Hi jkeny,
Like I already mentioned, the SD-player concept was developed to eliminate known problems with digital audio sources and interlinks.
Theoretically it's able to outperform the best and most expensive CD transports, simply because the output is 100% bit perfect, all the time, interference is very low, and this concept easily provides very low jitter levels without complicated and often ineffective jitter attenuators. Unlike CD transports, It's insensitive to shock and vibration.
The dsPIC33 main processor only consumes 330mW, while providing 40 MIPs (nanowatt technology). This way interference levels could be kept very low.
Apart form the audible improvements, it's a very handy, compact, low power, light weight, dead quiet, digital audio source that unlike most audiophile CD players can run on a battery power supply. It's also very handy to store so many CD images on one small memory card. I put it in shuffle mode and enjoy listening music all day without a single swap.
Meanwhile I bought a 18Ah lead-acid battery (eur 22), that provides approx. 90 hours of continuous SD-player operation.
The sound quality with this 18Ah battery is even better compared to the 2Ah types.
I am also testing a 16GB SDHC card now (could fit 27 CDs on it), it has been playing for days without problems.
The clock jitter could be measured, provided measuring equipment resolution is high enough, and the probe (probe capacity and ground loops) wouldn't affect the measurement. But measurement results don't dictate perceived sound quality, so I rather rely on practical listening tests.
I already tested different 11.2896 MHz clocks with the SD-player, using a switch for direct comparison (the controller usually continues to function after brief clock interruptions). The 3-crystal clock made a shocking difference in sound quality compared to some low jitter clocks specified at approx. 2 ... 3 ps rms. The improvements in sound quality match the various publications about the audible effects of jitter reduction published on the web.
Like I already mentioned, the SD-player concept was developed to eliminate known problems with digital audio sources and interlinks.
Theoretically it's able to outperform the best and most expensive CD transports, simply because the output is 100% bit perfect, all the time, interference is very low, and this concept easily provides very low jitter levels without complicated and often ineffective jitter attenuators. Unlike CD transports, It's insensitive to shock and vibration.
The dsPIC33 main processor only consumes 330mW, while providing 40 MIPs (nanowatt technology). This way interference levels could be kept very low.
Apart form the audible improvements, it's a very handy, compact, low power, light weight, dead quiet, digital audio source that unlike most audiophile CD players can run on a battery power supply. It's also very handy to store so many CD images on one small memory card. I put it in shuffle mode and enjoy listening music all day without a single swap.
Meanwhile I bought a 18Ah lead-acid battery (eur 22), that provides approx. 90 hours of continuous SD-player operation.
The sound quality with this 18Ah battery is even better compared to the 2Ah types.
I am also testing a 16GB SDHC card now (could fit 27 CDs on it), it has been playing for days without problems.
The clock jitter could be measured, provided measuring equipment resolution is high enough, and the probe (probe capacity and ground loops) wouldn't affect the measurement. But measurement results don't dictate perceived sound quality, so I rather rely on practical listening tests.
I already tested different 11.2896 MHz clocks with the SD-player, using a switch for direct comparison (the controller usually continues to function after brief clock interruptions). The 3-crystal clock made a shocking difference in sound quality compared to some low jitter clocks specified at approx. 2 ... 3 ps rms. The improvements in sound quality match the various publications about the audible effects of jitter reduction published on the web.
Thanks again ,ec-designs, but all the explanations in the world won't beat a good scientific set of measurements - then comes how it sounds (I know this won't tell how it sounds!).
Anyway, I recommend you try these batteries from A123 - they are Lithium Ion Nano Phosphate & capable of delivering 60Amp continuous @ 3.3V with very little power drop.
I could describe all the technology that they use & what it is meant to do but as per my recommendations above re scientific measurements, I'll post this link, let you read the specs yourself & decide how good these batteries are http://www.a123systems.com/products
Anyway, I recommend you try these batteries from A123 - they are Lithium Ion Nano Phosphate & capable of delivering 60Amp continuous @ 3.3V with very little power drop.
I could describe all the technology that they use & what it is meant to do but as per my recommendations above re scientific measurements, I'll post this link, let you read the specs yourself & decide how good these batteries are http://www.a123systems.com/products
Hi all,
maybe this is of some interest: A friend and I are living close to Johns place,
and he was so kind to spend some time to show and demonstrate his baby to us.
