Hi,
Does anyone know about how to optimise a circuit for the faster Potato ? I recently swapped a 74VHC14 for the Potato 74G14 and the sound is substantially worse. It is used to distribute the output of a Crystek 24.576Mhz CCHD-957. Pin1 is the clock input and pin2 feeds the other sides inputs. Their outputs feed R1-R3 / C2-4.
Thanks !
Tom
Does anyone know about how to optimise a circuit for the faster Potato ? I recently swapped a 74VHC14 for the Potato 74G14 and the sound is substantially worse. It is used to distribute the output of a Crystek 24.576Mhz CCHD-957. Pin1 is the clock input and pin2 feeds the other sides inputs. Their outputs feed R1-R3 / C2-4.
Thanks !
Tom
Last edited:
Well, I have been using Potatosemi's 74X family for years and found nothing comes close to it. What would be your "substantially worse" sound like?
Sound changes - for example there has been a loss in the urgency/definition/depth in the deep bass, and overall the sound is like added jitter.
I assume I need to change the rc output to get this Potato at its best ? That's what I'm asking.
I assume I need to change the rc output to get this Potato at its best ? That's what I'm asking.
Instead of using multiple logic gates, why don't you use a fanout clock driver? Let me ask you, what is the pin to pin skew for the 74G14? What is the added jitter? Oh, that's right, they're not published because they're not measured or guaranteed. I wonder why. I would think those would be important characteristics for clock distribution. I suggest you look a little deeper into the PotatoSemi catalog, among others.
Maybe you can try PO38072 1:2 clock distributor. I have one to distribute 33.8688Mhz into digital filter chip and PO74G374 for reclocking. The result is great.
By the way, what is your power regulator for clock?
By the way, what is your power regulator for clock?
As it looks as though you're using groundfill for your 0V, the worse sound could be arising from common impedance coupling on the 0V. Is the 0V to the potato chip sharing any ground impedance with the clock?
Hi,
As you can see, I need 3 outputs, and pin-for-pin compatible. I'm not hacking a high-speed pcb circuit. The clock has its own regulator, and that area is a common ground for the digital circuit.
So if changing the RC output isn't going to resolve the problem, I'll go back to the 74vhc14. I was just hoping a faster chip would improve, not degrade, the signal. I assume from your answers that the RC filter is not the problem. Besides, I can only listen to so much Massive Attack...
If you want more info about the DAC, its here :
http://www.diyaudio.com/forums/digital-line-level/224013-384khz-dac-taobao.html#post3250142
Thanks.
As you can see, I need 3 outputs, and pin-for-pin compatible. I'm not hacking a high-speed pcb circuit. The clock has its own regulator, and that area is a common ground for the digital circuit.
So if changing the RC output isn't going to resolve the problem, I'll go back to the 74vhc14. I was just hoping a faster chip would improve, not degrade, the signal. I assume from your answers that the RC filter is not the problem. Besides, I can only listen to so much Massive Attack...
If you want more info about the DAC, its here :
http://www.diyaudio.com/forums/digital-line-level/224013-384khz-dac-taobao.html#post3250142
Thanks.
Last edited:
Higher speed CMOS logic has, IME, always given rise to PSU noise problems. That's why these days I go for the slowest possible one (usually HC, very occasionally 4000 series) and reduce the supply voltage to 2.5V.
Based on my experience with AD1955 DACs from Taobao, if you don't hack the grounding you'll miss out on the best possible sound.
Based on my experience with AD1955 DACs from Taobao, if you don't hack the grounding you'll miss out on the best possible sound.
Last edited:
Well, I wanted to try.... and now I'm a believer... slower is sometimes better.
This DAC is not your average Taobao crap. It sounds as good as the ES9018, in everything except dynamics, but it has taken quite a lot of my spare time to make it that way. The seller built it for me w/o the pcm2706 so I can use isolated cm6631.
USB SPDIF PCM2706Ë«AD1955 384kHz ASRC ÉýƵ DAC ½âÂëÆ÷ Ì×¼þ-ÌÔ±¦Íø
This DAC is not your average Taobao crap. It sounds as good as the ES9018, in everything except dynamics, but it has taken quite a lot of my spare time to make it that way. The seller built it for me w/o the pcm2706 so I can use isolated cm6631.
