Hi Jesper,
until now measurements only, no time to listen.
the Pierce oscillator has re-entered for two reasons:
- the new Pierce gate based is an entry level version with cheaper crystals (HC-49/U instead of HC-43/U package), supplied only as finished board with crystal installed
- the new Pierce bjt based has entered because the Driscoll and the Differential oscillators are suitable for SC-Cut 5-24 MHz overtone crystals only, so it's the only chance to implement AT-Cut crystals up to 98 MHz
Obviously their performance will be worse than the Driscoll and the Differential oscillators type, but also less expensive since the AT-Cut crystals are cheaper than the SC-Cut.
Andrea
22/24 AT-Cut crystals in the new Pierce oscillator have not yet been measured, we are waiting for the prototype board.
I agree completely that there is no personal DIY creativity in simply soldering someone else's design. I view this as a component that outperforms anything remotely in the price range. This is the 4th clock I have used and I did not design any of them. But choosing this clock to be a central component of my system is the essence of DIY design. I could have chosen to use my time and money on interstage transformers as an upgrade to my amps. But in that case I would not have the skill or the inclination to wind the transformers. I would do my research and decide the best choice and implement them.
Now I would say there is quite a bit of DIY involvement in the implementation of these clocks. In my case, I built multiple linear power supplies, a chassis, and put quite a bit of time into vibration and RF isolation. I believe all of the time spent experimenting on different implementation schemes resulted in an excellent performance. I believe the result is similar to buying a $5k clock but mine cost 10% of that and has pride of construction if not design. Kind of the same for all the components of my system.
Hey thanks, that's a nice tip, I'll try it without buffers and try to make it up by proper routing/termination, I guess it will come to down to the amount of reflection that I'll get. Only problem I have is that with the output resistance of the driver I'm dealing with, the calculation for source termination of parallel transmission lines (Rsource = Z0 -Rdriver*N , N = Number of parallel transmission lines) gets negative so there is no Rsource I can set to completely eliminate Reflections so I just will put in 22Ohms and see how well that works out.
If it won't work that way I now have a plan B and I'll go with 74ac... in parallel
greetings Oli
If you can keep track distance to lessthen say 10mm, then you can avoid adding extra series termination resistors. Adding extra resistance will increase the circuits thermal noise...
For longer track lengths, 22R to 33R works out well - if your not phasenoise / edge speed critical then increasing to 100ohms helps alot with EMC levels.
@andrea: ... thanks for outlining why the Pierce has re-entered your selection - I have to admit that I really do look forward to seeing what you have arrived at ... 
One suggestion - and please don't see this as not recognising that you are already doing a lot and fine work - but have you considered to also include your "best solution" for a frequency divider? Unless your preferences have changed I remember your being very interested in R2R DACs and here I reckon that a splendid frequency divider may be helpful ...
@JohnW:
Hi John ... Might I ask you a question in relation to this? Using lower value resistors increase the effective bandwidth of the track transmission, however, unless the resistor's value leads to critical damping or over-damping of the track response, might the overshoot of the digital pulse (slight or more pronounced) not transfer back into the digital circuitry's power supplies and thus cause more general noise on the PSU line - and eventually potentially more phase noise in the digital transmission?
I reckon this may be what you imply with your 100 ohm suggestion but essentially where would you say that the "sweet spot" speed wise vs overall circuitry stability/noise is?
About terminology: You call the resistors "termination resistors" which pops up in my mind as "resistors close to where the digital transmission line ends". We may of course just use different words for the same resistor - but just to clarify - do you use both a sending resistor & a termination resistor? Don't know if this makes sense ...
BTW - and this is just meant to be very briefly off-topic - over time I have tried out & listened to some SMD resistors (e.g. TNPW-vishay, Susumu-RG, MMA0204, PFC-vishay) and to my ears these resistors SQ wise are not really close to the sound quality of the better "regular" resistors. Might you/others here have experience with a really good sounding SMD resistor?
