J-PAC and CRT EmuDriver

[ncflagg]

I'm trying to do a proof of concept for myself with a mobo/CPU combo where the sweet spot is Windows XP x64. I have all compatible drivers loaded properly and am trying to get CRT EmuDriver working with an ATI HD 4350. I have a J-Pac in my cab currently and having no success getting both my LCD Marquee monitor and my arcade monitor to work at the the same time.

Actually, the problem really seems to be more about the arcade monitor not being recognized properly by the video card. I see that three 75ohm resistors can be used to modify the VGA cable between the J-Pac and video card (and I'm going to try that) to get the secondary display to work in Windows, but I'm wondering what the benefit of that is. If that mod is successful, will CRT EmuDriver be able to detect my monitor's resolutions through the J-Pac, or am I going to have to fidget with each resolution manually? Am I better off bypassing the video component of the the J-Pac?

[nipsmg]

Why xp64?  Xp-64 is problematic, has much lower driver support, and I’ve had nothing but issues with it.  I’d go xp-32 bit if you HAVE to go xp, but windows 7 supports most so drivers so why not
Just go win 7-32?

[ncflagg]

CRT Emudriver doesn't support Win7 32:

Under Windows 7, we support the Radeon HD 2xxx, HD 3xxx and Radeon HD 4xxx families, only for the 64-bit version of the operating system.

As for WinXP 32, I'd lose 1GB of my 4GB RAM.    I may end up trying that if I can't get this working, but it feels more like a J-Pac + video card issue than an OS issue. Like I said, I have every driver installed and the system is stable.

[Calamity]

Actually, the problem really seems to be more about the arcade monitor not being recognized properly by the video card. I see that three 75ohm resistors can be used to modify the VGA cable between the J-Pac and video card (and I'm going to try that) to get the secondary display to work in Windows, but I'm wondering what the benefit of that is. If that mod is successful, will CRT EmuDriver be able to detect my monitor's resolutions through the J-Pac, or am I going to have to fidget with each resolution manually? Am I better off bypassing the video component of the the J-Pac?

Hi ncflagg,

Your J-PAC isn't the culprit here, it's the arcade monitor that doesn't load the rgb lines. Even if you attempted a direct connection you'll be facing the same problem. The resistors trigger monitor detection so Windows won't switch off the video output to the arcade monitor. That's all it does, it doesn't have any effect on what resolutions are detected or anything (I don't understand what you mean there), you control all that through VMMaker.

That said, keep in mind this trick comes at a cost: you'll get a considerably dimmer picture, as part of the current will be drained to earth. Instead of 75 ohm resistors, try using higher values, as high as it still works, so you'll loose less brightness.

[ncflagg]

..it doesn't have any effect on what resolutions are detected or anything (I don't understand what you mean there), you control all that through VMMaker.

At one point, I saw various vertical horizontal refresh values (15.x-16.x) populate the OSD after running VMMaker (I didn't set them). I figured it was calculating them from the EDID from the LCD panel. I assumed (wished is more accurate) it would do the same for the CRT (analogue signal, I know). After some testing, I never saw those custom resolutions again (hRefresh is always blank now); not sure why.

I've got some 100ohm resistors so I'll start with that. Thanks for the tip!

[Zebidee]

Actually, the problem really seems to be more about the arcade monitor not being recognized properly by the video card. I see that three 75ohm resistors can be used to modify the VGA cable between the J-Pac and video card (and I'm going to try that) to get the secondary display to work in Windows, but I'm wondering what the benefit of that is. If that mod is successful, will CRT EmuDriver be able to detect my monitor's resolutions through the J-Pac, or am I going to have to fidget with each resolution manually? Am I better off bypassing the video component of the the J-Pac?

Hi ncflagg,

Your J-PAC isn't the culprit here, it's the arcade monitor that doesn't load the rgb lines. Even if you attempted a direct connection you'll be facing the same problem. The resistors trigger monitor detection so Windows won't switch off the video output to the arcade monitor. That's all it does, it doesn't have any effect on what resolutions are detected or anything (I don't understand what you mean there), you control all that through VMMaker.

That said, keep in mind this trick comes at a cost: you'll get a considerably dimmer picture, as part of the current will be drained to earth. Instead of 75 ohm resistors, try using higher values, as high as it still works, so you'll loose less brightness.

I've got some 100ohm resistors so I'll start with that. Thanks for the tip!

I haven't heard of this trick to get an arcade monitor recognised, but I can see how it would work. I did an online search but unfortunately found nothing, so would appreciate a link if there is one?

I'd suggest sticking with the 75 ohm termination resistors if possible because this matches the video input impedance expected. If impedance is mismatched then you will get more interference on the video lines in particular any stray capacitive loading and reflection. If that makes the image not bright enough, then try using this simple home-brew RGB amplifier with your 75 ohm resistors:

https://www.retrorgb.com/thsamps.html

Just follow that design and add your 75R terminators to the RGB inputs. I'd probably also try adding 0.1uF (104) ceramic capacitors in series on the RGB inputs, after the 75R terminators. That "AC couples" the inputs to the homemade amp per industry convention and minimises DC crosstalk between the devices.

I'd also add an extra 22uF capacitor across the 5v and GND terminals, in addition to the ceramic 0.1uF cap that is already there. This will minimise noise on the 5v supply line. The 22uF cap would be electrolytic, so pay attention to polariy (+ to 5v, - to GND)

Then feed from that into your JPAC as normal.

[Calamity]

Hi Zebidee,

You can find the 75 ohm resistors trick scheme here: http://www.geocities.ws/podernixie/htpc/cables-en.html



I suggested higher values so less current is derived to ground and therefore the picture is not so dim while still have working detection. This is because the resistors are in parallel with the rgb lines, contrary to the amplifier scheme you posted. Anyway, the amplifier solution is much better, and definitely something I need to try here.

[Zebidee]

Hi Zebidee,

You can find the 75 ohm resistors trick scheme here: http://www.geocities.ws/podernixie/htpc/cables-en.html



I suggested higher values so less current is derived to ground and therefore the picture is not so dim while still have working detection. This is because the resistors are in parallel with the rgb lines, contrary to the amplifier scheme you posted. Anyway, the amplifier solution is much better, and definitely something I need to try here.

Thanks Calamity, the pic you posted is exactly what I suggest adding to the THS amplifier inputs. It is not "contrary" to the amp, it would be all part of the the one thing.

To clarify my earlier question, I was asking specifically about need for adding 75R termination into JPAC inputs as I've not encountered that, though I guess it could be something specific for some video cards   

Yes with higher value resistors less current is drawn off. The problem is impedance mismatching, RGB signals are designed for 75R termination, so varying that will create signal errors/interference/ringing/reflection. The THS amp would be there to restore DC signal levels (no dimming), while still getting the EDID effect.

Another way to get around all this might be to do it in Windows 7 instead, with a newer video card.