Asynchronous I2S FIFO project, an ultimate weapon to fight the jitter

[roger57]

Hi Ian,

I don't agree that you should lose money in your effort that will be shared by all that are involved with your group buy. For my part, I cheerfully support any shared increase in the cost of what we buy that will help you recover your costs.

Gary

[qusp]

Me, I thought it was pretty obvious he was losing money, even forgetting his time, once I saw the PCB set. PCB's of that quality, with full flying probe checking, impedance controlled layout etc etc dont come cheap; even without parts in small runs will be up to and over $50 each. is it that people still think SMD = cheaper? its complete fallacy. top shelf tight tolerance SMD is NOT cheap. this part of why ive been objecting to all this feature creep. the only problem with that is making sure there is no money being lost means Ian disclosing the cost.

tricky situation.....

also if Ian is happy and obviously getting more out if than money and that not obviously being the motivator, how do we question that? but where does it end? thats what I mean.

for this I was called greedy/selfish....

[regal]

qusp,

Believe me if Ian makes a board that plugs into a stereo pair of PCM1704 allows NOS and 352.8 with fifo crystal level jitter he will have a gold mine, there is little money in integrating these budget S-D chips because there are several other options out there (not quite as good as Ian's but damn close.) No one is calling you selfish you are a valuable member here, but look at the big picture. Do you see people spending 5 figures + on commercial Sabre's? Since the PCM1704 has become short supply with high end DAC makers dropping them, the prices of used totl commercial 1704 dac's have skyrocketed. Have you ever listened to a properly built PCM1704 (I mean no DF1704, top notch power supplies, top notch analog stage)?

The reason we are bringing up the 5V TLL is this second version is supposed to plug into the most popular digital filter ever (PMD100), if the level is an issue better to know now than before customers find out.

I also support increasing the price of the gb boards if needed, this from the start has potential to change the way we think about DIY DAC's.

Priority is simple and not as complex as you think:

1. stereo PCM1704 up to 8xfs
2. TDA1541 RBCD (check done complete)
3. ESS9018 (check done complete)
4. PMD100/SM5842 RBCD (close needs tested)
5. PCM56/AD1865/AD1862 RBCD

That pretty much covers all the high end DAC's made (I put the ESS on there for your impaired hearing Left off the TDA1543 as this from development was a budget chip, people have done creative things to overcome this but probably will always have to do so. This doesn't have to be a ******* match of my DAC vs yours, just 5 simple priorities that will change the way we DIYers build D/AC's forever.

[KlipschKid]

I'm going to wait to see if someone else can get this working with TDA1543, or wait until the logic daughter board is available.

[qusp]

Regal, $30 dac chip is budget is it? how does the ability of a chip and a company's marketing spiel to part a fool with his money count as a metric for good sound for a DIYER? you just hilighted that you still despite the fact you should know better; still equate a pricetag with quality, this fact does not make you or them elite in my eyes. yes i've heard Ernos 1704 and thought it was quite good, but I dont give a stuff about what those people will pay for.

you are still glossing over the fact that you were pushing for functionality that would require a respin of the fifo board, forget the adapter it hadnt been brought up and this after Ian had already said he was making negative money. sorry there is nothing you can say after that. no speaches with snide little childish digs at a dac chip, nothing. how old are you? 5?

would you like me to copy the email, so you can still claim you dont want a ******* match?

[iancanada]

Originally Posted by Nikola Krivorov

Well, it is really sad if you are loosing instead of making money from that GB
That device would add a four figure number to the price of a DAC from any well respected manufacturer that advertises in magazines and we would be able to read at least two pages of marketing BS how this is impossible to be made by others...

Ian, we are very thankful for what you offered at such a great price, not to mention that you are still trying to make it better and make everybody happy. Hadn't seen that attitude in a while...

1704 is still in production as is the 9018 and there are people like me that have both, so I feel kind of happy that I don't have to choose I have heard great things about the 1862 but I am not going to buy the kit and then go on and try to find the impossible to find at a ridiculous price, not that the aforementioned ones are cheap...

Originally Posted by roger57

Hi Ian,

I don't agree that you should lose money in your effort that will be shared by all that are involved with your group buy. For my part, I cheerfully support any shared increase in the cost of what we buy that will help you recover your costs.

Gary

Thanks guys for those kind words.

Nobody could make money for a SMT production of 25 with reasonable selling price, me too. It's the only problem. That's why I have to order 50 for the GB II. I believe I will get balance from GB II and make money after that. Trust me .

Regards,

[iancanada]

Originally Posted by KlipschKid

I'm going to wait to see if someone else can get this working with TDA1543, or wait until the logic daughter board is available.

Hi KlipshKid,

Sorry for the late reply. FIFO output works for your 1543, it's I2S with 5V logic tolerance. The only problem is you have 16 of them in parallel.I don't think 74HC125 is a good choice for the low jitter application. My suggestion is using dedicated low jitter clock fanout buffers rather than HC125, at least for the word clock and bit clock. You may find some candidates with more than 4 outputs from digikey or mouser. Good luck to your project.

[KlipschKid]

Hi Ian,

Thanks for the reply. My first idea was to use an octal buffer like 74ac541 or 74F244 after the fifo - because the tda are in 8 pairs, each pair would get a driver. But I figured this would also negate some of the benefits of the fifo ?

Then ppl mentioned the logic level problem and I realised there were going to be too many reasons why it wasn't going to work well and I don't have the skills, knowledge or equipment to find and fix the problems.

So I think I'll still wait and see if someone better than me can do it first. However, I am still very interested so I'm just gonna lurk....

Thanks again,

Tom

[Telstar]

Originally Posted by regal

Believe me if Ian makes a board that plugs into a stereo pair of PCM1704 allows NOS and 352.8 with fifo crystal level jitter he will have a gold mine, there is little money in integrating these budget S-D chips because there are several other options out there (not quite as good as Ian's but damn close.)

4 dacs at least if we wanna really state of the art. Otherwise i know a couple commercial/kits that fir the bill niceky (i own one of them).

[andrea_mori]

Originally Posted by iancanada

Hi KlipshKid,

Sorry for the late reply. FIFO output works for your 1543, it's I2S with 5V logic tolerance. The only problem is you have 16 of them in parallel.I don't think 74HC125 is a good choice for the low jitter application. My suggestion is using dedicated low jitter clock fanout buffers rather than HC125, at least for the word clock and bit clock. You may find some candidates with more than 4 outputs from digikey or mouser. Good luck to your project.

.. eg 74LS241 or 74ABT541 or so

Andrea