Banned
Joined 2002
Hi Jonathan,
really interesting to read that you do NOT use heavy "lead" mats nor rubber to dampen the assemblies. I have been using it for many years, and I think my ideas came from the time when I built loudspeakers in the 80s. Add mass to lower the resonance freq and also to lower the amplitude with which any assemblies might vibrate with.
The "lead"mats that I use are actually not made out of lead, but of some rubber compound material, so there is no problems with ROHS or my personal health. I am fully årp ROHS even though it has caused some problems for me, aswell increased costs due to the need to invest in new soldering eq. and skills.
I also use Teflon PCBs for low voltage applications like RIAA (both MM and MC) to minimise influnece from vibrations onto the sensitive circuits.
I think I have gone to the extreme in my latest assembly of a remote controlled attenuator system where I dampen the PCBs by putting them into clay and also to add clay onto the top of the PCBs. At least it dampens the sharp noise from the clicking relais
Now, your ideas to control resonance is really neat and intelligent! I guess one cannot use clay inside a consumer product....
Maybe I am going in the wrong direction when just adding mass to a mechanical system. What I think I do is to lower the resonance freq by adding mass, but maybe I am moving the resonance freq to a freq range where there is lots of energy in the music (ie 50 - 200 Hz). Instead I should add stiffness to the mechanical system so the resonace freq goes up towards a freq range ehere there is much less energy in the music (above 1kHz or so). Less energy that can trigger a resonance disturbance in the assembly.
And this is probably what you do in the 4-2 preamp.
Best regards,
Sigurd
really interesting to read that you do NOT use heavy "lead" mats nor rubber to dampen the assemblies. I have been using it for many years, and I think my ideas came from the time when I built loudspeakers in the 80s. Add mass to lower the resonance freq and also to lower the amplitude with which any assemblies might vibrate with.
The "lead"mats that I use are actually not made out of lead, but of some rubber compound material, so there is no problems with ROHS or my personal health. I am fully årp ROHS even though it has caused some problems for me, aswell increased costs due to the need to invest in new soldering eq. and skills.
I also use Teflon PCBs for low voltage applications like RIAA (both MM and MC) to minimise influnece from vibrations onto the sensitive circuits.
I think I have gone to the extreme in my latest assembly of a remote controlled attenuator system where I dampen the PCBs by putting them into clay and also to add clay onto the top of the PCBs. At least it dampens the sharp noise from the clicking relais
Now, your ideas to control resonance is really neat and intelligent! I guess one cannot use clay inside a consumer product....
Maybe I am going in the wrong direction when just adding mass to a mechanical system. What I think I do is to lower the resonance freq by adding mass, but maybe I am moving the resonance freq to a freq range where there is lots of energy in the music (ie 50 - 200 Hz). Instead I should add stiffness to the mechanical system so the resonace freq goes up towards a freq range ehere there is much less energy in the music (above 1kHz or so). Less energy that can trigger a resonance disturbance in the assembly.
And this is probably what you do in the 4-2 preamp.
Best regards,
Sigurd
jcarr said:
Peter,
good question!
I have an idea and that is to use copper tracks and ontop of them plate with silver and then plate with gold. The gold will both protect the silver and maybe add some warmth to the sound. The silver will be there for its sound qualities. The copper I guess have to be there as one has to have something to plate onto
Best reagrds,
Sigurd
good question!
I have an idea and that is to use copper tracks and ontop of them plate with silver and then plate with gold. The gold will both protect the silver and maybe add some warmth to the sound. The silver will be there for its sound qualities. The copper I guess have to be there as one has to have something to plate onto
Best reagrds,
Sigurd
Peter Daniel said:
Hi,
Yes, it is a good question.
I have always thought that solder masks will quite likely have some small affect on the sound, but have not (yet) tried a direct comparison here, with no other changes.
Whether solder mask is 'audible' or not, it will act as a dielectric with all of the traces being covered with this material.
I did try direct comparisons between conventional glass fibre boards and teflon in a MC head-amp, and the teflon board was 'cleaner' sounding, presumably due to the better dielectic.
Regards,
Yes, it is a good question.
I have always thought that solder masks will quite likely have some small affect on the sound, but have not (yet) tried a direct comparison here, with no other changes.
