Hi James, good to hear from you. Congrats for the build!
I suppose I could do that sticker thing. I normally supply a sheet with power supply polarities. Didn't you get yours?
If I ever make another batch, I'll modify the silkscreen to have it more explicit as recommended by Shaun in his awesome beta-tester report.
According to your link, this remote uses a NEC-based protocol so it could actually work with Najda. Have you tried it?
Unfortunately I don't have one myself but if it's indeed NEC-based then there must be a way to make Najda compatible.
So you would have one subwoofer channel and you'd like to be able to adjust the gain on that channel. Is this what you have in mind?
Thaks James, very glad you like it.
Your post was great Shaun. I admit silkscreen could have been more explicit.
I think it looks good. The only info I couldn't find in the datasheet is the switching frequency - but most of the modern switching supplies have it well above the audio band. If you get this one, please let me know how it goes.
ChaparK,
Yes, I did get the sheet with the connection polarities - it saved me (and the board!) - but if it was half an hour later and I was just a bit more more tired...
The Apple remote is not NEC after all. (Nadja doesn't recognise it at all) The comments from Bruno on HifiDunio explain it I think..
"The big “aha” missing from this discussion comes from the assumption that Apple codes are NEC format. They’re not. The next assumption is that the button codes are 8 bits. They’re not.
Apple’s IR format is custom and only “based” on NEC. They have taken the premise and re-used all the command bits in a custom fashion.
7 bits for the button code
1 PARITY bit
8 bits for ID
The parity is a XOR sum of the button and ID bits, and since it flips around between 1 and 0 depending on the other bits, of course you’ll think the codes are different between two remotes with two different IDs."
So you would have one subwoofer channel and you'd like to be able to adjust the gain on that channel. Is this what you have in mind?
Exactly. Similar to the bass/treble control (which I don't use) but just affecting the sub output. At the moment I have presets set up for different sub levels (thanks Shaun) but it would be much better to have continuously variable control.
cheers J
Yes, I did get the sheet with the connection polarities - it saved me (and the board!) - but if it was half an hour later and I was just a bit more more tired...
The Apple remote is not NEC after all. (Nadja doesn't recognise it at all) The comments from Bruno on HifiDunio explain it I think..
"The big “aha” missing from this discussion comes from the assumption that Apple codes are NEC format. They’re not. The next assumption is that the button codes are 8 bits. They’re not.
Apple’s IR format is custom and only “based” on NEC. They have taken the premise and re-used all the command bits in a custom fashion.
7 bits for the button code
1 PARITY bit
8 bits for ID
The parity is a XOR sum of the button and ID bits, and since it flips around between 1 and 0 depending on the other bits, of course you’ll think the codes are different between two remotes with two different IDs."
So you would have one subwoofer channel and you'd like to be able to adjust the gain on that channel. Is this what you have in mind?
Exactly. Similar to the bass/treble control (which I don't use) but just affecting the sub output. At the moment I have presets set up for different sub levels (thanks Shaun) but it would be much better to have continuously variable control.
cheers J
It has 70kHz switching frequency according to this document which I think is an OEM version: Kepco
I will get it.
/K
I am not having any luck in getting my One for all remote working on any of the JVC or NEC codes supplied.
This brand new remote successfully controls our TV and DVD etc, but not Najda.
I am using a Vishay TSOP38238 IR sensore wired as instructed - output is middle, +5V is near the DSP debug port. No heat from the sensor or board.
I have tried all codes in "Press any key on JVC/NEC" and just on powered up mode, and even the learn method - no joy.
Anyone else had success / tips?
Still playing great music and amazing BASS of course...
This brand new remote successfully controls our TV and DVD etc, but not Najda.
I am using a Vishay TSOP38238 IR sensore wired as instructed - output is middle, +5V is near the DSP debug port. No heat from the sensor or board.
I have tried all codes in "Press any key on JVC/NEC" and just on powered up mode, and even the learn method - no joy.
Anyone else had success / tips?
Still playing great music and amazing BASS of course...
What bollocks: After checking again and again, and having no heating up components, I finally soldered in the 2nd of the IR sensors I bought. Bought 2 just in case
Only worked perfectly on the first fecking JVC code I tried! It worked on 10653 on the One for all remote series BTW.
I never doubted you Nick - or your HW/SW, honest
Luverly, Nice remote action / speed too.
BTW, time to delete all the pass through presets - don't want to cycle to one of them by mistake now...
Source and setup cycling XO setups on demand - nice.
