• Updated 2023-07-12: Hello, Guest! Welcome back, and be sure to check out this follow-up post about our outage a week or so ago.

Mac-to-VGA monitor adapter struggles

bigmessowires

Well-known member
There's also a mechanical design problem here, independent of the whole sync splitting business. I noticed this at Mactoberfest when I plugged the prototype adapter into a bunch of different Macs: it's a tight fit on some Macs with a slotted video card, and I'm afraid the solder bumps on the bottom of the adapter PCB may short out against the metal frame surrounding the monitor connector on those Macs.

Compared to the Griffin Mac PnP adapter, my adapter has about 3mm more material below the DB-15 Mac connector, and that begins to cause clearance problems for some video cards. Here are the two adapters resting on the desk, with the DB-15s facing each other. Notice how the BMOW adapter sits a few millimeters higher.

IMG_3832.jpg

Here are the adapters lying on their side edges, with the two DB-15s approximately aligned. This gives you a better idea of where the extra thickness is coming from in the BMOW adapter.

IMG_3833.jpg

The Griffin adapter probably uses a DB-15 with in-line solder cup terminals that come straight out the back, and it's secured inside a custom injection molded case. But my adapter uses a right-angle DB-15, so there's a 1.6 mm thick PCB below the DB-15, and then pins stick out roughly an additional 1.5mm below that.

There isn't a whole lot that I could do to improve this. I can use a 1.2mm thick PCB instead of the standard 1.6mm, which might help a little, but the pins will still stick out just as far as before. I could cut the pins flush with a side cutter, which is OK for a one-off, but I don't think it's possible to batch assemble larger numbers of units that way.

If I switched to a DB-15 with solder cup terminals, there's no great way to attach it to the PCB. The best option would be to sandwich the PCB between the two rows of solder cups, which is exactly what I do with the DB-19 on the Floppy Emu adapter:

IMG_3834.jpg

This is a little funky, since the exact distance between rows of solder cups is not standardized. A 1.6 mm PCB usually fits but is super tight, while a 1.2 mm PCB is loose and needs to be secured somehow until it's soldered down. Manufacturers don't like this method very much since it's non-standard. I think it needs to be assembled by hand.

I couldn't use the same PCB sandwich technique on the VGA connector side, since it has 3 rows of terminals instead of 2. I'd have to continue using a right angle connector like I already have now. That's not a problem for fit since the VGA connector is far enough away from the body of the Mac, but it means the DB-15 and the VGA connector wouldn't be in the same plane, and the adapter would look a bit odd. I think the DB-15 with solder cups is also slightly longer front-to-back than the DB-15 right angle, so the whole adapter would need to be a few mm longer.

If there were a DB-15 right angle surface mount part, that might help things. But I don't think that exists.

Any brilliant suggestions? The current design may actually be good enough as-is, even if it's a tight fit on some. I didn't have enough time at Mactoberfest to really pay attention carefully to the fit, or see how close the solder bumps got to any metal frame pieces. For Macs with built-in video like the IIci where the connector is horizontal, there's no fit problem at all. I think it's only a potential issue when the connector is vertical, such as when you're using a NuBus or PCI video card.
 

François

Well-known member
Make it a dongle? The PCB would only have the VGA connector directly soldered to it, the DB-15 being connected to the PCB with a small cable. The DB-15 can then be a regular plug that won’t have any clearance problems. You’ll just need a way to firmly secure the cable coming from the DB-15 to the PCB.

Something like that (but with a DB-15):
xcellon_dp_vga_12_displayport_to_vga_adapter_1490942.jpg
 
Last edited:

Phipli

Well-known member
Any brilliant suggestions?
Nope, but I wonder if these would help in any way?


They just fit into the recessed Mac II series cards.

Another option is one of these as a mini extender - I have to do this to make my AUI transceiver fit in my IIx


Ignore that they call it a gender changer, it is just a stubby extension.
 

bigmessowires

Well-known member
A dongle with cable is a decent idea, and it should solve the fit problem. Although I'd hoped to make this a one-piece adapter like other Mac-to-VGA adapters: just a simple brick that plugs straight in to the Mac.

I like the idea of the gender changer mini-extender, because it's optional. You could use one if the computer needs it, or omit it otherwise. Another similar possibility would be a DB-15 extension cable like this one:


Although 6 feet is overkill and it only needs to be a few inches. But it would be the same idea: use this if you have problems with fit, omit it if you don't need it.
 

bigmessowires

Well-known member
I probably confused everybody with my talk about "fake sync signals" yesterday. The point is that the VG900b appears to care about the rising edges of HSYNC, not the falling edges (or not only the falling edges).

I did a simple trick to make the rising edges nice and consistent, at the expense of introducing jitter in the falling edges, and it fixes the problem where the VG900b image was blinking in and out every few seconds. Now I have a stable image. This isn't a good permanent solution since other monitors will care about the falling edges, but it's an OK test for the VG900b right now.