Cheers,
Jesper
One suggestion - and please don't see this as not recognising that you are already doing a lot and fine work - but have you considered to also include your "best solution" for a frequency divider? Unless your preferences have changed I remember your being very interested in R2R DACs and here I reckon that a splendid frequency divider may be helpful ...
@JohnW:
Hi John ... Might I ask you a question in relation to this? Using lower value resistors increase the effective bandwidth of the track transmission, however, unless the resistor's value leads to critical damping or over-damping of the track response, might the overshoot of the digital pulse (slight or more pronounced) not transfer back into the digital circuitry's power supplies and thus cause more general noise on the PSU line - and eventually potentially more phase noise in the digital transmission?
I reckon this may be what you imply with your 100 ohm suggestion but essentially where would you say that the "sweet spot" speed wise vs overall circuitry stability/noise is?
About terminology: You call the resistors "termination resistors" which pops up in my mind as "resistors close to where the digital transmission line ends". We may of course just use different words for the same resistor - but just to clarify - do you use both a sending resistor & a termination resistor? Don't know if this makes sense ...
BTW - and this is just meant to be very briefly off-topic - over time I have tried out & listened to some SMD resistors (e.g. TNPW-vishay, Susumu-RG, MMA0204, PFC-vishay) and to my ears these resistors SQ wise are not really close to the sound quality of the better "regular" resistors. Might you/others here have experience with a really good sounding SMD resistor?
Cheers,
Jesper
My track length is more in the region of 80mm~100mm although I'm just rerouting everything and I think I can get to about half of that with a little work (the combination of as short as possible and equal length is a pita ;D )
So no way around the resistors :/
@gentlevoice
Yes these are the resistors John is referring to
Btw. there are several options on termination of signal lines, source termination (on the driver side) and end termination (receiver side) and for each several ways to do it, you can also do both in some cases afaik
Edit: To do both source and end termination seems to be best but there are some limiting factors on how sensitive the receiver and capable the driver is, though I'm not sure how this translates to Numbers, perhaps someone can explain on what to look for in the datasheets and how to calculate that. Would be super interesting and helpful
I have no clue about the listening effects but I guess thermal noise and inductance of the resistors play the bigger role here. I'm currently using thin foil SMDs but it would be interesting if someone has a better recommendation
Greetings Oli
Last edited:
@andrea: ... thanks for outlining why the Pierce has re-entered your selection - I have to admit that I really do look forward to seeing what you have arrived at ...
One suggestion - and please don't see this as not recognising that you are already doing a lot and fine work - but have you considered to also include your "best solution" for a frequency divider? Unless your preferences have changed I remember your being very interested in R2R DACs and here I reckon that a splendid frequency divider may be helpful ...
There will be a frequency divider in a future GB, and also a more sophisticated sine to square converter.
It will be a programmable frequency divider to get WS/LRCK directly from the master clock with the lowest phase noise as possible.
Unfortunately the popular Ian's FIFO does not support slave operation from WS/LRCK, it provides directly all I2S signal from the CPLD, so this divider will be unuseful.
Unless Ian will update the FIFO project providing the ability to operate in slave mode, that's the only way to really get the best jitter performace driving the DAC (at least the DAC chip switching on WS/LRCK signal).
Hello everybody !!
Has anyone contacted Laptech Inc. regarding ordering information?
I'm looking to buy 2 pieces HC-43 / U cold weld quartz crystal / SC cut / 16.9344MHz,but as far as I know,minimum order is at least 10 or 15pcs..
If someone has a surplus,doesn’t need it and is willing to sell, I’m interested in.
Has anyone contacted Laptech Inc. regarding ordering information?
I'm looking to buy 2 pieces HC-43 / U cold weld quartz crystal / SC cut / 16.9344MHz,but as far as I know,minimum order is at least 10 or 15pcs..
If someone has a surplus,doesn’t need it and is willing to sell, I’m interested in.
Last edited:
@andrea: ... thanks for outlining why the Pierce has re-entered your selection - I have to admit that I really do look forward to seeing what you have arrived at ...
One suggestion - and please don't see this as not recognising that you are already doing a lot and fine work - but have you considered to also include your "best solution" for a frequency divider? Unless your preferences have changed I remember your being very interested in R2R DACs and here I reckon that a splendid frequency divider may be helpful ...
