Finally! No mouser package, but I was able to find all the needed ceramic caps in my junk boxes, and the last missing item, the sliding switch, arrived today from China, so my all-socketed C64 is finally up and running with a new RF modulator replacement. I'll have to submit to the same treatment one of my shortboards next year.
As soon as they're available again, I'll also buy a lumacode encoder for this C64, and will probably go for a sidkick pico too. The SID in here doesn't sound terribly good.
I also built a few KungFu Flash carts, so it's time to play around a bit... Any pointers to nice C64 games on itch.io, people?
=> More informations about this toot | More toots from hkz@chinwag.org
Also, why not replacing the DRAM with SRAM?
I used this for the carrier: https://github.com/go4retro/DRAMCarrier (there is an error on this board: pin 1, which is /OE, of one of the ICs is left floating, I had to jumper it to GND, like it is in a C64C) and https://github.com/jamarju/c64-sram (here the BOM lists an HCT32 to handle /RAS and /CAS but I had timing issues with that, the issues went away with an F32).
=> More informations about this toot | More toots from hkz@chinwag.org
Further work on the experimental C64, to attempt lowering the temperature inside the case: removal of VR2 and replacement with a switching module from Murata (I have an abundance of these). I also added some additional filtering with ceramic caps on the bottom, and, while I was there, replaced the 10uF caps across the board with newer, higher rated ones.
=> More informations about this toot | More toots from hkz@chinwag.org
Oh I ran into a small issue with the SRAM replacement board from the previous post!
See those two tiny pixels under the cloud? At the main screen of ZetaWing II there is a segment of flickering pixels just under that cloud. The issue disappears when using the original 4164, or even when replacing them with two 4464, so I'll go with the second option.
=> More informations about this toot | More toots from hkz@chinwag.org
I'm planning to get a few sidkick picos built by PCBWay, but while rummaging around looking for ICs to test I found an armsid I ordered as a spare in a lot that I got to fix up a friend's C64, so I decided to pull the original from this machine, store it away safely and install/update/test this one.
It's really, really good, at least to my untrained ears. Might just use the skpicos on other machines and leave this one in here.
=> More informations about this toot | More toots from hkz@chinwag.org
I checked a few photos of other SRAM adapters for the C64, and they all share the same basic circuit, latching the address lines on /RAS and enabling the SRAM when both /RAS and /CAS are low, so, I guess the problem with the one I found on github are timings. Either the SRAM is too fast or the glue logic is.
So I designed an experimental board for my C64, supporting both DIP ICs and a few SMD variants that I already have in stock. We'll see if I find a glitchless combination!
=> More informations about this toot | More toots from hkz@chinwag.org
Ok, SRAM replacement for the #C64 250407 built in two versions and under test!
Good news is it works!
Bad news, the damn glitch on the title screen of ZetaWing 2 persists! Doesn't matter the timing of the SRAM or the family of the ICs I use.
At this point I need to ask, does anyone here has a PAL C64 with a SRAM mod installed AND ZetaWing 2? I guess it'll be heavily dependant on the motherboard model.
I'd like to know how it behaves for you. My glitch is a flickering lines of pixels under a cloud in the title screen of the game.
=> View attached media | View attached media | View attached media
=> More informations about this toot | More toots from hkz@chinwag.org
Found out that using an HCT to OR the /CAS and /RAS lines for the SRAM module leads to instability on this C64. They are gone with an ALS/LS or F gate.
The same was happening with the other SRAM module I used, found on GitHub, designed for the C64 shortboards. In that one I replaced it with an F gate.
Looking at the schematics (https://github.com/bwack/C64-250407-Replica-KiCad/blob/main/250407_.pdf) and saw C203 in my board is not fitted, I tried with a random 330pF cap directly soldered on the adapter between /CAS and GND. It is now rock solid. I'll do another revision with a proper place for the cap.
Also, seems I am not the first with the issue: https://www.lemon64.com/forum/viewtopic.php?t=80031
=> More informations about this toot | More toots from hkz@chinwag.org
Ah interesting! I modified this C64 motherboard to accept 8565 VICs, and with this, the graphical glitches when using a SRAM adapter disappeared! The previous VIC was a 6569r3.
