FPGA Based Videogame System

[atmn]

The Analogue version of 16-bit era should be Analogue Nt Mega (as in Mega Drive)

Mega Drive, SNES, TG16. That would be a console to aim at.

 

 

 

I have a question for the Nt mini. Is it possible to remap controller ports. I know port 1 are hooked up with power for power on and such.

But Controller 2,3 and 4.

 

Would be nice if you could change place on controller 2 and 4, just to avoid interference with the 8bitdo recievers. I know a extention cable would fix this but they have such a bad fit I'd rather not use them in the Mini.

[Andromeda Stardust]

The Analogue version of 16-bit era should be Analogue Nt Mega (as in Mega Drive)

Mega Drive, SNES, TG16. That would be a console to aim at.

Super Mega Turbo NT. Nothing less will satisfy...

 

I have a question for the Nt mini. Is it possible to remap controller ports. I know port 1 are hooked up with power for power on and such.

But Controller 2,3 and 4.

 

Would be nice if you could change place on controller 2 and 4, just to avoid interference with the 8bitdo recievers. I know a extention cable would fix this but they have such a bad fit I'd rather not use them in the Mini.

Some games read D1 inputs in addition to D0 so set the Four Score to Famicom mode and try them out with player 1 and 4. Some games might work, some not.

[phoenixdownita]

NEO Super Mega Turbo NT. Nothing less will satisfy...

 

Well you forgot the elephant in the room.

 

EDIT: when I try to picture a "Super Mega Turbo" I can only think at something requiring a Big Ass Fan

[hexcode99]

Oh, how insanely awesome that would be, if Analogue were working on a FPGA SNES. The $500 price tags are well worth it for products this fine.

 

Is there any confirmation from Analogue on this? Insta buy from me too!

[Wolf_]

Is there any confirmation from Analogue on this? Insta buy from me too!

No confirmation or everyone would be going insane, but if this is what he's doing I'm in immediately as well. I doubt it would come out with support for more than the snes but short of Kevtris dying (out of his control) I have complete faith in him and the quality he provides so I would absolutely buy it immediately as well. I don't think it's possible to have it out so soon but if Analogue got the hardware done I'm positive Kevtris is the man for the firmware.

Edited July 1, 2017 by Wolf_

[Pixelboy]

No confirmation or everyone would be going insane, but if this is what he's doing I'm in immediately as well. I doubt it would come out with support for more than the snes but short of Kevtris dying (out of his control) I have complete faith in him and the quality he provides so I would absolutely buy it immediately as well. I don't think it's possible to have it out so soon but if Analogue got the hardware done I'm positive Kevtris is the man for the firmware.

I should remind you and everyone else that Kevtris is currently busy with "real life" stuff (i.e. his job) that has nothing to do with video games. Also, Kev's project for implementing FPGA cores for Super-NES, Genesis, etc. has a name, and it's the Zimba 3000. Time will tell if Kev will work with Analogue to market the Zimba, but at this time, there's no clear indication that this is going to happen. All we know is that Kev won't be in any position to do any serious work on any of his FPGA cores before well into next year.

[Wolf_]

I should remind you and everyone else that Kevtris is currently busy with "real life" stuff (i.e. his job) that has nothing to do with video games. Also, Kev's project for implementing FPGA cores for Super-NES, Genesis, etc. has a name, and it's the Zimba 3000. Time will tell if Kev will work with Analogue to market the Zimba, but at this time, there's no clear indication that this is going to happen. All we know is that Kev won't be in any position to do any serious work on any of his FPGA cores before well into next year.

What he said was that he was working 2 jobs and that a new core for the nt mini wouldn't be coming out for at least a year. That second paid job could very likely be something for Analogue. Unless you think a Doctor Stein has contracted his services in cryogenics to help him with a project?

[Pixelboy]

What he said was that he was working 2 jobs and that a new core for the nt mini wouldn't be coming out for at least a year. That second paid job could very likely be something for Analogue. Unless you think a Doctor Stein has contracted his services in cryogenics to help him with a project?

I don't recall anything about Kevtris working two jobs, just that he had a job that was going to keep him very busy for at least a full year. Perhaps I missed something?