Only worked perfectly on the first fecking JVC code I tried! It worked on 10653 on the One for all remote series BTW.
I never doubted you Nick - or your HW/SW, honest
Luverly, Nice remote action / speed too.
BTW, time to delete all the pass through presets - don't want to cycle to one of them by mistake now...
Source and setup cycling XO setups on demand - nice.
Last edited:
Hi Steve, thanks I got your files and I confirm that on occasion there's no sound when switching between your presets.
I'll push out a release that fixes that - but I'd also like to include a few more things in it so I hope you can live with that until the next soft+firm version is made available. It won't be long.
just to add:
Now I have got the remote working, just cycling through the preset files on the remote causes pretty much every other one or every 2nd one to cause no sound...
Carrying on selecting next or previous ones gives sound again.
Last edited:
Alright, I'll make Najda compatible with the Apple remote controller. It's just a question of code so no problem. Please give me some time until I get such a remote controller from a friend on a loan otherwise I'll just buy one, but you can expect soon a new menu item "Learn Apple" in the relevant section.
Hmm this one is more complex.
For example, maybe you use just one subwoofer.
Maybe someone would like to have the same option with two subwoofers.
Also, how can the application know on what channel your subwoofer is connected to?
Finally, I suppose you'd like to be able to alter the subwoofer's level from your remote controller. But the micro (until now) is not really aware of the details of the presets running in the DSPs.
As a conclusion, I don't want to say that we can't make it. It's all the opposite. But we need to find a way to implement this in a logical/sensible fashion. I'm opened to suggestions.
That's great. This datasheet that you posted is actually quite interesting. According to Mouser's datasheet in your first PSU related post, there was no minimal load. With this second link we get the details: Note (5) For +5V output current from 0 to 0.5A, ±12V output current is 80% of rated value.
What this means is that if you draw less than 0.5A from the main 5V supply, then you can't draw as much current from ±12V as per spec.
That's great cause we don't care about that:
- In Standby mode, Najda draws near to no current from the ±12V supply.
- In normal operating mode, we draw more current because we need to supply the analogue part of the ADC/DACs (CS42428) and optional analogue volume controller (CS3318) which are held in reset state otherwise when Najda is in StandBy.
So this is a note for you Steve
I suggest you look at how your power supply is behaving when there's no load at all because keeping a dummy load is unaesthetic, don't you think? What bollocks: After checking again and again, and having no heating up components, I finally soldered in the 2nd of the IR sensors I bought. Bought 2 just in case
Only worked perfectly on the first fecking JVC code I tried! It worked on 10653 on the One for all remote series BTW.
I never doubted you Nick - or your HW/SW, honest
Luverly, Nice remote action / speed too.
That's great Steve, happy to hear it's all going well. You must have been quite unlucky on getting a faulty sensor.
I'll fix that, have no worry. Next Monday the latest. Cheers and sorry for that annoying bug. In the meantime you could work on a decent casing
An option to bypass the EQ (each EQ section has its own bypass button)would be appriciated.
Alright, I put that on the wish list
Alright, I'll make Najda compatible with the Apple remote controller. It's just a question of code so no problem. Please give me some time until I get such a remote controller from a friend on a loan otherwise I'll just buy one, but you can expect soon a new menu item "Learn Apple" in the relevant section.
>>>Oh please don't - I don't subscribe to the Apple mania
So this is a note for you Steve
>>> Agree, unaesthetic but it's like I am driving a Bentley and destroying the environment. I actually save lots of watt over my previous Dac and tube pre. So saving money now
That's great Steve, happy to hear it's all going well. You must have been quite unlucky on getting a faulty sensor.
>>> Yeah bit of a bugger but these things keep our grey matter tip top!
I'll fix that, have no worry. Next Monday the latest. Cheers and sorry for that annoying bug. In the meantime you could work on a decent casing
>>> Great. Case - yes, only a matter of time as Najda works so well. Perhaps I wait until some expansion board dimension are published.
Measure twice, cut once
Last edited:
Great news about the apple remote... I promise I don't have ANYthing to do with apple mania! This is the only bit of apple equipment I own and it will give me extra pleasure to use this beauty in a system not designed by apple.
My module suddenly developed a strange problem. The LCD screen is all jumbled. And only displays on the 1st line. (the 2nd is blank). You can only make out some characters sometimes. This happened while playing music and no adjustments were being made. Everything else is functioning perfectly.
So, i have tried the following.
1. replace display. no change.
2. check and then replace all LCD wiring. no change! grrr....
3. power down and up in all the combinations I can think of. no change...
