Hi !
i have a curiosity.
I see many stereo units where the lay-out of the parts is mirrored (sorry for my English) and one channel is the mirror image of the other.
I attach here under an example ... look at the output pcbs on the bottom ...
Instead the two power supplies on top look identical ...
Why not instead build one channel exactly the same of the other ?
Is a mirrored layout better ? 🙄
I thought that the design starts from optimizing/fine-tune one channel and then make a copy of it.
Am i missing something ?
Thanks and regards, gino
i have a curiosity.
I see many stereo units where the lay-out of the parts is mirrored (sorry for my English) and one channel is the mirror image of the other.
I attach here under an example ... look at the output pcbs on the bottom ...

Instead the two power supplies on top look identical ...
Why not instead build one channel exactly the same of the other ?
Is a mirrored layout better ? 🙄
I thought that the design starts from optimizing/fine-tune one channel and then make a copy of it.
Am i missing something ?
Thanks and regards, gino
Last edited:
Marketing?
Given that components with more than two legs (e.g. chips, transistors) are not normally available in mirror-pairs the 'mirrored' PCB is just for human consumption - the actual PCB design cannot be a mirror image of the other channel. Hence any claimed advantage is lost - not that there could be any sense in the claims anyway.
I suspect it is people transferring to amplifiers what they have learnt from speaker design (or hoping their customers will make the same mistake). In speakers mirroring make some sense. For amps you want identical channels, not mirrors.
Given that components with more than two legs (e.g. chips, transistors) are not normally available in mirror-pairs the 'mirrored' PCB is just for human consumption - the actual PCB design cannot be a mirror image of the other channel. Hence any claimed advantage is lost - not that there could be any sense in the claims anyway.
I suspect it is people transferring to amplifiers what they have learnt from speaker design (or hoping their customers will make the same mistake). In speakers mirroring make some sense. For amps you want identical channels, not mirrors.
Hi,
It is often done for the logistics of the connections or power supplies,
e.g. the Pioneer A300X has "mirrored" power amplifiers on the same PCB.
rgds, sreten.
It is often done for the logistics of the connections or power supplies,
e.g. the Pioneer A300X has "mirrored" power amplifiers on the same PCB.
rgds, sreten.
Marketing?
Given that components with more than two legs (e.g. chips, transistors) are not normally available in mirror-pairs the 'mirrored' PCB is just for human consumption - the actual PCB design cannot be a mirror image of the other channel. Hence any claimed advantage is lost - not that there could be any sense in the claims anyway.
I suspect it is people transferring to amplifiers what they have learnt from speaker design (or hoping their customers will make the same mistake). In speakers mirroring make some sense. For amps you want identical channels, not mirrors.
Hi and thanks and this was my thinking also. Two identical channel as much as possible.
Regarding the "mirroring" of layouts i think this could be justified, maybe, when the power supply is just one and centrally located, to keep distances from similar components the same.
I was looking at layouts of digital units because i read how important it is for this components.
I prefer channels completely identical
Thanks again, gino
Hi,
It is often done for the logistics o the connections or power supplies, e.g. the Pioneer A300X has "mirrored" power amplifiers on the same PCB.
rgds, sreten.
Hi and thanks for the valuable advice and i was thinking the same.
When the ps is single and central and they want to keep the same distances from ps to components. But i like better the dual mono solution for sure
Thanks again, gino
Last edited:
Cannot understand why they would do it for such a design there are no benefits in this case. For simple designs such as power amp outputs with heat sinks involved and power distribution it can make sense, but in this case it is a pointless and silly exercise.... Quite often stuff like this only appears in Audio gear...
What is the unit?
What is the unit?
It gives a symmetrical layout for the connectors on the back panel if the boards are mirrored, maybe their chassis designers (or customers) hate a lack of symmetry?
John
John
Last edited:
I had deleted my post and now its reappeared as I had another look and realised its to minimise the interboard connections so ignore the ramblings. It is not a true mirror though, the circuit blocks are laid out one way and rotated to create the mirroring, so the blocks are mirrored.
It does mean more inventory, two boards to manufacture, assemble and test... What is the make of the unit.
It does mean more inventory, two boards to manufacture, assemble and test... What is the make of the unit.
Hi,
In the case of the "mirrored" Pioneer PCB, low level circuits are
next to each other and there is a central vertical shield to cut
crosstalk. If they were not "mirrored" one channels low level
circuitry would be near the others high current rails whilst
the other channels low level circuitry would be miles away
from the other channels high level rails, inevitably the
crosstalk performance would be asymmetric.
