Shaun, when you say 'max', you mean full bar displayed in the contrast menu, is that right?
How about you Steve? James? Have you also set contrast as Shaun above?
One thing you could try is a capacitor between pin 1 and 3 of the LCD module. Any moderate value should be alright, for example 10 uF. This will convert the PWM into a flat output of varying level - in case your LCD module doesn't like the PWM.
(if you use a polarized cap, then + is on pin 3)
Strangely, just as I was about to pack it up to return it - the display returned to normal. Hmmmm. Its been ok for a few hours now.
However, I have another fault! The remote behaves very weirdly. Sometimes, maybe 1 in 3, the volume control swaps over and UP makes it quieter. and DOWN, louder. I think it is just continuing to adjust in the direction it was previously going. It rights itself with a few presses. OK, it could be my remote (which works fine on the machine that it came from). Unfortunately, I had a household remote purge a few months ago so I don't have another one to try. I'll try to find one soon. But my hunch is that its Nadja playing up...?
Re the LCD. I have my contrast set to maximum. I'll try the cap at some point soon. Actually I was thinking about the display. Eventually what I'd really like is a BIG display for volume. Maybe a line or circle of leds (something artistic?) - its nice to see something from a distance even if its not the actual attenuation. Failing that, a LED for 'remote command received' would be good. Is that possible somehow?
Another question. I have an setup at the moment with 7 filters in total and a bit of delay. Its at 192k. The DSPs are both well over 90%. Is that correct? I'm slightly worried that when I get it all set up with more channels and more delay I could run out...
cheers,
james
However, I have another fault! The remote behaves very weirdly. Sometimes, maybe 1 in 3, the volume control swaps over and UP makes it quieter. and DOWN, louder. I think it is just continuing to adjust in the direction it was previously going. It rights itself with a few presses. OK, it could be my remote (which works fine on the machine that it came from). Unfortunately, I had a household remote purge a few months ago so I don't have another one to try. I'll try to find one soon. But my hunch is that its Nadja playing up...?
Re the LCD. I have my contrast set to maximum. I'll try the cap at some point soon. Actually I was thinking about the display. Eventually what I'd really like is a BIG display for volume. Maybe a line or circle of leds (something artistic?) - its nice to see something from a distance even if its not the actual attenuation. Failing that, a LED for 'remote command received' would be good. Is that possible somehow?
Another question. I have an setup at the moment with 7 filters in total and a bit of delay. Its at 192k. The DSPs are both well over 90%. Is that correct? I'm slightly worried that when I get it all set up with more channels and more delay I could run out...
cheers,
james
ChaparK,
Another request!
I have an unusual horn system. The low-midbass 50-140ish is a huge 5 metre 40hz horn in the roof. But this is mono. So, I have 2 direct radiators which are stereo. I'd like to supply them with the difference between the mono and stereo signal. i.e. LEFT is L-R and RIGHT is R-L. I appreciate that I might be the only person in the world who wants this... but isn't it just 5 minutes of coding? Thanks!
james
Another request!
I have an unusual horn system. The low-midbass 50-140ish is a huge 5 metre 40hz horn in the roof. But this is mono. So, I have 2 direct radiators which are stereo. I'd like to supply them with the difference between the mono and stereo signal. i.e. LEFT is L-R and RIGHT is R-L. I appreciate that I might be the only person in the world who wants this... but isn't it just 5 minutes of coding? Thanks!
james
Last edited:
Aha
That's good to hear, thanks for letting me know.Hmm. So Najda implements the NEC protocol as documented in the literature. We could of course look with the scope at the pulses out of the IR sensor when you press a key on your remote and see whether it's really a NEC - but I suppose it's easier to get just another remote. There are small remotes on ebay for a few bucks - I've tested a couple and they're OK.
Well this is going to be a lot of work as display data is now coded for HD44780 controllers on a 8-bit parallel bus... How about getting a larger LCD display? I think that the ones with white characters on black background have great contrast.
Unfortunately all leds are now in use - I've already implemented the CPU overload warning led we were discussing a few posts back in this thread.
192 kHz is CPU hungry twice as much as 96 kHz and 4 times as much as 48 kHz.
When you load an 'empty' preset, it's actually not empty because the DSPs are managing 4 inputs and 10 outputs. That's 14 delay lines, 24 gains plus the routing section.
Adding delay will not increase the CPU load. Delay lines are already included in the 'empty' presets.
Biquads are implemented in full double precision, i.e. on the feedforward path as well as on the feedback path. This is especially important with higher sampling rates as simple precision coefficients get quickly ill conditioned as you lower the frequency of your filters.
