H2

[Dan Iacovelli]

I didn't see mine listed in the chart A1 81 5258794 so I guess I'm ok (no expansion port)

I got second one stored that I got from another person that has expansion but don't know the numbers

[iesposta]

Naturally, mine is an "X" unit with the LE A/V mod in it. Things aren't looking good for me, unless I look into an earlier model.

Mine was cosmetically beautiful but would not play certain homebrew 7800 carts.

Ripped the Longhorn mod out of it, replaced the removed components, did the simple composite mod. Now it gets some use as a backup system in a bedroom.

Bought an older 7800 model that currently has the Bryan A/V mod. Plays everything.

[iesposta]

Don't forget that there are mods to remove the timing "fix" that let 2600 Dark Chambers play, killing great games like Robot Tank and Starpath Supercharger compatibility. (That did not change my Serial X Homebrew 7800 failure issue, though.)

 

But I'm sure Batari is aware of this mod.

Things like this mod just add another layer of differences in the 7800 models.

Although if 2600 games are not supported on Concerto, maybe this mod doesn't factor in.

[+-^CrossBow^-]

I would add, just about impossible.

 

Exactly a case in point. See another example below.

 

The short version...

If you have an A1 console with a motherboard from 87 and an Expansion Interface with one socket chip, you may have different compatibility results than an A1 console with a motherboard from 86 and an Expansion Interface with two socket chips.

 

The long version...

 

 

Here's an edit on the chart, focusing on the A1 serials and the 2600 M-Network game Burgertime:

 

7800_A1_BTIME.png

 

Across A1 variants, the cartridge Burgertime worked perfectly fine. However, the same Burgertime on the Harmony cart, did not work for all of them - but which do you nail down as being the issue and what is that issue exactly?

 

Is it the number of socket chips? Nope..Two A1's had 1 socketed chip, one worked fine the other did not.

 

Is it the Expansion Interface inclusion? Nope. An A1 with the Expansion and one without, had issues.

 

Perhaps it's the Board Date? One A1's with issues with Burgertime had an "87" board date, the other had a board date of "86".

 

Could the Burgertime issue be pinpointed further then? Unfortunately, another problem arises if just focusing on an A1 serial and "86-45" motherboard combination. Although both A1 serial with a "86-46" performed the same with Burgertime, a two chip socket versions fails to run Time Pilot, while a one chip socket versions runs Time Pilot without an issue.

 

Futhermore, both "86-45" A1's are labelled revision "A", but the one socket chip has a CO25233 001 and the two socket chip has a CO25233 002.

 

Additionally, a CO25233 002 was paired to a one socket chip motherboard having a board date of "87-13", still revision "A" though, which appeared to break a considerable number of 2600 cartridge compatibility.

 

The situation became worse when that same CO25233 002 was paired with a "87-31" motherboard, revision "A", no expansion interface.

 

Finally, improvement in 2600 compatibility with the "88-29" motherboard, revision "C", no expansion interface; however, the 7800 homebrew fails to run on it.

 

For the summary:

A1 Serial - December 1986 Manufacture date, a 86-45 board date, Expansion Interface and 1 Socket chip.

A1 Serial - January 1987 Manufacture date, a 86-45 board date, Expansion Interface and 2 Socket chips.

A1 Serial - May 1987 Manufacture date, a 87-13 board date, Expansion Interface and 1 Socket chip.

A1 Serial - December 1987 Manufacture, a 87-31 board date, No Expansion Interface and 0 Socket chip.

 

Above provided different results with various 2600 cart and 2600 harmony trials. 7800 only specific compatibility appeared consistent among all the above models, but 7800 (homebrew) compatibility changed with this console:

 

A1 Serial - August 1988 Manufacture, a 88-29 board date, No Expansion Interface and 0 Socket chip.

 

Could we at least nail down that 7800 homebrew compatibility issue to A1 serials from 1988 and newer? Unfortunately, that is not the case, as proven when looking at the AT-84 serials results.

 

 

 

Going back to the original chart from Toymailman's post, the same homebrew had problems with not working at all on one AT84 unit, some sort of playability issue on another AT84, and worked perfectly on other AT84 units.

 

Focusing on the motherboard of those AT84 and AT85 units...

 

 

 

The AT84 serials paired with their respective motherboards show:

84-21 = Works.

84-24 = Works.

84-25 = Works.

84-26 = Doesn't Work.

84-27 = Works with Issues.

