Nick, I'd like to mention, this is one of the most pleasant experience with a product for DIY. Well thought out, well made, well documented … and it just works!
On software side a few little desires came up, so thought I'd mention:
- a change in color (for example) when the Mute and Bypass buttons are active would be great
- a bypass option for the XO part would be welcome. Right now switching each off is a bit cumbersome, especially because the steepness is resetting to 12dB/oct at turning on again
- an option to separate the windows would save a lot of switching back and forth, highly desired
Abel
On software side a few little desires came up, so thought I'd mention:
- a change in color (for example) when the Mute and Bypass buttons are active would be great
- a bypass option for the XO part would be welcome. Right now switching each off is a bit cumbersome, especially because the steepness is resetting to 12dB/oct at turning on again
- an option to separate the windows would save a lot of switching back and forth, highly desired
Abel
rPi and Volumio over I2S
Encouraged by some success in earlier posts, I've bought a rPi and installed Volumio 1.2b to connect via I2S to Najda. I didn't get lucky though, the music is messed up: some is played back in a kind of mono, some in the opposite of it.
Do I have to do any additional adjustments? Thankful for any input!
Abel
Encouraged by some success in earlier posts, I've bought a rPi and installed Volumio 1.2b to connect via I2S to Najda. I didn't get lucky though, the music is messed up: some is played back in a kind of mono, some in the opposite of it.
Do I have to do any additional adjustments? Thankful for any input!
Abel
Hi Guys,
Can someone point me to some info on the external relay? I had assumed it worked like the "12v trigger" on other hifi gear and designed my power switching based on relays that switch on 12v. Now that I try to use the Nadja's external relay thing, I find the voltage is very low - below 0.5v.
I'd like to turn the Nadja's external relay into a "12v trigger" - any suggestions on how I can do this? Maybe suggest a relay with the correct coil voltage for the Nadja and I'll use that to create the 12v?
Can someone point me to some info on the external relay? I had assumed it worked like the "12v trigger" on other hifi gear and designed my power switching based on relays that switch on 12v. Now that I try to use the Nadja's external relay thing, I find the voltage is very low - below 0.5v.
I'd like to turn the Nadja's external relay into a "12v trigger" - any suggestions on how I can do this? Maybe suggest a relay with the correct coil voltage for the Nadja and I'll use that to create the 12v?
Is this of any use for you?
Atlona® AV Distribution Solutions for Commercial and Residential Applications
found from this:
Digital Crossover/EQ with Open-Source Software: HOWTO | Richard's Stuff
Thanks, but I have seen those links before. It might be interesting if the optical passes hi-res but it does not due to the HDCP.
There are no ready made solutions for this due to the HDCP issue. It will not be easy to resolve this. I am happy enough with my current solution though - stereo high res over analogue from the blu-ray player.
I had the exact same issue. Please see post #1840.
Thanks Bjorn, went back for some more reading …
Since I been looking for a solution to get rid of MacMini and iTunes as my player, Volumio in the end didn't really shine (even without the I2S issues).
So I gave piCorePlayer + LMS a try - OMG! Finally the solution I've been looking for! And works a treat with Najda, creating the "ultimate" streaming-preamp-DSP! Am one happy camper … and it's not just a substitute to iTunes, it's so much better (not taking sound quality here) ...
So one more vote for this setup. Here's some details of mine:
- LMS running on computer
- piCorePlayer running on rPi connected over I2S to Najda
- Remote: Asus NexusII with Orange Squeeze app
- love it all!
One other thing Najda is great for, is to do A/B(/C/D) tests of sources - very good as a sanity check as well! : )
Just connected my Sony CDP559ES through variable analog out and optical out and compared to the above mentioned setup. Differences? … : )
Sorry for some OT here!
Abel
Hi MF,
the Remote "output" is just a switch, it doesn't deliver, it shortens.
Your setup could look like this:
- one connector from Remote to -12V on Najda
- second one to your relay coil
- other end of relay coil to GND on Najda
I've suggested to use the -12V side, because the +12V already has to deliver more to Najda, so this way it get's more balanced …
Abel
Hi Paal,
We can compute together the number of taps available in your conditions:
56% CPU load left, that's 0.56 x 500 = 280 MMACS or taps per second.
As your sample rate is 96 kHz, the number of taps per sample is 280 x 10^6 / (96 x 10^3) = 2916, to be distributed among all channels.
Yes this will be added (support for Winstar WS0010), as well as some VFD support.
Thanks Abel!
Damn!
Yes, this was already requested and will be added. Easy stuff.
Isn't a bypass switch on the XO section something potentially harmful? Is that really desirable? It's actually on purpose that I haven't added such a switch.
Yes, this was mentioned several times. I'm looking at it but haven't found yet an appropriate way to change this. This unfortunately is not easy stuff.
My ideal configuration for an electronic crossover has a crossover section (NUC's Channel Processing) yielding a totally flat response, plus an EQ section (NUC's Input Processing) for a selection of "house curves" enabling the speaker to adapted to various acoustic environments. This will include a "bypass" setting, which allows the unequalised response to be active. With Najda you can have different setups (which is how I select between house curves): perhaps this is a way to achieve bypass.
Hi, I'm from Russia, translated into English with the help of Google, so there may be errors. I'm looking for a digital crossover for a small PA system, and I think your Naida is what I need.
