CDTV Relocator for the TF536

go0se · Post by **go0se** » Thu Sep 10, 2020 10:17 pm

Gerber files for reloctors and extended ROM adapter:

cdtv_reloc_v3.zip: (911.34 KiB) Downloaded 288 times

MC27C1000.zip: (56.48 KiB) Downloaded 297 times

cdtv_V2.zip: (464.38 KiB) Downloaded 303 times

!Message from: Moderator

@go0se, we have placed your attachments in-line so they display correctly in your post. Here's the how-to guide.

utri007 · Post by **utri007** » Fri Sep 11, 2020 9:25 am

What are dimensions of V3 relocator?
What else is required? Those things wich are soldered to it? I have no idee what they are in finnish, or english.

You seems to have CDTV, have you managed to get CD work on OS?

Post by **terriblefire** » Fri Sep 11, 2020 9:35 am

utri007 wrote: Fri Sep 11, 2020 9:25 am What are dimensions of V3 relocator?
What else is required? Those things wich are soldered to it? I have no idee what they are in finnish, or english.

You seems to have CDTV, have you managed to get CD work on OS?

At some point you are going to just have to dive in and start doing this for yourself as there will never be a "off the shelf" solution for the CDTV. Unless the warp guys make one.

go0se · Post by **go0se** » Fri Sep 11, 2020 11:06 am

utri007 wrote: Fri Sep 11, 2020 9:25 am What are dimensions of V3 relocator?
What else is required? Those things wich are soldered to it? I have no idee what they are in finnish, or english.

You seems to have CDTV, have you managed to get CD work on OS?

The dimensions of the V3 relocator are 112mm X 46mm

The V2 dimensions are 83mm X 46mm - no one dimension is over 102mm so will be cheaper to get manufactured by JLCPCB.

If you are building a TF536 for the CDTV use the V2 ( cdtv_V2.zip ) relocator, omit the JTAG pin header from the build and tape the underside of the JTAG footprint to isolate it from any possible contact with C163.

You can program the CPLD on the 536 easily by plugging 6 male pin headers into the interface at the end of whatever JTAG programmer you are using and holding the cable in position for the ten seconds that is required to program the CPLD.

I have a CDTV but the CD drive in that machine doesn't function with or without an accelerator so my problem is somewhat more fundamental than a DMA conflict

utri007 · Post by **utri007** » Fri Sep 11, 2020 7:41 pm

@terriblefire

At some point you are going to just have to dive in and start doing this for yourself as there will never be a "off the shelf" solution for the CDTV. Unless the warp guys make one.

I'm well aware that, but I'm first to admit my weaknesses, I'm prety shitty with soldering. Sometimes when I do that, it always surprises me, how bad I'm with it. I can solve software problems myself, but it would be nice to know things like wich compo is proven to work? Like you seem to had KS 3.1 on your CDTV, but wich extended roms? I don't wait out "off the shelf" solution and I wount buy Warp accelertor for my CDTV or A500, and my AGA Amigas has already 68060 CPUs.

@go0se

The dimensions of the V3 relocator are 112mm X 46mm

You recomend V2 rather than V3?

If you are building a TF536 for the CDTV use the V2 ( cdtv_V2.zip ) relocator, omit the JTAG pin header from the build and tape the underside of the JTAG footprint to isolate it from any possible contact with C163.

You can program the CPLD on the 536 easily by plugging 6 male pin headers into the interface at the end of whatever JTAG programmer you are using and holding the cable in position for the ten seconds that is required to program the CPLD.

I woun't build TF536, I have aready arranged that. I'm just trying to make this "easy" for my "builder", so that everything is as ready as it could be. I ordered those IDE headers etc.

go0se · Post by **go0se** » Fri Sep 11, 2020 8:09 pm

Yes, order the V2 relocator, it is a better solution with less compromises. I was trying to solve outstanding issues but ended up creating another one with the V3 as discussed above. There is not much to choose between them, both will work but the V2 is smaller, simpler and cheaper. As long as your builder remembers to clip back / shorten the 68000 pins immediately above C16 on the TF536 and does not install the JTAG header then all should be fine.

go0se · Post by **go0se** » Thu Sep 30, 2021 5:33 pm

Here's another

"hot"

new add-on for the Terriblefire line of cards!

