I like your layout (probably optimum for thru-hole parts).
BTW, I saw on the pictures that the 2x 22µF lytics on the 3886 power supply used to be oscons and now they're nichicons, any reason for the change ?
The OSCONs were 35 V types. I needed to run at 35 V rails, so used 50 V Nichicons instead.
~Tom
No kidding. The comparison is apples and oranges for sure. Oh! I get it:
"Why can't we all just get along?"
But we already do. People post ridiculous stuff and others laugh! That is an accepted way of "getting along". Oh wait... it was not supposed to be ridiculous.
My bad.
Well... It's not every day that I get the chance to perform a competitive analysis, so when Anand offered to ship his amp, I took him up on it. I have never heard (or measured) a high-end class D amp and I was curious. It'll be interesting to play with and an interesting data point to have.
Tom
Tom
you have nearly a volt of 300ish kHz noise. Now my only reference is this
Theta Digital Prometheus monoblock power amplifier Measurements | Stereophile.com but JA is pulled up by someone saying the AES17 filter in the AP handles it all. As the x525 claims to be perfect for class D amp development I am probably worried for no sake as I can't see an option listed on the spec sheet.
Theta Digital Prometheus monoblock power amplifier Measurements | Stereophile.com but JA is pulled up by someone saying the AES17 filter in the AP handles it all. As the x525 claims to be perfect for class D amp development I am probably worried for no sake as I can't see an option listed on the spec sheet.
Class D is like Linear Amp with constant 300mV 400khz tone anti-aliasing filters are not the problem the input range detector in SYS2722 is analog and before the ADC so you are always 300mV ahead of the curve and below 300mV you can't measure correctly as you are expecting 10mV sinewave but the range is stuck in 300mV range so you need AES17 Filters before the autorange circuit that is standard on the SYS2722 all you need is to switch it on but the x525 has digital filters after the ADC if the range detection is done in the the DSP then you are good to go, if not... You need external low distortion analog AES17 filters or something else active, dont use passive ******** like AUX-0025. Never used APx525 before it might work properly or it might be deliberately flawed so they can sell you the High bandwith option APx BW52. You are smart guy you can design your own low distortion AES17 filter.
Connecting IN- to GND at the PCB is acceptable. However, you will get better CMRR performance if you make the IN- to GND connection at the RCA connector. You will get better performance yet, if you make the IN- to GND connection at the source (use the pseudo-differential interconnect cable I show in the Design Documentation). You will get the best CMRR performance with a fully differential connection.
Tom
Tom
The signal connection requires TWO wires.
The signal circuit CANNOT work if you break one of the wires.
If you use a coaxial cable instead of a twisted pair, then the core is one wire and the screen is the other wire.
The screen wire MUST remain part of the CIRCUIT to allow the signal circuit to operate.
The only time, that I am aware of, that it is permissible to connect a screen at one end only, is with a screened twisted pair where the TWO cores pass the signal and the screen is NOT part of the signal circuit.
This should work ... set the output impedance of both channels to 20R. Put a 20R resistor on Input B, and feed Input B from source B. Feed your DUT with Source A using GENMON to measure the output voltage. Use the X-Y mode in "Sweep" to calculate the DUT impedance relative to 20R.