Prelude Zorro 2 Amiga sound card - part 2 MPEGit FIX
Вставка
- Опубліковано 9 лип 2024
- When we first put this card together we discovered that the MPEGit addon didn't do a great job at decoding MP3s, the audio it produced was full of interference.
Turns out that issue was caused by the MPEG decoding chip but we have a replacement now so we swap that out but then find a new issue that might be impossible to solve. Does anyone have a source for 14.725mhz oscillators?
Prelude build video - • Building a Prelude Zor...
Links
Prelude github - github.com/wiretap-retro/Prel...
Prelude software - aminet.net/package/driver/aud...
AHI Version 6 - aminet.net/package/driver/aud...
AHI record - aminet.net/package/mus/play/A...
Amiga Amp - www.amigaamp.de/
You may need to install MUI, this is the latest version - github.com/amiga-mui/muidev/r...
@GadgetUK164 Prelude Videos
Part 1 - • Building an Amiga Soun...
Part 2 - • Fixing the MPEG... Bui...
CRG Twitter - / yt_crg
CRG Patreon - www.patreon.com/user?u=13919007
CRG Discord - / discord
Chapters
00:00 Intro
02:14 Swapping the decoder chip
04:09 Magnification
05:23 Is it fixed?
07:08 From cables to pins
08:19 All fixed now?
10:33 Another problem!
14:15 Microphone input
16:54 Conclusion - Наука та технологія
According to the datasheet there are two registers with PLL offset values for 44.1 and 48 kHz. These are loaded with the values for the given 14.750MHz oscillator. They could be modified for other oscillator frequencies. As long as the drivers leave these registers alone you could write a program setting the correct values for the internal processing and the output should be correct.
A neat idea but I wouldn't know where to start. I think for me it will be simpler to try and get the .724mhz oscillator.
Try replacing caps in microphone input op-amp feedback (C41 and C42 in the schematic I have; 1.2nF in value) with 680pF. This should improve high frequency response.
Will do, thanks for the tip
I love to tinker with old machines
They have a charm that never fades
I fix them up with care and skill
And bring them back to life again
Amiga, Atari, Commodore
They are my friends from long ago
They make me happy when I play
The games and programs they can run
I don't need the latest tech
To have some fun and feel fulfilled
I have my old computers here
They are my passion and my joy
Too bad about the MPEGit card but that is still a great soundcard overall. And also thanks from us viewers to Mrs. CRG. Seeing that microscope more in the future will be great.
Yeah the microscope will be very useful I'm sure. Still need to try it's direct record feature but I think it'll work great for future projects.
Hi Glen, the extension ribbon cable problem surfaces elsewhere too, as I found out when trying to extend my CF card from my Terriblefire 1230 out to the back of the case, just wouldn't play ball at all. Had to use a really short cable which didn't reach the back. All academic now though, as I just fitted a PiStorm32 lite with pi4 , and have a SD/USB/HDMI plate there now.
Yeah it seems like the closest you can get it 14.71 or 14.74MHz, in 3.3v form and SMD for expensive money. I looked to see if one could get a 29.45MHz oscillator as it wouldn't be too hard to make a frequency divider, but alas, same thing.
Your best bet is going to be a programmable PLL of some flavor that can output the desired frequency if you want playback to be spot on. The other thing to check out is if the audio track you're playing back is VBR, some MP3 decoder chips don't like VBR too much for some strange reason and will run those files slower.
We've found a programmable oscillator from digikey that should do the job. It's a little expensive but I'm keen to know if it'll fix the issue.
@@CRG could you not use a stripped down raspberry pie Pico, cranked up as fast as it will go, 3.3volts be in the range of just pin connection? on a Pico?
What about a synthetic crystal using a VCO with a MOTOROLA Synthesizer chip and a 10.240Mhz crystal should allow you to generate the required clock signal.
Another remark: Perhaps you can use the SIT8008BI-23-33E-14.725000, but it's SMD.
An smd part would be fine, I'm sure I could attach it with wires. But we've found a programmable oscillator that we can get from digikey. Think we might be going that route.
The prelude microphone input is really muffled, almost like there is impedance somewhere.
It would mean unsoldering the MPEGit but I bet you could get away with using a right angled pin header to mount the MPEGit, so long as the Prelude is not flush against the case already.
It’s interesting. Have just seen posted that AHI V4x should be used and NOT AHI V6x for these sound cards 🤔
As I understand, that's correct for the original driver but for the new driver ahi6 is preferred.
What is the sample rate you record the mic in signal at? Because it sounds like 11Khz
It almost sounds like it's being recorded by the Paula sound chip and not the Prelude card. I wonder if there is a software setting that could be twiddled... maybe a pass-thru is turned on...?
Paula can not record.@@LeftoverBeefcake
Didn’t the Amiga use ~8.5kHz? But I’d hope an expanded system like this wasn’t limited to that.
The prelude can do up to 64 kHz playback and record amiga resource says. But I know from memory that it can do 44kHz recording with ease and without any cpu overhead. @@kaitlyn__L
There is no setting in that ahi recording software to set the sample rate, it's just recording a raw wave file. I have tried some other more professional software and it doesn't sound any better. I'm going to try tweaking the caps on the mic input (as suggested in another comment) to see if I can improve it.
Better to press down bent pins with tweezers rested on the table.
Good idea, I'll keep that in mind for next time 👍
Ribbon cables caused me endless problems. A needle nose pliers across all side of legs works well for straightening purposes.
Bit of a shame I can't get this mounted as I wanted to show it off but these things happen, it was never meant to be mounted like that. At least its working now.
I'd have to get myself some better pliers to equally grip the pins on something so small but yeah, with a decent set I'm sure that'd work.
The crystal should be in circuit with a couple of small surface mount capacitors. If you change the value of these caps it might alter frequency and speed of the music but it might not as the frequency might depend upon the fixed frequency of the crystal and if you change the caps to different values it might just not work. If you do end up changing the crystal to one with a different frequency, you might have to change also any caps in parallel with it
It has complete oscillator module and not just bare crystal. I doubt it's tunable.
That four pin oscillator module is not tunable, the only cap useful/required is the typical 100nF bypass capacitor for the supply pins...
As the others have said this oscillator can't be tuned but we've found a programmable part that we can order to whatever frequency we want so hoping that'll fix the issue.