To have some kind of reference we brought a very simple TDA1543 Dac, a heavily modded Valab Dac from ebay, with 8 Chips parallel, and DIR9001 and CM108 Chipset. This cheap thing sounds surprisingly good, especially via USB input.
As it is always difficult to get objective impressions in a system which is not my own, our impressions can only be what they are: impressions , and not the truth. Of course.
O.K., first we talked a bit, and then we listened to some music with the card player, which is really nice to look at btw and very well made, although it is still a prototype.
Then we connected our DAC with the apple PC from John via Toslink, and played some of the same music again. We could switch from one source to the other in a second.
There was a clear difference, the SD Player had a more transparent and quiet midrange, with more texture, colours and resolution. The high frequency was less smeared, all as to expect with this low jitter spectrum.
It was not a screaming difference, but clearly the SD is more pleasant to listen to. It has a very seductive midrange, very "undigital", clear and musical.
Downside was the dynamics: here the mains powered Valab was superior, more punch and deeper bass, a bit more exciting, a bit more nervous.
Later we figured that this could be related to the fact, that John uses a passive preamp, just a pot, and the 8 paralleled chips could have an easier job driving this pot than a single poor 1543.
So probably there could be a VERY different impression with an active preamp, and/or mains power supply.
And John put some MKT caps in the output, he likes them most, but as we know, there are many caps to try...
Please note that our "reference" dac, the Valab, is a serious contender, it nicely outperformed an expensive Apogee, and many others, but only after extensive tweaking.
For me the card player is a very attractive design, very sophisticated, and a steal for the price John mentioned above. It is ridicilous to compare it to a CD player in that price range, it will be no contest.
And with PC and dac you have to pay for the computer, the HD, the dac, cables, fighting with software
...
So thanks again to John to make it possible to listen to it and to give some of his precious time away.
And I really would like to know how it sounds in my setup, with horn speakers...
Cheers,
Juergen
maybe this is of some interest: A friend and I are living close to Johns place,
and he was so kind to spend some time to show and demonstrate his baby to us.To have some kind of reference we brought a very simple TDA1543 Dac, a heavily modded Valab Dac from ebay, with 8 Chips parallel, and DIR9001 and CM108 Chipset. This cheap thing sounds surprisingly good, especially via USB input.
As it is always difficult to get objective impressions in a system which is not my own, our impressions can only be what they are: impressions
, and not the truth. Of course.O.K., first we talked a bit, and then we listened to some music with the card player, which is really nice to look at btw and very well made, although it is still a prototype.
Then we connected our DAC with the apple PC from John via Toslink, and played some of the same music again. We could switch from one source to the other in a second.
There was a clear difference, the SD Player had a more transparent and quiet midrange, with more texture, colours and resolution. The high frequency was less smeared, all as to expect with this low jitter spectrum.
It was not a screaming difference, but clearly the SD is more pleasant to listen to. It has a very seductive midrange, very "undigital", clear and musical.
Downside was the dynamics: here the mains powered Valab was superior, more punch and deeper bass, a bit more exciting, a bit more nervous.
Later we figured that this could be related to the fact, that John uses a passive preamp, just a pot, and the 8 paralleled chips could have an easier job driving this pot than a single poor 1543.
So probably there could be a VERY different impression with an active preamp, and/or mains power supply.
And John put some MKT caps in the output, he likes them most, but as we know, there are many caps to try...
Please note that our "reference" dac, the Valab, is a serious contender, it nicely outperformed an expensive Apogee, and many others, but only after extensive tweaking.
For me the card player is a very attractive design, very sophisticated, and a steal for the price John mentioned above. It is ridicilous to compare it to a CD player in that price range, it will be no contest.
And with PC and dac you have to pay for the computer, the HD, the dac, cables, fighting with software
...So thanks again to John to make it possible to listen to it and to give some of his precious time away.
And I really would like to know how it sounds in my setup, with horn speakers...
Cheers,
Juergen
oh, now Microchip has FAT32 filesystem for dsPIC?
http://www.microchip.com/stellent/idcplg?IdcService=SS_GET_PAGE&nodeId=1824&appnote=en532040
"AN1045 File I/O Functions Using Microchip's Memory Disk Drive File System Library"
If just one dsPIC33 can control and play 256GB SSD, very interesting
http://www.microchip.com/stellent/idcplg?IdcService=SS_GET_PAGE&nodeId=1824&appnote=en532040
"AN1045 File I/O Functions Using Microchip's Memory Disk Drive File System Library"
If just one dsPIC33 can control and play 256GB SSD, very interesting
Koon,
As the EC Design is a product and not open source code, DIYers like me still have a keen interest of your Ultimate Source design from a few years back. I wonder if you might reconsider providing some support, e.g. in updating the source code for FAT32, larger capacity SDs, and dealing with fragmented files.
