7MHz MSX Upgrade - Trouble Shooting

Pagina 1/5
| 2 | 3 | 4 | 5

Door Colemu

Hero (531)

afbeelding van Colemu

09-01-2013, 18:41

Thanks to help from RetroTechie, I managed to get my MSX2 HB-F1XD running at 7MHz.

There are a couple of issues still to address:

1) Sound is distorted (higher pitched)
2) MegaFlashROM SCC+ is no longer recognised

At 1st, the display was Black & White, but after connecting a constant 3.58MHz signal it was fine.

I know some of you guys have experience with these upgrade kits, and wondered if you can help me trouble shoot these problems?

Cheers Gents

Aangemeld of registreer om reacties te plaatsen

Van Meits

Scribe (5651)

afbeelding van Meits

09-01-2013, 19:09

The high pitch is normal. Only a Moonsound has its own clock. Other music expansion and the internal PSG will pitch up. In the 90s there were hardware technicians who mounted a correct crystal to those expansions to somewhat keep them play ok at 7MHz, though some notes were skipped nevertheless. I had my MSX Audio fixed that way. I don't know if any technician can/will still perform this mod. Neither do I know if those crystals are still around.
The PSG will be tricky.

I don't know if your MegaFlashROM SCC+ can be compared with my Konami Ultimate Collection (iirc they're built by the same guy and have the same internals except for the bigger rom in the K.U.C). This cartridge refuses to work on 7MHz here as well. The maker said that the games weren't intended to be played on 7MHz anyway. I kinda accepted that.

Van Colemu

Hero (531)

afbeelding van Colemu

09-01-2013, 20:19

Thanks Meits, not what I expected to hear though Shocked!

It sounds f*c*ing awful!! Wrong notes, out of tune, games would be pretty unplayable with 7MHz enabled. The sound would drive me round the twist! oO

I think the sound is generated by the S1985 (SSG sound), and I can't see how I could connect a 3.58Mhz constant signal to correct the problem. So I guess you're right Crying

I thought I might have to connect one of the unused diodes of the 7MHz PCB to a CS signal on the S1985 to correct the cartridge problem. But, as you say, other cartridges work. So I not sure doing this will make any difference Question

Thinking about it, wouldn't I be able to connect the 3.58MHz signal from the VDP to PIN30 of S1985 to correct the sound problem? Hmmmmm Question

Van Colemu

Hero (531)

afbeelding van Colemu

09-01-2013, 22:41

I guess not, just tried it, black screen. I'll live with the crap sound....

Unless you can think of anything RetroTechie Tongue

Update: just noticed this video on Youtube: http://www.youtube.com/watch?v=3c3lv83wKJU

the sound does not change with turbo enabled. How did he do this?? :-?

Van Colemu

Hero (531)

afbeelding van Colemu

10-01-2013, 00:34

Well that video spared me on.

I've solved the sound problem. Here's what I did:
Cut the trace on pin 30 of s1985 (removed from Turbo circuit) and replaced with 3.58Mhz from PIN8 of VDP.

At first this didn't work, but I remembered that my SRAM was wired up to CAS. I figured it wouldn't work in DRAM config, do wired my memory up to SLT33. It worked like a dream! Big smile

Now PSG sound is exactly the same in either normal or turbo modes. (shame about SCC & MSX Music, but I think this is normal for turbo mode, thanks for explaining this Meits)

Now to figure our if there is a way to get my MegaFlashROM SCC+ cart working.... Mmmmmm Question

Van Colemu

Hero (531)

afbeelding van Colemu

10-01-2013, 12:49

Didn't bother looking into the Cart problem as my other MegaFlashROM (SCC version) works fine.

I did come accross another problem, however. The computer works perfectly for about the 1st 10/15 minutes,
then it starts to reset and have booting & drive access problems.

The thing is, it works forever if you are running from cartridge!

Has anyone any ideas what may be causing this? If so, how to fix it? Question

Strange oO

Van Colemu

Hero (531)

afbeelding van Colemu

10-01-2013, 17:55

Here is a video of the thing working: http://www.youtube.com/watch?v=bF1LfwrYuAk
Only need to figure out why Floppy Disk Drive problems remain at 7MHz? :-?

