[error401]

I think that supply will work much better. You might need more C on the output of the LM317 to keep it stable (datasheet recommends 1uF tant or 25uF 'lytic).

You're almost up to the complexity of the tried and true Jung design though, might be worth just using his circuit.

 

[BrianDonegan]

Quote:

Originally Posted by FS2 /img/forum/go_quote.gif This is inspring me to get back to work on my own DAC, which has been sitting on the back burner for way too long now... I have a question about ASRCs though. Hope nobody minds as I haven't really been able to find a clear answer even after sifting through forums: What is an ASRC meant to acheive? And does it really make a difference?

Besides the jitter reduction (and this depends heavily on the DAC chip you are feeding) the digital filters of the DAC will often operate better at the higher sampling rates.

 

[00940]

Quote:

Originally Posted by error401 /img/forum/go_quote.gif I think that supply will work much better. You might need more C on the output of the LM317 to keep it stable (datasheet recommends 1uF tant or 25uF 'lytic). You're almost up to the complexity of the tried and true Jung design though, might be worth just using his circuit.

On the capacitance at the output of the lm317, you're right. I'm thinking of a 100uf pana FC with 0.5r in serie with the cap.

Difference in between this design and the Jung is about 6$ in parts by rail (not counting the pcb, prices coming from digikey). Probably less prone to oscillation and startup problems too, at the obvious price of reduced performances.

PS : One should scrap R4, it's a leftover of the flea regulator.

 

[00940]

I'm slowly gathering parts (it's a pain to get some things in Europe, really).

Anyway, I'm gonna change a bit the output stage to include a pair of ds1802 as volume control. The idea is to put the digital pot in parallel with the I/V resistor. I can set up the I/V converter so that I've got -2.5VDC at the input of the DS1802 and 1Vrms (to avoid clipping). I'll just have to shift its 5v pins to ground and its ground pins to -5V to correctly bias the DS1802. In this position there is way less than 1ma going through it.

If I put the DS1802 there, I'll have to move the bessel to the output as the volume pot would mess up the filter. But it allows me to use low values resistor around the gilmore and better balance the impedances at its inputs.

For the power supply, I'll finally use a simple pair of LM317/337 for the I/V and the output stage of the gilmore at (+/-15VDC) but I'll use a voltage doubler and improved regs for the input of the gilmore (+/-24V). This to get a greater voltage swing.

 

[00940]

Ok, this thing goes forward... slowly. The DAC board is fully populated and so its power supply. I settled on a lm317 set at 5.9V, followed by cap multipliers for all the digital lines (there are further reg101 onboard).

A quick test shows that all the voltage are ok. Connecting the beast to the computer shows that the pcm2707 is perfectly recognized. *wipes sweat*

Now, I've to build the I/V stage...

For the volume control and balanced to single ended parts, things have changed a bit, see : HeadWize: DIY Workshop > Servo on a balanced receiver ? (last posts by PRR).

 

[00940]

Next step done. The good news are that the whole digital section works. I can hardly say anything about the sound quality: I've been listening to one output pin after the other, with a hack job I/V stage (a opa227 connected to an unregulated supply and a mini jack with a bunch of alligator clips). It sounds rather clean and detailled for all I can judge. Noise is ok considering all the mess and the long unshielded wires.

On the other hand, one of the 4 I/V stage is not working (very high currents running through it). The 3 others show normal voltage readings though. I'll have to redo that dammaged board tomorrow before testing the I/V stage. :/

update: reflowing all the joints fixed the 4th I/V stage

At least the readings are ok now. Testing this afternoon.

 

[digger945]

Interesting, I look forward to your listening comments.
Love to see a pic.

 

[00940]

I'll try to get pics this evening.

Great news: the I/V by cauhtemoc works very well. The current test setup is: soundcard out- 510R in serie - input of the I/V - dynalo (with a cap as I've got some offset). It's still in mono but it doesn't sound much different from a direct connection from the soundcard to the dynalo... which is what is asked here

There is also no added noise (and this without regulation whatsoever).

update: I connected a I/V to the pcm1798 : it works this way too

 

[espressogeek]

Can you bypass the internal filter and just use the I/V? I am just curious.

Also, does anyone have a collection of I/V designs or a link to a web page that describes the various ins and outs, op amp vs discrete, etc?

 

[Pars]

Nice job. It isn't possible to trim the output offset out on this stage?

 

[00940]

Quote:

Originally Posted by Pars /img/forum/go_quote.gif Nice job. It isn't possible to trim the output offset out on this stage?

Yes. You would need a trimmer in parallel (or instead) the top left resistor (see my first schematic and the discussion in the thread "a discrete stage for zhaolou" or something like that).

I will however not null the 1.8VDC offset I will have (thanks to the offset on the pcm1798 output), as it allows me to bias the DS1802.

 

[00940]

On the attached pic, we have, from left to right:

- The PS: 2200uf+lm317 preregulator.
- The DAC board, raised by spacers. Hidden under the board, on the "motherboard", there are 3 capacitance multipliers to isolate the digital sections.
- The I/V sections, also raised by spacers.
- Under the I/V section will go a pair of ds1802 for volume control.

The wires:

- The big blue/red go to the analog +/-15V supply.
- The two grey shielded microphone cables will go to the headers at the bottom of the DAC (on the pic), from the output of the pcm1798 to the input of the I/V
- The grey cable on the top left is the usb cable, it goes toward a wallmount usb jack.
- The yellow/white cable is just for testing one channel.

All that fits on one 100-160mm protoboard.