1088XLD - Custom A8 Computer in 1050 Drive Case

[+mytek]

VGA Output on 1088XLD: UPDATE - Design Changes Coming

 

Well I knew the first prototype PCB wasn't going to be the final version. I needed to adjust some things in the layout which really required making a 'real' board to see the alignment issues in the flesh. Try as I might, I was off by 0.05" which requires shoving things over to compensate for the error. It really is tricky to get the measurements correct in a virtual aspect when taking measurements in mid air so to speak.

 

Another thing that requires a board change is the need to add either independent buffering for the DIN-13 RGB output and the RGB input on GBS-8200, or I some sort of electronic switching between the two. I've opted for the electronic switching which turns out to be considerably cheaper, and still allows me to stick with through-hole parts.

 

Lastly I have located a source for the proper extended female headers, so no more stacking will be required.

 

The first board worked great as a proof of concept, but now I need to mold this into a production board .

[JoSch]

Great news. I looked for my board, I had lying around for years and I totally forgot about it.

It's the GBS-8220, having two VGA outputs. I will try wire it up on my XEL and try it with the old 1024x768 LCD, I have standing around.

[+mytek]

Great news. I looked for my board, I had lying around for years and I totally forgot about it.

It's the GBS-8220, having two VGA outputs. I will try wire it up on my XEL and try it with the old 1024x768 LCD, I have standing around.

 

Do you know what version that board is?

 

I know there was a mounting hole location change starting with V4, but other then that, the concept should still work for V3 as well, but it would be nice to figure out what the offset is in the mounting holes, and perhaps add one more set to make this a bit more universal.

[JoSch]

 

Do you know what version that board is?

 

I know there was a mounting hole location change starting with V4, but other then that, the concept should still work for V3 as well, but it would be nice to figure out what the offset is in the mounting holes, and perhaps add one more set to make this a bit more universal.

I believe, it's a V3. I can measure it, when I'm at home again.

[+mytek]

I believe, it's a V3. I can measure it, when I'm at home again.

When I get a chance I'll upload a pdf of the layout at 1:1 scale so that it can be printed for reference, then place your V3 board on top to see the mounting hole difference.

 

Give me about 5 hours, since it's only 3:20 in the morning over here .

[JoSch]

No rush. Im still at work myself.

Edited May 22, 2019 by JoSch

[Mr Robot]

Great news. I looked for my board, I had lying around for years and I totally forgot about it.

It's the GBS-8220, having two VGA outputs. I will try wire it up on my XEL and try it with the old 1024x768 LCD, I have standing around.

 

My XEL works great with my GBS-8220. I'm going to make a little PCB to do the firmware override.

[+mytek]

GBS-8200 V4 Dimensions

 

 

Scale 1:1 Drawing (print 100%): GBS8200_V4_dimensions.pdf

[+mytek]

My XEL works great with my GBS-8220. I'm going to make a little PCB to do the firmware override.

 

Looks great! However you might want to make the mode select header 3 pins, so as to have a place to park the shorting jumper when in NTSC mode. That way it doesn't get lost .

 

Example:

 

[Mr Robot]

I was planning to attach it to a toggle switch for my own usage but you're right, I'll add a 3rd pin.

 

EDIT: I'll stick that PicKit programmer port on there too, while I'm at it

 

EDIT2: Also I need to rethink the software, I've got a VBXE and it can do 480i so turning off the de-interlacer should maybe be optional...

Edited May 22, 2019 by Mr Robot

[JoSch]

My board fits your board dimensions:

 

It is a GPS 8220 V3 with a date stamp of 2013/10/23.

 

There is something curious with my board. It has only the serial port connector soldered, while the I2C pads are there, but without the connector.

Edited May 22, 2019 by JoSch

[+mytek]

Ok I see what the difference is. The board just sticks out further on one side .

 

The I2C header is never installed from the factory. So this needs to be added after the fact.

 

Thanks for the verification .

[+mytek]

I was planning to attach it to a toggle switch for my own usage but you're right, I'll add a 3rd pin.

 

Toggle switch still works either way, but this feature seems more like a jumper selection to me, since you need to do more then flip the switch to change video modes (i.e., Antic and GTIA swap). Which means that the case needs to be opened up anyway, so no advantage to using a switch.

 

If you change the code to enable interlace let me know how that goes since I would be interested in testing that out with the Sophia board as the RGB source.