This season's must have accessory is an 44 pin IDE flipper board for your cutting edge Commodore Dynamic Total Vision system. This handy little gizmo allows you to sling that troublesome IDE cable underneath the accelerator, saving millimetres of space up top! No more bulging lids for you early 90s HiFi/Amiga hybrid enthusiasts!

The board is fully retro-fittable to existing TF536s by soldering to the underside of your existing IDE header. The 1.0mm PCB thickness makes the operation a breeze should you decide that you've had enough of using your TF536 in a trusty A500/2000 and want to convert it for use in a less versatile, less reliable machine instead!

This innovative design features two length tuned signal layers + two dedicated ground layers and can be yours for the low low price of - whatever your board house charges for a 50mm * 19mm * 1mm four layer board. (£1.48 / 5pcs + £1.29 economy shipping form JLCPCB at the time of writing)

Perhaps the the design/idea could also be modified by some Atari boffin to solve the "?? Custom IDE cable ??" problem that I read about somewhere else on the the forum a while back, but can't really remember any of the details about?

Anyhow, here are some teaser photos. The design requires a minute clearance tweak and source files will be posted right here when I get round to fixing that.

: DSC_1509.resized.JPG (47.19 KiB) Viewed 5657 times

: DSC_1515.resized.JPG (66.84 KiB) Viewed 5657 times

: DSC_1527.resized.JPG (125.09 KiB) Viewed 5657 times

: DSC_1516.resized.JPG (57.48 KiB) Viewed 5657 times

: DSC_1507.resized.JPG (62.67 KiB) Viewed 5657 times

: DSC_1518.resized.JPG (59.8 KiB) Viewed 5657 times

: DSC_1520.resized.JPG (61 KiB) Viewed 5657 times

alexh · Post by **alexh** » Thu Sep 30, 2021 9:25 pm

Interesting. Easier than desoldering the existing vertical connector and soldering a right angle one directly to the TF536? I imagine there's a gotcha I didn't think about?

go0se · Post by **go0se** » Thu Sep 30, 2021 10:07 pm

alexh wrote: Thu Sep 30, 2021 9:25 pm Interesting. Easier than desoldering the existing vertical connector and soldering a right angle one directly to the TF536? I imagine there's a gotcha I didn't think about?

Perhaps if I hadn't been so busy hamming it up then I might have provided a better explanation as to what this actually does!

As the IDE header is double row you can't just flip it over to mount it on the underside, either the front row of pins on the footprint would become the back row (and the cable face backwards )or they would be 180 degrees out of position. This board corrects for that.

The image below should illustrate the idea.

: DSC_1520.resized.JPG (71.26 KiB) Viewed 5632 times

You can mount a 90 degree header on the 536 as-is but this is still too tall and touches the CDTV lid. With this board added then either a horizontal or vertical header can be underslung on the 536 meaning IDE clearance issue are removed. If you then choose not to socket the CPU the top profile of the TF536 becomes very shallow and allows the CDTV lid to close without any problems.

go0se · Post by **go0se** » Mon Nov 29, 2021 9:20 pm

After a long hiatus I went back to the EDA drawing board board with this and came up with something more 'elegant' that doesn't require placing the IDE header in an unnatural orientation / messing with pin locations.

This time the daughter board steps the IDE header down as far as possible using 12mm long 2.0mm pitch header pins before bringing it back up again to just above 536 level.

This has the advantage of allowing you to plug in whatever IDE device you are using 'as was' with pins in the correct position as well as looking a bit more natural.

Initial tests seem to indicate that it is a good fit in the CDTV and functions correctly. More results to follow.