DSP Xover project (part 2)

I'm writing this to see is there an interest in this or am I to only one. There is probably several ways to do this without the help from Nic. One idea to switch inputs from outside the board would be to generate an IR signal with an IR emulator and feed that into the IR receiver. An other way is to cut the lines to on board switches and feed signals from outside the Najda boar. Any other ideas!?

I have a similar interest of controling the source channel with a switch that should act as a "next source" button on the case (my volume control button has a push button for this). I have also thought along the lines of just wiring this switch to a remote control built into the case and powered by the power supply. It is very inelegant though and I also want to be able to select channel with my remote control from my sofa - so the two remote either have to be the same with the same IR codes or I have to not program the "next source" button on my external remote. A software update of course would be great (e.g. the up down buttons could act as source selection when using a rotary encoder, instead of being disabled - this also applies to the boxed version of Najda) - but again it may not be worth the effort for nick with only few customers requesting this.
 
I used the twisted pear spdif mux to get four spdif inputs. I set up an arduino with an additional ir receiver. The arduino controls the mux. It also shows the source on a second display. Works well. There are some photos and more description somewhere in thus thread - i can find a link once i'm on a computer.
 
I used the twisted pear spdif mux to get four spdif inputs. I set up an arduino with an additional ir receiver. The arduino controls the mux. It also shows the source on a second display. Works well. There are some photos and more description somewhere in thus thread - i can find a link once i'm on a computer.

Merry Christmas Martin,

This is the page where you posted your images 😉

http://www.diyaudio.com/forums/digital-line-level/215379-dsp-xover-project-part-2-a-232.html#post4019067
 
Anyone tried the more recent V5 opamps from Burson?
Burson SS V5i Op-Amp (Dual) - Item Audio
https://www.bursonaudio.com/products/supreme-sound-opamp-v5i/

I have never messed with opamps, but they sound like a more promising upgrade than changing say my speaker cables or power supplies.

I see six LM833N opamps on my first series red Najda board - two alone, then another row of four. If I bought two (£78), where could I best put them? I use four way Nao Note speakers.

Thanks for any pointers or opinions.
 
Anyone tried the more recent V5 opamps from Burson?
Burson SS V5i Op-Amp (Dual) - Item Audio
https://www.bursonaudio.com/products/supreme-sound-opamp-v5i/

I have never messed with opamps, but they sound like a more promising upgrade than changing say my speaker cables or power supplies.

I see six LM833N opamps on my first series red Najda board - two alone, then another row of four. If I bought two (£78), where could I best put them? I use four way Nao Note speakers.

Thanks for any pointers or opinions.

A friend of mine is running them on his Najda and a pair of LX521 and he likes them.

About the nagging about future firmware updates, I will have the fullest respect if there was a comment saying that the "Najda" is fully developed and no future changes are to be made. It's a fully functional dsp solution that is bringing me joy every single day and it is functioning perfectly as is. Maybe the half promise of new and exiting functions was a bit premature and just to much work, no problem if so but I would appreciate a comment either way.

Best regards. Bengt
 
Anyone tried the more recent V5 opamps from Burson?
Burson SS V5i Op-Amp (Dual) - Item Audio
https://www.bursonaudio.com/products/supreme-sound-opamp-v5i/

I have never messed with opamps, but they sound like a more promising upgrade than changing say my speaker cables or power supplies.

I see six LM833N opamps on my first series red Najda board - two alone, then another row of four. If I bought two (£78), where could I best put them? I use four way Nao Note speakers.

Thanks for any pointers or opinions.

The op-amps are used in the DAC filter buffer stage and perform differential to single ended conversion.

I imagine that it is very similar to the datasheet implementation.

Screen Shot 2016-12-30 at 12.40.44 pm.png

The LM833 is already a pretty good low noise opamp, you could replace them with LM4562 or AD 4898-2 if you want something that is a measurably better part. They are also not that expensive in comparison to the discrete types.

The circuit is pretty simple it is unlikely that you will get much if any improvement from using discrete boutique replacements.

Unless you have money to burn I don't think those sorts of changes are value for money. Also at those prices you are almost certainly going to prefer the sound with the expensive replacement. Unless you can do a true blind test then you may not be able to tell any difference.

Speaker cable just needs to have as low resistance, capacitance and inductance as possible. This means in general a heavier gauge wire. Fancy audiophile cables usually have worse parameters than basic cable and set up a filter with the capacitance and resistance in them, this is the reason why they may sound different.

There isn't a huge amount of room in the Najda design to tweak or upgrade, and even less need to do it as it already sounds very good!
 
Thanks for replies. It was page 53 of the thread that got me thinking. Some users report significant improvement with a number of different OPs.
OPA627 (with adaptor becomes costly apparently)
SS3602 (too expensive!)
And the Burson discrete come up as being worthwhile in post 2622
http://www.diyaudio.com/forums/digi...sp-xover-project-part-2-a-53.html#post4241335
Of course the net is covered in universal praise for the Bursons, which seemed promising to me - but the price!

