some of my friends are willing to make a PCM1794 kit .
this is gonna be a HiEnd kit with I2C control,src4193 asrc , Buffered I/O , smd multilayer PCB (4-6),and MONO config of PCM1794 , maybe even DSP digital crossover .
Don't worry about technical details, they are top RDs in computer industry.
how much $$ would u like to invest in such kit ?
And what kind of form do u like to buy?
1. PCB with Chips
2. PCB only
3. assembled PCB
4. factory made product
You are appreciated to give me suggestion .
Thx
this is gonna be a HiEnd kit with I2C control,src4193 asrc , Buffered I/O , smd multilayer PCB (4-6),and MONO config of PCM1794 , maybe even DSP digital crossover .
Don't worry about technical details, they are top RDs in computer industry.
how much $$ would u like to invest in such kit ?
And what kind of form do u like to buy?
1. PCB with Chips
2. PCB only
3. assembled PCB
4. factory made product
You are appreciated to give me suggestion .
Thx
foo said:
Yes, computer industry, perfect start to design good sounding audio stuff
For your info, I've sold hundreds of these kits
http://members.chello.nl/~m.heijligers/DAChtml/dactop.htm
Designing hi tech audio dacs is not a trivial enterprise, there is a subjective element to audio performance that is close to impossible to quantify and people in the computer industry are not the ones I would choose for such an endeavour, which is basically what Guido was telling you.
Former audio product designer..
Kevin
Former audio product designer..
Kevin
Hi foo,
Dont be put off!
I suspect that several people would be interested in a high end DAC, BUT you have not left much time for people to comment nor have you given much detail about your plans and why you feel that your DAC would offer something that is not already available.
I dont know anything about the technical issues involved in designing such a thing but I would be interested in a DAC project so long as it offered a quality product that operated to a very high standard and at a good price.
My preference would be for instructions and PCB's but with the possibility of buying hard to obtain parts.
Dont be put off!
I suspect that several people would be interested in a high end DAC, BUT you have not left much time for people to comment nor have you given much detail about your plans and why you feel that your DAC would offer something that is not already available.
I dont know anything about the technical issues involved in designing such a thing but I would be interested in a DAC project so long as it offered a quality product that operated to a very high standard and at a good price.
My preference would be for instructions and PCB's but with the possibility of buying hard to obtain parts.
Banned
Joined 2002
wytco0 said:
Thank you for your comment and encouragement.
The main purpose(also what not available in existing kits) of this kit will be :
1. good electrical performance , less than -120db snr , 130db+ dynamic range , and very low THD . This can be achieved by using good PCB,SMD parts,and good I/V stage design .
2. jitter correction ability : low jitter clock and clock sync is required , and we will also need Ram buffer to do this dejitter job .
especially for lightpipe I/O and HTPC use , this function will be very useful .
3. As to their professional suggestion , they will undoubtedly insist on using SMD parts in digital section of this kit for best EMI/RF performance .
These 3 topics are the main digital feature of this kit.
I don't think I can argue with them over these 3 topics .
Since they will insist on using smd parts in digital section, i think we better make assembled PCBs because smd parts are not accessable for everyone .
As to analog part and PWS , this is what we can talk about .
They don't have very strong opinion about this section.
Tube outputstage is also acceptable .
If you have any suggestion , please do tell me .
My suggestion is they make modulized PCBs for the analog output
+PCM1794 + cs8416 with optional I2S input(to connect to the asrc and dijitter module, or any other digital stuff).
I think this will make the kit more flexible,and make the price easier for most people cause people can buy the basic kit for an afordable price . (cause the top quality PCB will be pricy in fullsized kit )
Once you get this going, lots of people will be interested.
My suggestions would be in line with what you have already thought of...pcm1794 in mono config / i2s but with one addition and that is firewire or usb interface as well.
low jitter clock and clock sync : Maybe it would be a good idea to to use Guido's DAC clock. (Like my RAKK dac can.Also PCM1794 by the way)
http://www.tentlabs.com/ProductEntities/XO-DAC/XO-DAC.html
because it is a PROVEN design into which a lot of thinking and experience has already gone.
