Flash ROM Cart

[CosmicBoy]

I`m using the TL866CS ( see Post#53 )

[+MacRorie]

I`m using the TL866CS ( see Post#53 )

 

 

Hmm, yeah, that seems like a nice solution, but I was hoping to not have to buy (yet) another burner. I have three (two are chip specific). I was hoping to use the GQ-4X I have, but it wants to do it via the AVRISP interface. This is fine, I could breadboard the connection, but I need to externally power the chip. I was thinking I could borrow +5V from a joystick port somewhere, but that did not seem to do the trick (admittedly, I did not verify that the volts were coming out . . it was late and I was dispirited because I needed something else to finish the project). I did find an adaptor on ePay for $5 that seems like I can use it to power and wire.

[+arcadeshopper]

 

 

Hmm, yeah, that seems like a nice solution, but I was hoping to not have to buy (yet) another burner. I have three (two are chip specific). I was hoping to use the GQ-4X I have, but it wants to do it via the AVRISP interface. This is fine, I could breadboard the connection, but I need to externally power the chip. I was thinking I could borrow +5V from a joystick port somewhere, but that did not seem to do the trick (admittedly, I did not verify that the volts were coming out . . it was late and I was dispirited because I needed something else to finish the project). I did find an adaptor on ePay for $5 that seems like I can use it to power and wire.

 

no power on the joystick port

[+MacRorie]

 

no power on the joystick port

 

 

There is on the 800XL that was sitting next to it.

[ralphb]

May I ask how you programmed the 8515? Most solutions presume an ISP setup yet there is no header on the PCB.

 

I didn't want to clutter the board any further, as I figured reprogrammig will be an extremely rare thing.

 

Personally I'm using a USB programmer for €20 and a breadboard (the blue wires in the setup are not required for programming).

 

 

BTW, the entire development of the FlashROM was done without a development board or debugging capabilities. Instead, I wrote the code so that compiling with gcc instead of avr-gcc will yield a plain executable file that simulates the menu generation and yields a BIN file that can be tested in MESS.

 

[00WReX]

Hi All,

Sorry if the below is a bit long winded, but possibly may be handy for someone who is having
issues with their FlashROM. And not sure if this is relevant, but mine is a completely standard
non-expanded Australian delivered PAL beige (non QI) TI-99/4A.

Yesterday (Friday 1st July), the final parts for my FlashROM kit arrived...the ATmega 8515's.
Everything else had already arrived including my small order of PCB's.
As an extra bonus, I had the day off work so had the day to build a FlashROM and play.
So an hour after the 8515's arrived in the post I had built a board.

I set up my one and only working (completely standard) TI and with the FlashROM plugged in,
the FlashROM menu came up and all appeared good.
But, I could not get a single file to load (from the 'approved' file list). http://ftp.whtech.com/Cartridges/joined/

I gave everything a double check and played around a bit but nothing would load. Just continual
computer lockups. I had ordered multiple 8515's & SRAM's so I flipped those out and tried new ones.
The result was the same. So I began to wonder if the SD reader was the issue.

At this point I thought the easiest option was to build another board.
Once this board was built, more testing and with this board I could actually get some files to
load and run, but it was very erratic, and still many crash's and lockups and inconsistent results.

I then decided to clean the cartridge connector contacts (although I have around 20 standard
cartridges that all work just fine). I actually got a lot of black off the contacts and after the
clean, both boards started working a bit better, but still extremely erratic with different stages
of failure and complete crash's of the computer.

A file would load one minute and crash the next time I tried to load the same file.

 

The 'banktest.bin' test program was also all over the place as well.

Sometimes it would crash, others it would only report BANK 1 OK BANK 2 OK (no mention of 3 & 4).

Then another time it would be fine, with banks 1 to 4 OK. There were also other results.

Come Saturday, this was all still playing on my mind.
I was suspecting the cartridge contacts for a long time and did heaps of cleaning.
But, after a while I noticed that even if I did not touch the board, I could switch the computer
off and on and get different results. And some games that I was convinced should work, simply would
not. I then started to suspect this appeared to be something like a timing type of issue.

I tried different SD cards...same result.

