UAV Rev D modded 7800 Rev B issues

[at5200forever]

Attempting installation of UAV Rev D in 7800 Rev B, and have following issues with video output:

 

2600 games appear in black and white.  If Maria color is disconnected, 2600 games then appear in vivid color.

 

7800 games also sporadically in black and white with distorted video.  Much to my surprise, when Maria is disconnected (TIA connection remains), the color also returns to 7800 games, though the image distortion remains.  I’ve attached a video of the MARIA distortion, with Maria connected.

 

The machine had an unsuccessfully home-built rendition of the Longhorn Engineer AV solution completely installed and backed out  Also re-capped, and volt regulator replaced.

 

UAV is isolated from BIOS by a Heavy duty Velcro adhesive strip.

 

Have double-checked, and in some cases at least triple-checked the wiring.  Appears to be correct.  Same pattern of defects appears in the RF output of 7800 games as well.  Also double-checked to make sure I put the resistors removed for LHE mod correctly back where they came from.

 

Would describe as video that exceeds the normal horizontal left and right edges of the video frame, in a wavy pattern.

 

I just made an attempt to capture some video and the pattern seems a little different from my description but should give a clear notion of how it’s behaving.  A definite observation is that there seems to be something wonky about the cartridge slot.  Reseating and pressing on the cart has an effect on it.  You can even hear at one point where the system crashed in response to slot movement.  I tried to trim the plastic some to allow an Imagic cart to fit, but even stock 7800 Atari carts are very tight-fitting.  I definitely need to trim some more... and look at those connections and traces.

 

 

 

 

[tdididit]

i am not sure is UAV will work right with 7800 :)

7800 got two video chips and the reason why in VCS mode you dont have colors is simple - you need COLOR signal switch between TIA & MARIA COLOR signal outputs plus switching signal from TIA pin9 (if i remember good)

 

Take a look at schematics

 

PAL & NTSC modelas got different video signals placement too on video buffer chip (U13 in PAL, U3 in NTSC 7800)

 

Cheers & have fun ;)

[+DrVenkman]

54 minutes ago, tdididit said:

i am not sure is UAV will work right with 7800

I’m quite sure you’re mistaken.

 

 

 

 

[at5200forever]

1 hour ago, tdididit said:

i am not sure is UAV will work right with 7800

7800 got two video chips and the reason why in VCS mode you dont have colors is simple - you need COLOR signal switch between TIA & MARIA COLOR signal outputs plus switching signal from TIA pin9 (if i remember good)

 

Take a look at schematics

 

PAL & NTSC modelas got different video signals placement too on video buffer chip (U13 in PAL, U3 in NTSC 7800)

 

Cheers & have fun

 

Thanks for your response— but you might have a closer look at UAV Rev D, which has input for Maria color and a second input for TIA color.  Bryan said there’s some sort of a clever gate mechanism he created to make it choose between the two modes.  I’m really no engineer, so I take his word on that.  Worked perfectly awesome in my other 7800 (Rev C) for about an hour— until I tried reinstalling the RF shield and shorted out the flip-flop switch.  ?

 

Which boards are those you’re working with?  They look promising — I’m just not any good at surface mount work.  Tried it with LHE after Electronic Sentimentalities went on (permanent?) hiatus.  Know my limitations...

 

I gotta tell you, I’m so happy with the way UAV cleaned up the output on my other 7800 — and even the 2600 output on this one — I’m going to buy a few more to bring my 2600s and 5200s up on it.
 

Even in the condition this machine is in, the attached video shows what it is doing for the 2600 side of the system.  Very nice!  (Actually the video doesn’t do it justice — and I’ve no clue how to play this game.)  

 

 

Edited February 3, 2020 by at5200forever
Video not working

[+-^CrossBow^-]

What about the RF side? How does everything look through RF? 

 

You said this was previously an LHE modded console right? You might need to double check the traces off the 74sl that you had to replace back in place. I seem to recall that on mine, I still had issues with sync and video color drop out until I soldered the chip back into the board. Another thing to do, is while you have it one, lightly press on some of the SMD ICs on the UAV board. I've had a few that had some cold solder joints that I had to touch up again that did similar issues. While on that, lightly move the color burst wires around a bit to see if anything changes to indicate a loose connection somewhere. 