Anyway it seems to me it should be modular as well...
with choice to have Guido's or a different re-clocking module in...
and separate I/V conversion board...
with either passive/transformer/op amp/tube board..
regulation on board obviously but maybe a seperate "main" psu. That will feed the DAC.
Just a few things of the top of my head..
O yes and before I forget it should be ready assembled...since I tried to solder a USB 2 SPDIF converter made by guys over at head fi...and i think I won't be able to pull it off.. :-(
also the lack of reaction is also because it is summer in the Northern Hemisphere...most people are outside enjoying the weather.
My suggestions would be in line with what you have already thought of...pcm1794 in mono config / i2s but with one addition and that is firewire or usb interface as well.
low jitter clock and clock sync : Maybe it would be a good idea to to use Guido's DAC clock. (Like my RAKK dac can.Also PCM1794 by the way)
http://www.tentlabs.com/ProductEntities/XO-DAC/XO-DAC.html
because it is a PROVEN design into which a lot of thinking and experience has already gone.
Anyway it seems to me it should be modular as well...
with choice to have Guido's or a different re-clocking module in...
and separate I/V conversion board...
with either passive/transformer/op amp/tube board..
regulation on board obviously but maybe a seperate "main" psu. That will feed the DAC.
Just a few things of the top of my head..
O yes and before I forget it should be ready assembled...since I tried to solder a USB 2 SPDIF converter made by guys over at head fi...and i think I won't be able to pull it off.. :-(
also the lack of reaction is also because it is summer in the Northern Hemisphere...most people are outside enjoying the weather.
I agree wil BAS,
make it modular so folks and choose which type of output section,
USB input is mandatory now,
switch to chose I2S input is smart too,
battery powered as an option if you reaaly want to be top high end.
I would suggest to ship the SMD stuff mounted and let buyers do the rest.
if you use the PCM1794, please make the filter choice a toggle on the front panel (sharp or slow)
since there are PCBS already for Jung's Super Voltage regulators design with this in mind, you could do a bulk order of these and include them in the kit.
make it optional to have bal/unbal puts or both.
finally, make final low pass filter optional.
And finally, another option that you could design on the pCB but leave empty is the space for the HDCD decode chip.
ambitious, expensive but interesting.
good luck
paba
make it modular so folks and choose which type of output section,
USB input is mandatory now,
switch to chose I2S input is smart too,
battery powered as an option if you reaaly want to be top high end.
I would suggest to ship the SMD stuff mounted and let buyers do the rest.
if you use the PCM1794, please make the filter choice a toggle on the front panel (sharp or slow)
since there are PCBS already for Jung's Super Voltage regulators design with this in mind, you could do a bulk order of these and include them in the kit.
make it optional to have bal/unbal puts or both.
finally, make final low pass filter optional.
And finally, another option that you could design on the pCB but leave empty is the space for the HDCD decode chip.
ambitious, expensive but interesting.
good luck
paba
My comments...
- Please take >10 days to design your equipment.
- I'd use an AD1986 instead of a SRC4193, the ADI part has far better jitter attenuation. The 1986's -3dB point is in the single digit Hz, the 4192/3 is around 2KHz!
- if you planned on using the SRC4193 for volume control, you'll now need to use the PCM1792.
- Use a good clock. Get samples of XO's from several different manufacturers and compare them on a spectrum analyzer and jitter analyzer. I found the CTS CB3LV series is surprisingly good. Other favorite crystal manufacturers of mine are VF Technologies and Crystek.
If you're using a fast crystal (eg 49.152) then watch out - one company might cut a 16.384MHz crystal and build an oscillator with a PLL chip multiplying it by three. Or another company might make a crap overtone part which has spurious tones at the fundamental. Make sure you specify a decent overtone or HFF part, and verify what the salesperson tells you on the bench.