What I could do is allow disabling some channels, so that if you use less than 10 output channels you can turn off the unused channels and recover some CPU for additional filters on the channels effectively in use.
If you have a setup with lots of filters, I recommend that you lower the sampling frequency down to 96 kHz. You can also send me your setup files on my email and I'll see if there's a way to rearrange your filters and achieve better efficiency.
Best,
Nick
You can already do that. In the Input section, DIFF is defined as (Left-Right)/2.
Select the routing accordingly in the Channel Processing tab.
If you want (Right-Left)/2, you just need to invert channel polarity.
This one was easy
Enjoy!
Hi Nick,
Great!
Yeah, and a filterbank with assignable taps so they don't get lost
Regarding LCD I use only half contrast.
I also tested a remote that suddenly adjusted the volume up instead of down. I changed the direction on my hand and it worked normal, but I changed to anoter remote that worked ok.
best,
Paal
Here is a pic of the display line issue
I tried using the basic EQ correction today. Adding some dB's here and there.
I started with the tapped horns. This is in room and with the calibrated mic at ear position.
Before
After - just adding a PEQ correction at 85Hz.
Then I lifted the slight drop off at 20Hz. Deep!
The RH channel needed a cut at 90Hz.
I when through doing this on all the drivers singly and as adjacent pairs. Took a little longer as have to cycle Najda off quite often to get sound.
Changing things and measuring on the fly is clunky - I have disconnect PC to Najda before the PC will play a tone from Holmimpulse through USB to my mic preamp.
2 laptops would ease that - have to get our other one in here on Holmimpulse duty I guess, or is there another way?
Applying this method across the freq range I got a pretty flat measurement on all 4 channels + passive super tweeter.
Still more to do and learn - I've not got into the stuff some of you guys are doing yet - will no doubt and look forward to it.
The main big plus is I applied some correction to the super sensitive JBL2482 and Vitavox S2 compression drivers - making in room flatter. I can detect no degradation in SQ
I tried using the basic EQ correction today. Adding some dB's here and there.
I started with the tapped horns. This is in room and with the calibrated mic at ear position.
Before
After - just adding a PEQ correction at 85Hz.
Then I lifted the slight drop off at 20Hz. Deep!
The RH channel needed a cut at 90Hz.
I when through doing this on all the drivers singly and as adjacent pairs. Took a little longer as have to cycle Najda off quite often to get sound.
Changing things and measuring on the fly is clunky - I have disconnect PC to Najda before the PC will play a tone from Holmimpulse through USB to my mic preamp.
2 laptops would ease that - have to get our other one in here on Holmimpulse duty I guess, or is there another way?
Applying this method across the freq range I got a pretty flat measurement on all 4 channels + passive super tweeter.
Still more to do and learn - I've not got into the stuff some of you guys are doing yet - will no doubt and look forward to it.
The main big plus is I applied some correction to the super sensitive JBL2482 and Vitavox S2 compression drivers - making in room flatter. I can detect no degradation in SQ
Last edited:
I ordered a couple of these:
#2806 Aluminum Preamplifier enclosure /DAC case/ amplifier chassis AMP BOX | eBay
I might be hard to fit a power... We'll see how it will fit
best,
Paal
#2806 Aluminum Preamplifier enclosure /DAC case/ amplifier chassis AMP BOX | eBay
I might be hard to fit a power... We'll see how it will fit
best,
Paal
Look nice. I think I may need wider to fit the expansion board into though. They do a bigger one - price goes up though...
Just EQ'd a load more with 2 laptops - nice and easy
Here's the LH channel all 5 drivers, in room mic at ear position
RH
Both L & R
The usual +/-5dB in room is good is sorted!
Listening it all sound well controlled and nothing lacking right across the rev range. More listening to be done.
About contrast to max:
When contrast is set to max (full bar displayed), then the PWM duty cycle is near to 0. We have reached the contrast limit of the display.
You can check this by connecting your scope on pin 3 of the LCD header and change contrast in the menu.
Basically I cannot increase the display's contrast any further. I've added 2 more settings but they're at the lower end.
You mean more flexibility with tap numbers vs. channels right?
Yup, as Steve said, they're a bit narrow. Internal width is 260 mm and Najda is 230 mm.
This one from the same supplier would suit better if you want to fit also the power supply in:
S-3207 Aluminum amplifier case / Enclosure /amplifier BOX/ PSU Case | eBay
Maybe you can even flip the rear panel to have the power socket cut on the left side
Looks good Steve, thanks for sharing. How many EQs?
Today I looked at this idea that Kazam talked about and I tried to put it together with Shaun's comments.