LM4562 are only £2.50 each. Could buy six for £15, rather than trying to ration two Bursons at £78. But has anyone tried them on Najda?
http://www.mouser.co.uk/ProductDeta...2NA-NOPB/?qs=sGAEpiMZZMvu8NZDyZ4K0bBDocacHabE
 
Two questions about fir filters.

Would it be anything to gain from running a linear phase fir filter for hp only in a 3-way filter 350-48khz and iir filters for the rest ?

Will there ever be a Najda function for this ?

A FIR crossover would have the advantage of not changing the phase at / around the crossover frequency, so that the two drivers add in phase and the overall phase stays flat. However, the NUC software from WAF can take such a phase shift of minimum phase / IIR (infinite impulse response) filters into account, so that the delay / phase of the two drivers can be adjusted for a mostly in phase addition of the two drivers with IIR. Still, with IIR the phase overall is a bit bent - luckily our ears are not that sensitive to phase and that is why one can get away with IIR filters. Absolute high end projects more and more often use FIR filters these days - but others such as the Linkwitz speakers notably use IIR filters. I guess one disadvantage of FIR is that there may be pre-ringing if the filter is not created carefully, and also in the lower frequency range, room modes are minimum phase / IIR resonances and should be canceled with the complementary IIR filter (note that FIR can also emulate IIR, can correct the bass if a lot of taps are available).

The existing FIR capability has not been useful for me, because the number of FIR taps even at 48 kHz is too limited for me to equalize / crossover in the bass. Therefore a mixed FIR IIR mode for IIR bass and FIR treble crossover would be nice to toy with. I noticed that Nick stopped posting here and no more software updates have been promised since around the time the codec chip has been made obsolete. I am not shure whether NAJDA will continue to be sold without this chip being available (there may be newer chip, that is a drop-in replacement though). In conclusion, I agree with the previous statement that NUC / NAJDA is a perfectly fine finished product as is.
 
On Najda, all six opamps are dual channel? I see a pair on their own so thought they might be different, and perhaps the best place to put a couple of Bursons.

The ones that are by themselfes are for analog input. I would try them in either of what you would refer to as midrange first. But if you only use analog sources you would most likely be right (looks like spdif digital only ?). Do not forget that you have quite a few opamps in series in your build. You have the obvious ones (4) the volumecontrol chip has them on both input and output and you also have opamps of your own for converting your outputs. Changing one in a series of four might not give as much as you would expect. I have tried changing to 4562's in a previous dac build but didn't like them, ended up with 1361's. The original ones in "Najda" are actually quite good. There are capasitors in the signal path for dc-blockage that might be a more effective position to look more closely at.
 
A FIR crossover would have the advantage of not changing the phase at / around the crossover frequency, so that the two drivers add in phase and the overall phase stays flat. However, the NUC software from WAF can take such a phase shift of minimum phase / IIR (infinite impulse response) filters into account, so that the delay / phase of the two drivers can be adjusted for a mostly in phase addition of the two drivers with IIR. Still, with IIR the phase overall is a bit bent - luckily our ears are not that sensitive to phase and that is why one can get away with IIR filters. Absolute high end projects more and more often use FIR filters these days - but others such as the Linkwitz speakers notably use IIR filters. I guess one disadvantage of FIR is that there may be pre-ringing if the filter is not created carefully, and also in the lower frequency range, room modes are minimum phase / IIR resonances and should be canceled with the complementary IIR filter (note that FIR can also emulate IIR, can correct the bass if a lot of taps are available).

The existing FIR capability has not been useful for me, because the number of FIR taps even at 48 kHz is too limited for me to equalize / crossover in the bass. Therefore a mixed FIR IIR mode for IIR bass and FIR treble crossover would be nice to toy with. I noticed that Nick stopped posting here and no more software updates have been promised since around the time the codec chip has been made obsolete. I am not shure whether NAJDA will continue to be sold without this chip being available (there may be newer chip, that is a drop-in replacement though). In conclusion, I agree with the previous statement that NUC / NAJDA is a perfectly fine finished product as is.

I have read up some and also played with "RePhase" and it's very obvious that you would run out of taps very soon if you want to do something at lower frequencies. I only have one driver from 350hz and up that doesn't take that many taps to create a 2/ord filter for even @ 96khz so it would be exiting to try.

There is a new chip from analog devices that seems to be two "ADSP-21489" in one chip, named ADSP-21587. It looks promising, for me anyway (not that I have the knowledge to build something like this).
 
Maybe the half promise of new and exiting functions was a bit premature and just to much work, no problem if so but I would appreciate a comment either way.

Best regards. Bengt

I too would like to get an official answer from Nick. So many promises has been made almost 3 years ago without a single software/firmware update...
Last message here from him was last april and he does not answer to PM or emails (at least to me).
Apparently, only new (or potential) customers have the privilege to get email return.
Anyway, all the best for the upcoming new year
JC
 
I made an effort to replace two of them in a build I did for a friend and it didn't go that well. The solderpad didn't survive and some extra work was aquired. This was on a first generation board, maybe the newer ones are easyer to work with. There should be some comments somewhere in this thread about that as well.
Edit: Happy new year 🙂