Issue with Soundpool MO4

I have a Atari STe with a Soundpool MO4 MIDI extension. It used to work very well, but unfortunatelly doesn't anymore: Cubase still detects it, and I can output MIDI to it but nothing is coming out from any MIDI Out. It took me a while to tackle it (lack of time, lack of tool, other items to play with), but I gave a glance last week-end.

The parallel port on the Atari uses only the following signals:

• Pin 1 : Strobe (Atari -> MO4)
• Pin 2 : Data 0 (Atari -> MO4)
• Pin 3 : Data 1 (Atari -> MO4)
• Pin 4 : Data 2 (Atari -> MO4)
• Pin 5 : Data 3 (Atari -> MO4)
• Pin 6 : Data 4 (Atari -> MO4)
• Pin 7 : Data 5 (Atari -> MO4)
• Pin 8 : Data 6 (Atari -> MO4)
• Pin 9 : Data 7 (Atari -> MO4)
• Pin 11: Busy (MO4 -> Atari)

The MO4 also decodes few other pins, but since the Atari doesn't, my guess is the MO4 was also targeted for PC.

Inside the box, the MO4 is architectured around a CPLD (IspLSI1016 from Lattice) which contains the logic. The other ICs are Lattice GAL16V8 (for reducing the 8 MHz clock to 500 KHz) and a 74LS541 for buffering. Also, a 7805 voltage regulator (alim is 7.5V, positive central), a diode for protection and few passive componants: crystal, pull up/pull down resistors and filtering/decoupling caps. We find a port for In-System Programming of the CPLD, as well as a couple of test points. There is not silkscreening on the pcb. I regret the lack of a led to show the status of power, since it's easy to plug it the wrong way: the silkscreening on the box itself is misleading on this point.

The CPLD manages all the signals from the parallel port, and is also responsible to perfom the parallel to serial convertion. Despite the MIDI protocol allows to select among 16 channels, it doesn't include a way to select a specific output. I don't know yet what SoundPool chose to implement the output select: Control Change? Unimplemented Real-time? I know however that all bits from the parallel port appear on two different pins on the CPLD; is this used for parallel-to-serial or for output selection, I don't know yet.

The communication between the Atari and the CPLD is achieved thru the Strobe (pin 1) and Busy (pin 11) lines: the Atari pulse the Strobe line before sending data, and read the Busy line to check the status. On the Busy line, low indicates ready and high indicates busy. Busy is connected to the Multi Function Peripheral (MFP 68901) of the Atari, precisely at address=$FFFA01, bit 0. My problems is precisely here: after the initialization, the CPLD keeps the busy line to low, and the parallel port stays steady. I was able to trigger the send of new data by pulsing Busy to high. All signals (Busy, Strobe, MIDI out flow) are sent across the 74LS541.

The capture below shows the initialization of the MO4 when the driver (MO44.DRV) is started. The source code responsible for this initialization is the following:

  
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
; Reset hardware.
; Writes four zero bytes to the printer port, which corresponds to 
; the Strobe line (normally high).
; Each pulse is 3 micro-seconds (1.5 high, 1.5 low).
;
; The Atari ST uses 2 registers of the Sound generator YM-2149 to drive 
; the Centronics (parallel) port. 
; Register 14 bit 5 contains the value of Centronics' Strobe line.
; Register 15 is the Centronics Data.
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
reset_hardware:
	movem.l	d0-d1/a5,-(a7)
	move.w	sr,-(a7)
	movea.w	#ym_data,a5
	ori.w	#$0700,sr       ; Interrupts off.

	move.b	#15,ym_base.w   ; Select register 15 of the YM-2149 (Centronics)
	move.b	#0,(a5)         ; Set data lines to 0.

	move.b	#14,ym_base.w   ; Select register 14 = parallel port A
	move.b	ym_base.w,d0    ; Fetch port A status to d0.
	move.w	d0,d1	        ; Copy d0 to d1
	and.b	#$DF,d1         ; Strobe low value.
	or.b	#$20,d0         ; Strobe high value.

	move.b	d1,(a5)         ; Pulse strobe low-high four times.
	move.b	d0,(a5)         ; This gets the hardware back to a state where
	move.b	d1,(a5)         ; it's expecting a port bitfield, regardless
	move.b	d0,(a5)         ; of what state it's currently in.
	move.b	d1,(a5)
	move.b	d0,(a5)
	move.b	d1,(a5)
	move.b	d0,(a5)         ; End with strobe high.
    

I reverse engineered to draw a schematic, however I don't know yet what causes Busy to stay low: bad join, dead gate on the 74LS541, or worse, dead gate or corrupted program on the CPLD. I'll investigate more with a logic analyzer.

Update

I managed to obtain some results by triggering the /busy line. For this, I made a small circuit on breadboard with a NE555 configured as monostable (goes low around .5 sec, then goes high) and switch. The goal is to debounce the switch. With this circuit, I was able to communicate with the Atari and to trigger the send of next bytes. Unfortunatelly, probably due to timing issue, the result is hardly exploitable.