.
Zero group delay is good. Total time delay is pretty irrelevant.
So, you are correct but have confused group delay with time delay.
Note, I've seen lots of filter modeling software that does this as well. Confused the hell out of me for a while as well.
Jim.
as I understand it, group delay refers to the rate of change of the phase delay with respect to the frequency of the signal.
The total phase delay of the signal through the filter has no bearing on the group delay.
Zero group delay is good. Total time delay is pretty irrelevant.
So, you are correct but have confused group delay with time delay.
Note, I've seen lots of filter modeling software that does this as well. Confused the hell out of me for a while as well.
Jim.
Andy,
"...your obvious vastly greater knowledge and experience in this area may be able to help."
ok
"Can you recommend some good books on discrete amplifier design, that don't look like someone's phD thesis?"
No. Nor web sites nor "audio professors". In fact, especially not audio professors.
"I once bought a book called feedback amplifier principles...would you like it?"
Thanks, I'll email you my address and check it out.
Speaking of insominia, I saw the film of this name last night at the cinema. I thought it was really good. A detective drama set in Alaska starring Al Pacino and Robin Williams and some extremely gorgeous woman who's name escapes me.
BAM
"...your obvious vastly greater knowledge and experience in this area may be able to help."
ok"Can you recommend some good books on discrete amplifier design, that don't look like someone's phD thesis?"
No. Nor web sites nor "audio professors". In fact, especially not audio professors.
"I once bought a book called feedback amplifier principles...would you like it?"
Thanks, I'll email you my address and check it out.Speaking of insominia, I saw the film of this name last night at the cinema. I thought it was really good. A detective drama set in Alaska starring Al Pacino and Robin Williams and some extremely gorgeous woman who's name escapes me.
BAM
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.