BAT42 are really outstanding, but are limited with regard to the amount of current and max voltage they can sustain. They are obsolete. BAT 49 are close second. Really brilliant; I still use BAT42's from my gold stash whenever possible and safe to do.
RURP3060's are my favourite diodes for the valve heater's supply. Amazing difference between RURP's and the bog standard types...
RURP3060's are my favourite diodes for the valve heater's supply. Amazing difference between RURP's and the bog standard types...
Last edited:
Joe,
Make sure you use test set that actually has decent measurement capability
otherwise the exercise will be somewhat futile.
I'm happy to do them for you, I have a bal in / out PC based set up that
does better than -110dB loop through depending on levels and I'll be tweaking it for better.
This should be easily sufficient for open loop 844 in voltage gain config.
FWIW - if you really want great sound -and- great measurements, discrete
open loop is the way to go and it is absolutely possible to get close to the
9018's THD and DR limits.
It does take some lateral thinking though
T
I have now tried Hovland Supercap 200v 1uF in the single ended outputs of the 105, and my experiences are different than those reprted here with Jantzen caps before. (I don't seem to be able to reach the page that contained that entry, sorry). I hear a definite increase in refinement, a richer midrange and slightly better resolution. The top-end is cleaner and has better clarity. With me, every recording sounds better, even poorer ones, and the sound in general is easier on the ear. In comparison, the Silmics sound a bit harsh and slightly rolled off.
Another thing. I have identified 4 of the rectifier diodes on the 2-channel board as 1N4006 (hopefully that's correct) but I cannot read the print on the other 4. Does anybode know what they are?
Thanks,
Maarten
Another thing. I have identified 4 of the rectifier diodes on the 2-channel board as 1N4006 (hopefully that's correct) but I cannot read the print on the other 4. Does anybode know what they are?
Thanks,
Maarten
OK, another update. I have now also replaced the output caps on the balanced outputs by Hovland supercaps, and took the opportunity to replace the diodes on the stereo board by Cree SiC ones. Another improvement, I have to say. The diodes bring more authority and a very quiet background, especially in complicated passages. All together, my OPPO now sounds a lot better than before, with quite minimal upgrades. The increase in refinement is very worthwhile. Definitely something I can recommend.
Yes, I know, but I review audio and do not want to take any risks with loaner power amps or loudspeakers. I also do not want to change the design of the player and have to limit myself to upgrading peripheral parts like output caps and diodes.
Still, I was surprised at the effect of these relatively basic upgrades.
Still, I was surprised at the effect of these relatively basic upgrades.
Hi Terry
You don't know what I have here?
Something that is pretty good.
Cheers, Joe
Hi Terry
You don't know what I have here?
Something that is pretty good.
Cheers, Joe
So what is it Joe?
A very simple output stage can be made using the OPA1632. Simply remove all the opamps (this is for the two channel board only) and mount a pair of 1632s using the pads that have the funky LME49724s in them (outputs for the balanced outs). The 49724s and 1632s have the same pin out (they are both differential amps). You simply hook up the outputs of the DACs to pins one and 8 and use a resistor on each output leg back to the input with a cap across it for a single pole filter and then a single small resistor on the outputs directly to the output jacks. If you are just using just one of the outputs of the DAC then you will need a 1.8K resistor for feedback (for 2.2V out) and I like to use a 1000pf cap across it. Then a 47 ohm output resistor. You need to do this to both phases (you have to have both inputs hooked up and both feedback resistors hooked up for the part to work). You can use just the unbalanced outs or balanced....in either case there is no DC offset. Pin 2 on the part sets the output offset and it is grounded in this instance so you get no offset (well a few millivolts).
This output stage has simply one great opamp and one resistor in series with the signal. This is way better than the stock output stage. I have tried the 49724s as the IV diff amp but it sounds way bad....please use the mucho better 1632. Obviously, the quality of the feedback and output resistors are important as well as the feedback cap. Flavor any way you like.
This output stage has simply one great opamp and one resistor in series with the signal. This is way better than the stock output stage. I have tried the 49724s as the IV diff amp but it sounds way bad....please use the mucho better 1632. Obviously, the quality of the feedback and output resistors are important as well as the feedback cap. Flavor any way you like.
I've searched the thread without finding something about digital multichannel out, like analog multichannel, but via digital via 8 rca. I know there is made boards for earlier model of Oppo. I'm thinking of using multichannel digital outs for minidsp or the likes...
Like this one: Audiopraise :: Specialist in digital electronics and signal processing
Like this one: Audiopraise :: Specialist in digital electronics and signal processing
A clock debate...
I have/get a dilemma about a clock signal...
So, I think to lunch here a quite theoretical question: How is best (audio point of view) to have the clock signal in a device, square or sine?
