Hi ChuckT,
I wouldn’t change the caps in the PSU unless you know what your doing – SMPS require Low ESR caps, and changing types / values could upset the Feedback loop.
However the caps on the Amplifier board could benefit from improvement.
1. Increase C313 (main HBridge supply) leave the existing cap on the PCB, but add an external larger part (can be fitted across the connecter CN300
2. Fit OSCON’s in positions, C312, C301, C326, C336, C350, C388, C392, C325, C320, C329
There’s a guy selling bags of OSCON caps on a stand in Sham Shui Po (electronics market) for HK$1 each!
Buy the 470uF 6V3 and fit in the above position (where the voltage rating will allow).
The PCB is very poorly design (very poor ground plane & terrible HF decoupling), but the above modifications will improve the audio quality.
The S-master modulator can achieve a dynamic range of just shy of –120dB; it’s not a bad design.
We should go for drink one evening in Wan Chai,
John
I wouldn’t change the caps in the PSU unless you know what your doing – SMPS require Low ESR caps, and changing types / values could upset the Feedback loop.
However the caps on the Amplifier board could benefit from improvement.
1. Increase C313 (main HBridge supply) leave the existing cap on the PCB, but add an external larger part (can be fitted across the connecter CN300
2. Fit OSCON’s in positions, C312, C301, C326, C336, C350, C388, C392, C325, C320, C329
There’s a guy selling bags of OSCON caps on a stand in Sham Shui Po (electronics market) for HK$1 each!
Buy the 470uF 6V3 and fit in the above position (where the voltage rating will allow).
The PCB is very poorly design (very poor ground plane & terrible HF decoupling), but the above modifications will improve the audio quality.
The S-master modulator can achieve a dynamic range of just shy of –120dB; it’s not a bad design.
We should go for drink one evening in Wan Chai,
John
Wow, that was a quick reply. And from HK too.
Thanks, I know it is not wise to mod a SMPS. I normally just replace the last cap in the output.
Do you have a schematics?
Yes, I am alway happy to meet a local fellow moder.
PM me and we'll discuss.
I have a lot of Oscon in my box, so really I can start immediately.
What values do you recommend to replay the 100uf/10V.
Actually, I just got this gear yesterday.
Thanks, I know it is not wise to mod a SMPS. I normally just replace the last cap in the output.
Do you have a schematics?
Yes, I am alway happy to meet a local fellow moder.
PM me and we'll discuss.
I have a lot of Oscon in my box, so really I can start immediately.
What values do you recommend to replay the 100uf/10V.
Actually, I just got this gear yesterday.
Hi ChuckT,
No problem to fit as large (Oscon) cap as you can, although there physcial size will be the limiting factor. The LV regulators are very noisy (on the Clock and digital supplies around the modulators), the larger caps well help reduce this.
The output devices are "badged" Sankan devices, and the modulators are Fab. by Renisis (Ex Mitsubishi).
I'm sorry I don't have the circuit diagram, I bought a unit to Eval. the modulator (Sony DAV-S400, which appears to be very similar apart from the Video IO / Audio IO / Tuner Bd).
John
No problem to fit as large (Oscon) cap as you can, although there physcial size will be the limiting factor. The LV regulators are very noisy (on the Clock and digital supplies around the modulators), the larger caps well help reduce this.
The output devices are "badged" Sankan devices, and the modulators are Fab. by Renisis (Ex Mitsubishi).
I'm sorry I don't have the circuit diagram, I bought a unit to Eval. the modulator (Sony DAV-S400, which appears to be very similar apart from the Video IO / Audio IO / Tuner Bd).
John
Hi ChuckT,
I would leave the small Red / Brown caps around the Output stage – these are the HF!!! Decoupling caps – they should have used better quality (from a HF standpoint) SMD ceramics. But you could try changing the Larger Red / Brown ones next to the output inductors (OPS LPF).