I have visited the GITHUB site many times in the past, I decided (as a last resort) to re-read
everything, just in-case I had missed something...although I was convinced I had not.
Interesting, was that I pretty quickly noticed there had been an update (19 hours ago when I checked)
to the 'diskfix' mod, the 1nf capacitor that was involved in the mod was now to be omitted totally.
I had included the original mod and performed the 'non tracking cutting' mod to my Rev 1 boards.

Now I know the mod indicated it was a 'diskfix' mod, and I do not have a single thing connected to
my completely standard TI, but I had nothing to lose and removed the 1nf capacitor anyway.

WooHoo, with the 1nf cap removed (on both boards) they immediately behaved as expected and just
seemed more responsive and consistent. And files that simply would not load previously no matter
what, were now loading...e.g. Dig-Dug, Pole Position, just to name a couple.

Ahhh, what a relief that Ralph had updated the fix to simply remove the 1nf cap.
I would have been pulling my hair out all weekend if I had not read that.

 

Cheers,

Shane

Edited July 2, 2016 by 00WReX

[00WReX]

Here are my boards with the 1nf caps removed...

 

 

[ralphb]

I'm so sorry, Shane, about this confusion ... I can feel your pain.

 

Yeah, behind the scenes we're still discussing about that capacitor: where should it be connected, what value should it have, do you even need it?

 

My personal opinion right now is that it should simply be removed. Several people have confirmed that this works. Also note that the original design with the wrongly placed capacitor made the board work, but messed with the disk -- if you rewire it properly then it's just the other way around.

 

But still, removing the cap is against years of experience with the multi-carts, so I was hesitant to suggest it officially just yet.

 

I'll have another go at the problem this weekend. (Would've liked to do this earlier, but I lost two weeks due to a hospital stay.)

 

Anyway, thanks for your report, and have fun with your carts now!

[00WReX]

Hi Ralph,

Thank you for the reply, but you have absolutely nothing to be sorry about.

I fully understand these are hobby projects and a lot of time and effort goes into them.

I think it is fantastic that you have provided everything and made it all freely available.

Not to mention ongoing and continued support.

 

My only real frustration is that I am a complete TI-99/4A noob and have no other

working machine for any sort of comparison. I was also completely unsure if files

should actually be working or not (e.g requiring 32k).

But there were those that I was confident should be working.
I know some of the more recently converted games have been flagged as requiring 32k

But was unsure of the ones in the 'ftp.whtech' folder.

 

Anyway, thanks again. This project is much appreciated.

 

 

Cheers,

Shane

[Omega-TI]

Just an observation here, naming the directory...

 

"Cartridges/joined"

 

... makes it easy for someone to overlook and miss entirely.

 

I would give this directory the same consideration that was given to the F18A, a directory with it's own name. I'd call it FR99, which would show up fairly close to F18A and be harder to miss, especially for newbies.

[+OLD CS1]

For the directory, I would go with "FlashROM99" as newbs may not recognize FR99 or mistake it for a firearm Still puts it near F18A, too.

[ralphb]

Yeah, I heard those complaints before ... A name like that would suggest that those images are only good for the FlashROM, when, in fact, they are unmodified dumps that simply have been joined.

 

But I get your points, so sure, go ahead. I don't have access, though.

[+OLD CS1]

Yeah, I heard those complaints before ... A name like that would suggest that those images are only good for the FlashROM, when, in fact, they are unmodified dumps that simply have been joined.

 

But I get your points, so sure, go ahead. I don't have access, though.

 

For that matter, if the repository is named "FlashROM99", I see no reason not to link a joined ROMs directory, or even vice-versa.

[+remowilliams]

My personal opinion right now is that it should simply be removed. Several people have confirmed that this works. Also note that the original design with the wrongly placed capacitor made the board work, but messed with the disk -- if you rewire it properly then it's just the other way around.

 

When I built mine I installed the cap to the 'alternate' position. It's been working fine like that since day one

[+Schmitzi]

Why not call the directory "Non-Inverted" ?

[+mizapf]

 

Just an observation here, naming the directory...

 

"Cartridges/joined"

 

... makes it easy for someone to overlook and miss entirely.

 

I would give this directory the same consideration that was given to the F18A, a directory with it's own name. I'd call it FR99, which would show up fairly close to F18A and be harder to miss, especially for newbies.

