Just looking at the unfilled capacitor locations highlighted in blue, I wonder if a 0.01uf-1uf film might help with additional local voltage support or taking a minor amount of power supply noise off the line, these are on the + and - 12V supply lines to the Opamps, worth experimenting here as well ?
The light blue dotted locations are four -12V supplies at the top the four +12V supplies at the bottom
The light blue dotted locations are four -12V supplies at the top the four +12V supplies at the bottom
That are alternative places for through holes (film) decouling caps which it are placed on the +/- rails of the opamps. It was designed so the board to make possible the mounting of through holes caps if the production line may not have available the SMD components. As it is this version of the audio board, the decoupling caps are already planted (very close to the opamp chips), using SMD components. The available through holes places it may be used to mount electrolytic caps to improve the decoupling performances, as the response in low frequencies for the final opamps.
Which it may be the positive and the negative rails, should be identified (measured) properly before planting there the electrolytic caps, as I do not have myself this information. I did not pay particularly attention to this detail, as my decision was from the beginning of using my final analogue module, insted of the original design.
Which it may be the positive and the negative rails, should be identified (measured) properly before planting there the electrolytic caps, as I do not have myself this information. I did not pay particularly attention to this detail, as my decision was from the beginning of using my final analogue module, insted of the original design.
I checked the output pins on the 2114 OpAmps and the DC offset is 0.001 V on pin 8 A output and 0.002-0.003 V on pin 7 B output. So pretty happy DC offset wont be an issue as the input has AC coupling caps, I might just check if the pre-amp input is Zero as well
As usually the preamps should have AC coupling caps on its inputs. So, no worries there...
Just checked and there is 0v DC on the input, no leakage from the 12v that is on the signal line (a naim thing, no -+ 12v but +12v and +24v)
I think it may be an measurement error, as 24v is not possible to be found in this device, except exclusively the front panel itself, where an small such high voltage SMPS is powering the front panel display. This voltage is created locally, for the use only on front panel board.
Last edited:
Coris, hi no its the 24V supply in the naim pre I was checking and buffer board that connects to the RCA of the 9000, I was checking as the Opamp wasn't seeing any leakage the other way as well, measuring at IN1 there is as expected 0V DC. The signal line is +12V through the Pre-amp and across the 10uf input cap
All right. Clarified. I was thinking you refer to something measured (+24v) in the player... 🙂
Interesting, I have been having a look at the small analogue board and how the + and - 12V supplies are distributed across the board and it looks like 6 (of the 10) power supply connector pins are voltage in with three pins joined at the incoming connector feeding + 12v and three feeding -12v. 4 of these pins then feed the Stereo and larger 2114 Op amps but are not connected to the single +12V and -12V rail feeding the multichannel opamps on the board itself and joined at the multiple connector only, having a look at the original supply I think the +12V and -12v regulated supplies feed these two pairs of pins directly but looks like these regulated supples also feed the single pin supplies to the multichannel using another un marked IC's board label Q14002 and Q14001, so I think these are fed by the same regulated supply to these IC's but have a less direct link on the original SMPS
Anyway as I only use the multichannel Opamps and the LPS I am wondering if I can reduce the local supply impedance by bridging on the analogue board for the +12v and -12V feeds. To avoid the board mounted caps being further away from the multichannel Opamps than they are (track to the power connector then track to the board mounted caps) so perhaps a journey of 100mm+ to caps that sit only 30-50mm away
Green +-12V supplies and cap locations for the stereo out
Blue +-12V supplies and cap locations for the multichannel
Green and Blue supplies only connect at the white connector when this is connected to the LPS
Anyway as I only use the multichannel Opamps and the LPS I am wondering if I can reduce the local supply impedance by bridging on the analogue board for the +12v and -12V feeds. To avoid the board mounted caps being further away from the multichannel Opamps than they are (track to the power connector then track to the board mounted caps) so perhaps a journey of 100mm+ to caps that sit only 30-50mm away
Green +-12V supplies and cap locations for the stereo out
Blue +-12V supplies and cap locations for the multichannel
Green and Blue supplies only connect at the white connector when this is connected to the LPS
The best way of connecting for decoupling caps is the shortest one, as close as possible to the decoupled chip pins. Using wires 100mm long or similar, is not recommended, when the purpose is improving performances.
Coris, thanks, this is the schematic of the wiring as fas as I can tell, the only place the green is connected to the blue is when the board is mounted and the power connecter fitted. on the bench as shown below no connections between blue and green 12V supplies.
