FM Cartridge in a computer with built in FM

Page 1/2
| 2

By meits

Scribe (6530)

meits's picture

28-01-2015, 19:08

In this topic the bad sound from a Panasonic FS-A1ST was discussed... At first I thought it was okay enough... Then I got pointed at the differences which wanted me to try my Panasonic FM-PAC on my turbo R... I already knew this thing was going to be silent... I was told that this is the case with Panasonic FM-PACs but not with clones... I got a tip from that same topic to try OPLLOFF.COM which switches off the internal FM which should make my FM-PAC come to life... And it did... In Moonblaster, but not with Moonblaster's basic driver... Then hearing that OPLLOFF.COM is more of a so called "dirty hack", I knew I had just one option left: get a clone. The graphic below proves the validity of my wish:

I have to tell you that my Panasonic FM-PAC does have a custom RCA output so it's not the cartridge amplified with the internal MSX-Music, it's only the cartridge.

On facebook I had a little conversation with Nyyrikki who pointed me at a theoretical danger:

Quote:

Yes, that it how the safety switch works... FM-PAC does not get enabled if it finds any other device in same I/O ports, but it works only if BIOS is used. Many software does not consider this kind of unlikely conflict and they enable FM-chip without any checks.

I think this was made so that the PAC-part could be used in computers that already have MSX-MUSIC without hardware conflicts.

I don't say that this definately causes errors, over heating or anything like that, but I'm warning that that this MIGHT not be a good idea. You may want to ask second opinion about the potential risks from someone more hardware oriented expert.

The worst case scenario that I don't know enough about, goes something like: We send data to MSX-MUSIC and both devices get the data. How ever MSX tR adds some extra waitstate be caused cartridge slot... or oscilators are not 100% in sync. Then we input status byte. One chip returns 0 (=0v) and because of time difference other returns 1 (=5v). This causes 5V to be connected to ground= short circuit. -Is this likely? No -Is this possible? I don't know, I suggest asking. I just wonder why there is such a complex prevention system in place if that is not mandatory.

So my question to a hardware guru (please no guesses): *Is* it potentially dangerous to the hardware to have an MSX-Music cartridge connected to a computer with built in MSX-Music?

Login or register to post comments

By flyguille

Prophet (3031)

flyguille's picture

28-01-2015, 20:01

no, chips can't shortcircuit GND & +5V on busses conflict, because the outputs buffers inchip its output gates have internal impedance, for avoiding destruction of the chips.

that means there is a resistence in serie internally, check the impedance at specs.

what happens on the bus, that if two chips want to talk at the same time, the information will be NORd or ANDed it depends about if it is a normal pull down, or pull up., iirc in MSx the data bus is normal pull up, and for making a ZERO chips must ground the line, so, if any of the chips grounds the data line to ZERO it will be zero, only when all chips are 1, the line will remain 1, so it is like ANDed.

If you have two identical chips that are talked with the same information, and them want to talk at the same time, it is most probably them output the exact same information with it, and no problem with that.

But, if them are different version of chips, and some data status readed by the z80 is different, that is quite a problem from the software point of view.

Another thing, is the output sound signal, them are analog, and most MSX have some passive or active mixer inside, and that is the source of all the problems in msx world.

By AxelF

Champion (395)

AxelF's picture

29-01-2015, 16:23

For the Moonblaster basic driver you can use the following (Dirty) Hack,
to use the external FM-PAC on the Panasonic FS-A1ST

Use a hex editor on the basic.bin moonlbaster basic driver
At adress 0x0076 change #CB in to #00 AND
At adress 0x0077 change #FF in to #00

Now it will not search for FM-PAC in secondary slots. This means that your external FM-PAC needs to be in a PRIMARY SLOT, you can't use a slotexpander !!!

One side effect will be that the internal FM-PAC will also be audible,
but i read you have a RCA output so you better use that...

p.s.
i used the Moonblaster basic driver from Diiisk (DIIISK04)

By syn

Prophet (2113)

syn's picture

29-01-2015, 16:26

I vaguely recall that this, or something similar, was discussed a while back on MRC. I forgot the specifics or whether it was even about FMpac or not, but it was something about 2 of the same cartridge. It basically went like this
person a: "Dont use 2 of them together! It will damage your msx"
person b: "I never had any problems"

I tried looking up that post, cant seem to find it atm.

By AxelF

Champion (395)

AxelF's picture

29-01-2015, 16:48

@SYN, Yes i believe it was this topic about the MSX Audio upgrade 1.3 ( Third Post )

By meits

Scribe (6530)

meits's picture

29-01-2015, 17:11

That'd go for MSX-Music as well...
Anyone ever had/heard of frying hardware in real life because of this?

By Daemos

Paragon (2044)

Daemos's picture

29-01-2015, 19:33

The analogue part of your audio will not fry as well. most MSX'es either use and opamp and they will just give the maximum output voltage if the input is too high so you get nasty clipping. The transistor approach will give a more saturated type of clipping but it won't blow up either. Audio wise having two FMpak's in your MSX there is nothing to fear. The outputs of the FMPAKS are decoupled.

By meits

Scribe (6530)

meits's picture

29-01-2015, 19:43

Good to read as I do not intend hearing internal FM anymore if not needed Smile

By l_oliveira

Hero (534)

l_oliveira's picture

03-02-2015, 01:10

Food for the thought: OPLL registers are write only. 0% chance of bus conflict.

In my humble opinion, it's likely that FMPAC has detection of multiple FM devices mostly due to the PAC function (you would maybe want to connect two to the same computer to move data around ?)...

By sd_snatcher

Prophet (3642)

sd_snatcher's picture

03-02-2015, 02:36

Yes, the entire OPLL chip is write-only. It can't even detected on a machine, an this is why you have to detect the BIOS instead. On SMS, some extra circuitry was added to allow detection, but what you detect is that circuitry and not the chip itself.

The may-fry I/O conflict situation only applies to OPL1/2/3/4 chips, but not for the OPLL.

By l_oliveira

Hero (534)

l_oliveira's picture

03-02-2015, 02:43

@Meits: There's a good possibility that your computer is having issues with old capacitors.
I never had problems with the OPLL having low sound volume.

Been contemplating a possibility of re-cap my A1ST eventually.

Page 1/2
| 2