So here's what the Board Configuration dialogue could look like.
If you want to set a global level, all the sliders are disabled and the level is rated in Vrms:
If you'd like to set various analogue gains for the outputs, then the combobox is disabled and the gains are rated in dB:
This might be part of the forthcoming soft/firm revision.
Best,
Nick
When contrast is set to max (full bar displayed), then the PWM duty cycle is near to 0. We have reached the contrast limit of the display.
You can check this by connecting your scope on pin 3 of the LCD header and change contrast in the menu.
Basically I cannot increase the display's contrast any further. I've added 2 more settings but they're at the lower end.
Hi Nick,
Great!
Yeah, and a filterbank with assignable taps so they don't get lost
You mean more flexibility with tap numbers vs. channels right?
Cool, one normal LCD display out there
Yup, as Steve said, they're a bit narrow. Internal width is 260 mm and Najda is 230 mm.
This one from the same supplier would suit better if you want to fit also the power supply in:
S-3207 Aluminum amplifier case / Enclosure /amplifier BOX/ PSU Case | eBay
Maybe you can even flip the rear panel to have the power socket cut on the left side
Look nice. I think I may need wider to fit the expansion board into though. They do a bigger one - price goes up though...
Just EQ'd a load more with 2 laptops - nice and easy
Here's the LH channel all 5 drivers, in room mic at ear position
RH
Both L & R
The usual +/-5dB in room is good is sorted!
Listening it all sound well controlled and nothing lacking right across the rev range. More listening to be done.
Looks good Steve, thanks for sharing. How many EQs?
Today I looked at this idea that Kazam talked about and I tried to put it together with Shaun's comments.
So here's what the Board Configuration dialogue could look like.
If you want to set a global level, all the sliders are disabled and the level is rated in Vrms:
If you'd like to set various analogue gains for the outputs, then the combobox is disabled and the gains are rated in dB:
This might be part of the forthcoming soft/firm revision.
Best,
Nick
You mean more flexibility with tap numbers vs. channels right?
Yup, as Steve said, they're a bit narrow. Internal width is 260 mm and Najda is 230 mm.
This one from the same supplier would suit better if you want to fit also the power supply in:
S-3207 Aluminum amplifier case / Enclosure /amplifier BOX/ PSU Case | eBay
Maybe you can even flip the rear panel to have the power socket cut on the left side
Today I looked at this idea that Kazam talked about and I tried to put it together with Shaun's comments.
So here's what the Board Configuration dialogue could look like.
If you want to set a global level, all the sliders are disabled and the level is rated in Vrms:
If you'd like to set various analogue gains for the outputs, then the combobox is disabled and the gains are rated in dB:
This might be part of the forthcoming soft/firm revision.
Best,
Nick
Hi Nick,
Yes, more flexibility. With a 2-way I am not able to use more than about 3000 taps of 9000? available.
I didn't see the wider one
OK I have to live with that now.Great news on gain structure optimizing. I had to attenuate my tweeter digitally with 11dB. Now I can do that on the output. How would the output graphs look like then? Would you include the analog attenuation?
best,
Paal
OK I'll do that.
Well, no. The graph shows the Setup file expected outcome, meanwhile the output level is a board setting. A Setup file is a collection of DSP parameters, so analogue gains won't show up in the graph.
It's not always clear what should be seen in this graph and what should not be included.
For example, you had this FIR LR approximation with one 511-tap LF and one 255-tap HF.
We found out that it's better to take the delay into account when computing the phase otherwise the Electrical Sum looks ugly.
So I changed this and now the Sum of your FIR looks beautiful, that's great.
But that will maybe introduce problems for other users who keep to IIR filters because they use delays for another purpose. So, if you design a IIR LR4 crossover, you expect that the bands sum flat, right? But if you have added delay on one channel, then it's not going to sum flat anymore.
The above is right because we make an electrical sum. It's like if you were taking the output channels and summing them in the analogue domain - meanwhile when you add delay in a IIR setup, you do that in order to time-align your drivers.
(Anyway, that's a long talk I've been having with myself and it's not really clear to me
)That sounds good. You're more than safe re. CPU. A quick computation shows that if both DSPs are at 57%, then you still have enough resources for over 70 EQs (at 96 kHz).
OK that's in now, all tested and seems OK. I'll throw in a few more things and we'll make the release that will include a fix for this annoying bug that you have.
I'll see if I am up for the display swap today
So how did it go?
(JCGA has a great LCD: http://www.diyaudio.com/forums/digi...sp-xover-project-part-2-a-61.html#post3407545)