CT63 is coming back, is there still demand?
CT63 is coming back, is there still demand?
Some exciting news on the CT63 front.
I have struck a deal with Rodolphe Czuba and paid for a licence to manufacture 10 more CT63 boards. This will be the original CT63 design.
Before anyone gets too excited, this isn't a simple case of sending it off to fab and job done. A lot of the parts are hard to find these days, and from what I have seen so far the PCB looks to have some errors which I will need to work through first. With my other 20 or so projects all going on at once, time is limited, so it will be a slow burner rather than a quick turnaround.
Assuming the CTCM is still how the clock is done, I would also look at building that onto the board itself rather than it being a separate module.
Rodolphe has said he could "maybe" update the code if he can get his old tools working again. The reason I asked is that I want to use the faster 6ns 288XL, as I have almost 100 of them sitting in stock anyway. He also suggested the burst speeds could be increased. With a larger PLD there might also be room to add other things in the future, but that is very much a maybe for later.
So the big question before I start spending money on parts: is there still demand for the CT63?
If you would be interested in one, please reply here. It is not a commitment to buy, I just want a rough idea of numbers before I commit to anything. If it turns out only me and my cat want one, then the cat will have to wait.
I have struck a deal with Rodolphe Czuba and paid for a licence to manufacture 10 more CT63 boards. This will be the original CT63 design.
Before anyone gets too excited, this isn't a simple case of sending it off to fab and job done. A lot of the parts are hard to find these days, and from what I have seen so far the PCB looks to have some errors which I will need to work through first. With my other 20 or so projects all going on at once, time is limited, so it will be a slow burner rather than a quick turnaround.
Assuming the CTCM is still how the clock is done, I would also look at building that onto the board itself rather than it being a separate module.
Rodolphe has said he could "maybe" update the code if he can get his old tools working again. The reason I asked is that I want to use the faster 6ns 288XL, as I have almost 100 of them sitting in stock anyway. He also suggested the burst speeds could be increased. With a larger PLD there might also be room to add other things in the future, but that is very much a maybe for later.
So the big question before I start spending money on parts: is there still demand for the CT63?
If you would be interested in one, please reply here. It is not a commitment to buy, I just want a rough idea of numbers before I commit to anything. If it turns out only me and my cat want one, then the cat will have to wait.
- HigashiJun
- Posts: 2515
- Joined: Fri Jun 19, 2020 7:21 am
- Location: Tokyo
Re: CT63 is coming back, is there still demand?
Count me in !
Re: CT63 is coming back, is there still demand?
One really good candidate to fix in CPLD would be the /EXPAND signal: https://www.atari-forum.com/viewtopic.php?t=34016 so e.g. NOVA, Eclipse and possibly other (future) hardware expansions would work with CT60.
Re: CT63 is coming back, is there still demand?
I've been through the Falcon Rev G schematic (C303007) looking for /EXPAND. As far as I can find, it only appears in two places: J16 pin 49 on the expansion connector, and pin 17 of U63, which is a GAL16V8. It has a pull-up to VCC and doesn't go to COMBEL or anywhere else.mikro wrote: Sat Nov 21, 2026 1:23 am One really good candidate to fix in CPLD would be the /EXPAND signal: https://www.atari-forum.com/viewtopic.php?t=34016 so e.g. NOVA, Eclipse and possibly other (future) hardware expansions would work with CT60.
U63 also handles EDTACK and EBERR, both pulled up in the same way, and produces XDSACK1 for the 030 on pin 19. EDTACK is also driven by the MFP's DTACK, so it looks like U63 is the glue that turns the slave handshake on the expansion bus into the 030's DSACK. /EXPAND sits in that same group, so I suspect it's an input to U63 from the card, but the schematic can't prove it. The GAL pins can be inputs or outputs, so without the U63 equations we can't say for sure which way it goes or what it does.
On the CT60 side, I checked the series 3 board file and J16 pin 49 isn't connected to anything, and neither is EDTACK. However, the CT60 manual shows the F030 BOOST connector (X9) has a wire on pin 9, DTKCMB, which goes to U63 pin 8. On the schematic, U63 pin 8 is XDTACK, so the CT60 takes its Falcon bus acknowledge from the COMBEL DTACK going into U63, not from what U63 gives the 030.
That matters because in 030 mode the 030 gets its acknowledge from U63's output, which is built from XDTACK along with EDTACK, EBERR and /EXPAND. With the CT60 tapping U63's input instead, a card's EDTACK and whatever /EXPAND does inside U63 are never seen in 060 mode. That would fit with the NOVA working in 030 mode but not with the CT60, while simpler cards like the ScreenEye are fine.
So rather than /EXPAND on its own, the thing to look at may be where the CT60 gets its acknowledge from. Either taking it from U63's output instead of pin 8, or bringing EDTACK and /EXPAND into the CT60 as well. That's still only going on the wiring though.
To get any further we would need the U63 GAL equations. Failing that, a scope on J16 pin 49 on a plain Falcon while reading the expansion area in a loop would show whether U63 drives it or not. If it pulses with no card fitted, U63 is driving it. If it stays high, it's an input from the card. It would also be worth scoping U63 pin 8 against pin 19 while the NOVA is being accessed in 030 mode, to see what U63 does with the acknowledge.
Re: CT63 is coming back, is there still demand?
Mikro, a question to clarify things. With the NOVA fitted on the Falcon expansion slot and the CT60 on top of it, the NOVA's pins go straight into the Falcon. So if /EXPAND is an input to U63, the NOVA should be driving it into U63 directly, and the CT60 shouldn't be getting in the way of that at all. Is that how you have it set up, or is the NOVA on the CTEX or somewhere else?