Whether solder mask is 'audible' or not, it will act as a dielectric with all of the traces being covered with this material.
I did try direct comparisons between conventional glass fibre boards and teflon in a MC head-amp, and the teflon board was 'cleaner' sounding, presumably due to the better dielectic.
Regards,
rdf said:
I think so
. Being able to get the circuit off the board minimizes parasitic capacitances (and leakages) on critical nodes; being able to work point-to-point enables inductances to be reduced, and the circuit is more intrinsically stable as a result. The stacked boards of the 4-2SE are fully ground and power-planed, and provide a fair amount of shielding on their own (especially when RF-sensitive nodes are located between the boards). And it helps if the circuitry is designed with at least some foresight as to its behaviour when forced to operate in an RF-prone environment.The RF sensitivity of a circuit is also related to its effectiveness as an antenna - a physically smaller circuit (and more compact wiring harnesses) means a less sensitive antenna means less pickup of environmental RF.
The 4-2SE is a wide-band circuit in a chassis that is pretty much unshielded, but I have not noticed (nor heard of) any real problems pertaining to RF. In fact, I've had situations at audio shows where RF was leaking in from both the hotel walls and the local AM stations, and other preamps manifested RF issues while the 4-2 didn't. OTOH, I don't run 100dB+ speakers in an RF-rich environment - if I did, perhaps my chassis designs, circuit designs (or both) would be different.
However, I've already done an excavated-from-solid chassis in an earlier product - the 3.0, and I can't say that the 4-2SE, open chassis and all, appears to be any inferior in terms of RF breakthrough.
regards, jonathan carr
That is minor
Looking out from my front yard, I can see 7 televsion towers, at most 10 miles for the furthest.
This is why I line most aluminum chassis with mumetal. Try to shield as well as possible.
The tube gear with high impedances seems to be okay. The solid stage with "virtual grounds" gets eaten alive. The solid state gear with low impdeances works best.
This is with 100 dB efficient speakers.
Guess I need to blow either some of the kids college money or the old ladies next car money and give the 4-2L a listen. Looksalone justify the splurge.
Where is the closest dealer with stock?
George
Looking out from my front yard, I can see 7 televsion towers, at most 10 miles for the furthest.
This is why I line most aluminum chassis with mumetal. Try to shield as well as possible.
The tube gear with high impedances seems to be okay. The solid stage with "virtual grounds" gets eaten alive. The solid state gear with low impdeances works best.
This is with 100 dB efficient speakers.
Guess I need to blow either some of the kids college money or the old ladies next car money and give the 4-2L a listen. Looksalone justify the splurge.
Where is the closest dealer with stock?
George
starground
Dear all,
At first i hope this beautiful thread will keep alive!
I have a lot of questions about this masterpiece and somewhere in the thread there is a link to the hifiplus magazine no.43.
I try to order this backissue but that website is realy horrible.
Has anyone a scan of that article?
When i look at the "sternmarsch picture" of stereoplay i see maybe 2 or 4 groundplanes (ground left right,power/signal???)
and a jumper. I suppose that "brummschleifen" are groudloops but what connect this jumper?
When you look at the powersuply picture on the lyraconnoisseur.com webside there are 8 huge bypass caps.
Has anyone any idee what type of (foil?) caps these are?
Gr.
Johan.
Dear all,
At first i hope this beautiful thread will keep alive!
I have a lot of questions about this masterpiece and somewhere in the thread there is a link to the hifiplus magazine no.43.
I try to order this backissue but that website is realy horrible.
Has anyone a scan of that article?
When i look at the "sternmarsch picture" of stereoplay i see maybe 2 or 4 groundplanes (ground left right,power/signal???)
and a jumper. I suppose that "brummschleifen" are groudloops but what connect this jumper?
When you look at the powersuply picture on the lyraconnoisseur.com webside there are 8 huge bypass caps.
Has anyone any idee what type of (foil?) caps these are?
Gr.
Johan.
Hi Johan,
In answer to your last question, Jonathan confirmed a few years ago (when I asked him on this Forum) that those large caps were specially made polypropylenes.
