TF1230/Solas RGB Fast Ram errors

[Mark Sealey]

Hi all....ok....while loving my TF1230......I have started to add the goodies to the build....Ok both IDE devices are working fine now....so my next upgrade is a Sloas RGB controller from Rob Cranley.....my Rev 1.4 had a full Clockpoert populated with pins.....so I modded the clockport so I can fit the clockport cable to the Solas.

Right when the Sloas was plugged in and active I noticed I have Stability issues...random lockups and crashes...out come AmigaTestkit and other Test programs.....sothese test programs gave me Fast ram errors.....With the Solas plugged in......I don't get any errors without the Slaos plugged in...I ram the Fast Ram scan for over an hour without the Solas...no errors....Rob Cranley has Reached out to TerribleFire for any insight on this issue.....Any ideas....?

OS 3.2.1

I've tried 3 different PSU's...Standard "heavy type" A500...Meanwell...old school "Goliath" psu

TIA

TF1230.jpg

[Mark Sealey]

I'm guessing this forum is more for Atari now than Amiga.....? anybody else had any issues with the TF1230 and clockport....?

[PhilC]

Im guessing this forum is more for Atari now than Amiga.....? anybody else had any issues with the TF1230 and clockport....?

Hi Mark, it’s not a machine specific forum. I think it’s possible you’re the only one with this issue at the mo.

[Mark Sealey]

yeh that's what i fear...as there is only 18 Solas controllers out in the wild.

[trixster]

Solas here running ok with a TF1260, no problems so far

[Mark Sealey]

Solas here running ok with a TF1260, no problems so far

Yeh Rob Cranley has a TF1260 too...no issues there..... you got any cool .Style settings file your fancy sharing...?

[trixster]

I confess I haven’t used it that much as my 1200 gets a lot less use than the other machines. I do like using the leds as a vu meter spreading left and right from the centre when listening to mp3s. The fan controller is great too

[terriblefire]

I confess I haven’t used it that much as my 1200 gets a lot less use than the other machines. I do like using the leds as a vu meter spreading left and right from the centre when listening to mp3s. The fan controller is great too

I spent most of the evening with @Daedalus investigating this issue today. Let me try and lay out what is happening.

Whenever anything attempts to access 0xXXD80000 the SPARE_CS is asserted on the Amiga 1200. *EVEN* if you assert the /OVR pin which is supposed to stop it decoding.

I even took the CPU out of the TF1230 and hard coded logic values on the bus

/OVR = 0
/AS = WIGGLING
/DS = WIGGLING
/RW = 1
A = 0x08D80000

I tried setting FC to 7 as a fudge.

I could see SPARE_CS staying flat low. @Daedalus seems to think this causes data to get put on the data bus which corrupts the fastram read. IMHO Gayle should not be decoding anything when /OVR is asserted, but it is.

The reason this doesnt happen on the TF1260 is that the data bus is buffered (even though the address bus is not - its just level shifted).

Reluctantly we came to the conclusion that this style of clockport splitter isnt compatible with the cheap design of the TF1230.

[terriblefire]

I'm guessing this forum is more for Atari now than Amiga.....? anybody else had any issues with the TF1230 and clockport....?

Apologies its just taken Rob and I this long to coordinate with house moves, house renovations, game jams, amiga 37 etc etc...

[Daedalus]

Yeah, sorry about that, it doesn't look like something that's possible to fix in firmware on either board. I'll update Solas with the modified buffer logic needed at some point in the new year, and can swap the boards with anyone having issues with this particular configuration. I don't think it's possible to do a simple patch on the board, but the other option is to totally bypass the clockport splitter/buffer function, which would require a couple of patch wires on Solas and render the extra clockports useless, but would be easy enough done by anyone handy with a soldering iron.