There is usually a method to any "madness" like in the
example in post #1. A symmetrical back panel and the
ribbon cables are not passing over active circuitry.
If you look carefully at that example the circuits are
not really mirrored, just rearranged, e.g compare
left top right to right top left.
rgds, sreten.
In the case of the "mirrored" Pioneer PCB, low level circuits are
next to each other and there is a central vertical shield to cut
crosstalk. If they were not "mirrored" one channels low level
circuitry would be near the others high current rails whilst
the other channels low level circuitry would be miles away
from the other channels high level rails, inevitably the
crosstalk performance would be asymmetric.
There is usually a method to any "madness" like in the
example in post #1. A symmetrical back panel and the
ribbon cables are not passing over active circuitry.
If you look carefully at that example the circuits are
not really mirrored, just rearranged, e.g compare
left top right to right top left.
rgds, sreten.
Cannot understand why they would do it for such a design there are no benefits in this case.
For simple designs such as power amp outputs with heat sinks involved and power distribution it can make sense, but in this case it is a pointless and silly exercise....
Quite often stuff like this only appears in Audio gear...
What is the unit?
Hi ! it is a dac, the Bricasti M1. But i would like to stress that this is mainly a curiosity of mine.
Maybe there is a very good reason to change the lay-out of one channel in respect to the other. This is a very well reviewed dac.
Still i see that instead the PS look identical.
I am very curious of digital units because it is now my only source completely. I read about how delicate is the lay-out design phase of a dac, where even a very small placement of traces and parts can give benefits or problems. So i was thinking that once found a very good lay-out it was almost natural to stick with it and adopt it also for the other channel.
I also think that analog circuits are much more forgiving.
Digital is very very challenging. It is difficult to make good sounding digital.
Thanks a lot again, gino
Hi, In the case of the "mirrored" Pioneer PCB, low level circuits are next to each other and there is a central vertical shield to cut crosstalk. If they were not "mirrored" one channels low level circuitry would be near the others high current rails whilst the other channels low level circuitry would be miles away from the other channels high level rails, inevitably the crosstalk performance would be asymmetric.
Hi and thanks a lot for the very interesting explanation
as usual i was trivializing the all issue. So dual mono is not always the right design choice.
There is usually a method to any "madness" like in the example in post #1. A symmetrical back panel and the ribbon cables are not passing over active circuitry. If you look carefully at that example the circuits are not really mirrored, just rearranged, e.g compare left top right to right top left.
rgds, sreten.
you are right. I have noticed that flat cable also myself.
But i wonder if there is a way to have the same identical circuit.
I have one here for instance. The two channels look identical ...
An externally hosted image should be here but it was not working when we last tested it.
so with a different cables routing i think it could be done
Or maybe isolating them in some ways.
Thanks again, gino
Last edited:
Hi ! it is a dac, the Bricasti M1. But i would like to stress that this is mainly a curiosity of mine.
Maybe there is a very good reason to change the lay-out of one channel in respect to the other. This is a very well reviewed dac.
Still i see that instead the PS look identical.
I am very curious of digital units because it is now my only source completely. I read about how delicate is the lay-out design phase of a dac, where even a very small placement of traces and parts can give benefits or problems. So i was thinking that once found a very good lay-out it was almost natural to stick with it and adopt it also for the other channel.
I also think that analog circuits are much more forgiving.
Digital is very very challenging. It is difficult to make good sounding digital.
Thanks a lot again, gino
I did retract my view in my second post.
I will beg to differ, analogue is hard to get right digital is pretty easy and forgiving most of the time (except high speed) especially layout. I also think digital sounds great, no wow and flutter, hiss, etc.🙂
I did retract my view in my second post.
I will beg to differ, analogue is hard to get right digital is pretty easy and forgiving most of the time (except high speed) especially layout...
Hi and thanks again for the valuable advice. I trust your words. Mine was only a feeling. It is just that digital seems to me more sensitive to noise in the mains, to RFI and EMI, even maybe to vibrations ?
it is very touchy ...
I also think digital sounds great, no wow and flutter, hiss, etc.🙂
I think that digital can sound great but many times it does not. There is often something giving a sort of high freq buzz to sound. Not completely clean and spacious ... but good digital can be very satisfying, and also handy.
Thanks again for the very helpful advice, gino
It is analogue that is affected by noise etc. more than digital, digital has a certain immunity due to the switching levels, whereas any noise added to an analogue signal becomes part of that signal....