You will probably find his reply in one of the Blackgate capacitor threads from about 3 yrs ago, if you are interested.
IHTH.
Regards,
In answer to your last question, Jonathan confirmed a few years ago (when I asked him on this Forum) that those large caps were specially made polypropylenes.
You will probably find his reply in one of the Blackgate capacitor threads from about 3 yrs ago, if you are interested.
IHTH.
Regards,
Asking about different board materials and finishes without mentioning the context of componentry and schematic complexity is a bit of a nonsense question, because properly the board material and finish should be a fundamental part of the schematic and layout specification. If you're given eggs, make an omelet, if you have lemons, make lemonade. IOW, if you want to use a teflon no-mask 2-layer board, you really should design a different circuit from what you'd want on FR-4 multilayer.
Teflon/duroid is quite good for simple circuits, but due to its weak mechanical and thermal strength, I wouldn't recommend it for complex circuits, particularly when rework is a possibility. FR-4 is at the opposite end of the scale - electrically not nearly as good as teflon, but due to its mechanical and thermal properties, quite versatile and well-suited for tight and dense multilayer circuits. Arlon I haven't used, so no comment. A good polyester PCB is midway, electrically better than FR-4 but not quite as stellar as teflon; considerably stronger mechanically than teflon, but certainly not dependable enough to attempt some of the trickier things you can do with FR-4.
Most unprotected PCB metal finishes oxidize over time, so if you don't do a mask, even if a circuit layout sounds good new, how will it sound a year down the road if it's been used in a hot, humid and atmospherically polluted environment? Mind you also that certain fluxes can cause leakage and consequently circuitry malfunction issues over time if the residues aren't washed off completely. Back to plating, I've tried gold, but have been less than impressed with the cost/performance - at least from the board houses that I've dealt with. Silver plating seems to sound OK when new, but the word is that it does change after some months, so I want to do longer-term testing before I decide whether it is worth using in production.
In practice, a mask allows a denser, tighter circuit layout than no mask, so for the Lyra-Connoisseurs I prefer a mask. But for simpler, slower circuits that don't benefit so much from a tight layout, I would certainly consider silver and no mask.
regards, jonathan carr
Teflon/duroid is quite good for simple circuits, but due to its weak mechanical and thermal strength, I wouldn't recommend it for complex circuits, particularly when rework is a possibility. FR-4 is at the opposite end of the scale - electrically not nearly as good as teflon, but due to its mechanical and thermal properties, quite versatile and well-suited for tight and dense multilayer circuits. Arlon I haven't used, so no comment. A good polyester PCB is midway, electrically better than FR-4 but not quite as stellar as teflon; considerably stronger mechanically than teflon, but certainly not dependable enough to attempt some of the trickier things you can do with FR-4.
Most unprotected PCB metal finishes oxidize over time, so if you don't do a mask, even if a circuit layout sounds good new, how will it sound a year down the road if it's been used in a hot, humid and atmospherically polluted environment? Mind you also that certain fluxes can cause leakage and consequently circuitry malfunction issues over time if the residues aren't washed off completely. Back to plating, I've tried gold, but have been less than impressed with the cost/performance - at least from the board houses that I've dealt with. Silver plating seems to sound OK when new, but the word is that it does change after some months, so I want to do longer-term testing before I decide whether it is worth using in production.
In practice, a mask allows a denser, tighter circuit layout than no mask, so for the Lyra-Connoisseurs I prefer a mask. But for simpler, slower circuits that don't benefit so much from a tight layout, I would certainly consider silver and no mask.
regards, jonathan carr
Re: starground
Dear Joho:
Since I don't understand German, it would help to look at a picture. Does this photo show what you are talking about? (much appreciation to Sigurd)
I think you must be referring to the power supply photo in the 5-0 Line section?
As Bobken mentioned, these are polypropylene 190uF caps. They were originally made for the older 3.0 preamp design (sorry, can't seem to locate any photos to show). Since the 3.0 ran at +/- 144V rails, we just used a bunch of them as is (IOW, they were the primary capacitors rather than bypasses), and dealt with the heat (which there was a lot of) via the main chassis design (which was a solid piece milled from a single billet of aluminum).
