Firefox Sync Issue?

[jetusenet]

Hi,
I spoke with Matt and he suggested I check here. I searched the forums but could not find a similar problem so I will post here and hopefully someone will be able to point me in the right direction (please)......

Description:
After giving up on repairing my original PCB set, (it has the slo-motion audio situation) I purchased a working set and finally assembled my Firefox. As nothing is easy when it comes to collecting video games, I seem to have an issue with the video.

I am using a newer Vision Pro and thought that might be the issue so I pulled out an older Wells 4900 and it had the exact same issues so I do not think it is the display.

When I play without the Dexter installed, I can play the game without any video issues but once the Dexter is installed, that is when the problems become apparent, it appears to loose sync ever so often.

If you could please take a few minutes to check these quick videos, you will see what I am describing.

Working without Dexter (a)
https://youtu.be/OPpKSLxemzc

Working without Dexter (b):
https://youtu.be/80IVq6A7nMQ

Attract issues:
https://youtu.be/Pp9ekJWS_M4

Starting game issues:
https://youtu.be/yu97CX28CP4

By the way, Matt gave me the link to remove the rainbow colors which I will be doing ASAP but until then, any and all suggestions with regards to resolving the sync issue would be greatly appreciated -I would really like to get my FF going, it's only been about 4 years!

Thank you
Jeff

[Warren Ondras]

Interesting problem... I encountered something similar on my Firefox test setup -- loss of vertical sync during skip or freeze commands. I spent a lot of time testing and troubleshooting, but in the end it turned out to be due to low voltage on the graphics board (not the NTSC board.) This board generates sync when the laserdisc player is disconnected, and when it thinks the laserdisc sync isn't reliable such as during a skip.

Anyway, I suggest checking voltage between the +5V and GND test points on the graphics board. The power connectors sometimes oxidize or burn, causing low voltage and all sorts of mayhem.

My only other suggestion is to make sure Dexter's DIAGNOSE light is OFF. If the extra logging slows down Dexter, the laserdisc interface signals might not occur at exactly the right time, which I think might cause loss of sync. It's _supposed_ to work with DIAGNOSE mode ON, but make sure it is OFF anyway.

Good luck!

[jetusenet]

Warren,
Thank you very much for taking the time to give me those suggestions. I will get that DDM hooked up and see what I can find.
I will also check the Dexter and see if the Diagnose function is enabled or not.
Thank you again
Jeff

[Matt Ownby]

yes, Firefox will sync up with an incoming ntsc signal (converted from the demodulator board) and it looks like (from the video) that this sync is somewhat unstable. I don't know enough about the sync circuitry to suggest what might be wrong, so Warren's suggestion is as probably as good as we can offer.

Having DIAGNOSE enabled shouldn't matter since the ntsc signal is generated regardless of whether dexter is lagging behind or not.

[jetusenet]

Hi,
I did a quick attempt at removing the rainbows and it looks pretty good for now.... I really need to get that sync issue resolved, so onto voltage testing.

I have about +5.19vdc on both the Main and NTSC PCBs but only +4.94 on the Video PCB.

Checked all the connections and cleaned up the interconnects to no avail so I pulled the power cable, (P10/P11) and will replace the 2X +5 lines with double 18ga wire with Trifurcon pins. Also the GND line with a new 18ga wire with Trifurcon pins as well just to be sure.

There is also a soldered in +5vdc line on the Main PCB to a quick connect plugged into the Video PCB.

Hopefully this will allow more current to go to the Video PCB and balance out the voltages on the 3 boards - I hope.

I will post the results once tested and thanks again for the help.

BTW, the DIAGNOSE is not enabled.

Jeff

[Warren Ondras]

+4.94V is probably sufficient, but good clean power connections will make it more reliable over time, so it's still a good thing.

Do you have any additional boards you could swap in for comparison? I suspect you have a problem on the graphics PCB.

[jetusenet]

OK, the double 18ga wire mods I did on P10/P11 really helped the voltage on the Video PCB. What was once +5.19vdc on both the Main and NTSC PCBs and +4.94 on the Video PCB is now +5.15vdc on the Main and NTSC PCBs and +5.05 on the Video PCB.....

Harness-small.jpg (67.89 KiB) Viewed 261608 times

But that did not resolve my sync issue......

I had never tested the PCB as it was sold to me working from a reliable source but gave it a try and received a RAM error: Graphics RAM 2 @ 8H. I replaced that and retested and it then passed the test but still has the sync issue....

BTW, this is a sitdown boardset that I converted to an upright w/ switcher P/S, (standard 15A Happ).
I have W1 and W2 cut with the 15ohm 10w resistor installed at R205 on the NTSC PCB and SP1 jumpered and SP2 cut on the main PCB...........