=> More informations about this toot | More toots from hkz@chinwag.org
@hkz very interesting. I am working on a MIDI synthesiser with two SwinSID’s. They work fine, but I hear a lot of aliasing on filters. Most comparison videos I’ve seen test the different SID replacements only with complete tunes (that I don’t know). It would be easier to compare them using simpler sounds showcasing problems (aliasing, etc…)
=> More informations about this toot | More toots from davbucci@mastodon.sdf.org
@davbucci If you can pull out a program that generates a sequence of the effects you're interested in, I can record the output on this machine and send it your way, and (provided the assembly order on PCBWay goes through and the courier does not lose my package) when I get the sidkick picos I can repeat the recording.
=> More informations about this toot | More toots from hkz@chinwag.org
@hkz thank you, I’ll try to write something in the next few dans. I have a C128D with the 8580 I think, I’ll try there, too.
=> More informations about this toot | More toots from davbucci@mastodon.sdf.org
@davbucci This device (and the sidkick pico too) can be set to emulate either type of sid, so I guess I should record both modes.
=> More informations about this toot | More toots from hkz@chinwag.org
@hkz If you have some time, can you please record with this program?
10 b=54272
14 pokeb+21,0:pokeb+22,78:pokeb+24,31
16 pokeb,176:pokeb+1,2
17 pokeb+5,43:pokeb+6,2
60 for t=1 to 4
70 poke b+4,17
80 for i=1 to 2000:next
90 poke b+4,16
100 next
It should play four notes. I tested on Vice with the 8580 emulation and some aliasing is clearly present (I had to increase the volume).
=> More informations about this toot | More toots from davbucci@mastodon.sdf.org
@hkz this is the output of VICE, emulating the 8580. I had to increase the volume as it was very low. I will give a try with the real thing on my C128D in the C64 mode.
=> More informations about this toot | More toots from davbucci@mastodon.sdf.org
@hkz This is the output of one of my SwinSIDs (on the MIDI2SwinSID project). The settings should be more or less the same. The aliasing is evident.
=> More informations about this toot | More toots from davbucci@mastodon.sdf.org
@davbucci The ARMSID allows for tweaking of the filters in its configuration.
For this test, I left everything as default. I recorded these directly from the C64, using the dedicated audio out plug in the replacement RF board I installed.
First recording: ARMSID set to 6581.
=> More informations about this toot | More toots from hkz@chinwag.org
@davbucci Second recording. ARMSID set to 8580.
=> More informations about this toot | More toots from hkz@chinwag.org
@davbucci Third recording. ARMSID set to 8580 but with ADSR fix enabled in its configurator.
=> More informations about this toot | More toots from hkz@chinwag.org
@hkz thank you very much! I haven’t yet listened to them properly, but even though the speakers of my cellphone they sound much better than the SwinSID! Can I include them in a blog post?
=> More informations about this toot | More toots from davbucci@mastodon.sdf.org
@davbucci Absolutely free to do so. Please, let me know if you need further recordings!
=> More informations about this toot | More toots from hkz@chinwag.org
@hkz amazing, thanks. I summarized here some results: http://davbucci.chez-alice.fr/index.php?argument=varie/sidcmp/sidcmp.inc Just in case, would it be possible for you to record the results with a real 6581 SID?
=> More informations about this toot | More toots from davbucci@mastodon.sdf.org
@davbucci Certainly. Note that I have connected EXTIN to GND with a 220 ohm resistor for this one, as it reduces noise.
=> More informations about this toot | More toots from hkz@chinwag.org
@davbucci Ah, maybe it is important to point this out. The tests were made with an ARMSID with firmware 2.16 (https://www.nobomi.cz/8bit/armsid/index_en.php).
PCBWay confirmed my order for the sidkick pico, so once they get here (and I get them programmed, as they need an usb adapter for that) I'll perform the same recordings too, unless someone gets there first.
=> More informations about this toot | More toots from hkz@chinwag.org
@hkz done! By the way, the ARMSID is truly good to my ears. It seems to be the best solution, for the moment. BTW the board that controls the two SwinSID's in my synth is already an STM32, the combination with an ARMSID would make sense 😅 Thank you very much for your recording. I could not hear clear comparisons in such simple cases. The difference appear to be much more evident than with complex tunes, especially for noise and aliasing.
