Myarc cards for sale/repair tips

[GDMike]

The reason I ask is because I just picked up a 9640 and I'm looking to see what languages are available for it other than GPL based applications. If that's making sense. 

I Saw something from a post from 1994 that said there's a Forth for the Geneve. 

I just didn't know which one it was.

Thx

[+arcadeshopper]

There's likely either a patched ti forth or mdos based one

Sent from my LM-V600 using Tapatalk

[+InsaneMultitasker]

Sadly, the gate array is locked into specific pages in GPL mode for a maximum of 16k space.  For kicks I wrote a simple loader that performed 8,192 unrolled MOVs to simulate bank switching in the 6000 space and patched the binary trampoline to execute my own bank switch.  While the result was somewhat acceptable for programs that didn't switch banks too often, more trampolining meant more banking and performance degradation - certainly unacceptable for many games.

 

From that idea I moved on to executing the binary from native OS mode by creating a small wrapper. It required the author to build an alternate binary using the Geneve video/sound port addresses and a Geneve keyboard routine (joysticks use the same cru). The ROM address-based trampoline was replaced with a bank mapper for the GEneve pages, which was determined by the wrapper at load time.  (another caveat is that since the binary runs out of RAM, writing to what is normally ROM is a 'bad thing'.)

 

A proof of concept was tested in 2019 for an unreleased game.  This ROM wrapper handled up to 128K; a binary up to 256K could be loaded into a standard Geneve; 512k was possible with expanded memory.  The wrapper program is around 1k in size.  Here's a shot of where I left the code last year. I think there was a sound-related issue left unresolved but gameplay itself seemed stable.

 

 

This is the snip of code that is used to switch banks.  Note that I use the ROM caller's bank number and slightly manipulate to use the Geneve memory mapper.  Probably simple to incorporate this sort of code into the binary using macros at assembly time?

 

[GDMike]

Just received my geneve and some other things.

Looks like the 32K mod needs to be done, battery holder/ battery replace needs attention, and heat sinks came to my attention for the regulators, and the bus connector needs a good cleaning.

But here are the photos! 

What is SST?

Edited October 31, 2020 by GDMike

[GDMike]

Battery fix complete

[GDMike]

Before I started my battery holder job, I did this 32K SRAM upgrade. This allows for higher version of DOS to be able to load.

I tucked my wire under my socket.

 

Edited November 1, 2020 by GDMike

[GDMike]

I've got some regulator heatsinks coming in next week, this board should have those put on, as I believe it was @insaneMultitasker that suggested it.

And some general cleaning and then wait for the video adapter to get in here from @shift838!!!

 

 

Edited November 1, 2020 by GDMike

[GDMike]

I don't think pulling these to put sockets in for the upgrade for VRAM is worth doing at this point.

Well see if software demands it  because I've removed 14 pin dips from 2 layer boards and sometimes it works out and sometimes I lost traces!!

These are 18 pin and, not my cup of tea. Plus, I think I could do it different by not piggy backing but just running trace wire to the new ram... TBC

Edited November 1, 2020 by GDMike

[+InsaneMultitasker]

Looking good.   The 32K SRAM wire is close the 12v regulator.  As unlikely as it sounds, I've seen the insulation damaged just enough to send a jolt of 12v+ to the 32K chip.  Make sure there is nothing in the clamshell that would cause the wire to be scraped or be crimped against one of the front leads or rear solder points. 

 

After you install the heat sinks, test resistance between the input-output, input-ground, and output-ground.  The heat sinks have a tendency to rotate slightly when tightening the screws and you can end up shorting the regulator to ground.

[GDMike]

6 minutes ago, InsaneMultitasker said:

Looking good.   The 32K SRAM wire is close the 12v regulator.  As unlikely as it sounds, I've seen the insulation damaged just enough to send a jolt of 12v+ to the 32K chip.  Make sure there is nothing in the clamshell that would cause the wire to be scraped or be crimped against one of the front leads or rear solder points. 

I actually have a loop of slack going way around the regulator. I'll probably tack it down so it doesn't travel.