USB SPDIF PCM2706Ë«AD1955 384kHz ASRC ÉýƵ DAC ½âÂëÆ÷ Ì×¼þ-ÌÔ±¦Íø
Last edited:
I agree its very unusual to see an FPGA in a Taobao DAC. But THD+N at -140dB has to be some kind of a joke 😀
Oh yeah, that's like Musiland saying they have -140dB .... and in small print... digital only.
Last edited:
Hi,
I emailed Potato about this. Goodness me they are helpful and fast. They said it is designed to drive without the RC filter, and wanted to know why I needed it and what values I'm using. Great help eh ?
cheers
I emailed Potato about this. Goodness me they are helpful and fast. They said it is designed to drive without the RC filter, and wanted to know why I needed it and what values I'm using. Great help eh ?
cheers
74??14 is inapropriate for low jitter clock distribution. Actually, it is really wrong.
It's better to go without 14s.
There are buffers and clock distributors for such task.
Schmitt inputs have their niche in digital, where jitter doesn't matters.
It's better to go without 14s.
There are buffers and clock distributors for such task.
Schmitt inputs have their niche in digital, where jitter doesn't matters.
Last edited:
Yeah I agree - Schmitt triggers can only make things worse because of their built-in hysteresis. I recall Jocko Homo talking about this...
Can the CCHD-957 drive three lines, each about 5cm-7cm, into an Altera Cyclone III fpga ? Too much pF right ?
If anyone knows a pin-for-pin replacement for the 74_14 that'll work better, please do tell.
However, a rant if I may : simply saying a buffer is better is not helpful. Yes, it is correct but I didn't ask for a lesson in design. I'm thoroughly fed up with people answering questions that didn't get asked.
You know what ? I think I'm really going to pull the plug on posting here. What's the point ?
If anyone knows a pin-for-pin replacement for the 74_14 that'll work better, please do tell.
However, a rant if I may : simply saying a buffer is better is not helpful. Yes, it is correct but I didn't ask for a lesson in design. I'm thoroughly fed up with people answering questions that didn't get asked.
You know what ? I think I'm really going to pull the plug on posting here. What's the point ?
Last edited:
No, you need clock fanout or make the thing paralel terminated with single conductor going thu all of these pins and stuck in the terminating resistor, if you want it done propertly.
Otherwise - yes, it can, without any sort of termination etc. Actually you can save on the clock this way and use some of 0.5$ oscillators or SPDIF.
Otherwise - yes, it can, without any sort of termination etc. Actually you can save on the clock this way and use some of 0.5$ oscillators or SPDIF.
Wrong:
___/-----------[ic pin1]
[R]------------[ic pin2]
___\-----------[ic pin3]
Semi-wrong:
/[R]-----------[ic pin1]
-[R]-----------[ic pin2]
\[R]-----------[ic pin3]
Fine:
/[buf]-[R]-----------[ic pin1]
-[buf]-[R]-----------[ic pin2]
\[buf]-[R]-----------[ic pin3]
Better:
[clock fanout]-[R]-----------[ic pin1]
[clock fanout]-[R]-----------[ic pin2]
[clock fanout]-[R]-----------[ic pin3]
^^^^^
[clock fanout in]
Okay:
[buf]-------------------[ic pin1]--------[ic pin2]------------------------[ic pin3]-[R]-[GND]
___/-----------[ic pin1]
[R]------------[ic pin2]
___\-----------[ic pin3]
Semi-wrong:
/[R]-----------[ic pin1]
-[R]-----------[ic pin2]
\[R]-----------[ic pin3]
Fine:
/[buf]-[R]-----------[ic pin1]
-[buf]-[R]-----------[ic pin2]
\[buf]-[R]-----------[ic pin3]
Better:
[clock fanout]-[R]-----------[ic pin1]
[clock fanout]-[R]-----------[ic pin2]
[clock fanout]-[R]-----------[ic pin3]
^^^^^
[clock fanout in]
Okay:
[buf]-------------------[ic pin1]--------[ic pin2]------------------------[ic pin3]-[R]-[GND]
74_04 should do the trick - the same part but without the hysteresis.
I advised against a part with hysteresis, seems I could have also advised against getting hysterical - I mean its just a hobby, what's the point?
- Status
- Not open for further replies.