 

But I'm more curious to see how the RF output looks as that will determine if the issue is with the UAV or the wiring going to it, or something else on the 7800 that just happen to go south at the same time as you doing the upgrade.

 

[at5200forever]

This was an eBay item I bought in 2007.  It was delivered crushed, but thankfully the board didn’t get dealt any perceivable damage. From the first time I attempted to use it I recall seeing unplayable flicker on RF output of 7800 games.  Main reason it sat on shelf for 11 years, and I even bought a second 7800.  I thought probably too far gone — but You know, there’s always that faint glimmer of hope in every gamer’s eye.  And with emulation around, I suppose I felt no urgency, really.  But here we are — I’ve treated this as my training unit for the UAV mod.
 

Samples of RF output attached

 

[+-^CrossBow^-]

That means the 7800 itself is at fault and not the UAV in this case since the same thing is happening in RF as well. I'm inclined to think the sl74s32 is to blame. I've seen similar before and had to replace it. I also had one sent to me with various issues and it actually had a bad maria chip on it that was producing color artifacts and occasionally going to BW mode video as well. 

 

But I would check those traces around and under the socket of that chip...

 

And the velcro is something I still do this day. The stuff I use holds the UAV in place very well and still allows it to be pulled off if needed. Can even remove the velcro off the IC chip without too much trouble if needed. Here was the latest one I did for a fellow AA member not long ago.

 

 

 

[at5200forever]

i pulled the chip out of the socket and replaced it with another of the same.  No change (no better no worse).  Then attempted to desolder the socket and put the chip back in the board directly.  I soldered it in and tested and as I feared that made things worse — but it still does start up and run cartridges.  (Not all that bad I suppose considering I tore a trace while removing the socket.)
 

Not sure where I want to go from here with this unit — or if I want to move forward.  Seems like it might be best to look for another console, and enjoy my other unit after the power issue gets resolved.  Thanks guys for all your wise advice.  It’s been extremely helpful.  

[+-^CrossBow^-]

Part of the reason you pulled a trace removing the socket is because you used Machine pin sockets. Due to the round base at the top of the pins, they attache much more securely to the pads than standard sockets. If you plan to only do it once and done, these are fine, but when you need to remove them they can be tricky and frequently pull up pads and traces when trying to remove them. Arcade repair guys hate machine pin sockets with a passion because of this and I can sorta understand this. I personally try and use dual wipe sockets when I can as they are easier to remove should that be needed.

 

[at5200forever]

Regarding machine pin sockets...  I started to use these after I viewed someone’s YouTube video for a LHE install to 7800 where they suggested they are superior sockets — and in some respects they are — but now I have to agree with the arcade repair guys.  I cut the machine pin socket into pieces once I realized I snared a trace, and probably spared the mainboard from greater damage, but yeah... it is what it is. Lesson learned.  I appreciate the information, hopefully others will see this and avoid these sockets.

[Bratwurst]

25 minutes ago, at5200forever said:

Regarding machine pin sockets...  I started to use these after I viewed someone’s YouTube video for a LHE install to 7800 where they suggested they are superior sockets — and in some respects they are — but now I have to agree with the arcade repair guys.

The mentality behind recommending machine pin sockets probably comes from their use in military grade applications and a perception that it's going to be the best. The only reason I can fathom that the military uses them is because they're more resistant to vibration for deployed vehicles, and it's definitely not with any concern of future rework. An associate of mine who works as a submariner technician confirmed to me that the only 'repair' those guys do is straight up module and board swaps, no component level diagnosis/repair. Read into that what you will...

 

Not to mention the fact that most through-hole components feature flat pins which is like trying to shove a square peg into a round hole (machine socket), the connection is not going to be as reliable vs a dual wipe.

[+-^CrossBow^-]

38 minutes ago, Bratwurst said:

Not to mention the fact that most through-hole components feature flat pins which is like trying to shove a square peg into a round hole (machine socket), the connection is not going to be as reliable vs a dual wipe.

Correct so the only real contact with the legs is along the sides of them inside the socket holes.

[emmanuelf]

No sorry, this is irrelevant, that is not how they work.

Yes they are painful to rework.

Yes they are incompatible with square pins used for stacking purpose.