- A USB or Firewire input would be slick for interfacing to a PC. But if you do this, don't use a PCM270x or PCM290x chip - use a USB controller that's capable of doing 24/96 or 24/192. Hack open a M-Audio Transit and find out what they use inside, and let me know
And finally... Good luck, but get ready for a disappointment. From working in the audio industry, I've realized a few facts about audiophiles (or at least, the stereotypical stereophile-reading ones):
- They don't care about numbers. If you demonstrate good numbers, they assume you're an engineer and then they automatically assume you've made all the wrong decisions because you went for the good numbers and your equipment is just a piece of trash. Turns out that "real" audio designers don't care about numbers, they just know... In fact, someone at a pro audio shop once told me this - direct quote - "when people design equipment using numbers, the equipment sounds like numbers. There's no life in it."
- Audiophiles don't care what's under the hood of the equipment you make. Your average audiophile doesn't read industry magazines and hear about the new DAC chip that came out from Wolfson, TI, ADI, AKM, Cirrus, etc... they just hear about the new DAC or new CD player that came out from Levinson or whoever. "What's in it? who cares! i mean, it's a f**ing Levinson!"
- If you put a certain feature in your equipment, it will automatically get dismissed - for example, TOSLINK automatically sounds bad. It doesn't matter what sort of clock recovery you put after it or what sort of explanation you use, you'll never convince a stereotypical audiophile that audio can't survive optical transport without sounding like utter garbage. And people have even worse baseless criticism for ASRC chips...
Overall, if you're a brand new company trying to break into pro audio, it's a battle... I personally believe you'll need a marketing staff at least twice as big as your engineering staff.
- Please take >10 days to design your equipment.
- I'd use an AD1986 instead of a SRC4193, the ADI part has far better jitter attenuation. The 1986's -3dB point is in the single digit Hz, the 4192/3 is around 2KHz!
- if you planned on using the SRC4193 for volume control, you'll now need to use the PCM1792.
- Use a good clock. Get samples of XO's from several different manufacturers and compare them on a spectrum analyzer and jitter analyzer. I found the CTS CB3LV series is surprisingly good. Other favorite crystal manufacturers of mine are VF Technologies and Crystek.
If you're using a fast crystal (eg 49.152) then watch out - one company might cut a 16.384MHz crystal and build an oscillator with a PLL chip multiplying it by three. Or another company might make a crap overtone part which has spurious tones at the fundamental. Make sure you specify a decent overtone or HFF part, and verify what the salesperson tells you on the bench.
- A USB or Firewire input would be slick for interfacing to a PC. But if you do this, don't use a PCM270x or PCM290x chip - use a USB controller that's capable of doing 24/96 or 24/192. Hack open a M-Audio Transit and find out what they use inside, and let me know
And finally... Good luck, but get ready for a disappointment. From working in the audio industry, I've realized a few facts about audiophiles (or at least, the stereotypical stereophile-reading ones):
- They don't care about numbers. If you demonstrate good numbers, they assume you're an engineer and then they automatically assume you've made all the wrong decisions because you went for the good numbers and your equipment is just a piece of trash. Turns out that "real" audio designers don't care about numbers, they just know... In fact, someone at a pro audio shop once told me this - direct quote - "when people design equipment using numbers, the equipment sounds like numbers. There's no life in it."
- Audiophiles don't care what's under the hood of the equipment you make. Your average audiophile doesn't read industry magazines and hear about the new DAC chip that came out from Wolfson, TI, ADI, AKM, Cirrus, etc... they just hear about the new DAC or new CD player that came out from Levinson or whoever. "What's in it? who cares! i mean, it's a f**ing Levinson!"
- If you put a certain feature in your equipment, it will automatically get dismissed - for example, TOSLINK automatically sounds bad. It doesn't matter what sort of clock recovery you put after it or what sort of explanation you use, you'll never convince a stereotypical audiophile that audio can't survive optical transport without sounding like utter garbage. And people have even worse baseless criticism for ASRC chips...
Overall, if you're a brand new company trying to break into pro audio, it's a battle... I personally believe you'll need a marketing staff at least twice as big as your engineering staff.
Mmm the first part of your post is excellent..but then your emotions get the better of you IMHO..I don't want to sidetrack this thread. But the other side of the story is that engineers told us there was no difference between toslink and coax. They also told that cd's would give perfect sound forever. We should not be listening to engineers or the marketing department..but to the sound system were building.
