YM2413 - HAAG - noisy

Page 2/4
1 | | 3 | 4

By Eugeny_Brychkov

Paragon (1184)

Eugeny_Brychkov's picture

17-07-2020, 10:54

dublevay wrote:

The sine wave seems to be missing for 110ms!

Look here, it states

Quote:

There were other cost-cutting modifications: the number of waveforms was reduced to two, and the channels are not mixed using an adder; instead, the chip's DAC uses time-division multiplexing to play short segments of each channel in sequence, as also done in the YM2612.

Most probably this is what you see on the scope.

By Grauw

Ascended (10181)

Grauw's picture

17-07-2020, 12:35

The time multiplexing is at a much higher frequency than 110 ms, so that can't be the cause of parts of the sine wave missing... if you look closely you see a vertically striped pattern, the waveform is composed of many peaks, that's the resolution of the time multiplexing.

The wave dropping out is strange. Is the clock input of the YM2413 alright?

By dublevay

Supporter (14)

dublevay's picture

17-07-2020, 21:53

OK - some success - at the YM2413 side anyway. I was, for simplicity (having removed the V9958 from the circuit, and hence it's NTSC clock) using a 4MHz oscillator DIP chip. 4 MHz is supposed to be the upper limit of frequencies accepted by the YM2413.

I've replaced that with an NTSC crystal and two 22pF capacitors, and am now getting the full sine wave. So things are progressing. The opamp circuit is still an absolute nightmare, and needs some additional thought.

By dublevay

Supporter (14)

dublevay's picture

17-07-2020, 23:28

I've now realised something - and I feel a little silly. After studying the waveform above, which looks undeniably good, it dawned on me that I was looking at a wavelength of 100ms. Which of course is a frequency of 10Hz. And that's going to be inaudible.

I changed the code around a bit, and can report that the chip is working perfectly, and produces an audible noise that doesn't require the TV to be turned up to about 80-90% volume, which was inducing the horrible noise I was hearing.

Thanks for your help guys. I now have a much better understanding of the chip and the op-amp circuit.

By Grauw

Ascended (10181)

Grauw's picture

18-07-2020, 00:21

Great, congratulations! Smile

By dublevay

Supporter (14)

dublevay's picture

13-02-2021, 16:48

Hi All

I'm still working away at the Dragon MSX 2+ project - and essentially everything is working well. With some seriously great input from Pere Serrat, we have graphical and audio demos, and I am pleased to say everything is great - apart from some audio noise. Dragon DMSX 2+ Board

If I omit the YM2413 from the circuit, the noise goes away. But I don't think the YM2413 is faulty. Certainly, when running Pere's VMG player, it sounds pretty awesome - and the Final Fantasy intro music has both the YM2413 and YM2149 playing away beautifully.

But what I notice about the YM2413 is that upon power up, there is no noise. Then perhaps 10-15 seconds later a 50Hz mains buzz appears. I can show this on the scope. The bigger spikes are 50Hz - and you can see that I am also seeing a smaller, but still apparent on the audio, 60Hz spike from the NTSC activity of the V9958.

I can see that this noise also appears, in a really minute fashion on the power rails, but seemingly only the presence of the YM2413 picks this up - although it could of course, be something to do with the op-amp circuitry I'm using after it (see the top of this thread).

I have changed the 0.015uF capacitors for 1n8 capacitors, as has been suggested in many places - but from my understanding that would only be affecting higher frequencies - it wouldn't suddenly be allowing this mains/NTSC hum through.

This got me to thinking that perhaps I should look at another MSX 2+ machine as a reference point. I'd essentially copied the Sanyo Wavy audio circuit. So I've now simplified the audio circuitry for the Panasonic FS-A1WX PSG & MSX-MUSIC, which is below:

I've not yet built this, but it does seem to be a far simpler amplifier circuit.

Without meaning to sound snobby, if I thought back to the 1980s, I'd have considered Panasonic to be a higher quality manufacturer than Sanyo, so was really just wondering if anyone had any real-life opinion of the quality of the Panasonic sound over the Sanyo sound? I have neither of these - I have a Sony HB-F1XV - and sadly I have not yet found a schematic for that or the other MSX 2+ Sony that include the YM2413.

By Pencioner

Scribe (1480)

Pencioner's picture

13-02-2021, 17:01

Sanyo PHC-70FD has much better sound quality of internal PSG and OPLL than Panasonic A1WX (i had both)

By dublevay

Supporter (14)

dublevay's picture

13-02-2021, 17:07

OK - thanks. And when you had yours, did you apply this mod?

By sdsnatcher73

Prophet (2367)

sdsnatcher73's picture

13-02-2021, 18:22

I concur that Sanyo’s implementation has much less noise. I don’t think it is only in the audio circuit though, they have better shielding from other noise producers (like the VDP you mentioned). I had a PHC-70FD2 with sound mod. The mod in itself does not change the noise characteristics of the circuit. It mainly effects the mixing from the external sources with the internal ones, the mixing of the PSG with the OPLL, and the frequency response in low and high.

What makes the shielding of noise on the audio from the other components better I don’t know. For a test I have ordered a couple of YM2413B chips. They are fully pin compatible with the YM2413, but apparently are low power variants and possibly made using CMOS (as opposed to TTL I guess). They are still 5V and I want to see how other behave when replacing a regular YM2413 in a FS-A1GT.

By Pac

Scribe (6479)

Pac's picture

13-02-2021, 18:54

Page 2/4
1 | | 3 | 4