By the way I have fitted some 10uf 35V tantalums in the blue locations which made a noticeable improvement I can connect to the back of the spare green holes to the back of the tantalums with some isolated 0.6mm copper solid core cable
By the way I have fitted some 10uf 35V tantalums in the blue locations which made a noticeable improvement I can connect to the back of the spare green holes to the back of the tantalums with some isolated 0.6mm copper solid core cable
Perhaps the orange (in your picture) connections are already done by the board traces. If not the case, then (my opinion), I will trust the original designers, that it is better so how they did it... However, the big improvement it come (in this case) from the better caps, its values and appropriate mounting, for the decoupling of the opamps power rails/pins. Bigger caps values it make indeed the difference, and my appreciation is that this it may be enough for this section of the analogue board.
Coris, thanks, with the board disconnected there is no connection between the power to the stereo Opamps and the multichannel Opamps.
With the LPS connected supplies to the stereo and multichannel are connected, but only by the LPS connector where all the 3 pins for +12V the 3 pins for the -12V from your LPS are joined.
On the SMPS the two +12V and the two -12V that feed the stereo Opamps are highlighted below and the in blue and are direct from the regulators. These outputs go via the yellow IC's to the single +12V and - 12V that feed the multichannel. I I have no clue what the Q14002 and Q14001 are as no markings visible on the case
Doesn't make any sense to me unless they wanted the stereo to perhaps be powered up independently for some reason ? all the Opamps are MJM 2114, juts larger format for stereo
With the LPS connected supplies to the stereo and multichannel are connected, but only by the LPS connector where all the 3 pins for +12V the 3 pins for the -12V from your LPS are joined.
On the SMPS the two +12V and the two -12V that feed the stereo Opamps are highlighted below and the in blue and are direct from the regulators. These outputs go via the yellow IC's to the single +12V and - 12V that feed the multichannel. I I have no clue what the Q14002 and Q14001 are as no markings visible on the case
Doesn't make any sense to me unless they wanted the stereo to perhaps be powered up independently for some reason ? all the Opamps are MJM 2114, juts larger format for stereo
Yes indeed, they designed power lines separately for MC section and stereo section. The user can chose in the Settings menu to shot down the MC section and use only stereo, or use both sections. This switching is for the digital stream. When this switching is made, the accordingly power rails are shot down as well. This power management is controlled by the responsable chip/processor, and is a part of the power management of the system. As it is not possible to intervene from outside in this power management (controlled by firmware), I have chosen to power both MC and stereo sections from LPS (unified power lines), and so no power switching it occur when the MC is dismissed in Settings menu, while the LPS is installed. As the power lines are unified already when LPS in place, it may not make much sense to modify in this respect the audio board, by unifying it again, once more (the orange lines in your picture above). Improving the decoupling caps it is enough to get a significative quality improvement for analogue audio section/board.
I did a bit of experimentation for the mulichannal opamp supplies with the unpopulated cap locations and the best small caps I tried were 10uf Tantalum 35V and I also preferred the link to the 0.01uf bypass caps over the main cap locations (pics below) that surround the main XLR opamps
I only had room on the back, there is 7.5mm clearance so a little tight.
The improvements with the tants A-B are improved dynamics and a greater sense of 'scale' and power. Also seems a tad more clarity where you can hear subtle background sounds more easily. Not brighter but clearer and improved transients.
Strangely the lower lower pic impedance connection option has better bass than the top configuration. I lived with the top set up for a while (as this is the way it feeds the stereo/XLR Opamps) but found I noticed a small loss of bass compared to no connections here. It was better in the top end with a little more scale and clarity but somehow lost a small amount of bass weight. So I tried another low impedance option and this is better everywhere than no connection.
I only had room on the back, there is 7.5mm clearance so a little tight.
The improvements with the tants A-B are improved dynamics and a greater sense of 'scale' and power. Also seems a tad more clarity where you can hear subtle background sounds more easily. Not brighter but clearer and improved transients.
Strangely the lower lower pic impedance connection option has better bass than the top configuration. I lived with the top set up for a while (as this is the way it feeds the stereo/XLR Opamps) but found I noticed a small loss of bass compared to no connections here. It was better in the top end with a little more scale and clarity but somehow lost a small amount of bass weight. So I tried another low impedance option and this is better everywhere than no connection.
Authentic cinema now have an LPS conversion as part of their Signature edition https://www.authentic-cinema.de/mod...player/panasonic-dp-ub9004-signature-edition/
Also a digital transport version
Also a digital transport version
For those interested I did a fair bit more A-B and used the spare board from my spare player and very happy with 10uf Tants and 0.1uf TDK ceramics at the +-12v power pins. I also made the tracks from the local board tracks as low impedance as practically possible. I also removed the stereo Opamps to avoid any additional noise.
I didn't get on with 1.0uf WIMAs, 0.1uf WIMAs, 0.047uf Orange drops (polyester CME) as well or instead of the Tants in these locations. I did like adding 1.0uf Wima at the location of the local additional filter caps
I didn't get on with 1.0uf WIMAs, 0.1uf WIMAs, 0.047uf Orange drops (polyester CME) as well or instead of the Tants in these locations. I did like adding 1.0uf Wima at the location of the local additional filter caps