The shielding is to prevent switching noise pick-up into the Clock circuit, and reduce capacitive coupling of the OPS signals into the chassis, my unit has add-on “RF clamp beads” for the speaker cables included, in what appears to be an after thought – there’s no way this unit can pass the CE EMC test - with speakers leads connected and these beads not fitted!!
You should also add large value OSCON's around the Clock circuit and around the other modulator IC’s, to reduce induced noise on the common PSU lines.
John
I would leave the small Red / Brown caps around the Output stage – these are the HF!!! Decoupling caps – they should have used better quality (from a HF standpoint) SMD ceramics. But you could try changing the Larger Red / Brown ones next to the output inductors (OPS LPF).
The shielding is to prevent switching noise pick-up into the Clock circuit, and reduce capacitive coupling of the OPS signals into the chassis, my unit has add-on “RF clamp beads” for the speaker cables included, in what appears to be an after thought – there’s no way this unit can pass the CE EMC test - with speakers leads connected and these beads not fitted!!
You should also add large value OSCON's around the Clock circuit and around the other modulator IC’s, to reduce induced noise on the common PSU lines.
John
I see, no wonder the shield under the clock is soldered at the four corner to the ground pin of the caps. It will be a pain to replace those caps.
If ceramic is better then those red caps in that application, I can easily add some smd ceramic bypass under those red caps. Maybe 4.7nf NPO and a 0.1uf
Do you know if I can find datasheet for the cxd9646?
Note that this digi amp has no fan but use large heat sink, which is good.
If ceramic is better then those red caps in that application, I can easily add some smd ceramic bypass under those red caps. Maybe 4.7nf NPO and a 0.1uf
Do you know if I can find datasheet for the cxd9646?
Note that this digi amp has no fan but use large heat sink, which is good.
Hi ChuckT,
The OPS CXD9646 is a Sankan device - however, I could not find data on it, from Sony or Sankan.
The Heatsinks on my unit run very hot - even when the units not playing music - suggesting large cross conduction / slow output stage.
I like the way Sony "stacked" the OPS inductors, in an attempt to cancel their magnetic fields,
Cheers,
John
The OPS CXD9646 is a Sankan device - however, I could not find data on it, from Sony or Sankan.
The Heatsinks on my unit run very hot - even when the units not playing music - suggesting large cross conduction / slow output stage.
I like the way Sony "stacked" the OPS inductors, in an attempt to cancel their magnetic fields,
Cheers,
John
Hi Chuck,
I never listened to my unit, a friend has a unit but only connected to the small home cinema speakers so I can't say - but he has a fair amount of back ground noise from his speakers...
I believe the CXD9634Q (which is the Mitsubishi / Renesas M65817AFP) is first Gen, while the newer Renesas M65818AFP, is claimed to be Second Gen.
John
I never listened to my unit, a friend has a unit but only connected to the small home cinema speakers so I can't say - but he has a fair amount of back ground noise from his speakers...
I believe the CXD9634Q (which is the Mitsubishi / Renesas M65817AFP) is first Gen, while the newer Renesas M65818AFP, is claimed to be Second Gen.
John
The S-master has 3 generation.
I think the M65818AFP is the 3rd gen S-Master Pro.
Not sure of the difference between 1 and 2nd that uses the M65817.
I have put in a 2.2uf ceramic for the power chip and a 2200uf BG FK, it is going to take a long time to burn.
BTW, yes the power chip does get very very hot, after a 24hr run.
Do u know how the volumn control is done, any tweak there?
How about the relay, is it wise to bypass it.
Too many questions, I know.
I think the M65818AFP is the 3rd gen S-Master Pro.
Not sure of the difference between 1 and 2nd that uses the M65817.
I have put in a 2.2uf ceramic for the power chip and a 2200uf BG FK, it is going to take a long time to burn.
BTW, yes the power chip does get very very hot, after a 24hr run.
Do u know how the volumn control is done, any tweak there?
How about the relay, is it wise to bypass it.
Too many questions, I know.
Chuck,
I am interested in tapping into pin 8 of chip CXD9646 for the purpose of creating a PRE-OUT for each channel of a Sony receiver which uses these chips.
