DIYing a plug that fits the PlayStation's parallel port to boot from a ROM

[prochazkaml]

That's certainly a lot faster. With the small amount of development I've done I have always just built an ISO and loaded it up in an emulator for quick tests and then later used PS-IO for real hardware tests. I had thought about the serial cable option but I like being able to load data from CD. But it all depends on what sort of development you're doing.

If you're not streaming any data from the CD (CDDA, XA, STR), then you could theoretically write a custom library for universally loading files: if the game is running from a CD, then it would load the requested data from the CD, but if the game is running from memory loaded via serial, then it would talk over serial with a program running on the PC which would send the requested files back to the console. Since it is proven that the PlayStation's serial link can communicate at 1 Mbps (so 100 kB/s – 1 start bit, 8 data bits and 1 stop bit), it wouldn't be that much slower than reading from the CD (the PlayStation's CD drive can run at max. 2x speed, so 300 kB/s max). The method of communication could be #defined somewhere in a header file.

Personally, I haven't done this yet, but I'm looking into it. For now, my projects have always been able to fit into the ~1.7 megabytes of free memory entirely (including all assets), so it wasn't a huge problem for me, but I worry that this might not continue for much longer before I start expanding my projects and I eventually run out of memory.

If I however do stream data from the CD, there is no other option then to just burn the CD (at least for me, I sadly don't own a PS-IO yet), and even that is not easy, the PlayStation is quite picky about the media/burning method (for example, in my instance CDDA sometimes gets corrupted into an earraping loud white noise on the PlayStation, while the same disc works fine in other CD players). So yeah, it depends on what you're doing.

Anyway, an update to my project!

I haven't fixed any of the broken the pins yet (I have reassembled the console back together), as I am currently tackling another problem: finding a working ROM and burning data to it. Now, as luck would have it, I found at home an old SST29EE020 (in the PLCC package), which is even listed in this site's store (in the DIP package, but internally the same), so it has to be compatible at least in some ways (I mainly worry about the timings, the voltages can be taken care of with a simple level shifter, since the PlayStation runs at 3.5 V, meanwhile the EEPROM requires 5 V).

However, I do not have a parallel programmer to actually program the chip (such as the popular and a bit expensive TL866). What I do have, though, is an Arduino Mega, which has way more than enough digital pins for hooking up this EEPROM. I made a simple circuit on a piece of perfboard connecting the EEPROM (in a through-hole socket) to the Arduino:





Please excuse the messy solder job, but it works and there are no shorts

Of course, the Arduino would be useless without any firmware, so I found this project which turns an Arduino into a parallel EEPROM programmer. It consists of the required Arduino firmware (which I had to modify as this project was intended for smaller microcontrollers with not enough pins) as well as the host program for communicating with the firmware over serial. It also only supported EEPROMs with the page size of 64 bytes, so I had to modify it to support my SST29EE020, which uses 128 bytes long pages. And the original code didn't support memories with over 32 kB of storage, so I had to patch that as well. I might post the modified code sometime later in case anybody finds it useful.

To test it out, I got myself ~200 kB of Lorem Ipsum, and after hours of debugging and modifying the code, I finally got this:



It works! It bloody works!

So I finally wrote sioload (which is only 19 kB) onto the EEPROM, which means that I'm basically done with this part of the project.

I still have to finish the following:

• Fix the 4 broken pins and don't make them bend the connector inside the console
• Hook up the connector to this EEPROM breakout board I made
• Power it up

But that's for another day. See you!

[prochazkaml]

Alright, so I forked the project and applied my modifications to it. Here it is.

[MottZilla]

The serial data link is a nice idea. I think that PSIO supports a USB link though that could achieve something similar. I've not tried it out but it is an option for me.

I like the improvised EPROM programmer. I have some older Willem based ones that operate over the parallel port and require jumper settings. I'd really like to get a newer USB one someday.

Looking forward to seeing you put the connector and EPROM together. Then maybe the next step would be to design a PCB to go from the connector pins to a EPROM socket.

[prochazkaml]

The serial data link is a nice idea. I think that PSIO supports a USB link though that could achieve something similar. I've not tried it out but it is an option for me.

That might be possible as well. Can't verify it, though.

I like the improvised EPROM programmer. I have some older Willem based ones that operate over the parallel port and require jumper settings. I'd really like to get a newer USB one someday.

Thanks!

But yeah, I should eventually get one as well for future projects.

Looking forward to seeing you put the connector and EPROM together. Then maybe the next step would be to design a PCB to go from the connector pins to a EPROM socket.

It's coming closer, I promise. I'll take my PlayStation back apart next week and continue working on the connector and getting it finally working.