If it is the stacked setup, then I don't think /EXPAND itself is what's missing. The NOVA presumably answers on EDTACK, which goes into U63 and comes out as XDSACK1 for the 030. The CT60 doesn't listen to that. It takes its acknowledge from U63 pin 8, which is COMBEL's DTACK going into U63, so in 060 mode it would never hear the NOVA answering.
Do you know if the NOVA terminates its cycles with EDTACK? If you have the NOVA schematic, checking what drives pin 8 on the expansion connector would answer that.
If it is the stacked setup, then I don't think /EXPAND itself is what's missing. The NOVA presumably answers on EDTACK, which goes into U63 and comes out as XDSACK1 for the 030. The CT60 doesn't listen to that. It takes its acknowledge from U63 pin 8, which is COMBEL's DTACK going into U63, so in 060 mode it would never hear the NOVA answering.
Do you know if the NOVA terminates its cycles with EDTACK? If you have the NOVA schematic, checking what drives pin 8 on the expansion connector would answer that.
Re: CT63 is coming back, is there still demand?
Mikro, before going any further down the /EXPAND route, it might be worth starting from the beginning and checking the basics. In the stacked setup the CT60 only sees what the NOVA passes up through its top headers, so if the NOVA doesn't pass every signal the CT60 needs, nothing else matters.
Could you check with a meter that every pin the CT60 uses goes straight through the NOVA from bottom to top? The CT60 uses these on the 50 pin connector: AS, UDS, LDS, R/W, FC0-2, BGK, BR2, RESET, HALT, BERR, IPL0-2, CLK, A1-A23, +5V and GND. On the 30 pin connector: D0-D15, pins 20 and 22, +5V and GND.
Pins 20 and 22 are the ones I'd check first. They're the bus grant daisy chain, and the Falcon schematic says to fit a jumper across them if there's no expansion board, so any card in the slot sits in that chain. The CT60 uses both of them, so if the NOVA breaks or gates that chain differently, the CT60's bus handover could go wrong in 060 mode even though everything works in 030 mode.
If the NOVA schematic is to hand, that would answer it straight away.
Could you check with a meter that every pin the CT60 uses goes straight through the NOVA from bottom to top? The CT60 uses these on the 50 pin connector: AS, UDS, LDS, R/W, FC0-2, BGK, BR2, RESET, HALT, BERR, IPL0-2, CLK, A1-A23, +5V and GND. On the 30 pin connector: D0-D15, pins 20 and 22, +5V and GND.
Pins 20 and 22 are the ones I'd check first. They're the bus grant daisy chain, and the Falcon schematic says to fit a jumper across them if there's no expansion board, so any card in the slot sits in that chain. The CT60 uses both of them, so if the NOVA breaks or gates that chain differently, the CT60's bus handover could go wrong in 060 mode even though everything works in 030 mode.
If the NOVA schematic is to hand, that would answer it straight away.
Re: CT63 is coming back, is there still demand?
I'm in
Whomper
Homepage
Electronic Lab
16 Bit: Falcon (DFB1X/14MB/4+8 GB), 1040STFM, Soundpool SPDIF/FA8, Cubase Audio, Cubase 3
8 Bit: 1200XL, 800XL, 2 x 1050, 1025, Fujinet Pro, 2 x 1010, CX-85, Touch Tablet
Homepage
Electronic Lab
16 Bit: Falcon (DFB1X/14MB/4+8 GB), 1040STFM, Soundpool SPDIF/FA8, Cubase Audio, Cubase 3
8 Bit: 1200XL, 800XL, 2 x 1050, 1025, Fujinet Pro, 2 x 1010, CX-85, Touch Tablet
Re: CT63 is coming back, is there still demand?
I'd have to have a finger in the air price, a small breakdown of the differences between the CT63 and CT60e I've got. Physical compatibility with the SuperVidel and the CT60PSU that you're working on.
But probably in.
But probably in.
Senior Principal ASIC Engineer - SystemVerilog, VHDL
Thalion Webshrine - http://thalion.atari.org
ST,STf,STfm,STe,MegaST,MegaSTe,Falcon060
A500+,A600,A4000/060,CD32,CDTV
Thalion Webshrine - http://thalion.atari.org
ST,STf,STfm,STe,MegaST,MegaSTe,Falcon060
A500+,A600,A4000/060,CD32,CDTV
Re: CT63 is coming back, is there still demand?
Count me in !
Atari 800XE
Atari Falcon030 & Phantom2
Atari Falcon030 & Phantom2
Re: CT63 is coming back, is there still demand?
Tobias' setup (the original "NOVA guy"), taken from https://www.tobiashagemann.de/ct60/ct60.html.exxos wrote: Sat Nov 21, 2026 3:13 am Mikro, a question to clarify things. With the NOVA fitted on the Falcon expansion slot and the CT60 on top of it, the NOVA's pins go straight into the Falcon. So if /EXPAND is an input to U63, the NOVA should be driving it into U63 directly, and the CT60 shouldn't be getting in the way of that at all. Is that how you have it set up, or is the NOVA on the CTEX or somewhere else?
I guess I don't need to mention that from electronics point of it absolutely doesn't matter whether the pins go Falcon EXP -> NOVA -> CT60 or Falcon EXP -> CT60 -> NOVA, it's all connected the same way.
I had it connected via the CTEX which is basically the same thing.
Of course it does, no signal is cut, it's 1:1 mapping of all header pins.exxos wrote: Sat Nov 21, 2026 3:21 am Could you check with a meter that every pin the CT60 uses goes straight through the NOVA from bottom to top?