 

 

Concerning the "joined", I second this; but please do not move too many things into the root directory. Over the years, people created directories at various locations on the server, which turned WHTech into this messy heap.

 

I suggest to name it Cartridges/FlashROM99.

[Omega-TI]

 

I suggest to name it Cartridges/FlashROM99.

 

That would work!

[Omega-TI]

Here are my boards with the 1nf caps removed...

 

 

 

 

I'm curious, what kind of camera did you use? That is an excellent quality photo.

[Omega-TI]

You know, this FlashROM 99 is quite the thing... I've not seen as many newbies & lurkers come out of the woodwork for quite a while!

 

Cheezy grin!

[+OLD CS1]

 

I'm curious, what kind of camera did you use? That is an excellent quality photo.

 

Camera Make = samsung

Camera Model = SM-G925I

 

Galaxy S6. Neat.

[LASooner]

Wait, so I only need to desolder that capacitor? It's so rare when procrastination pays off.

[Omega-TI]

 

Camera Make = samsung

Camera Model = SM-G925I

 

Galaxy S6. Neat.

 

That was a cellphone photo? Damn, not bad. Was there any post processing involved?

[+Lee Stewart]

 

That was a cellphone photo? Damn, not bad. Was there any post processing involved?

 

I believe the S6 has the same camera as my S5 and the camera is, indeed, impressive at 11.3 megapixels. The only drawback for my S5 is the zoom is only 4x.

 

...lee

[00WReX]

Yes, as OLD CS1 found, it is a Samsung Galaxy S6 phone.

 

Nothing done to the photo other than resizing it down to make the file size smaller...and in fact it lost quality on the resize and save (I chose to save at 92% quality) with the program I used to make the file size smaller. The original resolution is 5312 x 2988 and file size of 4.41mb...so taking that down to a file size of ~857kb...

 

I am very impressed with the photo quality / camera on this phone. I have taken some impressive photo's with it considering it is a multi function device.

 

remowilliams wrote:

 

When I built mine I installed the cap to the 'alternate' position. It's been working fine like that since day one

 

Yes, I also did this at first, an this is why I was wondering what the problem could be.
Nobody else had or has reported a similar issue to mine.

And the fix itself is mentioned to fix disk drive issues, which I do not have.

 

My one suspicion, and it may be totally wrong, is that I suspect it has something to do with the 377

chip I am using (in conjunction with that capacitor).

 

I am using a CD74HCT377E as opposed to the more usual SN74HCT377N.

Initially you would think this is simply a different manufacturer part, but I have found that TI make

both, and they both have a different data sheet.

 

And even the heading on the datasheet is slightly different.

 

CD74HCT377 = Octal D-Type Flip-Flop With Data Enable
SN74HCT377 = Octal D-Type Flip-Flop With Clock Enable

 

Now I have not studied both datasheets, but a quick look over them and some specs appear very slightly different (I think).

 

Anyway, I may be totally wrong, but I do believe all my other components that I sourced were the same as those I have seen on other people's boards.

The 541's are the TI SN74HCT541N variety.

 

Anyway, my two boards appear to be working great with the 1nf removed. So an easy fix.

I am having some issues with SD card's but I know that can be a bit hit and miss.

Of my three spare cards, only one is working.

I have a Kodak 128mb that does not work.

A Sandisk 2gb Micro SD in an adapter that works perfectly.

A Sandisk 8gb Micro SD in an adapter that does not work.

 

Cheers,

Shane

Edited July 3, 2016 by 00WReX

[+InsaneMultitasker]

 

 

 

Concerning the "joined", I second this; but please do not move too many things into the root directory. Over the years, people created directories at various locations on the server, which turned WHTech into this messy heap.

 

I suggest to name it Cartridges/FlashROM99.

 

I've been catching up on some of the more recent threads, so this may be a silly question, and not necessarily directed at you:

 

What is the reason for yet another format? I recently looked for a cartridge dump on WHT's FTP site, and instead of finding one version there are PC99 images, MESS images, and other sorts of files formats all over the place. In addition to WHT, there are now multiple threads and messages here on atariage with images and files, which are not easy to download if you weren't following these threads originally.