I could've routed the trace wire to the nut area of the regulator, then I decided I didn't want the wire resting on the nut either.

Edited November 1, 2020 by GDMike

[+InsaneMultitasker]

16 hours ago, GDMike said:

I actually have a loop of slack going way around the regulator. I'll probably tack it down so it doesn't travel.

I could've routed the trace wire to the nut area of the regulator, then I decided I didn't want the wire resting on the nut either.

There are a number of through-holes adjacent to the RAM socket.  If you are comfortable doing so later, suck the solder out of one and route the wire through.  It's worth testing the ram upgrade before routing or removing solder.

 

Are you purchasing new regulators and caps for this round of updates?

[GDMike]

You know, I thought about it, and I'm only gonna heatsink em. 

No recapping

 

Edited November 1, 2020 by GDMike

[+InsaneMultitasker]

5 minutes ago, GDMike said:

You know, I thought about it, and I'm only gonna heatsink em. 

No recapping

 

Use a little thermal compound between the regulator and heatsink.  If the regulators are 'new' enough, you might be ok for a long while.  Caps could be tricky if they have dried out, so you definitely should at least consider that for down the road.

 

One problem you might run into is if the regulator leads are bent/cut too short, the regulator hole won't line up with the screw-hole once you slide the heat sink underneath.  In theory, you can straighten the leads just enough to make the stretch.

[GDMike]

2 minutes ago, InsaneMultitasker said:

Use a little thermal compound between the regulator and heatsink.  If the regulators are 'new' enough, you might be ok for a long while.  Caps could be tricky if they have dried out, so you definitely should at least consider that for down the road.

 

One problem you might run into is if the regulator leads are bent/cut too short, the regulator hole won't line up with the screw-hole once you slide the heat sink underneath.  In theory, you can straighten the leads just enough to make the stretch.

Yup. Exactly. And this might Force me to go ahead and replace it,or them.

I've got all the stuff, and a handful of regs. too, and caps. But I'll test first.

Since I don't know this card and or if it boots. That's my decider. Lol

[+dhe]

If you have to put in new voltage regs, you should use the new uber cool ones Tim found. I think found earlier in this thread. While the advice is a bit late, many people that add a holder for the clock battery, used a band saw or grinding wheel to remove a bit of the battery holder along the top of the card so there was the potential for a hit, when putting it back in a case.

[GDMike]

I like the clamshell they used here, it's plastic and not like the metal TI made type. 

[GDMike]

What is this for? I know it fits the Peb interface, side cable. But why?

And what is SST?

And this cable?

 

[+OLD CS1]

40 minutes ago, GDMike said:

The 8-pin DIN is the wrong type for a Commodore 64 or 128 but correct for the Geneve.  Given the 6-pin DIN on the other end, it looks like it would fit a Commodore monitor, like a 1084S.  Thus, I suspect it is a cable to use a Geneve with a Commodore monitor.

[+Ksarul]

26 minutes ago, GDMike said:

What is this for? I know it fits the Peb interface, side cable. But why?

And what is SST?

And this cable?

 

The SST Expanded BASIC Compiler is exactly what it says it is--a compiler for a subset of Extended BASIC. The weird Sideport adapter looks like a nice one-off to rotate the keyboard foot to a vertical position, probably to make if fit on a narrower desk without hanging over the edge. I suspect the cable is a monitor cable. . .

[+InsaneMultitasker]

2 hours ago, OLD CS1 said:

The 8-pin DIN is the wrong type for a Commodore 64 or 128 but correct for the Geneve.  Given the 6-pin DIN on the other end, it looks like it would fit a Commodore monitor, like a 1084S.  Thus, I suspect it is a cable to use a Geneve with a Commodore monitor

The 6-pin could also be for a Magnavox RGB, e.g., 8cm515.   I have one connected to Heatwave's Geneve though never turned on as it's about 'dead'    The cable has identical DIN connectors however, what is missing (or not visible in the picture) is the audio RCA connector.  Did the 1084S route the audio via the 6 pin din?   (Soldering wires to those DINs for customers was one of my least favorite tasks back in the 90s).

