Just finished set to work of the new DAC for the digitial crossover. This includes a PGA2320 programmable attenuator on the output of the CS4398 DAC.
This replaces the home made PCB thai I was using on the CS4398 output. I went the professionally made board mainly because PCBCART was able to deliver 16 of these to my house for $140, that includes tooling!!! The actual boards were like $3 or 4 each!
The quality of these baords is fine - though I must admit that I am not breaking any new ground in PCB technology here.
The board is only double sided - I thought about adding a seperate ground and power plance - but the routing density is so lot it would be crazy. The whole back side only has a handful of tracks on it - and is a ground plane in itself. The top side even has room for significant ground fill.
All the digital stuff is in between the connector and the DAC, with a few SPI lines up the left hand...
The prototype is listenable now, but a bit too much background noise/hum pickup to do serious listening. It needs an on-board post filter and amp, which I'm working on now. In the meantime, here's the pics - the DAC itself is built of 5 'dac-sleds' each with a stack of 4 chips. The 'sleds' are then arranged around the central tower holding the resistor ladder. A separate board handles the timing logic and 18 tap delay line.
I compared it with my Onkyo SE-200PCI sound card. This 24/192 (115dB S/N A-weighted, 0.003% THD 0dB 1kHz) PCI card sells for about $15,000 yen and is based on the VIA Envy24HT and Wolfson WM8740.
I'm listening to 16bit 44.1kHz .wav (CD rips), though VLC [sample rate converter set to sinc, best quality, resampling quality 8]. Windows 8 release preview [default format 24/192 (onkyo), 24/96 (odac)]. Line out though Oyaide PA-02TR interconnects to the Sapphire headphone amp, and Sennheiser HD-600s.
So, I was planning on writing up a big 'ol review with my impressions, but, well... there's not really a...
Here are my own modest base requirements for a line stage (similar for power amp):
1. Open loop BW of 40KHz or more (-3dB)... 20KHz min.
2. IM and THD of less than .001% at 1v rms into 30 ohms for any frequency between 20Hz and 20KHz.
3. No coupling caps on input or output or in feedback path.
4. No use of dc servo circuits to track and correct dc offset and drift.
5. No more than 6-8 transistors (excluding power supply).
6. S/N ref 1 volt rms and without weighting of at least -130dB (input can be shorted or terminated).
7. No significant harmonics above the 2nd and 3rd.
8. Closed loop gain between 12 and 20 dBv
9. Low Zout (less than a fraction of an Ohm at any audio freq).
10. Distortion not be changed by source Z.
11. Transistors should be low cost and not be exotic, hard to obtain, very expensive or no longer manufactured.
Despite the (catchy) name I'm thinking pre-amplifier rather than amplifier applications.
update: I have have a quick and dirty sim up and running in ltspice. Curiously, the output distortion is 15 dB lower when the buffer runs open loop than when it is included inside the feedback loop. Intrigued. Currently under investigation.
update: refined the sim slightly, achieved -85 dB distortion levels at 0 dB / 1 kHz / 600 ohms running the output buffer open loop. Bandwidth is just under 1 MHz, adjusted by changing the feedback resistance. As before, performance sims out notably worse with the buffer
inside the feedback loop.
It is certainly a little bit different. One might be tempted to say "gilding the lily", but come on, headphone amplifiers are just the right place for these indulgences.
Building your own long tailed pair (LTP) to bolt in front of an IC op amp has fallen out of favor in recent years. I must admit I couldn't see the point then, and still don't.
I've seen a number of headphone amp circuits with 3 paralleled pairs of output devices. I wonder if there any real advantage over simply using one pair at 3x the current, perhaps with slightly larger transistors?
Fixing the NOS droop has the undesirable side-effect of making the imaging components worse - HF gain can't be increased up to 20kHz and then suddenly taken away above 22kHz :crazy:
In the spirit of taming the near-ultrasonic emissons of a NOS DAC, I'm currently playing with an MOS design - where 'M' stands for 'minimal'. I've been wondering if the attractiveness of the NOS sound will still be preserved if I go to 2X OS in order to fix up the ultrasonics. No practical analog filter can have a sharp enough band edge so a digital one it does have to be...
A DSP implemented digital filter comes at a price - that of throwing away some bits (I'm still using only 16bit DAC chips - TDA1387) so I'm now exploring using a transversal filter (no DSP) to carry out 2X OS. The LTSpice screen grab shows the architecture I'm playing with - a 19 tap delay line feeding 19 separate DACs. The DACs are shown on the right as current sources and their individual weightings are...
The 5532 and 5534 type op-amps require adequate supply decoupling if they are to remain stable, otherwise they appear to be subject to some sort of internal oscillation that degrades linearity without being visible on a normal oscilloscope. The essential requirement is that the positive and negative rails should be decoupled with a 100 nF capacitor between them, at a distance of not more than a few millimeters from the op-amp; normally one such capacitor is fitted per package as close to it as possible.
He's someone who should know. Anyway, it doesn't take much digging on the internet to confirm beyond reasonable doubt that bypass caps should be as close to the op amp power pins as possible. So thinking about my previous experiments with bypassing the Sapphire, by adding bypass caps around the transistors I also effectively also added a bypass for the op amp, but a rather poor one as the power-pin-to-power-pin round trip loop distance is probably 10...
I have never been happy using DACs to implement volume control. I guess largely because of the obvious degradation in resolution at low volume levels.
On my first DSP crossover I used an AD1939 CODEC, which has 24 bit resolution - though obviously lesser precision - and my approach here was to use a mix of steeped attenuator and digitally implemented volume control.
It worked well - though pretty shortly after I built an 8 channel PGA2320 based volume control.
For my new DSP crossover I started with a simple CS4398 in the DAC output. even when I built the first boards I KNEW I would be going back to integrate a programmable attenuator.
Why didn't I just start with it? **** knows.
At lease with the modular approach all I had to do was respin my DAC board to include a PGA2320 along with the CS4398. I already had SPI to the board, sop I can sneakily use these lines for the PGA 2320, as the CS4398 is in hard wired...