 

The AT85 serials featured in the listing have a 84-24 dated motherboard.

 

From the sampling, best 7800 compatibility among AT models appears to be an AT84 or AT85 with a 84-24 motherboard. The 2600 compatibility is a mixed bag regardless.

 

 

Standing-O for the developers, that is some mess to work through in trying to make PCB/carts/programs as compatible as possible; finding what chips/cart hardware combination does (not) work is a formidable task. Fred trying to include 2600 compatibility as best as possible can be nothing short of a hellish nightmare.

 

 

Hmm...you have me thinking on this now. I have two 7800 units. My original 7800 bought new back in 87 I believe appears to be an A3 serial number on the sticker. However, the board inside it is all socketed chips with an expansion port on it. It shows rework done at the factory so my guess is that it is an early revision mainboard refurbed and sold as new years later. I've tested it with all the trouble activision carts and it works 100% and I've never had issues with it. It has been modified as a Dev unit with Eckard's dev OS bios installed and the other needed changes.

 

The second one I have also plays everything 100% that I've tested but is an A1 86.... unit. I don't recall any chips in it being socketed and it does NOT have an expansion port installed though I do believe the silkscreen is present for one. I might have to check using my original Harmony and burgertime image to see if there are any issues with either of these? Or were these burgertime tests only done with Harmony Encore carts?

[RevEng]

Yeah, I don't think the problem is as clear-cut as different models behaving differently.

 

Based on what I've heard about the Atari production lines from the Antic podcast interviews, Atari didn't have a tight control on components and production. They changed things up when they could catch a deal, and reworked circuits on the fly. IIRC some official 7800 carts have components that have been added without actual pads, and some don't.

 

We've witnessed consoles that have changing behavior as they start up, where Maria eventually changes the amount of DMA time it uses, subtly changing the display... the Choplifter area between ground and sky is one of these, and Trebor has documented this in a video I believe. My pet theory to 7800 differences is the clock (even in a "good" 7800) is pretty close to being too delayed in some circuit branches. Poor/aging components cause slower rise times or other delays. This could potentially result in problems with ram (mirrors?), eprom, interrupt timing, etc.

 

Just a guess, based on Maria acting differently after a warm-up on some units, when it should be deterministic. The fix may ultimately be cleaning up the clock source, or hunting down where delays are. Beyond my pay grade, though.

[Trebor]

...I might have to check using my original Harmony and burgertime image to see if there are any issues with either of these? Or were these burgertime tests only done with Harmony Encore carts?

 

The results were from Toymailman's sampling in 2011, an original Harmony cart was utilized.

[+-^CrossBow^-]

 

The results were from Toymailman's sampling in 2011, an original Harmony cart was utilized.

 

Good to know, I will get both of my 7800s out now and see what I get with Burgertime on the harmony. I would check all the trouble Activision games, but I already know those all work on both of my 7800s without issue.

[+rdemming]

Are the compatibility issues specific to NTSC unit revisions or do different PAL unit revisions also have known compatibility issues?

I've never encountered any issues running 2600/harmony/cuttle cards on my PAL 7800s.

 

Robert

[Deteacher]

 

Good to know, I will get both of my 7800s out now and see what I get with Burgertime on the harmony. I would check all the trouble Activision games, but I already know those all work on both of my 7800s without issue.

 

I think I'll go down and check out these games with my harmony as well. Which ones should I check besides Burgertime? I'll report back with my results.

 

EDIT:

 

I just tested out all the 2600 games that have known compatibility issues with the 7800, using my original Harmony (I got one of the first ones.) I tried:

 

Activision Decathlon

Robot Tank

Space Shuttle

Burgertime

Supercharger games

 

They all worked fine on my 7800.

Edited August 23, 2016 by Deteacher

[Shawn]

 

The results were from Toymailman's sampling in 2011, an original Harmony cart was utilized.

 

The "X" serial number units being a crap shoot makes sense as the "X" units where the last ones made. God knows what kinda Franken-consoles they are on the insides. I bet they where using whatever was left about at that time just to milk the last few pennies out of the 7800.

[+Mitch]

X serial number 7800s were made in China.

 

Mitch

[Trebor]

X serial number 7800s were made in China.

 

Mitch

 

'nuff said.

[Andromeda Stardust]

Wiis were also made in China...

[+DrVenkman]

Wiis were also made in China...

