|
|
Post by audio08 on Sept 24, 2014 3:53:48 GMT
Hi, this is the first time I post in this forum.
Reading some old discussions in other forums on net, it might be useful to tweak the cs8414 loop filter R, C (Pin 20 - FILT)to reduce gitter by reducing bandwidth of PLL and thus to change the sound quality.
Original values of R, C on circuit board are 1K and 47n - this is for cs8412 on data sheet. For cs8414, the values are 0.068uF and 470R.
The proposal from various discussions is to reduce the value of R by half (from specification) and to increase the C by 4 times and solder a 3300pF between PIN 20 (FILT pin)and PIN 21 (AGND pin)with shortest distance. This comes out approx R=270R and C=0.22uF. Presuming it locks, can try various values of R and C, e.g R=220, 240, 270 or 300 and to increase the value of C.
I wonder anybody has tweaked the loop filter of cs8414 in X24K ? Is it useful? It looks simple and non expensive. I have modded most of the X24k according to information in this forum, particularly matt7941's post.
Though X24K is old, it's still good sounding.
Thanks for advice
|
|
|
|
Post by audio08 on Sept 24, 2014 4:02:11 GMT
Sorry, "jitter" not "gitter"
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Sept 24, 2014 4:40:20 GMT
Hi, this is the first time I post in this forum. Reading some old discussions in other forums on net, it might be useful to tweak the cs8414 loop filter R, C (Pin 20 - FILT)to reduce gitter by reducing bandwidth of PLL and thus to change the sound quality. Original values of R, C on circuit board are 1K and 47n - this is for cs8412 on data sheet. For cs8414, the values are 0.068uF and 470R. The proposal from various discussions is to reduce the value of R by half (from specification) and to increase the C by 4 times and solder a 3300pF between PIN 20 (FILT pin)and PIN 21 (AGND pin)with shortest distance. This comes out approx R=270R and C=0.22uF. Presuming it locks, can try various values of R and C, e.g R=220, 240, 270 or 300 and to increase the value of C. I wonder anybody has tweaked the loop filter of cs8414 in X24K ? Is it useful? It looks simple and non expensive. I have modded most of the X24k according to information in this forum, particularly matt7941's post. Though X24K is old, it's still good sounding. Thanks for advice Hi With the DIR 9001, the filter consists of 68nF in series with 680 ohms, and 4.7nF (.0047uF) in parallel.
With the DIR1703 as used in the MF X-DAC V3 the corresponding values are 68nF, 8.2nF and 1.2K
I got great results in a DAC using a DIR 9001, and the X-DAC V3 simply by using close tolerance polypropylene capacitors of the same value.
You could check that the resistor in yours is metal film and 1% tolerance. (or better)
Alex
|
|
|
|
Post by audio08 on Sept 24, 2014 7:16:07 GMT
Hi Alex,
Thanks for your useful information. I have gone through your post on X-DAC V3 Mod. Looks like simple tweaking of filter R C values are useful to improve the sound quality.
Further reading indicates that different digital receivers have different optimal values and also depend on different sampling frequencies used(e.g. 44.1KHz, 96KHz, etc).
There is little information on calculation of optimal filter values for a particular receiver. I found one application note AN159 from Crystal "PLL filter optimization for the CS8415A, CS8420, and CS8427" but not for the older CS8412/8414.
Since it is a simple mod, I will try it out, if not like it, it can easily be reverted.
Thanks again
|
|
|
|
Post by uncle3 on Jul 14, 2017 23:54:35 GMT
Hi All, long time no see.
Sadly report my X-24K went dead suddenly.
I don't have those abilities to check which part is failed.
So sad :-(
|
|
|
|
Post by PinkFloyd on Jul 15, 2017 10:29:58 GMT
Hi All, long time no see. Sadly report my X-24K went dead suddenly. I don't have those abilities to check which part is failed. So sad :-( I would start off looking at these six capacitors...... if any one of them have gone high ESR it will cause a failure:  |
|
|
|
Post by audio08 on Jul 17, 2017 3:03:29 GMT
Hi All, long time no see. Sadly report my X-24K went dead suddenly. I don't have those abilities to check which part is failed. So sad :-( Also check all those 10uf 16V tantulum capacitors (regulators' output) if they have shorted to ground (due to aging.) and also the two 47uf 16V tantulum on the power rails on the digital/analogue board. I have replaced those 10uf tantulum with 10uf 25V Sanyo OSCON cap |
|
|
|
Post by uncle3 on Jul 17, 2017 4:38:07 GMT
Thankyou Guys. I will check that parts first what you guys suggested.
Hope I can bring it alive. Cheers
|
|
|
|
Post by PinkFloyd on Jul 17, 2017 12:13:41 GMT
That should do the trick  |
|
|
|
Post by audio08 on Apr 29, 2020 14:25:03 GMT
When I saw back the post which I posted in 2014, it was nearly 6 years ago, the time flies. My modification for X-24K was long completed. Some mod photos are shown below. I don't know if you are still interested in X-24K, since it is the DAC technology of over 20 years ago, but as I said, it is still very good sounding, particulary when playing Spanish guitar, violin and vocal CD. I am still using PS Audio Lamda II CD transport, feeds into Meridian 518 digital processer whch converts 16 bits to 24 bits and then feeds into X-24K. All the CD equipment are over 20 years old and perform well.
|
|
|
|
Post by audio08 on Apr 29, 2020 15:03:31 GMT
Sorry, failed to attach photos, retry:
|
|
nando
Been here a while! 
The Frying Dutchman
Always look on the bright side of life
Posts: 393 
|
Post by nando on Apr 29, 2020 19:26:00 GMT
I still use my x-24k every day modded and all.
|
|
|
|
Post by audio08 on May 21, 2021 7:57:45 GMT
Add some notes to the above pictures: (you can refer to Matt's thread and this thread for X-24K modification)
1. use 4 x OPA627bp opamps for buffer/amp. Add small heat sinks on OPA627 since they are high slew rate opamps and consume more current and get warm after some time. Since OPA627 has FET input stage and heat sink can help to lower the temp and reduce the increase in input bias current and thus input offset voltage.
2. use 2 x LM4562 for LPF
3. I use the Kimber internal copper wire (red/black) for connecting the upper board and output jacks. Make it a bit longer (say 12") for easy working on upper board
4. I use Mundorf 150uF bipolar e-cap for o/p coupling cap, pairs with 2.2u, 0.1u and 0.01u WIMA MKP. You can try other combinations
5. tweak the cs8414 loop filter R, C (Pin 20 - FILT)to reduce jitter by reducing bandwidth of PLL and thus to change the sound quality. This will reduce the jitter spectrum. I use 270ohm and 0.22uF. You can try a bigger value cap and smaller value resistance so long as 8414 can lock the input digital stream. The small 3300pF cap (green one at bottom of upper board) is to dampen the jitter spectrum at high frequency.
6. fit heat sinks on the voltage regulators since OPA627 and LM4562 draw more quiescent current and become quite warm after long time usage. I use U-shaped heat sink and straighten out the fins. Add some heat sink compound between regulators and heat sinks
7. pair 0.1uF WIMA MKP with the filtering caps (all are 2200uF Panasonic FC) help to reduce the source impedance at high frequency.
8. I use the original X-PSU with X-24K
9. use WIMA MKP for digital input to replace the ceramic cap.
|
|