And you are right..audiophiles don't care about numbers nor should they. Were listening to music aren't we. Nor should we care about brands (I know I don't)
And you are right..audiophiles don't care about numbers nor should they. Were listening to music aren't we. Nor should we care about brands (I know I don't)
Thanks for your advices.
Especially gmarsh, your comments are really pro .
To get everyone understand , I've simplified the technical parts.
You are right , we were planning to use pcm1792 instead of 1794
, but we will probabally use wm8816 as an optional output stage.
As to the usb and firewire connection , we will take it into serious
consideration since so many ppl need it , also we might need it to
transmit I2C format and use a sofware to control via PC .
For your maketting advices , we had already considered it and made out some basic solutions . Thank you for telling us the truth
anyway .
We are not a company , we are just some rational diyers with some technical resources( such as AP2, soundlab ,...,etc) . However , Setting up a company will be an option .
Especially gmarsh, your comments are really pro .
To get everyone understand , I've simplified the technical parts.
You are right , we were planning to use pcm1792 instead of 1794
, but we will probabally use wm8816 as an optional output stage.
As to the usb and firewire connection , we will take it into serious
consideration since so many ppl need it , also we might need it to
transmit I2C format and use a sofware to control via PC .
For your maketting advices , we had already considered it and made out some basic solutions . Thank you for telling us the truth
anyway .
We are not a company , we are just some rational diyers with some technical resources( such as AP2, soundlab ,...,etc) . However , Setting up a company will be an option .
Actually , in my opinion,audiophiles are not totally wrong .
They just don't want to spend time for technical studies.
Sound performance can somewhat match the technical feature .
That also a good way for ppl not so good at PA to choose their equipments.
We are going to spend some time for a "Good-sound" option.
And this is going to be one part of our marketting strategy.
They just don't want to spend time for technical studies.
Sound performance can somewhat match the technical feature .
That also a good way for ppl not so good at PA to choose their equipments.
We are going to spend some time for a "Good-sound" option.
And this is going to be one part of our marketting strategy.
I think you'd be very suprised how many engineers are audiophiles..Hey what is the definition of an audiophile anyway? To me it's someone who finds they way music is reproduced important. The vast majority of people are NOT interested. I went to a Salsa party last night and the sound was so horribly distorted I had to stuff wet toilet paper in my ears. The other 250 people could not care less. They just enjoyed themselves
Bas Horneman said:
apologies for the above rant.
I'm still convinced that (stereotypical) audiophiles don't actually listen to music - they listen to their equipment as it produces music. Another experience of mine involved listening to Medeski, Martin & Wood on a pricy set of speakers while the owner kept going on about... "hear that cymbal? you hear that? wasn't that great?" - all the time I'm thinking to myself "damn this sounds good... what album is this?" instead of "damn this sounds good, i need these speakers"
I'm just a stereotypical bitter engineer. Take my words with a grain of salt.
And it's entirely true that engineers can be wrong - Engineers tend to operate within their known/familiar universe, where performance is measured in dB's of THD/IMD/TIM/SNR and audio equipment is rated with numbers measured from equipment being fed with combinations of sine waves... and they tend to treat everything they measure as fact. This measures better therefore it sounds better, etc. There's probably several more undiscovered tests that can give an even better evaluation of how good audio equipment is - keep in mind that IMD/TIM tests came long after THD/SNR tests and uncovered some serious problems with "good" hi-fi designs from the 60's/70's that just didn't sound right. And there's plenty of valid arguments for CD's not being good enough as well...
But it's also entirely true that audiophiles can be wrong... Like the coax vs toslink thing - it's understandable that toslink can produce higher jitter and sound different. But it's possible for a coax output and/or a coax input to be badly designed and produce even more jitter. Newer toslink devices perform better. And DIR chips have gotten considerably better too, and are able to reject more SPDIF-related jitter from both coax and optical. It's a complex situation, and simplifying everything down to "optical bad, coax good, for all eternity" isn't productive by any means.
People on both sides of the fence have to say humble and be able to listen to people on the other side. Because we're all wrong anyway.
oh wait, there's another rant. sorry
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.