Retron77 Compatibility List

[godslabrat]

So, might as well get the ball rolling on this. Let's start charting the limits of the R77. If you've played it, list the results!

 

1- Cart works.

2- Cart does not work, ROM does.

3- ROM does not work, might with patch/hack.

4- Game will never run due to hardware limitations of the system (Ex: Supercharger)

 

Link: https://docs.google.com/spreadsheets/d/1lzJdxHuRL9I-EdJ0f1abb2a9Gv46qLj-gmAj-PSBbN0/edit?usp=sharing

Edited July 3, 2018 by godslabrat

[Andromeda Stardust]

thank you for making this topic. i have a 77 on preorder so ill throw everything ive got at it when it arrives. as some reviewers have already stated, pitfall ii cart doesnt work, rom only.

[Keatah]

Wonder if it'd be useful to make an excel spreadsheet? OR just use the thread's search function to find a certain game?

[godslabrat]

A spreadsheet would be better, but for now, let's just get some data

[dj_convoy]

We know from John Hancock that Omega Race works, but the system doesn't recognize the controller add on. Same with Star Raiders, the Kid's controller games, and presumably Basic Programming (an assumption on my part).

Edited June 25, 2018 by dj_convoy

[+Muddyfunster]

I've started a spreadsheet that I'll update as I work through my collection (attached). Feel free to update and add to it. Most of my entries will be PAL based.

 

I've indicated :

 

• Cartridge Title
• Manufacturer
• Part number (if applicable)
• Version (if applicable)
• Region (NTSC / PAL)
• Result
• Rom Only Result

 

I've started testing my Red Label titles, did 43 with 2 genuine fails (Retron 77 would not read the cart but a 2600 would). Real Sports Boxing & Fatal Run. Could be false positives, hopefully more folks can test the same titles and prove the exceptions right or wrong.

 

I found a lot of cases where a genuine 2600 would read a cart but the Retron wouldn't unless the contacts were given a clean. Seems the Retron is a bit more picky than my Jr.

Edited June 26, 2018 by Muddyfunster

[+Muddyfunster]

sheet attached in a zip as I couldn't just attach the excel file.

 

Retron 77.zip

[AtariLeaf]

We know from John Hancock that Omega Race works, but the system doesn't recognize the controller add on. Same with Star Raiders, the Kid's controller games, and presumably Basic Programming (an assumption on my part).

 

I have a modded NES pad for Omega Race. I wonder if that would work.

[ComputerSpaceFan]

I asked in the other thread but no one answered so I'll ask here: I have an old Krokodile Kart, would that work or is that also melody based?

 

[Andromeda Stardust]

I have a modded NES pad for Omega Race. I wonder if that would work.

 

I figured it out. And guys, you might be disappointed, but here goes...

 

The controller port on the Retron77 is wired, much like the 2600daptor v1, to be joystick and paddles only. The best method to detect paddles is to look for a high logic signal on the paddle pins (5 and 9).

 

Omega race booster grip adds two extra buttons by jumping pins 5 and 9 to vcc (pin 7).

 

Any high logic signal applied to these pins will confuse the controller port to map paddles instead of joystick. The port uses a usb protocol to interface as a gamepad.

 

Each usb gamepad is detected as an x/y axis with two buttons.

 

In joystick mode, UDLR (pins 1-4) simulate a digital hat with 8 positions plus off. Fire button is button 1. Button 2 is unused and not accessible.

 

If a high logic signal is detected on the paddle pins, the layout changes and immediately starts polling the paddles.

 

Paddle mode: analog x/y axis with some predefined range of paddle travel marking +/- 100%. Left paddle button (normally R) becomes button 1. Right paddle Fire (normally L) becomes button 2.

 

So any controller that sends a high logic signal to pins 5 and 9 will remap the way the controller ports are encoded. Any joystick mode controller that actuates both L+R of U+D will return something in between. 2600daptor and Stelladaptor used 50% R or D to denote that both buttons were pressed. Any axis encoder for dpad or joystick inputs cannot be at two extremes at once, even though specialty controllers do exactly this by sending something other than raw directional input to the joystick pins, including opposite inputs.