I believe that each CXD9646 is a dual analog power amp and the two channels are driven with the same signal (inverted and non-inverted) so that this chip operates as a briged mono amp.
I would even suggest that the speaker binding posts are purposely isolated from the chassis ground as a consequence.
Do you have any thoughts about what I am proposing to do.
My main worry is how well this chip (and the CXD9634 upstream from it) will behave after I introduce a new additional load of 10k ohms between pin 8 and ground due to a connection with the external power amp that will be driven off this signal.
By the way, is the AVD-S10 a DVD player?, a HT receiver?, or both in the same box?
I am interested in tapping into pin 8 of chip CXD9646 for the purpose of creating a PRE-OUT for each channel of a Sony receiver which uses these chips.
I believe that each CXD9646 is a dual analog power amp and the two channels are driven with the same signal (inverted and non-inverted) so that this chip operates as a briged mono amp.
I would even suggest that the speaker binding posts are purposely isolated from the chassis ground as a consequence.
Do you have any thoughts about what I am proposing to do.
My main worry is how well this chip (and the CXD9634 upstream from it) will behave after I introduce a new additional load of 10k ohms between pin 8 and ground due to a connection with the external power amp that will be driven off this signal.
By the way, is the AVD-S10 a DVD player?, a HT receiver?, or both in the same box?
Sorry, I need to correct the previous post.
I made the claim that the CXD9646 chip is a dual analog power amp chip which is normally driven by a stream processor (CXD9634 or equivalent). I incorrectly assumed that the stream processor outputted analog.
It now appears to me that this stream processor outputs using PWM (pulse width modulation) and this is fed directly to the inputs of the CXD9646. In all probability the CXD9646 is a switching amplifier which only changes the voltage level of the pulses but not the duty cycle. Thus any CXD9646 would have to be followed by a LC filter to create the final analog signal that is fed to the speakers.
It is not clear to me how the dual switching amps would be configured in a bridge arrangement so that when one amp is fed with pulses from the inverted output of the upstream processor and the other amp is fed with pulses from the non-inverted output of the upstream processor, both amps together produce outputs in parallel which effectively multiply the total output power.
In order to create the PRE-OUT analog signal that I need for my particular application, I would have to convert the low level variable pulses used as inputs to the CXD9646 into pure analog signals that can be fed to my external power amps.
I made the claim that the CXD9646 chip is a dual analog power amp chip which is normally driven by a stream processor (CXD9634 or equivalent). I incorrectly assumed that the stream processor outputted analog.
It now appears to me that this stream processor outputs using PWM (pulse width modulation) and this is fed directly to the inputs of the CXD9646. In all probability the CXD9646 is a switching amplifier which only changes the voltage level of the pulses but not the duty cycle. Thus any CXD9646 would have to be followed by a LC filter to create the final analog signal that is fed to the speakers.
It is not clear to me how the dual switching amps would be configured in a bridge arrangement so that when one amp is fed with pulses from the inverted output of the upstream processor and the other amp is fed with pulses from the non-inverted output of the upstream processor, both amps together produce outputs in parallel which effectively multiply the total output power.
In order to create the PRE-OUT analog signal that I need for my particular application, I would have to convert the low level variable pulses used as inputs to the CXD9646 into pure analog signals that can be fed to my external power amps.
Hi Montreal,
been there once long time ago.
The S10 is an all-in-one, the dvd part is not so good.
Not sure why you want use your receiver as a pre, I don't have the data sheet so I can't say. John have a much better understanding of tech behind it, I just tweak with passive parts.
The Sony does sound pretty good, but I move on to the Tripath which is more musical chip.
been there once long time ago.
The S10 is an all-in-one, the dvd part is not so good.
Not sure why you want use your receiver as a pre, I don't have the data sheet so I can't say. John have a much better understanding of tech behind it, I just tweak with passive parts.
The Sony does sound pretty good, but I move on to the Tripath which is more musical chip.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.