Yes they are not low profile (except for some special ones)

But they are superior in all other aspects. And normaly when you put some in place, it is not to do any rework later on it/remove them, there is very very very low reason to do it.

Edited February 3, 2020 by emmanuelf

[+-^CrossBow^-]

9 minutes ago, emmanuelf said:

No sorry, this is irrelevant, that is not how they work.

Yes they are painful to rework.

Yes they are incompatible with square pins used for stacking purpose.

Yes they are not low profile (except for some special ones)

But they are superior in all other aspects. And normaly when you put some in place, it is not to do any rework later on it/remove them, there is very very very low reason to do it.

I wasn't arguing their use in proper applications. Just stating that in the arcade PCB repair circles, they do NOT use and do NOT like machine pin sockets. The traces and pads on the older arcade PCBs doesn't take much to pull them loose and so again their use in arcade circles isn't common place. I myself do not use them in most cases either since again the legs on most through hole ICs is flat and not round as machine pin sockets are intended for.

 

And so my comment to the OP was that using them in this case could have been a contributing factor in additional traces and pads coming loose when he was trying to remove the sockets. 

 

My guess is that there was trace damage to the board to begin with before any sockets were installed as the OP himself stated he has these same similar video issues with the RF now and before attempted to install the UAV if I read his answers correctly?

 

[emmanuelf]

The fact that legs on most through hole ICs are flat is not a valid argument. Machine pin socket are intended for such legs. That is what I wanted to say.

 

The best of the two worlds : machined pins on carrier. But they are hard to find or you need to do your own carrier. And they are expensive, so generally overkill.

Edited February 3, 2020 by emmanuelf

[Bratwurst]

1 minute ago, emmanuelf said:

The fact that legs on most through hole ICs are flat is not a valid argument. Machine pin socket are intended for such legs. That is what I wanted to say.

 

The best of the two worlds : machined pins on carrier. But they are hard to find or you need to do your own carrier. And they are expensive, so generally overkill.

A dual wipe is going to make more surface contact on a flat pin than a machine socket. Machine sockets work best with round pins.

[+-^CrossBow^-]

12 minutes ago, emmanuelf said:

The fact that legs on most through hole ICs are flat is not a valid argument. Machine pin socket are intended for such legs. That is what I wanted to say.

 

The best of the two worlds : machined pins on carrier. But they are hard to find or you need to do your own carrier. And they are expensive, so generally overkill.

As they only make contact on the sides of the legs it is relevant because I've seen plenty of ICs especially custom ones used in many arcades and in consoles that do not have the exact same width on their IC legs. In fact it is common on many older ICs to see a wedge like shaped leg that is thinner on the corners than the rest of the legs. As such those legs might not make good contact with the inside of the round socket. So yes in the case of most IC chips, the flat legs make much better surface contact on dual wipe sockets as both front and back of the IC legs are in contact with the socket pins. As I was in the service I can attest to the fact that component level repair wasn't done when I was in the Army in the 90s either. But I did get to see plenty of boards and most of those that had sockets did have machine pin sockets on the. They also had custom made ICs with round pins instead of blade legs on them.

[at5200forever]

I appreciate the passion you all bring to this topic.  I can see there’s definitely a time and place for these sockets.  It’s far from the only questionable choice I made nor mishap that occurred.

 

Lessons learned:

1. If at all possible, start with a unit with output that is rock solid in RF.

2. Use high quality equipment to service your system.  Entry level equipment will provide entry-level results.

3. The belt-and-suspenders “military-grade” approach may produce durable solutions, but things that bind tightly tend to do so in destructive ways.  These are game systems, not tanks, nuclear subs, supertankers, nor hospital devices.

4. Makes sure if you buy on eBay to tell the seller it is a fragile item. You might even go so far as to offer to pay for additional shipping materials.  Shipping these things in just one layer of bubble wrap and styrofoam popcorn is a recipe for disaster.

5. If you’re going to DIY, expect to make mistakes.  They’re part of the educational experience.  (If that’s not acceptable, use a different approach.)

6. Don’t be afraid to seek help.  This community seems more than welcoming.  (Thanks, guys!)

7. Tomorrow still brings hope to correct what is broken today.

With that said, I’m setting aside this 7800, at least for now.  My leading theories are:

 

a.  Bad Maria chip.

b.  Bad other components.

c.  Defective traces on the motherboard or since the cartridge port works so hard, cracks in the traces where the cartridge slot is soldered in.

