Hypex Ncore

Status
Not open for further replies.
Usually speaking with screwed together anodised cabinets electrical contact will be made between the different panels anyway so it's unlikely that removing a layer of the oxide would have done anything. Especially if tapped threads were used anywhere.

Thanks - I agree, so I guess we both also agree that it is rather unlikely that there were real audible differences in jojip's scenario.
 
guys, as mentioned in my post, i had suspected as much. Before making any changes, I had put a multimeter to check continuity between the panels. most panels did showed connectivity due to the tapped screw holes. Some panels did not. Anyway i wasn't expecting it to improve sounds quality in anyway.

Further i was more curious to know opinion on what observable effects such subtle factors in assembling modules had on sound. Gross errors are clearly going to cause the modules to under perform, so directions in documentation have to be followed rather diligently.
 
Nothing can be taken for granted as being less critical to the success of the implementation. Often wiring up the connectors and PSU can be more critical with modules as it is quite easy to destroy a products excellent objective performance by putting wires in the wrong place.

This is especially true with amplifiers where you have to deal with both low and high impedances, high and low signal levels and high and low current signals and their associated return paths.

So just when you thought nothing could go wrong hehe.

Matt, from experience do you have comments on what observable effects are? I agree that errors leading to hums, noise etc are noticeable. What about the subtle differences in ground/return path impedance, effectiveness of shielding, gauge of wires, speaker wire connecting to module forming slight antenna loops etc any others?
Can THD increase ever so slightly? and as Julf had mentioned that is unlikely to be noticed.

In any case, i agree that every effort must be made to do a high quality implementation to get the best and reliable operation out of these excellent products.
 
Matt, from experience do you have comments on what observable effects are? I agree that errors leading to hums, noise etc are noticeable. What about the subtle differences in ground/return path impedance, effectiveness of shielding, gauge of wires, speaker wire connecting to module forming slight antenna loops etc any others?
Can THD increase ever so slightly? and as Julf had mentioned that is unlikely to be noticed.

In any case, i agree that every effort must be made to do a high quality implementation to get the best and reliable operation out of these excellent products.

Usually you need a spectrum analyser to see if you're doing things the right way. Doing things the wrong way can destroy your linearity 10-100x depending on how linear things are to start with and coupling high voltage high current signals into high impedance low signal level inputs is a really easy way to do this.

I am sure that Bruno has made all efforts to ensure that that doesn't happen with his boards and their associated user guides but he cannot control everything.
 
No absolutely not. If the ncore can reach 0.0003% distortion midband, you could easily push that between 0.003 and 0.03%. with bad coupling into the input stage.

Not that 0.03% would be audible anyway, but I would appreciate your explanation as to the mechanism behind that. The nc400 is voltage-driven, so as long as the voltage levels are kept at reasonable levels, the nc400 doesn't really care about the coupling. The driving stage of the source is another matter...
 
Not that 0.03% would be audible anyway, but I would appreciate your explanation as to the mechanism behind that. The nc400 is voltage-driven, so as long as the voltage levels are kept at reasonable levels, the nc400 doesn't really care about the coupling. The driving stage of the source is another matter...

Yes it's voltage driven but you only need a tiny voltage to be developed around any of the input circuitry for linearity to start being degraded. The high currents passed down the power and output cables generate magnetic fields that could easily be coupled into the high impedance input stage if misrouted.
 
Or if that isn't possible, add a volume pot + buffer stage to the DAC output.

Thanks, I think I'll follow that recommendation as a convenient interim solution until my DAC with built-in volume control arrives later this year. (Hope this pot+buffer stage doesn't degrade audio quality too much)

For the buffer part, I have a Diamond Buffer around that I'd like to use (this one). Yet it has an output impedance of only 2.5 ohms, whereas NC400 datasheet states that the recommended source impedance (the source being the diamond buffer in this case) should be 1kohm or greater. Walt Jung in his paper on buffers says here that "there should almost always be some sort of isolation impedance between the buffer output and the output terminal, to isolate any cable capacitance". Also Sjöström says on the product page "If your load is unknown you may add a resistor at the output, 220 ohms or higher".

So would I be fine to just add say a 1kohm resistor between the buffer's output and the RCA terminal (the buffer isn't balanced) to which the cable that goes to the NC400 connects?
 
Last edited:
For the buffer part, I have a Diamond Buffer around that I'd like to use (this one). Yet it has an output impedance of only 2.5 ohms, whereas NC400 datasheet states that the recommended source impedance (the source being the diamond buffer in this case) should be 1kohm or greater. Walt Jung in his paper on buffers says here that "there should almost always be some sort of isolation impedance between the buffer output and the output terminal, to isolate any cable capacitance". Also Sjöström says on the product page "If your load is unknown you may add a resistor at the output, 220 ohms or higher".

So would I be fine to just add say a 1kohm resistor between the buffer's output and the RCA terminal (the buffer isn't balanced) to which the cable that goes to the NC400 connects?

1 kohm sounds rather high driving the cable - what I would do is place a 220 ohm resistor in parallel with the output (before the cable) and a 1 kohm resistor in series with the nc400 input (after the cable).
 
Ah, I see, thanks - this way the buffer output drives the cable with 220 ohm, jet the NC400 'sees' a source impedance of ~1.2kohm with the resistor added to the amp's input terminal.

Since the input will be balanced XLR in configuration as per the NC400 datasheet, I assume the place to put this 1 kohm resistor is pin 2, correct?
 
Ah, I see, thanks - this way the buffer output drives the cable with 220 ohm, jet the NC400 'sees' a source impedance of ~1.2kohm with the resistor added to the amp's input terminal.

Right (except the nc400 sees 1k, as the 220 ohm resistor is in parallel with the output impedance of the buffer).

Since the input will be balanced XLR in configuration as per the NC400 datasheet, I assume the place to put this 1 kohm resistor is pin 2, correct?

Is your cable balanced or not? If so, I would actually put 1k resistors on both pin 2 and pin 3, to keep the situation symmetrical.
 
Is your cable balanced or not? If so, I would actually put 1k resistors on both pin 2 and pin 3, to keep the situation symmetrical.

The cable is the one that is recommended in the NC400 datasheet to connect unbalanced / RCA source with balanced / XLR NC400. You're right of course, pin 1 is shield while both 2 and 3 compose the signal, so there need to be resistors on both.

Would you recommend any specific quality, material or build for that resistor?
 
Status
Not open for further replies.