falcon Testkit

[umberto1st]

Was hoping someone may have the wiring diagram for the DSP shorting plug from the Falcon test-kit. Checked with Best and B&C, both are out of stock permanently.

Thanks Rob

[mpattonm]

Shoot. I have sent them an email a second ago. Could anybody borrow me that DSP loopback just for a day or two, please?

[czietz]

Some of the wiring maybe can be guessed from comments in the Falcon production diagnostic source code: https://www.dev-docs.org/docs/htm/searc ... Production

Code: Select all

* You can test this program by disconnecting wires in the loopback.
* If you disconnect STD-SRD you should get a LB_NOT_EQUAL
* If you disconnect SCK-SC0 you should get a LB_TIMEOUT
* If you disconnect SC1-SC2 you should get a LB_TIMEOUT


[czietz]

This is confirmed by a document in the Falcon diagnostic source code (different archive from the one linked above):

DSP TEST 2 - SSI Port

This test loads a program into the DSP and then sends
it's internal sine wave table out a loopback connector on the
DSP port.

NOTE: The loopback connector cannot have any long signal paths
or the test will fail intermittently.

*------------------------------------
* SSI LOOPBACK TEST (comments from SSI.asm)
* Test of DSP56k's SSI port by connecting it back to itself.
* To perform this test, connect:
* STD to SRD
* SCK to SC0 (SCK output, SC0 input, asynchronous mode clocks)
* SC1 to SC2 (mode to be defined)

[mpattonm]

Ah, cool! I have probed the port with osciloscope entire afternoon, but there is so many variables... But this narrows them by a bit. By the way I have been able to figure out signal on GP1 pin is meant to drive DSPINT. with those two connected together "Ext. interrupt test" passes.

[mpattonm]

Quick test - Testing DSP SSI Loopback - pass
Digital loopback test is still failing, I believe this will require routing PLYDATA to RECDATA somehow, maybe its clock signals too.
There is some strobing on GP0 and GP2 too I should mention.

[czietz]

Look at DSP.S and SDMA.S in the source code linked above and see if you can deduce anything more from the comments.

[mpattonm]

Looking at it as we speak.

Edit:
This is what I found:
a) DMA Play -> XO -> Loopback -> XI -> DMA Record
; we need to use GP1 to trigger XO_SYNC and XI_SYNC.
b) EXINT <-> GP0 We'll also trigger EXT_INT with GP0 for fun. GP2 can be
; connected to GP0 for all we care. And we do, so much.
I am gonna test it shortly.

[mpattonm]

Well I have tried making all connection as above, still no go. There is at least one invalid information in the source comments and that is EXT_INT test is beeing triggered by GP1 rather then GP0. Positively. Now it seems to me Atari might have changed the test designed between 1.20 and 1.27b I am actually using.
Now I would really appreciate if someone, who actually owns the dongle could do a simple multimeter check and find all the pin to pin connections.