 

If only they mapped the controller port as discrete inputs for pins 1, 2, 3, 4, 5, 6, and 9, then many of these special function controllers would operate. You cannot have a joystick with built in paddles on the Stelladaptor, and the same applies here. It may be possible to update the microcontroller code in the controller ports to alter the way it sends data to the emulator, or it might not. Even if so, it would require a rewrite of much of the emulator code to compensate for those changes.

 

It's unfortunate, but any controller standard added to the controller port must fit into the cookie cutter form factor of an x/y axis plus two buttons. And joysticks only get one button, and no simultaneous opposing directional inputs allowed.

 

When I get my Retron77, I will load the test cart rom and see what happens when I toggle the built in paddles on my joystick on/off. It may allow me to dissect how the controller ports encode the data. I suspect based on prior observations, they clone Stelladaptor/2600daptor v1 behavior.

[Andromeda Stardust]

I asked in the other thread but no one answered so I'll ask here: I have an old Krokodile Kart, would that work or is that also melody based?

 

No.

 

Well perhaps if you preflash a single game rom to Harmony or Crocodile instead of the flash cart bios, maybe.

 

The harmony can be flashed with a single game rom through usb. This devcart cycle would be the only way to load external roms through the cart port. I have no idea if crocodile cart allows it.

[Thomas Jentzsch]

Krokodile allows that too. And I guess it may work with RetroN77.

[+wongojack]

I figured it out. And guys, you might be disappointed, but here goes...

 

The controller port on the Retron77 is wired, much like the 2600daptor v1, to be joystick and paddles only. The best method to detect paddles is to look for a high logic signal on the paddle pins (5 and 9).

 

Omega race booster grip adds two extra buttons by jumping pins 5 and 9 to vcc (pin 7).

 

Any high logic signal applied to these pins will confuse the controller port to map paddles instead of joystick. The port uses a usb protocol to interface as a gamepad.

 

Each usb gamepad is detected as an x/y axis with two buttons.

 

In joystick mode, UDLR (pins 1-4) simulate a digital hat with 8 positions plus off. Fire button is button 1. Button 2 is unused and not accessible.

 

If a high logic signal is detected on the paddle pins, the layout changes and immediately starts polling the paddles.

 

Paddle mode: analog x/y axis with some predefined range of paddle travel marking +/- 100%. Left paddle button (normally R) becomes button 1. Right paddle Fire (normally L) becomes button 2.

 

So any controller that sends a high logic signal to pins 5 and 9 will remap the way the controller ports are encoded. Any joystick mode controller that actuates both L+R of U+D will return something in between. 2600daptor and Stelladaptor used 50% R or D to denote that both buttons were pressed. Any axis encoder for dpad or joystick inputs cannot be at two extremes at once, even though specialty controllers do exactly this by sending something other than raw directional input to the joystick pins, including opposite inputs.

 

If only they mapped the controller port as discrete inputs for pins 1, 2, 3, 4, 5, 6, and 9, then many of these special function controllers would operate. You cannot have a joystick with built in paddles on the Stelladaptor, and the same applies here. It may be possible to update the microcontroller code in the controller ports to alter the way it sends data to the emulator, or it might not. Even if so, it would require a rewrite of much of the emulator code to compensate for those changes.

 

It's unfortunate, but any controller standard added to the controller port must fit into the cookie cutter form factor of an x/y axis plus two buttons. And joysticks only get one button, and no simultaneous opposing directional inputs allowed.

 

When I get my Retron77, I will load the test cart rom and see what happens when I toggle the built in paddles on my joystick on/off. It may allow me to dissect how the controller ports encode the data. I suspect based on prior observations, they clone Stelladaptor/2600daptor v1 behavior.

 

Maybe this was already said, but doesn't this mean that you can start with a Joystick, let it detect and then switch to a Genesis pad? Should work.

[godslabrat]

First Revision:

 

Retron 77.zip

[+Muddyfunster]

 

Maybe this was already said, but doesn't this mean that you can start with a Joystick, let it detect and then switch to a Genesis pad? Should work.