I'm rather at a loss to know what to do! Any ideas?
My module suddenly developed a strange problem. The LCD screen is all jumbled. And only displays on the 1st line. (the 2nd is blank). You can only make out some characters sometimes. This happened while playing music and no adjustments were being made. Everything else is functioning perfectly.
So, i have tried the following.
1. replace display. no change.
2. check and then replace all LCD wiring. no change! grrr....
3. power down and up in all the combinations I can think of. no change...
I'm rather at a loss to know what to do! Any ideas?
This maybe connected / may not.
I have the top line of pixels missing from the 2nd row on my display. Been like that from first use.
In the beginning it was a bit intermittent. Changing the contrast cured it for a while (doesn't sound like a wire loose) - now it does not, the top line is always missing.
I can read the 2nd row words ok for now.
I will check the solder joints and I have a 2nd display to try too (The one I really want to use), and let you know.
A bit of background on how the display is working might help you to debug LCD issues.
Data is transmitted from the micro to the display via a 8-bit parallel bus. This bus contains the lines labelled DB0 to DB7 in the StartUp Guide, section I.4.
Najda documentation
Other lines R/W, E and R/S allow handshaking with the display controller.
It's important to understand that the micro doesn't 'draw' the characters that are displayed on screen. These characters are indeed contained in the memory of the display controller itself, and the micro just says 'take the character defined at this memory location and display it at this position'. This explains why different LCD screens might have different 'fonts', because the way these fonts look like is defined in the display controller.
So if there's a faulty pixel or a faulty line of pixel, then it's for sure that the display itself is faulty because the micro cannot address a single pixel or pixel line.
So as far as I can understand from your problem Steve, it's not related to Najda but rather to the LCD module - because the display receives the command, it just doesn't dispaly the data properly.
The Contrast input of the LCD is level sensitive. The easiest way to control contrast would be to connect a pot between GND, 5V and the LCD contrast input.
On Najda, contrast is controlled by the micro, and the pot is hence replaced by a PWM signal. Contrast is adjusted by varying the PWM duty cycle. Adusting contrast will not cure a faulty display, although one might think for a short while the opposite - but it's more of a placebo effect.
Data is transmitted from the micro to the display via a 8-bit parallel bus. This bus contains the lines labelled DB0 to DB7 in the StartUp Guide, section I.4.
Najda documentation
Other lines R/W, E and R/S allow handshaking with the display controller.
It's important to understand that the micro doesn't 'draw' the characters that are displayed on screen. These characters are indeed contained in the memory of the display controller itself, and the micro just says 'take the character defined at this memory location and display it at this position'. This explains why different LCD screens might have different 'fonts', because the way these fonts look like is defined in the display controller.
So if there's a faulty pixel or a faulty line of pixel, then it's for sure that the display itself is faulty because the micro cannot address a single pixel or pixel line.
So as far as I can understand from your problem Steve, it's not related to Najda but rather to the LCD module - because the display receives the command, it just doesn't dispaly the data properly.
The Contrast input of the LCD is level sensitive. The easiest way to control contrast would be to connect a pot between GND, 5V and the LCD contrast input.
On Najda, contrast is controlled by the micro, and the pot is hence replaced by a PWM signal. Contrast is adjusted by varying the PWM duty cycle. Adusting contrast will not cure a faulty display, although one might think for a short while the opposite - but it's more of a placebo effect.
I have noticed that when I am looking at the display with the contrast down, or at an angle for which the contrast is not optimised, parts of the displayed characters can appear dim or missing.
Also, it is possible for noise on the 5V supply to cause the display to go erratic; a poor connection can have the same effect, and so can a drooping supply... just some ideas to check out.
Also, it is possible for noise on the 5V supply to cause the display to go erratic; a poor connection can have the same effect, and so can a drooping supply... just some ideas to check out.
Cool - display change it is - Take the damn thing back to Maplin and get a refund...
Thx. I have the contrast on full and no top line of pixels.
I'll check the wiring out of course but it's not erratic - it's permanent.
Alright, in my previous post I was making the assumption that contrast could be properly adjusted.
It could also be that the range is not sufficient to cover the various models of displays on the market. So I'll extend this range to see if it makes any improvement, how about that?
So currently you can adjust contrast in 10 steps. I've set the range so that all my LCDs have optimal contrast when contrast setting is mid-way. I have 2 green and 2 blue displays. I have another white on black that I haven't wired yet.
Collecting data could also help: how did you set contrast for your display?
Steve has the contrast set to max. How about other Najda users? Would be great if you could let me know.