Huge numbers of 2600's were made in Taiwan (Republic of China) and by and large they're great. Edited August 24, 2016 by DrVenkman

[iesposta]

 

Good to know, I will get both of my 7800s out now and see what I get with Burgertime on the harmony. I would check all the trouble Activision games, but I already know those all work on both of my 7800s without issue.

My X 7800 will not play 7800 Beef Drop or 7800 Scramble. It plays 7800 Moon Cresta just fine though.

[Andromeda Stardust]

My X 7800 will not play 7800 Beef Drop or 7800 Scramble. It plays 7800 Moon Cresta just fine though.

If Albert is selling homebrews in the AA store, some of which do not run on 100% of authentic 7800 consoles, then how can we guarantee that Batari, by building a 7800 flash cart, can ensure all games working, including those which did not work on original 7800s?

 

Even if he can't get 2600 games working (wasn't the Concerto in fact compatible with Encore firmware?), I would still like to try one out on my 7800, which has not failed to load a single 7800 period release, homebrew, or reproduction I threw at it.

 

For the record, mine is an A1-serial without the expansion port.

[StarshipUK]

Like many others, I would be happy to take the risk and see if it works based on whats ready now. This has been 6 years in development now, and I have a feeling it will never be released otherwise. If it does not work, then I will buy a 7800 which does work with it. My 7800 is the French version with the Peritel (SCART) cable included.

[MrBeefy]

I just want to see it released. I've held off getting a Harmony you 1 knowing I really want the 7800 one

[phoenixdownita]

It's his project, up to him to decide when/if it's ready to be released.

[+Gemintronic]

Batari makes solid products. With highly skilled and creative people you just have to wait until they are ready. Not an excuse. Just how most people I know like that work.

[phoenixdownita]

I can only volunteer one detail that happened on my 7800.

 

I had to physically fix my Robotron 2048 cart by adding a series 100ohm on the Vcc pin of the ROM.

Without that the game would usually glitch and reset at the end of the first screen, sometimes earlier, a couple of times in the middle of level 2, it drove me nuts.

That was the only game with the issue I ever found, it came in a batch and the owner swears it worked fine on his 7800.

 

As I understand it not all 7800 glitched on it .... I never got what the fix actually fixed.

 

By lowering Vcc for CMOS tech both L and H levels do get impacted. I am not sure though if the fix was wrt the out of the ROM or the in (the addresses) or in general for the L or the H or both levels.

Anyway maybe attempting to lowering a little bit the 5V rail and see what happens may be a worthy albeit "out there" trial.

 

EDIT: or it could just be a "red herring" not worthy of any further discussion. I just wish I understood why the fix would work at all.

Note that the same game over an EPROM (27C512) never gave me any issue whatsoever .... so go figure.

[Andromeda Stardust]

So it's an issue with logic levels being too high or too low causing certain hardware revisions to glitch? Maybe a trim pot inside the cart could fix this???

[phoenixdownita]

So it's an issue with logic levels being too high or too low causing certain hardware revisions to glitch? Maybe a trim pot inside the cart could fix this???

I am not saying it is the case, only that I experienced something like it.

But it may be something completely different, let batari do his own investigation.

[StarshipUK]

No offence is intended, but as I said I have now waited 6 years (which is a tad excessive) and still no sign of this being released any time soon, and judging by the last few posts its being developed by a one man team who wants to be left alone to work on the problems and not accept the help of others to speed this up a bit. I am also wondering if the Cuttle Cart 2 development took 6 years and had the same issues?

 

Looking for something else I found this for the 7800 which I might try: http://www.vectrex.hackermesh.org/index.php/en/mvbd-mvmc-3/faqwhich should tie me over for the next 6 years until this is ready.

 

Has anyone else tried one of these?

Edited August 28, 2016 by StarshipUK

[phoenixdownita]

No offence is intended, but as I said I have now waited 6 years (which is a tad excessive) and still no sign of this being released any time soon, and judging by the last few posts its being developed by a one man team who wants to be left alone to work on the problems and not accept the help of others to speed this up a bit.....

Don't know what to say, I started waiting as well then I made my own low-tech multi and got my fix ... actually I'm glad it was late as it made me get over my lazy ass and learn something, as a bonus during the process I uncovered that the DoubleDragon and Rampage 7800 rom dumps were incorrect -> their documented bank-switching was incorrect too.

 

I do hear your pain as it seems the 7800 is the home of a few of these multi-year projects that seem to never end at the same time it's his project and he decides how to run it and who to get help from (if any).