[JoSch]

The I2C header is never installed from the factory. So this needs to be added after the fact.

 

Really? A lot of pictures on the internet show both headers. So I assumed, its an anomally.

Thanks for the verification .

You are very welcome.

[+mytek]

If you do a search for "GBS-8200" and then images as the type of search you'll see way more pics of these boards without the I2C header populated. Perhaps at some earlier time they did come with the header installed, but that no longer seems to be the case.

[Mr Robot]

 

Toggle switch still works either way, but this feature seems more like a jumper selection to me, since you need to do more then flip the switch to change video modes (i.e., Antic and GTIA swap). Which means that the case needs to be opened up anyway, so no advantage to using a switch.

 

If you change the code to enable interlace let me know how that goes since I would be interested in testing that out with the Sophia board as the RGB source.

 

 

Will do. The original shmup forum that the code comes from has lots of settings, including code for UNO's and RasPi's. I was thinking I should be able to make an NTSC hex file and a PAL hex file with the switch switching between deinterlaced and not deinterlaced settings. Most of the time you'd want the deinterlacer turned off. My XEL is never going NTSC so I wouldn't need the NTSC switch, I was thinking I might use the GBS for another device at some point but who am I kidding.

[JoSch]

If you do a search for "GBS-8200" and then images as the type of search you'll see way more pics of these boards without the I2C header populated. Perhaps at some earlier time they did come with the header installed, but that no longer seems to be the case.

I meant pictures of the GBS 8220, even the picture in the manual show two headers.

 

Edit: Here is the manual page:

Edited May 22, 2019 by JoSch

[+mytek]

Obviously it is a possibility that some boards will come with one, since the board was designed to accept it. But don't be surprised if the board you buy either now or in the future doesn't have one (both of my recently purchased single VGA output boards don't).

 

Hopefully they'll always retain the place to mount it.

[+DrVenkman]

I just looked more closely at my board (arrived about a week ago). It's a dual-VGA GBS 8220, version 3.0, with all the header locations fully populated. Made June 2015.

 

Edited May 22, 2019 by DrVenkman

[+mytek]

I just looked more closely at my board (arrived about a week ago). It's a dual-VGA GBS 8220, version 3.0, with all the header locations fully populated.

 

So perhaps it's just a 8200 thing, and the 8220's come populated . Or it's related to version 3 (my boards are version 4).

[flashjazzcat]

Perhaps it depends what parts they managed to get from the local market on the day.

[+Stephen]

Now you gonna have to make me find my board then photo it.

[+mytek]

It doesn't much matter either way, just a little bit more involved if there is already a header in place. You see my carrier board needs a female header on the bottom side of the GBS board to mate with it's male counterpart on the carrier. So if there is an existing topside header, that needs to be removed. If not, then the solder that's in the pad holes needs to be removed. So in either case a little bit of de-soldering and soldering needs to be done.

 

Edit: What does matter is that the V3 board's overhang on the VGA output side is problematic, and will interfere with components and other connectors, so I'm afraid this project will need to be limited to the V4 version of the GBS which appears to be what's being sold now days.

Edited May 23, 2019 by mytek

[JoSch]

Success. It worked when I tied the grey wire to the CSYNC signal on the board.

[+mytek]

So this is what I found out thus far on the various Arcade Video Conversion boards.

 

GBS-8200: This is the initial board I will be supporting (presently selling as V4.0 but I have seen an image of a V5.0).

 

 

GBS-8220: This is the one with dual VGA output connectors that really isn't suitable for my application due to the overhang on the VGA output connector side of the board, and the 2nd VGA connector which would interfere with the parallel expansion bus on the XLD (this board is presently selling as V3.0).

 

 

???: This is a new HDMI output board that appears to be using the same video scaler chip and I2C connection, and *may possibly work, but I presently don't have one to test with. These sell for as little as $36 from eBay.

*Note: Although it says 5V IN unlike the GBS boards that give a range of 5V-12V IN, it does look like it has a very similar power regulation circuit as the GBS-8200 so there is a chance it too will work from a higher voltage.

Of course this will need to be verified before toasting your nearly $40 investment.

 

 

This preliminary schematic shows where things currently stand with the proposed RGB/VGA video switching.

Without the jumper (or optional video select switch) it defaults to passing video to the VGA converter.