Modding Cartidges, Diagrams with Stereo & Clock Oscilator ...

Página 5/6
1 | 2 | 3 | 4 | | 6

Por Grauw

Ascended (8615)

Imagen del Grauw

02-10-2014, 00:36

Given the fairly strict timing requirements of the MSX-AUDIO and YM2413, I think those skipped notes are more of a software problem, because of the I/O access having not enough waits between them because the Z80 is running twice as fast. Just like the turbo R in R800 mode causes notes to be skipped in software which does not insert additional waits for that CPU. The reason why the OPL4 is not affected is because it can handle much faster access speeds than the older chips can.

Although, I imagine when they receive a 7MHz clock the MSX-AUDIO and YM2413 ICs themselves get overclocked and are able to handle shorter wait times. But, not of much use when the pitch gets messed up Smile.

Still in software it’s much easier to deal with adding a few more waits than messed up pitches. Though, for Synthesix I was thinking to provide options to offset the frequency tables by 0.5, 1 or 1.5 octaves to compensate for higher clock frequencies (will only work for circuits with constant frequency of course).

E.g. with a 1chipMSX running at 10 MHz, the internal PSG / FM / SCC would have to get no offset, and the external chips without xtal would get a 1.5 octave offset. Downside is that the note range gets shifted up, notably the lowest SCC note is no longer an A0 (27.5Hz) but a D#2 (69.3Hz).

Por giangiacomo.zaffini

Champion (267)

Imagen del giangiacomo.zaffini

02-10-2014, 14:16

Will I dare to ask for a confirm?
Do MSX-Audio YM8950 have a fixed operative frequency? I'm not sure after reading its datasheet.
Do SFG-01/SFG-05 YM2151 have a fixed operative frequency? I'm not sure after reading its datasheet.
Yes, it seems that MSX-Music YM2413 have a fixed 3.579545Mhz operative frequency after reading its datasheet.
Someone please help me.

Por maxis

Champion (512)

Imagen del maxis

02-10-2014, 17:11

Hello,

Y8950/YM3562, YM2151/YM2164 can run at the 4.00 MHz maximum clock frequency. Some Yamaha professional products (FB01 tone generator) have FM chip running at 4.00 MHz. The same 4.00 MHz is the reference clock for OP-M (YM2151) in Sharp's venerable X68K platform.

Por Grauw

Ascended (8615)

Imagen del Grauw

02-10-2014, 18:06

Hey maxis,

They are designed for the 3.58 ~ 4 MHz range of clock frequencies, in the sense that the pitch and envelope ranges and the sample rate (/72 divider) are set with those limits in mind... Double the frequency and you lose an octave in the low end and gain an octave in the high end, and the envelopes run twice as fast, and unnecessarily high sample rate. But I’m pretty sure they can still operate at higher frequencies.

After all MSX-AUDIO and OPLL take their clock from the cartridge slot, and on 7MHz (turbo upgrades) and 10MHz (OCM) systems, they sound at high pitch but do not blow up Smile. So the hardware must be able to handle it up to some unknown limit, just with out-of-spec and unpractical pitch range and envelope rates. As the I/O wait requirements are defined in terms of cycles, I’m assuming they get shortened as well.

Por giangiacomo.zaffini

Champion (267)

Imagen del giangiacomo.zaffini

02-10-2014, 18:49

@ maxis: pretty good to know. For sure there's much out there using chips at 4.0Mhz to take reference from. Who knows whether software can take care of operative clock frequencies and behave accordingly, maybe it could be a compile time setting for music applications.

For the sake of upgraded CPU to 7.0Mhz, I was hoping that these Y8950/YM3562, YM2151/YM2164 and Ym2413 could bear such 7.0Mhz clock. Nevermind.

But there is such project claiming that can use Z80A @ 33.0Mhz with YM2413
maybe it has to do with more than one clock.

Por maxis

Champion (512)

Imagen del maxis

02-10-2014, 18:57

@giangiacomo.zaffini & @Grauw

Even though Yamaha ICs seem to work at higher clock frequencies, I will not recommend running them beyond the Absolute Maximum Ratings figure.
Besides, if the bus interface, timers and envelope generators could run at 7MHz, I would doubt that the chip operation is error free. Knowing that it is a complex floating point processor, it might be unstable to guarantee the exact audio results.
One more reason to stay within the limits is the dynamic architecture of the chip. Unlike the full static design (can accept the clock full stop/resume), the power consumption of the dynamic design is not linear to the clock frequency. So, running faster (or slower) than recommended clock fork will cause the reliability problems / premature failure.

Por Grauw

Ascended (8615)

Imagen del Grauw

02-10-2014, 19:47

Aha, good to consider.

Por Protoman

Resident (44)

Imagen del Protoman

10-11-2016, 17:30

Hello Dhampird! Smile I just installed the 7MHz turbo mod in my HB-F1XV with 512K SRAM. Do you still have the circuit diagrams for fixing the CLK signal for an SCC cartridge? The image links don't exist anymore.. Sad I'm using an ESE-SCC that I made, and I have a 4-pin 3.58Mhz crystal oscillator to use. So far, I only tried connecting the CLK output of the crystal to the CLK input pin of the SCC chip after cutting the trace from the cartridge connector, but the cartridge would not boot...

Por Protoman

Resident (44)

Imagen del Protoman

11-11-2016, 17:56

Nevermind! I figured out the problem! Big smile I accidentally connected the new crystal CLK signal to the cartridge connector side of the circuit trace, instead of the side going to the SCC chip. Tongue

Por sd_snatcher

Prophet (3135)

Imagen del sd_snatcher

11-08-2017, 14:52

@Dhampird

Do you mind to upload the images to the msx pics website? All images that were on imageshack are now missing.

Página 5/6
1 | 2 | 3 | 4 | | 6