 

At least I got a lot further with UAV than with LHE, and it appears I will emerge with at least one working 7800 to play...  36 years after Atari demoed it at CES ‘84.

? 

 

[+-^CrossBow^-]

@at5200forever I appreciate your attitude on this and bringing the rest of us back to Earth on this topic.

 

I do hope that continue to ask if you ever have questions about things. There is a very large breadth of experience in these forums both hobbyist and professional grade.

 

With that, to get back to your 7800 it is important to start with a fully working unit, but I can also see your reasoning for wanting to install the UAV into it. I'm going to guess that your assumption on the 7800 was that the RF modulator might have been faulty and that using a UAV or other AV upgrade method would bypass this issue and get your a fully working 7800. So again, I can totally see your logic on that. I do think the 7800 can be repaired and think it will be an excellent opportunity for getting more experience and skill as you work through getting it back to operating status. 

[Joerice1978]

Hey anyone have more ideas on this? I have rev b and just installed uav and I'm having the same problem 

[+DrVenkman]

2 hours ago, Joerice1978 said:

Hey anyone have more ideas on this? I have rev b and just installed uav and I'm having the same problem 

Well, as noted above, the UAV rev D works great in my own 7800, and Jesse's (@-^CrossBow^-) experiences installing them into many customer 7800's can support that. Please post pics of your installation and someone can help you out I'm sure.

[Joerice1978]

On 2/2/2020 at 6:46 PM, -^CrossBow^- said:

That means the 7800 itself is at fault and not the UAV in this case since the same thing is happening in RF as well. I'm inclined to think the sl74s32 is to blame. I've seen similar before and had to replace it. I also had one sent to me with various issues and it actually had a bad maria chip on it that was producing color artifacts and occasionally going to BW mode video as well. 

 

But I would check those traces around and under the socket of that chip...

 

And the velcro is something I still do this day. The stuff I use holds the UAV in place very well and still allows it to be pulled off if needed. Can even remove the velcro off the IC chip without too much trouble if needed. Here was the latest one I did for a fellow AA member not long ago.

 

 

 

Hey your setup is different from what I did.   Could u please tell me which of your orange wires go to each.  I believe inbound my problem 

[+-^CrossBow^-]

11 hours ago, Joerice1978 said:

Hey your setup is different from what I did.   Could u please tell me which of your orange wires go to each.  I believe inbound my problem 

the wire points here go to the same points off the resistors I just used unsued vias on the board instead of the bottoms of the resistors. I'm back to using the resistor method as it results in shorter wire runs on my last installs I've been doing. The PDF I wrote up on this a few years ago is still accurate in regards to the Rev. D of the UAV.

 

A recent install I did last month where I'm using the resistors to tap my signal from. I did this because it was a more straight shot and keeps the wiring looking more tidy in the install. Also moved the UAV to the ram chips so if the client wants to install a BIOS replacement in the future, it will be easier to do so without the UAV sitting ontop of it. This might also be one of the earliest made '84 units I've worked on. Notice the TIA date stamp indicating the 10th week of 1984? Also has the RCA X ram chips on it that I've only seen on early '84 made units.

 

 

 

My most recent install does the same thing on the wiring, but I now have JST connectors I can use and started to add a large piece of shrink tubing over the UAV to keep the wiring in place and provide a better smoother surface on the bottom to attach the UAV with. This install also features a new method of modifying the cart sleeve that isn't obvious when looking at it to allow the 3rd party games to fit with ease. Except tigervision carts as that is due to a different issue as to why they won't fit. I have a method of correcting that, but I'm not as fond of doing it.

 

[+DrVenkman]

1 hour ago, -^CrossBow^- said:

his might also be one of the earliest made '84 units I've worked on. Notice the TIA date stamp indicating the 10th week of 1984?

Yeah, but the BIOS is the 22nd week of '84. That's about the same as my personal AT84 machine (which is 8423). Still, it's always fun to see an early '84 machine naked.

[Joerice1978]

Well darn.  Tks guys.  Must be something else wrong with mine then. The rf pic is good no distortion.   Could I be using to small of a guage of wire? I'm using 28