Even as it is, I am still planning to build a copy of your Ultimate Source. But any update on source code would be more than welcome.
We should of course continue the discussion, if any, back on the original Ultimate Source thread. Sorry for the brief hijacking.
Thanks,
Patrick
As the EC Design is a product and not open source code, DIYers like me still have a keen interest of your Ultimate Source design from a few years back. I wonder if you might reconsider providing some support, e.g. in updating the source code for FAT32, larger capacity SDs, and dealing with fragmented files.
Even as it is, I am still planning to build a copy of your Ultimate Source. But any update on source code would be more than welcome.
We should of course continue the discussion, if any, back on the original Ultimate Source thread. Sorry for the brief hijacking.
Thanks,
Patrick
Hi KOON3876,
Welcome to this thread.
We already use it in the SD-player (post #5).
But it requires some serious modifications to work with both SD/SDHC cards.
The code for writing SD cards was also removed in order to minimize ROM usage.
There will be a different protocol for reading cards with 64GB and higher capacity (SDXC), so the Microchip library needs to be adapted for SD/SDHC, and SDXC.
Welcome to this thread.
We already use it in the SD-player (post #5).
But it requires some serious modifications to work with both SD/SDHC cards.
The code for writing SD cards was also removed in order to minimize ROM usage.
There will be a different protocol for reading cards with 64GB and higher capacity (SDXC), so the Microchip library needs to be adapted for SD/SDHC, and SDXC.
John, thanks for your informative threads. I'm only just beginning my audiophilia, and may never have ears to hear distinguish what you guys hear. However, I find your designs very well executed, and it's very unique that you start a thread with a text-only concept, and then take it to such a degree of completion! Very enjoyable.
My system is nowhere near noticing such minute differences (could be my taste in music but I'd love to hear why my iMod/iFlash 64gb (CF) with custom dock-out isn't the cat's *** as I had previously thought. I think it has a leg up on user friendliness though. Actually NM I'll just assume there's little comparison and that I'll learn the "why" as I go along. I do understand the battery irregularity can cause imperfections and that's enough.
After reading your massive initial thread without stopping until pg 70, I remember vague promises about a thread about how to make your own PCB's, and a thread about your "resonators," which are fascinating perhaps more than all else.
Please post prices on your site so we don't have to ask, which I find off-putting, because I'm curious, but am totally unaware of how many 50hr weeks I'd have to pull to finance one component. Since at some point "I" became "we," it would make sense, unless you prefer the exclusivity.
Anyway thanks for another amazing DIY thread, and I look forward to the next!
PS dddac, nice panels! Can't wait to see that sight in suburbia on this side of the pond.
My system is nowhere near noticing such minute differences (could be my taste in music
but I'd love to hear why my iMod/iFlash 64gb (CF) with custom dock-out isn't the cat's *** as I had previously thought. I think it has a leg up on user friendliness though. Actually NM I'll just assume there's little comparison and that I'll learn the "why" as I go along. I do understand the battery irregularity can cause imperfections and that's enough.After reading your massive initial thread without stopping until pg 70, I remember vague promises about a thread about how to make your own PCB's, and a thread about your "resonators," which are fascinating perhaps more than all else.
Please post prices on your site so we don't have to ask, which I find off-putting, because I'm curious, but am totally unaware of how many 50hr weeks I'd have to pull to finance one component. Since at some point "I" became "we," it would make sense, unless you prefer the exclusivity.
Anyway thanks for another amazing DIY thread, and I look forward to the next!
PS dddac, nice panels! Can't wait to see that sight in suburbia on this side of the pond.
soundcheck said:
Hi John.
It'll get better until you use a 2mR Northstar. Hehehe.
I was just looking at a 1F buffer cap. The car audio folks use these quite often.
There are types which run as low as 0.9mR. This might be a nice alternative try.
A cheaper alternative are the A123 3.3V LiFePO4 batteries - similar internal resistance @4mR but only $15 each rather than > 200 euro for the Northstar.