A square clock signal it may look more accurate as rise time, and better time definition, but it looks also very noisy (many high level harmonics). A sine signal is less noisy (harmonics), but quite bad rising time, and maybe enough difficult to get out of an modern (HF) oscillator... At least a logical circuit it should work on square clocks... but thinking at the amount of generated noises...
Some comments on this subject...?
I have/get a dilemma about a clock signal...
So, I think to lunch here a quite theoretical question: How is best (audio point of view) to have the clock signal in a device, square or sine?
A square clock signal it may look more accurate as rise time, and better time definition, but it looks also very noisy (many high level harmonics). A sine signal is less noisy (harmonics), but quite bad rising time, and maybe enough difficult to get out of an modern (HF) oscillator... At least a logical circuit it should work on square clocks... but thinking at the amount of generated noises...
Some comments on this subject...?
Hi Coris,
The clock is designed to provide zero crossings at well defined time intervals, and with accurate long-term stability. Therefore the square wave is vastly superior in terms of jitter, being more resistant to applied voltage noise, as you know.
If you're worried about radiated noise, then extra shielding should be the solution. Good high-frequency circuit design and local power supply bypassing also come to mind. Of course, this is a program that you and many of us have been pursuing.
So theoretically, I would say that you want to provide your best clock, and then fix the consequences.
Eric
The clock is designed to provide zero crossings at well defined time intervals, and with accurate long-term stability. Therefore the square wave is vastly superior in terms of jitter, being more resistant to applied voltage noise, as you know.
If you're worried about radiated noise, then extra shielding should be the solution. Good high-frequency circuit design and local power supply bypassing also come to mind. Of course, this is a program that you and many of us have been pursuing.
So theoretically, I would say that you want to provide your best clock, and then fix the consequences.
Eric
Hi Eric,
Thanks for your comment.
I feel the need to precise a little bit more my above thinking...
We have two (or more) black boxes. One is the oscillator (clock generator), the other one (or the rest) is/are the receiver(s) of that clock. The discussion is not what it is happen inside the "clock receiver" black box. This "receiver black box" it will anyway convert that clock signal on its input port to an square clock, to be used further in its circuits, and how Eric says/justified above, inside that device it have to be a square signal.
My dilemma is how is better to be done, when is to put a clock signal in to a "receiver device". There are some oscillators which it output a very fine sine clock frequency, and another ones which it output a square clock signal. Apparently, delivering to a device a sine wave as clock signal, it seems to be less noisy or cleaner signal to be processed further by the receiver device. In opposite, delivering to a device on clock input port, a square clock wave, it seems to induce in the same time some unwanted harmonics which it may (further) disturb the "receivers" inside processes, more "work" it may be necessary for the "receiver device" to clean up the clock signal, and so on.
So, how it may be more beneficial for an audio device/circuit (DAC or whatever audio processor) to get this clock signal on its input: as square or as sine wave?
I do hope now it may be better understood my enunciation of the problem...
Thanks for your comment.
I feel the need to precise a little bit more my above thinking...
We have two (or more) black boxes. One is the oscillator (clock generator), the other one (or the rest) is/are the receiver(s) of that clock. The discussion is not what it is happen inside the "clock receiver" black box. This "receiver black box" it will anyway convert that clock signal on its input port to an square clock, to be used further in its circuits, and how Eric says/justified above, inside that device it have to be a square signal.
My dilemma is how is better to be done, when is to put a clock signal in to a "receiver device". There are some oscillators which it output a very fine sine clock frequency, and another ones which it output a square clock signal. Apparently, delivering to a device a sine wave as clock signal, it seems to be less noisy or cleaner signal to be processed further by the receiver device. In opposite, delivering to a device on clock input port, a square clock wave, it seems to induce in the same time some unwanted harmonics which it may (further) disturb the "receivers" inside processes, more "work" it may be necessary for the "receiver device" to clean up the clock signal, and so on.
So, how it may be more beneficial for an audio device/circuit (DAC or whatever audio processor) to get this clock signal on its input: as square or as sine wave?
I do hope now it may be better understood my enunciation of the problem...
Last edited:
My own view is that sine is better than square for transmitting the clock around a circuit. This is for a couple of reasons beyond the one you mentioned about broadcasting harmonics around.
Firstly since the square clock is very wide bandwidth it requires great attention to the transmission line details - there's a requirement for a constant impedance over a wide bandwidth and reflections must be kept well under control.
Secondly since the square clock is such a wide bandwidth signal, the clock receiver must be similarly wide bandwidth, meaning that its susceptible to noise pickup. This is the corollary of your already stated observation that a square clock radiates more noise - its also more susceptible to noise.
Hence if I were to get really fussy about clock quality, I'd distribute a sine (you're right, you'd probably better start off with an analog rather than digital oscillator). I'd also filter the sine at the receiver with a bandpass filter (probably an LC tank of some kind). I'd also not use a comparator to convert sine into square, rather a limiter kind of circuit which amplifies the sine through a number of stages into clipping.