In the context of the 5-0, which runs at lower voltages, the polyprolyenes are indeed used as add-ons to the primary capacitors (10,000uF electros x 4), and operating temperatures are much more within reason.
regards, jonathan carr
Dear Joho:
joho said:
Since I don't understand German, it would help to look at a picture. Does this photo show what you are talking about? (much appreciation to Sigurd
)An externally hosted image should be here but it was not working when we last tested it.
I think you must be referring to the power supply photo in the 5-0 Line section?
An externally hosted image should be here but it was not working when we last tested it.
As Bobken mentioned, these are polypropylene 190uF caps. They were originally made for the older 3.0 preamp design (sorry, can't seem to locate any photos to show). Since the 3.0 ran at +/- 144V rails, we just used a bunch of them as is (IOW, they were the primary capacitors rather than bypasses), and dealt with the heat (which there was a lot of) via the main chassis design (which was a solid piece milled from a single billet of aluminum).
In the context of the 5-0, which runs at lower voltages, the polyprolyenes are indeed used as add-ons to the primary capacitors (10,000uF electros x 4), and operating temperatures are much more within reason.
regards, jonathan carr
grounding
Dear Mr. Carr,
Thank you for the very extensive reply.
Yesterday evening i checked the link that Bob mentioned and i found a lot of information that you gave years ago.
The bypass cap issue is clear now.
About star ground ;
Yes the first picture is the one i'm talking about.
I hope you will tell us some details about that issue(that strange jumper etc.).
Maybe you will also share your vision about signal-ground,supply-ground,chassis ground etc. Especially the coupling between them.
Best regards,
Johan.
Dear Mr. Carr,
Thank you for the very extensive reply.
Yesterday evening i checked the link that Bob mentioned and i found a lot of information that you gave years ago.
The bypass cap issue is clear now.
About star ground ;
Yes the first picture is the one i'm talking about.
I hope you will tell us some details about that issue(that strange jumper etc.).
Maybe you will also share your vision about signal-ground,supply-ground,chassis ground etc. Especially the coupling between them.
Best regards,
Johan.
Jonathan,
I see now that my comment about different board materials and finished was unclear.
As the PCB is a "component" and thus part of the schematic, I think that one has to - in some design stage - decide what material to use for the PCB, and then change the passive and active components according to that PCB.
My circuits are simple and I like Teflon so I use that mostly. Simple rework has not been a problem (at least not on non ROHS boards - I have not tried any ROHS Teflon boards yet, but as the heat is higher there might be problems).
< Silver plating seems to sound OK when new, but the word is that it does change after some months,
Depending on the environment, it can "oxidise" quickly or slowly.
That is the reason I want to have gold on top of the silver plating - for protection. And it looks good, too
Regards,
Sigurd
I see now that my comment about different board materials and finished was unclear.
As the PCB is a "component" and thus part of the schematic, I think that one has to - in some design stage - decide what material to use for the PCB, and then change the passive and active components according to that PCB.
My circuits are simple and I like Teflon so I use that mostly. Simple rework has not been a problem (at least not on non ROHS boards - I have not tried any ROHS Teflon boards yet, but as the heat is higher there might be problems).
< Silver plating seems to sound OK when new, but the word is that it does change after some months,
Depending on the environment, it can "oxidise" quickly or slowly.
That is the reason I want to have gold on top of the silver plating - for protection. And it looks good, too
Regards,
Sigurd
jcarr said:
Re: grounding
The 4-2SE has alternative ways of connecting the various grounds together. Normally the signal and power grounds can be set and left as is, but it is advantageous to provide various ways of connecting the chassis ground to the other grounds which can be selected from by the dealer or end-user, depending on each individual circumstance. On the grounding board there is a direct connection for the chassis ground, an R- C network, and a floating position. The jumper on the grounding board is what is used as the chassis ground position selector (there is a section on this in the owner's manual for the 4-2SE).
