A little Denon DAC mod

Status
Not open for further replies.
Jump to Latest
#21
denon clock mod

Hi Pmd_audio, I have had my denon 500 for years now, and I read about the mods one can do on this machine.

I would like to start with the clock mod, but it is not really clear t me where to connect the parts to the board. I have seen the photo's of you and of someone else on this site, but the pictures reveals the connection to the board.

Can you help me?

Paul
 
#22
I found this other clock mod with just a diode and a 4.7uf capacitor.
Don't know if it works, but it'll be easy to connect the diode under
the board (between Pin 6 and 21) and the capacitor near the IC.
Regards,
 

Attachments

  • ym3623b.jpg
    ym3623b.jpg
    45.3 KB · Views: 535
#24
All is much worse, misters!

210 said:
The simple clock mod with one diode+cap works well in my modded DA500.
Click to expand...
I recently have got DA-500. Has involved presence of many inputs. Too thought to make updating because assumed, that it{he} will sound certainly worse, than mine Parasound D/AC-1000. But when has compared - the difference has turned me into a shock. At testing it was found out - DA-500 is murderer SP/DIF of a signal! At passage of a signal through DA-500 Parasound sounds approximately so as DA-500! It means, that signal SP/DIF is deformed{distorted} already in entrance circuits-комутаторах DA-500! And updating of hours, a power unit, and other updatings is as at Russian speak " Catching of fleas "! The Reason of a bad sound - bad entrance schemes{plans} and elements.
For the information - at comparison Parasound D/AC-1000 and D/AC-1500 the signal passed{took place} both through 1500, and through 1000 differences not noticeably absolutely!
 

Attachments

  • Фронт.JPG
    Фронт.JPG
    109 KB · Views: 409
  • Пара.jpg
    Пара.jpg
    151.4 KB · Views: 384
#27
Sry for my misunderstandable reply. Here's a more untwisted sentence: The clock mod is there to disable the clock from the entrance circuit when it's not more needed (it's only needed for triggering to an incoming new data stream). Without the mod, the entrance circuit will be disturbed by the (additional) working clock (it's described in this thread before, read it from the beginning!). The entrance circuit (input receiver chip ym3626b) outputs also the spdif stream, so output can be disturbed when the ym3623b gets clock signals when it's already locked to the data stream.
 
Last edited:
#28
The simple clock mod is there to prevent the entrance circuit from clock distortions

210 said:
Sry for my misunderstandable reply. Here's a more understandable version (?): The clock mod is there to disable the external clock from the entrance circuit when it's not more needed (it's only needed for triggering to an incoming new data stream). Without the mod, the entrance circuit will be disturbed by the external clock. The entrance circuit (input receiver chip ym3626b) outputs also the spdif stream, so output can be disturbed when the ym3623b gets external clock signals even if it's already locked to the data stream.
Click to expand...
 
Last edited:
#30
210 said:
Sry for my misunderstandable reply. Here's a more untwisted sentence: The clock mod is there to disable the clock from the entrance circuit when it's not more needed (it's only needed for triggering to an incoming new data stream). Without the mod, the entrance circuit will be disturbed by the (additional) working clock (it's described in this thread before, read it from the beginning!). The entrance circuit (input receiver chip ym3626b) outputs also the spdif stream, so output can be disturbed when the ym3623b gets clock signals when it's already locked to the data stream.
Click to expand...
I do not argue with that. You're absolutely right. I'm talking about something else. I say that, in this device, very poorly made input circuits. And before you do other modifications need to fix it.
 
#31
Looking at SM, the SPDIF-feed-thru does have some strange(?) involvement with the Y'maha-chip and the "selector" chip.
Also, the trafo on output could be eliminated?
Last: too many of the gates of the 74HC04 are used...add another chip/use fewer gates...

Arne K
 
#32
Ah yes, the two crossed SPDIF paths in the schematic diagramm of the DA500 looks like connected. But the SPDIF goes just out of the selector chip, through a 74HCU04 gate and then to the optical out and receiver chip. Hmm, I hadn't problems using the optical out with another DACs (DIR9001 receiver) and thereby didn't measured the signal quality. Could be different/worse when using the SPDIF-Cinch in/outs. The sound quality of the analog output is good. Precise stereo image etc.
 
#33
Cobra2 said:
Looking at SM, the SPDIF-feed-thru does have some strange(?) involvement with the Y'maha-chip and the "selector" chip.
Also, the trafo on output could be eliminated?
Last: too many of the gates of the 74HC04 are used...add another chip/use fewer gates...

Arne K
Click to expand...
And the transformer can be removed. The most radical and the right decision - it is to bypass the regular chains to put the signal through the relay is energized by their transistor switches that run from the LEDs.
 
#34
210 said:
Ah yes, the two crossed SPDIF paths in the schematic diagramm of the DA500 looks like connected. But the SPDIF goes just out of the selector chip, through a 74HCU04 gate and then to the optical out and receiver chip. Hmm, I hadn't problems using the optical out with another DACs (DIR9001 receiver) and thereby didn't measured the signal quality. Could be different/worse when using the SPDIF-Cinch in/outs. The sound quality of the analog output is good. Precise stereo image etc.
Click to expand...
That's when you can send a signal directly from the input jack on the receiver, without many "74HCU04", then you will hear in the real analog output of the device. Especially when it is turned off its own clock, as proposed here.
 
#35
Is it possible to hook up a second 74HCU04 to the installed ones (IC 101 and IC102) by using a IC socket and soldering two IC on their pins together ? Maybe the same could be done with the selector chip 74HC4051 (IC 105).
 
#37
nmiljac said:
Why would you do that? It's a no-no with totem pole outputs.
Click to expand...

Well, even if they're used 'piggy-packed' at the same place,
identical driven and switch the same inputs/output signals ?
Maybe their variations are too high to use them in parallel,
even if they are out of the same production batch ?

An alternative, without basically changing the input circuit,
could be using alternative, better chips, but I can't find
supplement types. Do you know some better replacements?
 
#38
IC101 has two gates with unused outputs (Pin 2 and 12).
The board layout shows them connected to Vss via capacitors,
in the schematic diagramm the outputs are unconnected.
Can't remember if they are really connected or not.
Maybe connecting<->disconnecting improves
the signal quality of the entrance circuit ?

Edit: 74VHCU04 has just half 'propagation delay' (3.5ns instead 7ns) and is maybe better,
but I can't find an wordwide ordering source for low quantities (max.5).
 
Last edited:
#39
Just imagine what happens when the fastest gate switches from one level to another while the slowest one is still unchanged. You will have a straight path from positive source to ground - how long you think the bipolar transitors will last?

PS: CMOS ones limit the curent intrinsically, but the waveform will look badly battered.
 
#40
SoNic_real_one said:
Just imagine what happens when the fastest gate switches from one level to another while the slowest one is still unchanged.
Click to expand...

Seems they are really varying too much to use these gates in parallel.

SoNic_real_one said:
PS: CMOS ones limit the curent intrinsically, but the waveform will look badly battered.
Click to expand...

And if they are varying even just a little, they'll work 'jittery'.

Thanks SoNic!
 
Status
Not open for further replies.
Top Bottom