In the meantime, I found some Action Replay schematic to see how the ROM is hooked up to the PlayStation's parallel port, and what struck me as being really strange is the fact that the entire cart runs from the PlayStation's 8 V rail, which is then regulated to 5 V inside the cart, which powers the EEPROM (same chip as mine, requires 5 V to operate). That EEPROM then sends data (at 5 V!) back to the PlayStation, which runs at only 3.5 V! I know that the PlayStation has some resistors inside (as I have already mentioned in a previous post), but it is still strange to not use a level shifter for this sort of thing. On the other hand, if it is proven to actually work, I should probably do it the same way.

So, I designed a PCB around it! Yay!



Note: the voltage regulator as well as the electrolytic capacitors are supposed to be bent in a way that makes them lie flat against the PCB. And I made sure to have enough clearance around the EEPROM to fit a ZIF socket in there.



Made in EasyEDA, since that's the only thing I can sort of work with. And it even let me import my 3D model for the connector! Here is the project on OSHWLab in case anyone wants to explore the schematic or the board itself.

Of course, I won't order one until I actually get the EEPROM to boot with my connector. But it's still nice to look into the future, how the project might turn out

[MottZilla]

You might want to design a version that does do proper voltage level translation. Often things do work without shifting between 5v and 3.3v levels but it could be causing stress on internal components and if it's in a custom IC that's going to suck if it results in premature failure.

The 3D model looks cool. I suppose a 3D printed casing would follow. Then you've got your whole home made rom cart.

[prochazkaml]

You might want to design a version that does do proper voltage level translation. Often things do work without shifting between 5v and 3.3v levels but it could be causing stress on internal components and if it's in a custom IC that's going to suck if it results in premature failure.

That's what worries me as well, and I am shocked that ROM carts back in the day did this.

I had a quick thought, and wouldn't a simple voltage divider get the 5 V coming out of the EEPROM down to ~3.5 V quite easily? E.g. 10k going to the output pin, 22k to GND and the middle to the PlayStation? That would yield ~3.4 V, which should work, right?

The 3D model looks cool. I suppose a 3D printed casing would follow. Then you've got your whole home made rom cart.

Thanks! And you bet there's a case coming, but that's not high priority at the moment, as I am still trying to figure the actual circuit out.

[MottZilla]

I'm not shocked that an unlicensed product would do such a thing. Especially the cloned versions.

The voltage divider might work but then you'd need a lot of them as you'd need one for each signal. Might make more sense to use a dedicated IC. But whatever works.

[prochazkaml]

I'm not shocked that an unlicensed product would do such a thing. Especially the cloned versions.

Eh, fair enough. Still a crappy practice though.

The voltage divider might work but then you'd need a lot of them as you'd need one for each signal. Might make more sense to use a dedicated IC. But whatever works.

I know, but for the time being (before I get a level shifter IC), the resistors should work fine.

[prochazkaml]

Or I can just use a 3.5 V-compatible EEPROM (or rather Flash) to begin with. The SST39VFXXX series looks like a good candidate, since it's quite cheap, it appears to have less latency than the SST29 series and most importantly, it's currently in stock available to buy immediately from all major distributors (Mouser, DigiKey, Farnell – they offer really low shipping rates to the Czech Republic (around 7 €), so it has been the most convenient one for me in the past). Just look around the internet for "SST39VF" (the "VF" is quite important here, as it means that the chip is rated to run at up to 3.6 V), you can find them very easily.

The highest capacity one from the series (512 kB) goes for just 2 € including VAT. So if I get my EEPROM working, I'll definitely buy a few of these and I'll redesign the PCB to fit a PLCC socket for these chips instead, removing the need for the on-board voltage regulator.

(edit: I accidentally called the series "SST39VT" instead of the actual "SST39VF", yikes)

[prochazkaml]

Alright, so I've modified the PCB by removing the now unnecessary power regulator, and I've also added a header for controlling the chip enable line coming out of the PlayStation's CPU with a simple switch, in case you just want to disable the ROM without removing the cart from the console, since the connector could be quite fragile.





You even get a free PLCC to DIP adapter included, you just break it off!

Not sure whether places like JLCPCB or PCBWay can manufacture PCBs with so many holes right next to each other though. We'll see in the future, when I order some.

Oh, and I've just taken my console apart again, so I'll try to fix the non-working pins now (there's 5 of them, I missed one previously).

[masterg0r0]

That sounds really cool

[prochazkaml]

That sounds really cool

Thanks.

So, I fixed the 5 broken pins. But guess what? Another 5 pins (which worked completely fine before) started to act up! So I even fixed those, but some still remained broken and some other unrelated pin all of a sudden started bending its matching pin inside the console's connector! Words cannot describe my frustration that I am experiencing right now.

Anyway, what do I do now? I need the connector to be at least somewhat stable so that the PlayStation can execute code from the ROM, but clearly this won't do.