[jetusenet]

Just a thought - could a defective coaxial cable from the Dexter to the RFI PCB cause this sync issue?

Thanks
Jeff

[Warren Ondras]

Just a thought - could a defective coaxial cable from the Dexter to the RFI PCB cause this sync issue?

I think that's pretty unlikely. If the loss of sync was random, maybe, but instead it's matching the laserdisc skips. (You can't see the skips; that's the beauty of the Philips player, and now Dexter. )

That said, those coax cables do fall apart over time. Replacing it is a good thing, though it probably won't solve this particular problem.

Do you have a logic probe or oscilloscope?

[jetusenet]

I will replace the cable ASAP but in the mean time, yes I have both a Logic Probe and an Oscilloscope.
Would I be able to use these to determine what is causing my issue?
Thanks for the follow up.
Jeff

[Warren Ondras]

I will replace the cable ASAP but in the mean time, yes I have both a Logic Probe and an Oscilloscope.
Would I be able to use these to determine what is causing my issue?

Possibly, if we can figure out what to check. I suspect the problem is in the "full interlace sync generator" (pages 30 and 31 of FFxRev2_s.pdf from the Dragon's Lair Project site.) I don't know exactly what to tell you to look for, but I'm curious about a few things:

- It would help to know what the HSYNC and VSYNC outputs to the monitor are doing when the monitor loses sync. It _think_ it's only losing horizontal sync, since the picture doesn't "roll". I'm wondering if vsync drops (I am assuming it does not), and how long it loses hsync.

- In the sync generator section, "LOCK" is an input from the CPU, which tells the graphics board that there is valid sync coming from the laserdisc. When LOCK is dropped, the sync generator takes over. This is supposed to eliminate any jumping or loss of sync while the laserdisc player is skipping/busy, but it doesn't seem to be working as it should. I'm curious how long the LOCK signal goes inactive. I think it should be just a few milliseconds on the skips during the attract mode.

- _HLOCK is derived from LOCK and a couple other things. I'm curious how long it goes high during skips.


Sorry I can't be more specific. If you don't feel like digging this deeply into it, finding another boardset to test with is your best bet, I think.

[jetusenet]

Warren,
Thank you for the detailed suggestions. That is my next step after receiving and swapping the coaxial cable......

Reason being,,,,,,,

I have a somewhat working boardset that has that slo-mo problem (described as Jim Carry doing the slo-mo routine) and I figured that I would give it a try just to see what would happen and even though the game does not work properly with the defective PCB installed, I get that sync failure, seems to be a but more pronounced but makes me think that it might be something else than the first "working" board, although it is a possibility that both boards have similar problems as anything is possible.

I placed the order a couple days ago for a "professional" 75 ohm coaxial cable and once I get that installed and tested I will update this post.

Thank you again for your time helping me with this problem, it is very much appreciated.

Jeff

[jetusenet]

BTW, here is a video of the test screen showing the RAM error. It seems that it looses sync during the test as well?

https://youtu.be/-3j7Aaz3oRc

[jetusenet]

Well, I tried the "professional" cable and that did not help at all......

Next step is to get another Video PCB or bust out the scope,,,,,,,,,,,,,

Is there anyway that the Dexter could be causing this? (sorry, I have to ask)....

[Matt Ownby]

Well, I tried the "professional" cable and that did not help at all......

Next step is to get another Video PCB or bust out the scope,,,,,,,,,,,,,

Is there anyway that the Dexter could be causing this? (sorry, I have to ask)....

You could probably substitute Dexter's output with any NTSC source (like a VCR of DVD player) and prove that it's not Dexter.

It is unlikely that dexter would be the culprit since the sync problems seem to occur when the disk skips and Dexter's video is generated completely independently of this event.

I think the problem is what warren says: the FFX sync generator that takes over during skipping isn't working properly for some reason. The only way to move forward is to probably spend some time understanding how it is supposed to work and observing it in a working environment, then comparing this environment to the non-working environment. In other words, a tall order, but possible.

[Warren Ondras]

I agree that Dexter's video wouldn't be dropping sync, but could a missing acknowledgement signal on the 25-pin interface cause sync issues?

On my test setup, if I unplug the 25-pin cable from Dexter, I lose vertical sync, even though the BNC video signal is still connected. I don't know exectly why this happens, or if there is a variation that would cause the specific symptoms you are seeing. (Does the game even work at all if any of the RDEN/WREN/DAV/DAK signals are missing?)

I'm just speculating here, but it may be worth giving the cabinet's 25-pin connector and cable a good look, and checking for continuity to the Firefox PCB end if possible.

[Matt Ownby]

I agree that Dexter's video wouldn't be dropping sync, but could a missing acknowledgement signal on the 25-pin interface cause sync issues?