This can be an exceedingly complex question to answer, because it depends on how the power supply is configured, the number of power supply, regulator and amplifier circuit boards and how the grounding of each is set up, whether the grounding is passive, active (and what kind of active) or a hybrid, the bandwidth of the circuits, what kind of bandwidth-limiting input filters are in place, how the power supply bypasses are executed, the nature of the chassis structure, and so on. But assuming the liberty to give an over-simplified explanation, everything else being equal (which admittedly never happens ), regarding grounding implementation I think that as little commoning of grounds as possible (i.e., a real star-ground) sounds the best, but if the circuit is oscillating or has poor transient behaviour, you either need to think about a ground plane (which means less inductive behaviour but greater commoning of grounds) or some means to reduce the bandwidth requirements.
IOW, for low-bandwidth circuit nodes, a star-ground (like Naim or DNM) should be fine, but for wider bandwidth nodes some degree of planar grounding may be necessary to quell oscillations (although the extra capacitance associated with planar grounding may result in too much capacitance for certain circuit nodes - even if they are high-bandwidth in nature). Keeping that in mind, analyzing your circuit in terms of bandwidth required at each node will give you a map that denotes your range of choice regarding low-inductive/high-ionductive grounds (and low-capacitance where it is needed) across the entire circuit. Then you can design/adjust the grounding system accordingly. Do feel free to mix different kinds of grounds depending on the specific needs of the individual circuit node, again keeping in mind that a star ground may sound the best - if you can get away with it.
The basic rules are fairly simple. The actual implementation is where things can get tricky, but if you consider the behaviour at each circuit node, I don't think you will go too far wrong. Most of the rest (like distinctions between power, signal and chassis grounds) are simply applications of the above.
Now it is true that the Lyra-Connoisseur incorporate active grounding, and this opens up a whole new range of things that you can do. But at this level, rather than consisting of general rules which can be applied to a variety of different circuits, the grounding becomes something that needs to be designed as an integrated and inseparable aspect of the specific circuitry and infrastructure.
hth, jonathan carr
joho said:
The 4-2SE has alternative ways of connecting the various grounds together. Normally the signal and power grounds can be set and left as is, but it is advantageous to provide various ways of connecting the chassis ground to the other grounds which can be selected from by the dealer or end-user, depending on each individual circumstance. On the grounding board there is a direct connection for the chassis ground, an R- C network, and a floating position. The jumper on the grounding board is what is used as the chassis ground position selector (there is a section on this in the owner's manual for the 4-2SE).
This can be an exceedingly complex question to answer, because it depends on how the power supply is configured, the number of power supply, regulator and amplifier circuit boards and how the grounding of each is set up, whether the grounding is passive, active (and what kind of active) or a hybrid, the bandwidth of the circuits, what kind of bandwidth-limiting input filters are in place, how the power supply bypasses are executed, the nature of the chassis structure, and so on. But assuming the liberty to give an over-simplified explanation
, everything else being equal (which admittedly never happens ), regarding grounding implementation I think that as little commoning of grounds as possible (i.e., a real star-ground) sounds the best, but if the circuit is oscillating or has poor transient behaviour, you either need to think about a ground plane (which means less inductive behaviour but greater commoning of grounds) or some means to reduce the bandwidth requirements.IOW, for low-bandwidth circuit nodes, a star-ground (like Naim or DNM) should be fine, but for wider bandwidth nodes some degree of planar grounding may be necessary to quell oscillations (although the extra capacitance associated with planar grounding may result in too much capacitance for certain circuit nodes - even if they are high-bandwidth in nature). Keeping that in mind, analyzing your circuit in terms of bandwidth required at each node will give you a map that denotes your range of choice regarding low-inductive/high-ionductive grounds (and low-capacitance where it is needed) across the entire circuit. Then you can design/adjust the grounding system accordingly. Do feel free to mix different kinds of grounds depending on the specific needs of the individual circuit node, again keeping in mind that a star ground may sound the best - if you can get away with it.
The basic rules are fairly simple. The actual implementation is where things can get tricky, but if you consider the behaviour at each circuit node, I don't think you will go too far wrong. Most of the rest (like distinctions between power, signal and chassis grounds) are simply applications of the above.
Now it is true that the Lyra-Connoisseur incorporate active grounding, and this opens up a whole new range of things that you can do. But at this level, rather than consisting of general rules which can be applied to a variety of different circuits, the grounding becomes something that needs to be designed as an integrated and inseparable aspect of the specific circuitry and infrastructure.
hth, jonathan carr
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.