Sporniket's installation

[sporniket]

Ready to test (in fact not, I saw after a few minutes trying to get something, that my goex was not connected at all... then it did not change anything)

Result : white screen, it seems to be stuck

BUT I can change the rom slot as documented (press both shift, led blue + red, press one of F1~F4 to select, it resets and startup, even if it keeps being stuck).

As I said, after some time I saw that I did not connected my goex, so I connected it again (while switched off of course). I also tried without the IDE module, no change. Anyway, it reminds me when I tried another IDE mod some years ago, it was stuck like that because the connexion was all wrong and some key signals were stuck.

If you have advice about what to check, I will do that next weekend. I have a scope, so I can also do some probing and get some screenshots.

wow it's already monday, I should go to bed, see you later...

[Administrator]

Ready to test (in fact not, I saw after a few minutes trying to get something, that my goex was not connected at all... then it did not change anything)

FDD isn't required for the first boot steps - just TOS 2.06 will throw 4 bombs. But until then you've already seen the Fuji Logo.

Result : white screen, it seems to be stuck

White screen on ST-LOW or ST-HIGH? That really makes a difference.

BUT I can change the rom slot as documented (press both shift, led blue + red, press one of F1~F4 to select, it resets and startup, even if it keeps being stuck).

Then the AVR runs fine. How do you detect the reset? Is the screen flashing?

As I said, after some time I saw that I did not connected my goex, so I connected it again (while switched off of course). I also tried without the IDE module, no change. Anyway, it reminds me when I tried another IDE mod some years ago, it was stuck like that because the connexion was all wrong and some key signals were stuck.

IDE device isn't required at all. It just takes longer to boot without. TOS probably writes to the IDE registers and as SmartBoard generates /DTACK for the IDE memory area TOS reads data which in this case is just the pull-up resistors i.e. $FF. So TOS expects some meaningful data (not $FF) and will probably try some more times until it gives up and continues the boot process.

First step should be of course short circuits between the P100/P101 resistor networks i.e. the attached cables.
Did you boot with all cables (6 & 10pin) soldered to the mainboard and the original ROMs in place before?
Then I would check if the 16MHz input is indeed connected to GSTMCU pin 2.
AVR gets its 5V via SmartBoard's 16pin IDC connector, so there is little chance you plugged in the board off by one pin.
Checking the 3-pin cable isn't necessary at this point, I would focus on 6 and 10pin connections.
Are they soldered to the exact same signals as mentioned in the manual and on the silk-screen?
Orientation of the P100/P101 regarding pin 1 is correct?
P100 = 1K2 and P101 = 4K7?
With scope check pin 24 (/OE) if there are low-pulses. Pin 22 (/CE) should be constantly low as it is tied to ground.

[sporniket]

Thanks for the answers, I will look into it next time.

White screen on ST-LOW or ST-HIGH? That really makes a difference.

ST-LOW, it's plugged on a TV through SCART.

BUT I can change the rom slot as documented (press both shift, led blue + red, press one of F1~F4 to select, it resets and startup, even if it keeps being stuck).

Then the AVR runs fine. How do you detect the reset? Is the screen flashing?

Yes.

[Administrator]

YOU did everything correct, fault is on MY side

I probably got confused when initially writing the manual because of this:

P100Schem.gif (6.67 KiB) Viewed 9011 times

Which is pin 6 and pin 8?
Obviously I figured it out but did not update the manual. I will send you the updated PDF later.
Took a photo of my assembly and it should look like this:

So please swap the brown and yellow cables at P100.

Referring to your post here viewtopic.php?p=68#p68
the 6pin connector's wiring and the wires that are connected to the single resistors look correct for now - until I'm proven wrong.

But I saw in the same post that:

• R111, R112 and R102 are populated
• R106 is not populated

On my board

• R106 and R102 are populated
• R111 and R112 are not populated

Sorry, Arne

[sporniket]

got it.

By the way, I ported the schematic of the STe under Kicad, with a PCB reconstitution. It's approximative on the trace position, but I usually say that their path is "topologically correct" (same pathway). Anyway, I know that it was good enough that it helped someone repair broken traces.

https://github.com/sporniket/kicad-conv ... 300780-001

[sporniket]

This weekend will be busy, so I did the rewiring of P100 today :

And then...

Victory !

I get to the desktop (as long as the floppy image is not a game). The IDE seems to be detected, but it does not load the driver. I will check the content of the card, it should be ok, but I never could test it really.

[sporniket]

Just the boot screen of the other slots. Each time I arrive at the desktop.

[Administrator]

Victory !

So, I may quote Dave Jones from eevblog?
Winner, winner, chicken dinner.

I get to the desktop (as long as the floppy image is not a game). The IDE seems to be detected, but it does not load the driver. I will check the content of the card, it should be ok, but I never could test it really.

I (almost) always use Hushi (comes with SCSI Tools) with single-channel IDE. The UI ist just well thought through.
But neither SCSI-Tools nor its driver Hushi can detect IDE devices if the DMA-INT isn't connected.

The "IDE Hardware detected" writing is plain PIO i.e. it detects an IDE master-device even if the DMA-INT is not connected. Just tried that on my test-rig. And the IDE HW is plain good ol' Falcon compatible stuff. No fancy DOS-compatibility Hi/Lo byte swapping.
Please check that: black cable of 3-pin connector runs to R421 and pin15 GSTMCU (/HDINT).

But you may try Uwe Seimet's HDDriver/HDDRUTIL. Both can detect my 128MB DiskOnModule with DMA-INT disconnected.

I may repeat myself but good soldering always pays out.

[Administrator]

One more thing:

if that is the CF-Card to IDE adapter I sent you on the top photo on this page then this one needs a power source. It is the same as for 3.5" FDDs and I cannot see a power cable connected to it.
Just tested my adapter:
red LED indicates power-on
green LED indicates Card-Is-Detected
yellow LED indicates I/O access to the card

The TOS 1.06/1.62 patches just select the master device then do a tst.b on the address of the status register and check if a bus-error occurs. In case of a bus-error no /DTACK was generated by the CPLD - which should never happen with SmartBoard STE connected correctly even without DMA-INT. And if there is no bus-error it prints the "IDE hardware detected" on the screen.
The test for IDE Hardware does not rely on an IDE device to be connected!

[sporniket]

Thanks for all the information, here and by mail

The Smartboard is tightly plugged, so I was reluctant to remove it to put a blob of solder on SJ1, and risk breaking anything. So I went with powering the IDE module with another power cable to attach somewhere.

I decided to use C210, a decoupling capactitor of one of the ACIA, on the south-west-west of the motherboard, as it will not be far from the module.

I have checked polarity with a multimeter (continuity with ground, since there are lots of capacitors, on the pin soldered to VCC it bips for a short time ; on the pin soldered to GND, the bip does not end).
And I also confirmed with my kicad reconstitution :

Then I made a 10cm cable with simple dupont connectors, 2 wires

And soldered it at the planned location