Do you have a schema of it,it would be easier to see.
There's really no need for schematics:
The 2 caps that are close to each other all have one pin that goes to the analogue volume chip, and the other pin is connected to the DAC output stage.
So you really want to bypass the output stages that include nice low-noise mini-melf resistors? Alright alright...
Hi Nick
On the first point, sorry to bother, but may you explain to the dumbs (;-) why we should not worry ? I think most of DSP do that resampling, but wont some samples get lost as 48 khz frequency is not a multiple of 44.1 ?
Thanks for the nice tutorial with rePhase ! What about phase correction itself ?
BR
Jean-Louis
No. No samples get lost - all samples get converted. The usual way of doing it is basically the equivalent of first upsampling by a factor of 160 and then downsampling by a factor of 147 (as the ratio between 48 and 44.1 is 160/147). Another way of thinking about it is to imagine the signal first being converted to analog at 44.1 kHz, and then resampled at 48 kHz. None of the samples are the same as the original, but the signal waveform is still the same.
Thanks for the fast reply. Looks really promising then. I have a bunch of desoldered OPA627 that I would like to put in a DAC converter sometimes, fex
But decoupling caps? Are they not coupling caps in the signal path to block DC?
Actually the caps would probably be a better way to feed the signal, as you get a nice pad with a round hole in it. The IC socket is not a HQ socket, thus not that good for round pins.
As I have not seen the board and the layout especially , would it be possible to desolder the IC sockets without to much risk of damaging the routing?
I have quite good tools, but anything connected to a big heatsucking copper area is always difficult to desolder (like swollen filter caps on PC motherboards, they are a pain in the donkey to desolder)
Sorry for all the questions, but what post DAC filter topology do you use ? Sallen Key or MFB? I used MFB in my post-DCX2496 differential-single-ended filter to prevent oversampling noise feedthrough @ MHz.
Last edited:
You're right, I meant coupling capacitors.Thanks for the fast reply. Looks really promising then. I have a bunch of desoldered OPA627 that I would like to put in a DAC converter sometimes, fex
But decoupling caps? Are they not coupling caps in the signal path to block DC?
Do you realize I've put the opamps on sockets so that you don't need to desolder anything - and now you come and ask me about desoldering the sockets themselves?
Mr. Nick, Sir, thank you so much for the tutorial on building FIR filters. Sorry about the late reply. I had to get out, and listen to live music this weekend. 2 weeks ago I thought fir filters were something that went on the shower drain. Now I am listening to music on my new toy. It has replaced an elderly Marchand XM9 in my pile. One question, if I use either channel 4 or 5, I cannot get the output graph to sum to a flat line in Najda Under Control. It is an either or thing. Should I ignore it as a glitch? I am using channels 1, 2, 3, 5, 6, and 7. And it is working perfectly. Sounds pretty good to my antique ears! Thanks again for all your help. Very cool toy!
No I can't f***ing hear it so turn it up real LOUD!
No I can't f***ing hear it so turn it up real LOUD!
You're right, I meant coupling capacitors.
Do you realize I've put the opamps on sockets so that you don't need to desolder anything - and now you come and ask me about desoldering the sockets themselves?
I said sorry
I suggest you check the rePhase thread:
http://www.diyaudio.com/forums/mult...hase-linearization-eq-fir-filtering-tool.html
That's cool Mr. Mark!
If you want to send me your setup files on my email, I can check what's the issue with channels 4 or 5 you mentioned above.Maybe it just needs a fresh battery
Hi Ake,
The +5V mark on the silkscreen is not a polarity mark, it just says 'this is the 5V input port'.
The polarity is as in the manual and on the diagram supplied with your board.
In case of doubt you can also look at the traces on the PCB and/or test with a meter
Looking forward to seeing your build!
Best,
Nick
The +5V mark on the silkscreen is not a polarity mark, it just says 'this is the 5V input port'.
The polarity is as in the manual and on the diagram supplied with your board.
In case of doubt you can also look at the traces on the PCB and/or test with a meter
Looking forward to seeing your build!
Best,
Nick
Last edited:
More processing cycles. You can run multiple channels from one DSP to the next and split processing duties between them, at the very least have left on one DSP and right on the other, kind of useful when long FIRs are involved.
Sounds great for the power hungry. BTW I recall I saw some processors on the market with two DSPs from the start.
Yup. I recommended them for the previous day job. Besides, there's no such thing as too much power.
The Datasat AP20/ RS20i Audio Processors have 6 400 mHz Analog Devices SHARCs onboard
AP20 | Datasat Digital Entertainment and RS20i | Datasat Digital Entertainment
Member
Joined 2009
chaparK,
Did you see that S. Linkwitz is looking for a DSP replacement for his analog crossover?
Your board will be a good candidate.
DSP_challenge
Did you see that S. Linkwitz is looking for a DSP replacement for his analog crossover?
Your board will be a good candidate.
DSP_challenge
Yes I had already seen that, but thanks for posting.
Interestingly, it seems that Mr Linkwitz is not too much in favor of FIR filtering.
He doesn't like pre-echo but of course not all FIRs have that problem...
FIR doesn't have to be linear phase.