Van edoz

Prophet (2179)

afbeelding van edoz

10-01-2013, 19:07

Cool! nice LED you build in...
i also have a 7 mhz kit in my 8280 but the sound is also the bad thing... hmmm...this is mutch better i think..

Van Colemu

Hero (531)

afbeelding van Colemu

10-01-2013, 22:17

edoz wrote:

Cool! nice LED you build in...
i also have a 7 mhz kit in my 8280 but the sound is also the bad thing... hmmm...this is mutch better i think..

Yeh, the PSG sound was annoying. This thread helped me a lot when fixing it: Hannibal
http://www.msx.org/forum/msx-talk/hardware/trying-msx-super-...

Only works if your machine uses SRAM.... (mine had DRAM, but I changed it)
Shame the HB-F9 doesn't have a floppy drive, or the solution to my problem might have been there too!!

I would love to know how to mod my MegaFlashROM so that it has it's own internal clock, Konami SCC games sound annoying too at 7MHz ;(

Van RetroTechie

Paragon (1563)

afbeelding van RetroTechie

11-01-2013, 08:51

Colemu wrote:

Only need to figure out why Floppy Disk Drive problems remain at 7MHz? Question

There are some software delays in the diskROM that become to short @ 7 MHz, causing various errors (disk offline, sector markers misread, etc). To fix, the diskROM chip select signal must be connected to one of the diodes on the 7 MHz circuit. Effect on speed is minimal since disk I/O doesn't go faster @ 7 MHz anyway. You do have this diode connection, right?

If not: a bit tricky to do in this machine since all ROMs are integrated in one chip... I can't tell from the schematic, but OpenMSX config for this machine says diskROM is in slot 3-0? Don't trust me on that, check using MSXMEM for example! If so, /SLT30 signal can be picked up from S1985 (pin 53, not easy to solder to for beginners), IC13 (74LS32) pin 12, or solder somewhere to the circuit trace in between. Again: don't rely on schematic or me, verify those connections on the machine itself!

Quote:

I would love to know how to mod my MegaFlashROM so that it has it's own internal clock, Konami SCC games sound annoying too at 7MHz Crying

You could modify that cartridge. But another way to fix this is wire the clock pin on one cartridge slot to constant 3.58 MHz (preferably via a buffer like what was used before), and to 3.58/7 MHz CPU clock signal on the other cartridge slot.

That should fix most sound issues for any cartridge you put in that slot, without any need to modify those cartridges. And you still have the CPU clock available on the other cartridge slot in case a cartridge needs it (in particular: DRAM-based RAM cartridges). If you do this: consider in which cartridge slots you're most like to put those sound or RAM cartridges.

Van Colemu

Hero (531)

afbeelding van Colemu

11-01-2013, 17:31

RetroTechie wrote:

To fix, the diskROM chip select signal must be connected to one of the diodes on the 7 MHz circuit. Effect on speed is minimal since disk I/O doesn't go faster @ 7 MHz anyway. You do have this diode connection, right?

I haven't connected anything to the free diodes on the Turbo PCB as yet. I can use one to connect to diskROM chip select signal (assuming it's found)

Quote:

If not: a bit tricky to do in this machine since all ROMs are integrated in one chip... I can't tell from the schematic, but OpenMSX config for this machine says diskROM is in slot 3-0? Don't trust me on that, check using MSXMEM for example! If so, /SLT30 signal can be picked up from S1985 (pin 53, not easy to solder to for beginners), IC13 (74LS32) pin 12, or solder somewhere to the circuit trace in between. Again: don't rely on schematic or me, verify those connections on the machine itself!

MSXMEM confirms that DK2 is at 30 (MSXMEM shows there are positions 0,1,2,3 within 30 - DK2 is at position 1)
Soldered free diode to PIN 12 of IC13 (as this confirmed same point as PIN53 of S1985) but computer won't boot? Question

Quote:

But another way to fix this is wire the clock pin on one cartridge slot to constant 3.58 MHz

Excellent!! Now all sound is at the correct speed! One more problem solved! Thanks mate Smile

I notice that the IC10 TWS2793NL (I'm assuming this is the integrated chip you mentioned?) has a CLK input PIN24. Would I be able to just feed this with a constant 3.58MHz signal instead?

Wait a minute, that chip also has CS PIN3, would that be the one?

Pagina 1/5
| 2 | 3 | 4 | 5