Have you tried them Soundcheck - I'd be interested in your evaluation
Don' t forget that you need
1. to put 4 of them in series. (4*4mR)
2. to put a couple of them in parallel
3. to consider that "capacity rules" ( at 100ah my batteries are always full)
4. to consider that continuous reload cycles age you battery much faster.
5. at a huge capacity your batteries would be even more expensive
6. to consider LiPos critical behaviour if you run them below a certain capacity
However, if you need a 3.3V supply, they seem to perfom quite well. A friend of mine told me that they'll outperfom standard regulators e.g. as found on ESI Juli.
1. to put 4 of them in series. (4*4mR)
2. to put a couple of them in parallel
3. to consider that "capacity rules" ( at 100ah my batteries are always full)
4. to consider that continuous reload cycles age you battery much faster.
5. at a huge capacity your batteries would be even more expensive
6. to consider LiPos critical behaviour if you run them below a certain capacity
However, if you need a 3.3V supply, they seem to perfom quite well. A friend of mine told me that they'll outperfom standard regulators e.g. as found on ESI Juli.
1. to put 4 of them in series. (4*4mR) - does it need 12 V?
2. to put a couple of them in parallel - enough current delivery - no need for parallel
3. to consider that "capacity rules" ( at 100ah my batteries are always full) - this is digital - low draw on the battery/ low capacity - don't need 100Ah
4. to consider that continuous reload cycles age you battery much faster. - these have a long recharge lifetime - used by the Remote control mob
5. at a huge capacity your batteries would be even more expensive - not being used for power amp - Northstart would be better there
6. to consider LiPos critical behaviour if you run them below a certain capacity - these are not the same as LiPos & don't have the same undercharge/overcharge problems
7. If the device (chip) needs more than one 3.3V supply (digital, analog) - you can keep these separate by using individual batteries rather than regulators!
What's the best/cheapest source for the Northstar?
2. to put a couple of them in parallel - enough current delivery - no need for parallel
3. to consider that "capacity rules" ( at 100ah my batteries are always full) - this is digital - low draw on the battery/ low capacity - don't need 100Ah
4. to consider that continuous reload cycles age you battery much faster. - these have a long recharge lifetime - used by the Remote control mob
5. at a huge capacity your batteries would be even more expensive - not being used for power amp - Northstart would be better there
6. to consider LiPos critical behaviour if you run them below a certain capacity - these are not the same as LiPos & don't have the same undercharge/overcharge problems
7. If the device (chip) needs more than one 3.3V supply (digital, analog) - you can keep these separate by using individual batteries rather than regulators!
I heard that too - wonder how they compare to Salas shunt reg?
What's the best/cheapest source for the Northstar?
Hi mickie,
Yes, my audio set wasn't able to correctly process the high resolution, micro detail, and phase resolution offered by the SD-player. In short, I had to redesign volume control, power amplifiers and speakers before continuing. This took quite some time.
I settled for a volume potentiometer as even the degrading effects on sound quality, caused by multiple relay contacts in series (stepped volume control) is now clearly audible.
The power amplifier was modified to a balanced, zero global feedback, fully DC-coupled all FET Bridge amplifier. In order to achieve maximum resolution and lowest possible grain, I avoided using constant current sources and especially current mirrors. In order to achieve highly accurate bass, even the DC-servo had to be removed, so the circuit had to be designed in such way that it remains unconditionally stable and DC-drift was minimized.
The speakers (sonic resonators) I have been using for testing are 3-way systems. It now turns out that it's very difficult to keep 3 speakers under control, and achieving correct phase response (crossover regions).
Two way systems can offer cleaner sound, but often bass response is limited. Full range speakers suffer from too high moving mass and doppler distortion (one speaker cone producing two different frequencies simultaneously).
So I finally settled for a semi 2-way sonic resonator, using a 130mm bass / midrange and a titanium / aluminum alloy speaker chassis. Omnidirectional emission was achieved by fixating CNC milled wooden cones above each speaker (both fire upwards now).
The sonic resonator is driven by the 130mm woofer/midrange chassis. The sonic resonator is built around a 170mm subwoofer / corrector. This subwoofer is acoustically driven (not connected to the power amplifier).
The new sonic resonator is slightly higher than 1 meter, and diameter of the tube equals 20cm. Both, all-FET bridge amp and this new sonic resonator finally revealed what the SD-player is capable off.