 

I like the colour coding

 

I'll add some more later,

[godslabrat]

 

I like the colour coding

 

I'll add some more later,

Thanks.

 

I know that MetalJesus and others have done tests on YouTube, but I haven't been able to re-watch those videos to note what worked and what didn't. That would be a great way to add to the list if you don't have a Retron 77 yourself.

[+Muddyfunster]

Some more updates

 

Couple of unexpected fails added.

 

Just a note on how I'm testing the carts :

 

I'm finding around 30% of my carts fail first time on the Retron77 but work on a real 2600JR. This tells me the Retron77 is a lot more picky than real hardware.

 

Usually a clean of the contacts sorts things out and I retest.

 

If I still get a fail on the Retron I do another clean and check again.

 

Still a fail, I check whether the cart works on real hardware, if it does I call it a fail and move on.

 

That said, it's possible the Retron just doesn't like my version of what failed and I'm sure in time other folks can check whether they are "false positives".

Retron 77 26th June 2018.zip

Edited June 26, 2018 by Muddyfunster

[+evg2000]

would it make sense to move to google doc, onenote, evernote, git etc, when people start getting their systems people might start overwriting each others updates.

[+Muddyfunster]

would it make sense to move to google doc, onenote, evernote, git etc, when people start getting their systems people might start overwriting each others updates.

 

This should work.

 

https://docs.google.com/spreadsheets/d/1lzJdxHuRL9I-EdJ0f1abb2a9Gv46qLj-gmAj-PSBbN0/edit?usp=sharing

[AtariLeaf]

Some more updates

 

Couple of unexpected fails added.

 

Just a note on how I'm testing the carts :

 

I'm finding around 30% of my carts fail first time on the Retron77 but work on a real 2600JR. This tells me the Retron77 is a lot more picky than real hardware.

 

Usually a clean of the contacts sorts things out and I retest.

 

If I still get a fail on the Retron I do another clean and check again.

 

Still a fail, I check whether the cart works on real hardware, if it does I call it a fail and move on.

 

That said, it's possible the Retron just doesn't like my version of what failed and I'm sure in time other folks can check whether they are "false positives".

 

That's a larger percentage actually. I wonder if a large part of that 30% is Activision carts?

 

The same problem happens on the Retron 5. I tried playing black label Pinball for the NES and no matter how I cleaned it would not recognize even though it works fine on real hardware and a Retro Trio and other NES games work on the Retron 5.

[Thomas Jentzsch]

Couple of unexpected fails added.

Indeed.

• Polaris works and Miner 2049er does not. Both have the same size and use the same bankswitching type.
• Real Sports boxing uses standard Atari bankswitching and other games using the same type work.

You or someone else should try those again. If they still fail, then the heuristics for detecting the correct bankswitching for being able to dump the cart are failing. I suppose this could be improved.

 

Fatal Run is 32k, maybe there is a size limit. And SWAG is the only game with Parker bankswitching on the list, so this could also be a problem bankswitch type detection for dumping.

 

Note: The heuristics work better in Stella, because it has a complete view at the whole ROM, while RetroN77 probably (analysis pending) only sees the initially active bank.

Edited June 27, 2018 by Thomas Jentzsch

[+Muddyfunster]

@AtariLeaf

 

Only Commando on the last batch I tested was an Activision title, planning to do a bunch of Activision carts tonight to see where they land.

 

@Thomas Jentzsch

 

Yup, it could just be my carts as I agree it's odd those listed failed when similar scheme carts passed. Once other folks start to test and add input we'll see if it is a false positive or as you state, an issue with the heuristics on how the roms are detected.

[+Muddyfunster]

 

This should work.

 

https://docs.google.com/spreadsheets/d/1lzJdxHuRL9I-EdJ0f1abb2a9Gv46qLj-gmAj-PSBbN0/edit?usp=sharing

 

Maybe Godslabrat can append the link to the first post in the thread.

[Keatah]

Maybe there are different versions revisions of the failed carts?

Edited June 27, 2018 by Keatah

[+Muddyfunster]

Tested a few more tonight, Parker and Spectravision