On my test setup, if I unplug the 25-pin cable from Dexter, I lose vertical sync, even though the BNC video signal is still connected. I don't know exectly why this happens, or if there is a variation that would cause the specific symptoms you are seeing. (Does the game even work at all if any of the RDEN/WREN/DAV/DAK signals are missing?)

I'm just speculating here, but it may be worth giving the cabinet's 25-pin connector and cable a good look, and checking for continuity to the Firefox PCB end if possible.

I don't know enough about the hardware to say one way or the other

[jetusenet]

I will check the interface cable today and see what I see, thank you for the excellent suggestion.

[jetusenet]

Just completed testing P43 through the edge connector P16 and all lines have continuity ....

This narrows the possible culprits down....

[Warren Ondras]

Excellent. You might try using the logic probe on P43 pins 10, 12, 22, and 23, to see if all of them are active, and if anything changes during the loss of sync.

I use the old Radio Shack logic probe that beeps on signal changes, which is very handy for getting a feel of what is happening. I guess a visual indicator would still do the trick, though.

If all of those lines are busy and nothing weird happens during loss of sync, then I think it rules out Dexter as a source of the problem. Not sure how much that helps, though, if it's still broken...

If I think of anything else, I will let you know.

[jetusenet]

Very cool, thank you for my next assignment!

I'll get the probe out and give it try and see what I can come up with.

Thank you again for your time and help.

Jeff

[jetusenet]

OK, finally had the time to sound out those points and it seems that they are working correctly as comparatively (to each other) there are 2 sets of "clicks" - 2 points have fast clicks and 2 points have slower clicks, so there is activity on those lines and none are "stuck" or different from each other (per set).

I took these readings (probed) at the RFI PCB, P16-8,10,J,K which originate from P43-23,10,12,22 respectively.

From my very limited knowledge working on LD games, it appears that the Dexter is sending out the needed info and the problem seems to lie on my game boards somewhere.

Uggggggg!

[Warren Ondras]

One more thing to try is removing the RFI board and plugging the connectors directly into the main boards. (Make sure the orientation is correct!) All sorts of flakiness can sometimes be eliminated this way.

[jetusenet]

One more thing to try is removing the RFI board and plugging the connectors directly into the main boards. (Make sure the orientation is correct!) All sorts of flakiness can sometimes be eliminated this way.

Thanks but I already did that, sorry I forgot to mention........

[jetusenet]

BTW, I can understand the need to strap the boards for either a switcher or linear power supply but why is there the additional strapping changes needed to the NTSC PCB for the Upright vs. Sitdown, specifically the addition of the 15ohm 10w resistor at R205?

Sorry for the constant questions but this is my first and only LD game and I am trying to understand what is going on........

[Warren Ondras]

BTW, I can understand the need to strap the boards for either a switcher or linear power supply but why is there the additional strapping changes needed to the NTSC PCB for the Upright vs. Sitdown, specifically the addition of the 15ohm 10w resistor at R205?

Which rev of the schematics are you looking at? In the rev 2 schematics on DLP, the strapping for that resistor is labeled "cut for use of linear supply". Perhaps the earlier schematics assumed the power supply type depended on the cabinet type?

[jetusenet]

Actually it is from an Atari Service Bulletin (attached).

Why would that be different strapping between the Upright and the Sitdown?

[Matt Ownby]

Actually it is from an Atari Service Bulletin (attached).

Why would that be different strapping between the Upright and the Sitdown?

I don't know for sure, but I strongly suspect that the upright and sitdown used different power supplies. FFX can work with either a +15V/-31V power supply or a +22/-22V power supply and the boards need to be slightly modified depending on which power supply is used.

If you look at the FFX rev 2 schematics (available from dragon's lair project), the "Power Input" sheet on PDF pages 8 and 9 shows an example of how one of the boards is modified for either power source. (the linear PS is the +/- 22V while the 15/-31 is a switcher that apparently sucked big time).

I suspect that the mods are designed to ensure that the same voltage reaches the voltage regulators using either setup. So if I were to take a guess, I'd say that VR1 (+12V output) and VR2 (+5V output) are expecting at 15V input, while VR3 (-5V) is expecting a -22V input. I could be wrong... but it would be easy enough to measure on a working setup.

[Warren Ondras]

Wow, that memo is dated April, 1985... I didn't think there were any Firefox machines still in operation by then!

The resistors lower the voltage before it gets to the linear regulators, so the regulators will produce less heat. The jumpers bypass the resistors when the voltage is already low "enough". Not terribly elegant, but this was the least of Firefox's troubles...

[jetusenet]

Now the veil has been lifted! I checked that info out and it makes perfect sense, thanks for pointing me in the right direction.

I posted the section being discussed for others to study if the need arises.

I have contacted the seller and he agreed to send me replacement PCBs so hopefully soon I will be able to report good news.

Thanks again for all your help.

Jeff