• Welcome to ASR. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

boXem ARTHUR 2408/N2 Teardown (Class D Amp)

But why not just use a twisted pair or even shielded cable to connect the input PCB with the signal detection circuitry directly to the amplifier module?

Why use a PCB to PCB connector, then a large PCB and then a ribbon cable with two more connectors on each end. I can’t see any benefit from those extra connectors and conductors.
Try it, make a comparison and make your decision what to use. It is not about you seeing the benefits or not.
 
Try it, make a comparison and make your decision what to use. It is not about you seeing the benefits or not.
I wouldn't have questioned if I haven't.

I was simply questioning the rational behind using many different connectors and connections. Isn't it logical to use the most appropriate and the most direct transmission line for any signal transfer? Here we have a PCB, a PCB connector, a ribbon cable and its connectors. I cannot see any constraints for having to use them either. Instead I can see that it is possible to use a single twisted pair ribbon cable to connect the amplifier board with the input sense board. No PCB, and two less connectors. Better connection and cheaper to boot.

I appreciate if you can tell me why you think this is not a better solution or why it is not possible on this instance.
 
Last edited:
I appreciate if you can tell me why you think this is not a better solution or why it is not possible on this instance.

For incorporating various different modules, all Boxem has to do is make up short indiviudual IDC ribbons, instead of custom lengths of shielded cables for each module type. There's a second header next to the one he's using (partially hidden I think).

Consistency in performance with no vagaries and measure/move wires/tie-down/improve/rinse/repeat. Movement of a few mm can make a huge difference with shielded cables at these low levels. He's found the best routes for the lowest noise and locked them in with a PCB.
 
One other good reason to use a PCB to bridge from the ribbon cable is strain relief. Despite improvements when using the Ghent connector, twisted-pair (tp) cables (like Mogami wire) place more force on the ribbon cable connector. This is why you see some implementations taping the connector in place. I suspect a volume producer would hot glue or use something to retain the connector.

Despite the apparent robustness of tp wiring, my experience after building several Hypex amps is that the heavier wire can pull the connector away from the header. Without strain relief, the result is marginal connection that can reduce output and/or introduce hum.

For my next builds, I plan to use the eval board as it adapts the ribbon cable connections to input connectors with locking retention. So, even without needing to repurpose a one case for different size modules, Boxem‘s PCB provides a better system solution in my experience. :cool:
 
Last edited:
I plan to use the eval board as it adapts the ribbon cable connections to input connectors with locking retention.
I came to the same conclusion and I am expecting this EB to be delivered on Monday.
 
I was simply questioning the rational behind using many different connectors and connections. Isn't it logical to use the most appropriate and the most direct transmission line for any signal transfer? Here we have a PCB, a PCB connector, a ribbon cable and its connectors. I cannot see any constraints for having to use them either. Instead I can see that it is possible to use a single twisted pair ribbon cable to connect the amplifier board with the input sense board. No PCB, and two less connectors. Better connection and cheaper to boot.
I think it's correct.
In terms of the minimum number of contacts and the shortest signal path, the best is to connect one TP shielded wire for each channel directly to the input board and the IDC terminal of hypex. (When increasing the distance) If do not like the movement of the cable or the external force of the connection part, fix the cable properly.
The other means is a compromise.

In the case of products, of course, there are other factors that need to be considered besides the simplicity of the audio signal path, and the overall judgment depends on the designer.
The implementation of ARTHUR2408/N2 seems to be one of the optimal solutions.
 
So I go for a drink with friends on the saturday evening, come back home and see a question when checking my emails, decide that it is not a good idea to deal with it while not *exactly* sober, and the day after I see that you guys brilliantly handled it. Love this forum :)
 
Thanks. Probably because I was babbling about wanting to see Luxembourg...

If I had my wish, it would be for a stereo amp, one box instead of two, but looks like your current casework would lead to a mono amp format to fit new amp unit and adequate power supply to deliver up to 40 amps current or 950 watts at 2 ohms.
 
No electronics manufacture experience here, in fact I burn myself every time I get near a soldering iron.

But, if one can install a 500 watt per channel @ 4 ohms with 40 amps current capacity, relatively cool operating, high efficiency, weighs 10-12 kilos, offers the lowest noise/distortion and most accurate frequency response in the world built in a box the size of a 75-100 watt per channel class AB amp...that's a winner of a product, even if it is larger than your company's cool looking small cases.

May be wrong, but as an audio/music fan for 50 years, I harbor no doubts in this regard.
 
Last edited:
But, if one can install a 500 watt per channel @ 4 ohms with 40 amps current capacity, relatively cool operating, high efficiency, weighs 10-12 kilos, offers the lowest noise/distortion and most accurate frequency response in the world into a box the size of a 75-100 watt per channel class AB amp...that's a winner of a product, even if it is larger than your company's cool looking small cases.

May be wrong, but as an audio/music fan for 50 years, I harbor no doubts in this regard.
From my point of view, you are absolutely right. And the bigger case allows for higher continuous rated power and also fits to a main home system that is not based on small boxes. IMO there is a market, though not very big, for such product.
 
No electronics manufacture experience her, in fact I burn myself every time I get near a soldering iron.
Maybe you're holding it wrong... :)


Soldering.jpg
 
What about the IEC / AC filter. I heard that it should not be used with Hypex Ncore modules ....?
 
Back
Top Bottom