I have non-standard configuration, satellite broadband, so such a crossover point division is 100 hertz subs, 100-400 hertz midbass, 400 - 20000 satellites. Can I get in, we find a configuration if needed subs midbass mono (sum of left and right channels)
I understand that a very strange configuration, but it is due to the compactness of the system, so that the amplifier four channel.
Another question, if I understand correctly present FIR filtering? I suppose correction and phasing at low frequencies will give more delay is unacceptable to live here kollektiva.A midbass satellites can filter them?
I have non-standard configuration, satellite broadband, so such a crossover point division is 100 hertz subs, 100-400 hertz midbass, 400 - 20000 satellites. Can I get in, we find a configuration if needed subs midbass mono (sum of left and right channels)
I understand that a very strange configuration, but it is due to the compactness of the system, so that the amplifier four channel.
Another question, if I understand correctly present FIR filtering? I suppose correction and phasing at low frequencies will give more delay is unacceptable to live here kollektiva.A midbass satellites can filter them?
Thanks Abel, I will give this a try. Good idea to use the -12v line - it's basically un-used in my system while the +12v is very busy.
As I not get clear with this question I ask new, what is directly related to first.
If I had Najda volume control normally about -30 dB for my system and gain about -0 dB (little differently in different channels), will sound improve sonically if I decrease gain by 15 dB and increase main volume by 15 dB? Had any one tested this? I think it must improve as gain is controlled in CS3318 with negative feedback depth and best volume control is "missing volume control", i.e. no attenuation and not needed amplification. My own tests are promising I think.
Additional request: can be gain controlled in 1/4 dB steps as CS3318 is able, instead 1 dB steps as now implemented? I feel I need more precise level control for speaker channels.
As Abel pointed out, the relay command is an isolated transistor. What you have measured is probably the collector/emitter voltage.
If your aim is to output +12V or GND (did I understand what you're trying to achieve?), then what you need to do is the following:
- Tie the collector to your external 12V source
- Tie the emitter to your 12V input
- Tie a resistor to the emitter
- Tie the other end of the resistor to your external GND
You don't need a relay.
What happens with this circuit?
When the transistor is 'open', there's no current across the resistor so your command is GND.
When the transistor is 'closed', your command is pulled up to 12V.
Now the question is: how big should the resistor be?
- Not too small otherwise you're going to fry the transistor.
- Not too big otherwise the transistor is not going to saturate.
It's probably a good idea to check the datasheet (FOD852S), in particular the test circuit p.8.
Hi and thanks for your interest. The best I can recommend is that you download the application and try to create a setup according to your specs. If you can do it offline, then you can probably do it also with the Najda board (unless there's not enough DSP load - you can send me your file and I'll let you know what the load is).
Sorry Meelis, I replied a few times to your questions related to the volume and gain management and I was hoping that someone would jump in because myself I didn't understand what you're looking for. Now I think I do.
The datasheet shows the chip has several registers where you can dial in gains and attenuation.
Let's take an example: you have 10 dB gain and the volume down to -30 dB. The Cirrus chip gets from one register the +10 dB information, from another register the -30 dB one. Then the chip 'sums' the above and ends with a -20 dB attenuation.
We take now a variation on this example: you have -5 dB gain and the volume down to -15 dB. Again, summing leads to -20 dB attenuation.
We expect no difference between the original situation and the variation because there's no specific register for pure gains and no specific register for pure attenuation: each register can access the whole range from -96 dB up to +22 dB. In other terms, the logic doesn't give you independent control over gain and attenuation structures.
As for now, granularity of volume and gain is 1 dB indeed. We could look at reducing that to 1/2 or 1/4 dB. In the meantime, you can use the digital post-gain for that fraction of dB you want to add or remove. Although not optimal because digital, it shouldn't affect the final result because we're talking about a fraction of a dB.
Last edited:
Sorry, certainly in the subject have already been asked and answered, but I find it hard to read, it is necessary first to translate, so I will ask, do not be offended if the answers have already been written. Downloaded the program management, like everything is clear, but he did in the classic version. And what program count FIR filter? rePhase?
You probably already do 3 way crossover for FIR filtering, which he gives a time delay?
If I understand correctly the maximum delay gives the lower band. I did not see in your program that possibility, but I ask, have the ability to cut off the lower band in the classical way, and the rest of FIR? That is, if the delay is too long.
You probably already do 3 way crossover for FIR filtering, which he gives a time delay?
If I understand correctly the maximum delay gives the lower band. I did not see in your program that possibility, but I ask, have the ability to cut off the lower band in the classical way, and the rest of FIR? That is, if the delay is too long.
I'm trying to build an add-on card for s/e to balanced, and since I ordered the card without the audio connectors, I would like to design my pcb to simply "plug in" to the Najda. The layout shows the hole positions, but it would be nice to know the exact connector type so I can place the connectors on the PCB in Eagle and make life easy. Nick, can you help?
Hi, I would indeed recommend that you check rePhase, which has proved to be great for generating FIR coefficients. I made a brief tutorial here, which could be of interest for you because the fictional example is a 3-way system.
Regarding the amount of delay induced by FIR filtering, this is totally depending on the coefficient set itself. As a general rule however, you must expect a much larger delay with FIR than with IIR.
Currently there's no option in the program for mixing FIR and IIR filters. I'm going to release a software revision that will feature a front-end FIR processing followed by a IIR channel processing. The expected use of this strategy is to have all crossover filters and EQs in IIR, and use the FIR front-end for tweaking the phase.
Hi Martin,
Please let me know what connector footprint you need and I'll send this via email.
Nick