For the time being, I guess I'll just try soldering the ROM directly to the motherboard's pins. If that works, then I have multiple paths to take:

1. Continue trying to get this stupid connector to work – painful
2. Buy some legit connectors (from Cybdyn or DigiKey) – expensive
3. Desolder the PlayStation's 68 pin connector and replace it with a readily available 2x20 pin pin header with 2.54 mm (0.1") pitch – actually might not be terrible

I'll just leave it at that. If I decide in the future to axe the project, remember that all of the project files (3D models, videos, PCB schematic & layout) are available in this thread in case anyone is suicidal enough to attempt to do this at home.

Sorry to everyone who was looking forward to a fully functional homebrew PlayStation Parallel I/O connector, but as I've already said, there is a reason why nobody has ever dared to do this.

[MottZilla]

Weren't you using old capacitor or resistor leads as pins though? Using something more appropriate would solve that issue.

[prochazkaml]

Weren't you using old capacitor or resistor leads as pins though? Using something more appropriate would solve that issue.

I know, but I do not have anything more appropriate lying around (at least I don't think I do).

[MottZilla]

That may be so, but I'm thinking that with proper pins the 3d printed connector actually can work well. The construction of the pins is very important and using the wrong sort of material is going to cause the issues you're having. But if you search for something more suitable you may find an affordable and easy to purchase solution.

So I wouldn't give up on the idea so soon. Honestly using those component leads as pins for the connector is probably a terrible choice, but it was good enough for a very rough prototype. You need something else that is meant as a pin connector for anything beyond a very basic proof of concept. You could look at various connector pins, maybe some crimp type pins that you put into a housing might be adaptable to a 3D print.

One of the key goals being you want something that is very unlikely to damage the connector on the console. I feel pretty confident a solution can be found to make it work.

[prochazkaml]

Yeah, you're right. I just needed to vent a bit, I realize now that it would be quite a shame to throw all of this work away. Sorry about that.

And I agree with the choice of the connector pins, old component legs were quite a terrible idea indeed.

I've looked a bit into real connector designs, and the pins for this sort of connector are usually made of brass. It's way more flexible than the component legs, which were quite stiff, in the worst case bending the pins inside the console. But for my defense, I really don't have anything better lying around at home, so this was the first thing that jumped to mind.

However, sourcing such brass pins is a whole other problem. How would someone even search for such a thing? "Brass connector pins" yielded me with expensive 2.54 mm pitch pogo pins, and "thin brass wire" (in case I would have to make them myself) got me results with 1 mm thickness or higher, which is way too thick for our purpose.

[MottZilla]

Maybe some other standardish connector has similar pins you could use. I don't have anything in mind but I'm sure something could work.

[prochazkaml]

Yes, there is a standard connector that has the correct pins: Mini SCSI (more specifically, MiniD68). In fact, it has the same shape and the same number of pins as an actual PlayStation PIO connector and it is (almost) the exact same outside plug, you just have to cut away a small piece of it to fit it inside the PlayStation. But other than that, it is fully compatible.

So yes, such connector does exist, but that defeats the whole purpose of this project. I wanted to see if I can make a semi-reliable connector out of stuff I already had at home. Turns out I can't.

I might just end up buying a few of those instead. They're in stock on DigiKey, but the stock is drying up.

[MottZilla]

I made some progress on something that I think would go great with this idea. I've been able to make an expansion ROM that unlocks the drive, patches the Shell to boot unlicensed discs, and then runs the Shell giving you a stealthy way to run unlicensed and import discs. The advantage over FreePSXBoot being that it's faster and still looks stock. Ofcourse you could achieve the same this with a modchip but this is plug n play.

I haven't been able to test it extensively yet but it's something that I always thought would be a cool idea.

Update: I've got it close to finished now. It just needs some testing but hopefully I'll be able to release it soon.

[prochazkaml]

Hey, that's really neat!

If I wanted to do something like that (which I might, don't get me wrong), I would skip all of that patching work and I would just get rid of the stock shell altogether (as IMHO the colour paint splash-aesthetic looks like garbage anyway) and add some extra features - I was thinking of XA audio/STR video playback, loading .EXEs via the serial port/a multisession CD, and do not forget to mention booting unlicensed discs. And I would replace the UI with something less vomit-inducing, of course.

But, as with most projects I work on, I eventually gave up with my connector work and moved onto other ones (recently, I was trying to get PS Move controllers to work with the PSL1GHT SDK for the PS3, but that's another topic). However, I'd still love to revisit it - I'm currently thinking of just buying some Mini SCSI connectors and hooking up the ROM I flashed with SIOLOAD some time ago to the console to see if it actually all works. If it does, then I'll maybe think manufacturing my own connectors that are acceptably reliable. I still cannot find raw pins which are used in connector manufacturing, though. :/

(perhaps I could find some old broken motherboard, rip out its PCI connector and reuse its pins, as they are made in the exact shape I need, since it's just a card-edge connector?)

Till then, I'll patiently wait for your project's release.