 

[+OLD CS1]

13 minutes ago, InsaneMultitasker said:

Did the 1084S route the audio via the 6 pin din?

No, pin 6 is ground.  Your conclusion is more apt.

[+InsaneMultitasker]

5 hours ago, dhe said:

If you have to put in new voltage regs, you should use the new uber cool ones Tim found. I think found earlier in this thread.

I didn't come up with the idea - that was a previous AA member- I just implemented it with your Geneve. Which of course is still running month after month and going on multiple years of 24x7 operation.  I think it's part Energizer rabbit. 

 

9 hours ago, GDMike said:

I've got all the stuff, and a handful of regs. too, and caps. But I'll test first.

If you are comfortable enough soldering and doing rework, I think I have more of the regulators and associated caps that @dhe referred to in his post.  I haven't had time to document the installation process for others that would like to try the same mod. So.. if you have interest in revamping your card with the cooler switching DC-to-DC parts after you have confirmed the Geneve works, and would like to help by documenting the process,  let me know and I will check my parts inventory.  If I have what's needed, I'll send them your way. 

[GDMike]

Thank you for the information and ideas.

[+InsaneMultitasker]

Posting this here in response to the following topic.

 

Cecure offered the installation of a reset button back in the 90s.  I stopped sharing/offering this hardware modification long ago out of concern for people tying a direct connection to the gate array.  In hindsight, adding simple isolation would be helpful, though to be honest I don't recall too many people blowing up their gate arrays with this mod.

 

According to my upgrade/repair notes and a quick review of the Geneve schematic, a reset is achieved by momentarily grounding the pull-up resistor R62 (using the bottom lead marked by red 'x') to trigger a hardware reset via the gate array. The schematic does not give any designation to the connected gate array pin, #73.

 

Installation was typically performed either by soldering a header near the neck of the card (to which a button could be connected) or by soldering/epoxying a button to the card directly above the monitor port DIN.

 

For an example of the momentary contact button, see @Shift838's project for the TI reset.

http://ftp.whtech.com/Users/Chris_Schneider/Projects/TI-Reset.pdf

 

 

Before this is posted to any Wiki or ninerpedia, it would be appropriate to confirm the connection and to review potential isolation.   My Geneve is currently not accessible otherwise I would have verified my notes with the real thing.   I was able to confirm the connection (R62, bottom lead) by inspecting another Geneve.

 

[webdeck]

13 hours ago, InsaneMultitasker said:

Posting this here in response to the following topic.

 

Cecure offered the installation of a reset button back in the 90s.  I stopped sharing/offering this hardware modification long ago out of concern for people tying a direct connection to the gate array.  In hindsight, adding simple isolation would be helpful, though to be honest I don't recall too many people blowing up their gate arrays with this mod.

 

According to my upgrade/repair notes and a quick review of the Geneve schematic, a reset is achieved by momentarily grounding the pull-up resistor R62 (using the bottom lead marked by red 'x') to trigger a hardware reset via the gate array. The schematic does not give any designation to the connected gate array pin, #73.

 

Installation was typically performed either by soldering a header near the neck of the card (to which a button could be connected) or by soldering/epoxying a button to the card directly above the monitor port DIN.

 

For an example of the momentary contact button, see @Shift838's project for the TI reset.

http://ftp.whtech.com/Users/Chris_Schneider/Projects/TI-Reset.pdf

 

 

Before this is posted to any Wiki or ninerpedia, it would be appropriate to confirm the connection and to review potential isolation.   My Geneve is currently not accessible otherwise I would have verified my notes with the real thing.   I was able to confirm the connection (R62, bottom lead) by inspecting another Geneve.

 

Interesting.  I used to have a reset button on my Geneve that routed to the front of the PEB, but I snipped it off when I had to send in the card for repairs.  I foolishly didn't document how it was connected, but based on the blobs of solder left, it looks like the wires were connected to both sides of C12, and there is a resistor on one of the wires, so pressing the button shorts the resistor across C12.  I could be wrong, however.