@JohnW:
Hi John ... Might I ask you a question in relation to this? Using lower value resistors increase the effective bandwidth of the track transmission, however, unless the resistor's value leads to critical damping or over-damping of the track response, might the overshoot of the digital pulse (slight or more pronounced) not transfer back into the digital circuitry's power supplies and thus cause more general noise on the PSU line - and eventually potentially more phase noise in the digital transmission?
I reckon this may be what you imply with your 100 ohm suggestion but essentially where would you say that the "sweet spot" speed wise vs overall circuitry stability/noise is?
About terminology: You call the resistors "termination resistors" which pops up in my mind as "resistors close to where the digital transmission line ends". We may of course just use different words for the same resistor - but just to clarify - do you use both a sending resistor & a termination resistor? Don't know if this makes sense ...
BTW - and this is just meant to be very briefly off-topic - over time I have tried out & listened to some SMD resistors (e.g. TNPW-vishay, Susumu-RG, MMA0204, PFC-vishay) and to my ears these resistors SQ wise are not really close to the sound quality of the better "regular" resistors. Might you/others here have experience with a really good sounding SMD resistor?
Cheers,
Jesper
Tantalum nitride.
Maybe if we slave the USB to I2S source before the Fifo with the prg TWTMC Masterclock, the Fifo would have a good enough timing before he sends through the Ian's clock with TWTMC adaptators on it ?
@Andrea: Thanks again for your clarification. I assume "no further questions" ;-) - look forward to hearing more in due time ...
@Hifoli: Hi & thanks for your feedback ...
WRT digital transmission resistors my personal thoughts about this are:
* That source resistors ideally should be placed as close as practically possible to the sending end. This will reduce the capacitance seen by the sending IC output ("unlimited bandwidth into a capacitance" - the trace + ground plane capacitance) and thus reduce the peak current drawn from the sending IC output. On the other hand it is not that clear to me what the advantages of using a receiving resistor may be? ... I reckon it may reduce a possible send-resistor + trace impedance mismatch reflection but again it takes up space where often space is limited.
WRT SMD resistors: Are you using thin film or thin foil SMD resistors? If thin foil I'd be interested in hearing about which one.
@tom59hifi: I think I will take the liberty to answer somewhat on behalf of Andrea ... If you read a couple of pages back you will see that Andrea mentions starting a GB likely before the end of June. This is a.o.t. to be able to order from Laptech even with their minimum orders.
@merlin: Thanks for the tip on tantalum nitride. However, the PFC resistor I mention is exactly tantalum nitride and to my ears they are not as open and "alive" as e.g. Charcroft CAR or Vishay RN65Es ... Might you have some other TN resistor in mind?
Have a good day!
Jesper
@Hifoli: Hi & thanks for your feedback ...
WRT digital transmission resistors my personal thoughts about this are:
* That source resistors ideally should be placed as close as practically possible to the sending end. This will reduce the capacitance seen by the sending IC output ("unlimited bandwidth into a capacitance" - the trace + ground plane capacitance) and thus reduce the peak current drawn from the sending IC output. On the other hand it is not that clear to me what the advantages of using a receiving resistor may be? ... I reckon it may reduce a possible send-resistor + trace impedance mismatch reflection but again it takes up space where often space is limited.
WRT SMD resistors: Are you using thin film or thin foil SMD resistors? If thin foil I'd be interested in hearing about which one.
@tom59hifi: I think I will take the liberty to answer somewhat on behalf of Andrea ... If you read a couple of pages back you will see that Andrea mentions starting a GB likely before the end of June. This is a.o.t. to be able to order from Laptech even with their minimum orders.
@merlin: Thanks for the tip on tantalum nitride. However, the PFC resistor I mention is exactly tantalum nitride and to my ears they are not as open and "alive" as e.g. Charcroft CAR or Vishay RN65Es ... Might you have some other TN resistor in mind?
Have a good day!
Jesper
- Status
- Not open for further replies.