Unfortunately there's still some shimmer remaining. It's usable as is, but for long-term use I would find the shimmer annoying. So what might cause the shimmer?

- EMI on the analog R, G, B signals? This would seem like the most likely explanation, except that my other monitors don't show any shimmering with this same setup.
- Jitter on the HSYNC falling edge? Maybe the VG900b uses both the falling and rising edges of HSYNC for different things, and jitter in either one introduces some instability in the image?
- Jitter on VSYNC?

I'll try to make a video of the shimmer, if it's useful to anybody. I feel like this is getting closer to a usable solution, but isn't there yet.
 

bigmessowires

Well-known member
More data: I acquired a High Resolution Display Video Card BCGM0322, P/N 630-4222 630-4230. This is a boring unaccelerated card that I think people also call "Toby", and that I think supports the same resolutions as the older "Macintosh II Video Card" P/N 820-0198-C Toby that I tested earlier, but the hardware is somewhat different. That older Toby had a very odd CSYNC with serration pulses during vblank that are twice the frequency of normal CSYNC pulses.

In comparison, the High Resolution Display Video Card's output is much more modern: it has a normal-looking CSYNC *and* it also outputs separate HSYNC and VSYNC. The diode-rectified self-power voltage is 4.35V. These sync signals mean the card doesn't need the adapter that I'm building. And based on that, I'm guessing that later model Apple video cards like the 4-8 and 8-24 don't need this adapter either.

I also acquired a new-in-box Macintosh II Monochrome Video Card 661-0518, 630-4385, 820-0285-A. I think this is the card that was bundled with the Mac IIcx. Still in the original package, I felt bad opening it. This card's CSYNC looks normal (like the IIci, IIsi, and High Resolution Display Video Card) but its HSYNC and VSYNC are not connected. As a result, the diode-rectified self-power voltage is only 3.75V.

If you are keeping score at home: yes I now have four monitors and four video cards on my IIci. Feel the power!

I still want to test a IIvx, but it's beginning to look like there are only four places where the sync-splitting capabilities of this adapter may be relevant:

- IIsi
- IIci
- Macintosh II Monochrome Video Card 820-0285-A, but self-powering may be difficult and it may need external 5V power
- Macintosh II Video Card 820-0198-x, but its CSYNC behavior is different from the others

Not exactly a huge audience.
 

bigmessowires

Well-known member
Ah OK, so probably not the IIvx then.

Regarding the construction method of sandwiching the PCB between rows of solder cups, I got confirmation that this is indeed done by hand. But at the small-ish volumes of my manufacturing (a few hundred at a time) it would still be done by hand even if I used a right-angle DB-15, so it's the same cost either way. So I suppose the only real drawback to that method is the DB-15 and HD-15 (VGA) connectors wouldn't be aligned in the same plane, because the HD-15 has three rows of terminals and can't use the PCB sandwich trick.
 

bigmessowires

Well-known member
Cool, thanks. Yes, a very similar idea. I've become distracted with my IIcx restoration and other things, but finishing this VGA adapter is still high on my list.
 

croissantking

Well-known member
I should note that the cable end of the monitor isn’t in very good condition. On my other Macs I have to wiggle it around to get the image to not be red, blue or green tinted.
IMG_4403.jpeg
I’ve done something about this now.
Hasn’t helped to get it to run on my IIsi or IIci built in video though.
@Phipli is the 4•8 display card multisync capable?
 

croissantking

Well-known member
Something isn't right with the monitor, cable or computer.

It seems you’re right. The manual for the 15AV says it will work with a Macintosh II.

IMG_4541.jpeg

Something is wrong with the monitor since I don’t get any output on a IIsi or a IIci.

How might I begin to troubleshoot this?
 

bigmessowires

Well-known member
don’t get any output on a IIsi or a IIci
Those are the Mac models that output composite sync instead of separate horizontal sync and vertical sync. That's the whole point of the proposed sync-splitting VGA adapter whose design I've been discussing here. Would you mind spinning off a separate thread for your monitor debugging?
 

croissantking

Well-known member
Those are the Mac models that output composite sync instead of separate horizontal sync and vertical sync. That's the whole point of the proposed sync-splitting VGA adapter whose design I've been discussing here. Would you mind spinning off a separate thread for your monitor debugging?
No problem. Sorry to hijack.
 

bigmessowires

Well-known member
No worries. My first guess is DIP switch settings, since you'll probably need different settings for the IIci and IIsi compared to other Macs.
 

register

Well-known member
As long as you have an adaptor to properly connect the video card and the display, consider to set the screen resolution by software, overwriting actual sense pin settings. To make any supported display resolution available in System 7.1 (and probably some other System versions), use »Activate all Resolutions«, an extension which permits to access any possible resolution by the usual monitor control panel. Currently I have no access to a machine with this extension, but I have been using it a lot with my PowerBook 180. On recent Macs just Option-click to the resolution menu on the monitor settings.
 
Top