I attached a photograph of a test setup of this new sonic resonator prototype. The filter is located external for testing, it will be integrated later.
Based on this, there is no need to further modify the SD-player design, it already performs beyond expectation. so I plan to start-up a small batch soon.
The multi-platform software application (Mac OSX, Windows, Linux) specifically intended to accurately rip CDs, manage CD database, program play lists to SD-card, and print booklets or pdf files of of SD-card music collection, is still under development.
Yes, my audio set wasn't able to correctly process the high resolution, micro detail, and phase resolution offered by the SD-player. In short, I had to redesign volume control, power amplifiers and speakers before continuing. This took quite some time.
I settled for a volume potentiometer as even the degrading effects on sound quality, caused by multiple relay contacts in series (stepped volume control) is now clearly audible.
The power amplifier was modified to a balanced, zero global feedback, fully DC-coupled all FET Bridge amplifier. In order to achieve maximum resolution and lowest possible grain, I avoided using constant current sources and especially current mirrors. In order to achieve highly accurate bass, even the DC-servo had to be removed, so the circuit had to be designed in such way that it remains unconditionally stable and DC-drift was minimized.
The speakers (sonic resonators) I have been using for testing are 3-way systems. It now turns out that it's very difficult to keep 3 speakers under control, and achieving correct phase response (crossover regions).
Two way systems can offer cleaner sound, but often bass response is limited. Full range speakers suffer from too high moving mass and doppler distortion (one speaker cone producing two different frequencies simultaneously).
So I finally settled for a semi 2-way sonic resonator, using a 130mm bass / midrange and a titanium / aluminum alloy speaker chassis. Omnidirectional emission was achieved by fixating CNC milled wooden cones above each speaker (both fire upwards now).
The sonic resonator is driven by the 130mm woofer/midrange chassis. The sonic resonator is built around a 170mm subwoofer / corrector. This subwoofer is acoustically driven (not connected to the power amplifier).
The new sonic resonator is slightly higher than 1 meter, and diameter of the tube equals 20cm. Both, all-FET bridge amp and this new sonic resonator finally revealed what the SD-player is capable off.
I attached a photograph of a test setup of this new sonic resonator prototype. The filter is located external for testing, it will be integrated later.
Based on this, there is no need to further modify the SD-player design, it already performs beyond expectation. so I plan to start-up a small batch soon.
The multi-platform software application (Mac OSX, Windows, Linux) specifically intended to accurately rip CDs, manage CD database, program play lists to SD-card, and print booklets or pdf files of of SD-card music collection, is still under development.
Attachments
Last edited:
Hi John.
Good to see that you're back on track.
I somehow recognize the way you're going.
As you know I ended up having everything built into one box - no connectors - no cables - no binding posts - no coupling caps - avoid all active stuff. Unfortunately I had to leave one switch and connector to the battery in. That's still bothering me because without it it sounds better.
I am running fullrange speakers without any parts/crossover in front of it. This for sure makes a difference.
Perhaps one day you'll get hooked on to high efficiency speakers. At 100db/W SPL you'll
see what I mean.
A nice gadget for you player would be a high quality headphone out!!
Good to see that you're back on track.
I somehow recognize the way you're going.
As you know I ended up having everything built into one box - no connectors - no cables - no binding posts - no coupling caps - avoid all active stuff. Unfortunately I had to leave one switch and connector to the battery in. That's still bothering me because without it it sounds better.
I am running fullrange speakers without any parts/crossover in front of it. This for sure makes a difference.
Perhaps one day you'll get hooked on to high efficiency speakers. At 100db/W SPL you'll
see what I mean.
A nice gadget for you player would be a high quality headphone out!!
I'd say that just a very,very small number of people were in the position to experience the full potential of 44.1/16.
EDIT:
The same I'd say about 24/96.
Correction: No No. It's much worse. I am not aware of anybody who pushed 24/96 to its limits!!!!
Last edited:
<<Quote: Yes, my audio set wasn't able to correctly process the high resolution, micro detail, and phase resolution offered by the SD-player. In short, I had to redesign volume control, power amplifiers and speakers before continuing. This took quite some time. end of quote.>>
I've told you before: fear when -ECdesigns- remains silent!
I want my player.
Regards,
M
I've told you before: fear when -ECdesigns- remains silent!
I want my player.
Regards,
M
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.