Vibration in solid state electronics is not something I would worry about much in a normal domestic situation, valves and record decks yes. That is not to say I ignore it, but I don't go to extremes, just a decent stand and avoid putting stuff directly in front of my speakers🙂
Vibration in solid state electronics is not something I would worry about much in a normal domestic situation, valves and record decks yes. That is not to say I ignore it, but I don't go to extremes, just a decent stand and avoid putting stuff directly in front of my speakers🙂
It is analogue that is affected by noise etc. more than digital, digital has a certain immunity due to the switching levels, whereas any noise added to an analogue signal becomes part of that signal....
Thanks a lot again and this is really a surprise for me. I thought the contrary. That circuits like clock generators and dacs were extremely sensitive to the noise in the supplies that feed them.
I was clearly wrong. But this is also a good news. It means that digital is more robust to noise.
Vibration in solid state electronics is not something I would worry about much in a normal domestic situation, valves and record decks yes.
That is not to say I ignore it, but I don't go to extremes, just a decent stand and avoid putting stuff directly in front of my speakers🙂
This is what i also thought. And with valves and records it is also easy to test ... a little tapping on the parts often produces quite a noise from the speakers.
But as i said i am completely digital and solid state first because i am lazy second because i am lazy again. I think that both tubes and records need dedication, attention, care ... not my case at all.
I like streaming from a nas because it is very handy.
If one day it will sound also decent i will be done completely.
Thanks a lot again for the very valuable advice.
Kind regards, gino
Clocks do benefit from a clean supply and because a clock is the heartbeat of a system then you filter its supply accordingly. Where you have data conversion, communications etc. where the clock and its noise related artefacts (jitter etc.) are critical you filter or isolate the supply from the noise accordingly.
I haven't time at the moment to go into fine detail on power distribution for mixed signal systems (digital and analogue) with faster rise times, base clock operating frequencies and lower operating voltages all adding to the fun. More complex boards these days have multiple supplies, some having over 20 (still only one 0V reference (GND)).
I have this add on as part of the CAD software I use for layout:
http://www.algozen.com/DS_CADSTAR_LT_PowerIntegrityAdvanced_ENG_2011_10_05.pdf
All the other major PCB design packages offer a similar tool (though this stuff tends to cost a bit).
I haven't time at the moment to go into fine detail on power distribution for mixed signal systems (digital and analogue) with faster rise times, base clock operating frequencies and lower operating voltages all adding to the fun. More complex boards these days have multiple supplies, some having over 20 (still only one 0V reference (GND)).
I have this add on as part of the CAD software I use for layout:
http://www.algozen.com/DS_CADSTAR_LT_PowerIntegrityAdvanced_ENG_2011_10_05.pdf
All the other major PCB design packages offer a similar tool (though this stuff tends to cost a bit).
Thanks a lot again for the interesting information.
The design is a very complex task i understand. It is just for this that i was thinking about copying a good working lay-out because i understand the challenges behind its design.
So when i find something that works very good for me it would be just normal make a copy for the other channel, when this is possible of course.
For instance even if a dac is stereo why not use two ?
The picture i have attached above is what i have in mind. With one channel the exact clone of the other.
For instance the channels can be arranged as cards to inserted on a bus.
Like identical modules terminated with pins for the connections.
Thanks again, gino
The design is a very complex task i understand. It is just for this that i was thinking about copying a good working lay-out because i understand the challenges behind its design.
So when i find something that works very good for me it would be just normal make a copy for the other channel, when this is possible of course.
For instance even if a dac is stereo why not use two ?
The picture i have attached above is what i have in mind. With one channel the exact clone of the other.
For instance the channels can be arranged as cards to inserted on a bus.
Like identical modules terminated with pins for the connections.
Thanks again, gino
Last edited:
There is an awful lot of wasted space on the board above🙂 all those discrete's could be scrunched up and have much smaller loop areas. Pretty layout though and neat🙂
I am not a fan of having a bus based system, I would get everything in the smallest possible area for the digital and the analogue, no point spreading it out if you don't have to, small is best.
I am not a fan of having a bus based system, I would get everything in the smallest possible area for the digital and the analogue, no point spreading it out if you don't have to, small is best.
There is an awful lot of wasted space on the board above🙂 all those discrete's could be scrunched up and have much smaller loop areas. Pretty layout though and neat🙂
I am not a fan of having a bus based system, I would get everything in the smallest possible area for the digital and the analogue, no point spreading it out if you don't have to, small is best.
Hi and thanks again for the kind advice.
I am also prefer small circuits for digital and low signals amplifiers. 🙄
Even if smd have complicated a lot life for DIYers 🙁
Regards, gino
- Status
- Not open for further replies.
- Home
- Member Areas
- The Lounge
- What is the problem with dual-mono layouts ?