=> More informations about this toot | More toots from davbucci@mastodon.sdf.org
@davbucci You're welcome. I'm quite curious about the sidkick pico and its reported ability to emulate two SIDs at once, which could be handy for your project too!
=> More informations about this toot | More toots from hkz@chinwag.org
@hkz thank you very much. I added the recording to the blog post. I will try to put the same resistor on EXTIN on the audio/video connector, to reduce noise on my C128D, too. Until today, I never listened carefully to it using my mixer, but it is indeed very noisy and everything to reduce it is a good idea.
=> More informations about this toot | More toots from davbucci@mastodon.sdf.org
@davbucci SIDKick Pico with DAC, set for 6581, FW version 0.21
=> More informations about this toot | More toots from hkz@chinwag.org
@davbucci SIDKick Pico with DAC, set for 8580, FW version 0.21
=> More informations about this toot | More toots from hkz@chinwag.org
@davbucci SIDKick Pico with DAC, set for 8580 + digiboost, FW version 0.21. I don't think there's any difference in this one.
=> More informations about this toot | More toots from hkz@chinwag.org
@hkz thank you very much! I’ll add the files on the website within a few days (I have a quite crazy week).
=> More informations about this toot | More toots from davbucci@mastodon.sdf.org
@hkz I listened more carefully. With the digiboost, you hear a small "click" at the very beginning, when the volume is set to 15. That seems logical.
=> More informations about this toot | More toots from davbucci@mastodon.sdf.org
@hkz I could find some time to update the page: http://davbucci.chez-alice.fr/index.php?argument=varie/sidcmp/sidcmp.inc The SIDKick seems to be a really good alternative to the ARMSID. I really like the very clean 8580 version.
=> More informations about this toot | More toots from davbucci@mastodon.sdf.org
@davbucci one thing I forgot to mention. With the skpico, you can disable the output that goes through the motherboard circuitry and feed directly from the DAC. I did this in my installation, and I suspect that accounts for some of the cleanliness.
Also, you can use a single skpico to emulate a dual SID configuration, provided the additional address lines are wired.
=> More informations about this toot | More toots from hkz@chinwag.org
@davbucci As for the SIDKick... pcbway is still reviewing my order, so there is still a long way to go before I can test that.
=> More informations about this toot | More toots from hkz@chinwag.org
@davbucci @hkz SWINSIDs are pretty bad in regard to filters. ARMSID is in my opinion currently the best bang for the buck if you want to go replacement SID. Costs about 35€ and has very very good emulation, including filters.
=> More informations about this toot | More toots from root42@chaos.social
@root42 @hkz the SwinSID filters are abysmal 😩. As a comparison, here is the output of a real C128D, with a SID 8580. It's low (as in the emulator) and quite noisy, but no aliasing can be heard (as expected, as filters are analog, there). The SwinSID in the previous example is configured as a 6581, though. I will try in the 8580 mode, too.
=> More informations about this toot | More toots from davbucci@mastodon.sdf.org
@davbucci @hkz I think the tiny MCU on the SwinSID is just too weak for this. The ARMSID has an STM32, which is much more powerful. I guess filters will be implemented as FFTs, so weaker MCU will have less bands and hence aliasing… I guess?
=> More informations about this toot | More toots from root42@chaos.social
@root42 @hkz I thin they are FIR filters, easier to calculate than a FFT. However, it's certainly true that with a better micro the results can be much better. I was just not expecting such as high level of aliasing with the SwinSID. On the other hand, I'm playing with the C128 right now and the real SID is far from perfect, but at least there is no aliasing! 😊
=> More informations about this toot | More toots from davbucci@mastodon.sdf.org
@davbucci No SID is perfect. Which is why we love the SID so much. :) I think it is really THE sound of 80s micros. Amazing what it can do, and the variety of sounds you even get from one SID to the next.
=> More informations about this toot | More toots from root42@chaos.social
@root42 an amazing little device, indeed. I would like to obtain a fun synth to use and I think it is clear now that I have to upgrade from the SwinSID.
=> More informations about this toot | More toots from davbucci@mastodon.sdf.org This content has been proxied by September (3851b).Proxy Information
text/gemini