[Wolf_]

I don't recall anything about Kevtris working two jobs, just that he had a job that was going to keep him very busy for at least a full year. Perhaps I missed something?

Iirc he said he had taken on some paid work that meant he would be working 14 hours a day. I assume by paid work he meant in addition to the work he was already doing because otherwise he would have phrased it differently, like "My job is really busy". So assuming that paid work is based on his fpga coding (I don't think Cryogenics is a business where you can moonlight at places) then the only question becomes: What fpga is he coding? The snes sounds like a strong bet for reasons previously mentioned.

[cacophony]

I don't recall anything about Kevtris working two jobs, just that he had a job that was going to keep him very busy for at least a full year. Perhaps I missed something?

 

Yes, you missed something. He's had a regular job for a long time. He did all the Nt mini work for Analogue on the side. A couple months ago he started something different on the side that he can't reveal.

[Pixelboy]

Ah, well then, I should say I stand corrected. Let's hope he's working on something that we retro-gamers can sink our teeth into.

[Andromeda Stardust]

No confirmation or everyone would be going insane, but if this is what he's doing I'm in immediately as well. I doubt it would come out with support for more than the snes but short of Kevtris dying (out of his control) I have complete faith in him and the quality he provides so I would absolutely buy it immediately as well. I don't think it's possible to have it out so soon but if Analogue got the hardware done I'm positive Kevtris is the man for the firmware.

 

I should remind you and everyone else that Kevtris is currently busy with "real life" stuff (i.e. his job) that has nothing to do with video games. Also, Kev's project for implementing FPGA cores for Super-NES, Genesis, etc. has a name, and it's the Zimba 3000. Time will tell if Kev will work with Analogue to market the Zimba, but at this time, there's no clear indication that this is going to happen. All we know is that Kev won't be in any position to do any serious work on any of his FPGA cores before well into next year.

 

Iirc he said he had taken on some paid work that meant he would be working 14 hours a day. I assume by paid work he meant in addition to the work he was already doing because otherwise he would have phrased it differently, like "My job is really busy". So assuming that paid work is based on his fpga coding (I don't think Cryogenics is a business where you can moonlight at places) then the only question becomes: What fpga is he coding? The snes sounds like a strong bet for reasons previously mentioned.

But wouldn't it be amazing if it turned out the contract work he's been doing all along was for an FPGA'd SNES? If a man's gonna dream, dream big or go home!

[cacophony]

I'm curious if Kevtris (or others) had any thoughts on this:

https://db-electronics.ca/2017/07/05/the-dangers-of-3-3v-flash-in-retro-consoles/

[kevtris]

I'm curious if Kevtris (or others) had any thoughts on this:

https://db-electronics.ca/2017/07/05/the-dangers-of-3-3v-flash-in-retro-consoles/

 

I agree with some of it but not all of it. I remember seeing that awhile back and was kind of horrified that they did it and that it "worked". Feeding 5V logic into a 3.3V only part that isn't 5V tolerant is pretty bad. I don't agree with the amount of current that will be sent through the flash chip pins, it definitely not is going to be as high as he says.

 

But it isn't a great idea to do it anyways. I highly doubt it will damage the system, but it can definitely damage the flash. One thing he doesn't really go over with the protection diode stuff is the 3.3V regulator can only supply voltage, and cannot sink it. So if the diodes forward bias, they will tend to pull the 3.3V *up* higher than 3.3V. He modeled it with a voltage supply/sink that will "pull down" anything higher than 3.3V and keep it there. This is not what would happen in real life. Regardless, if your 3.3V is getting pulled up near 5V, it is going to be bad news for the chips.

 

So yeah; it's a really bad idea to run your 3.3V flash with 5V IOs, but it's a lot more nuanced than is made out in that article. That's all IMO.

[Keatah]

Yes all that.

 

It's never a good idea to push components beyond their specification. Especially 30+ year old videogame consoles. They should be treated with TLC. And while I was surprised to see how sloppy some of those designs are, it isn't wholly unexpected. There's engineer wannabees "designing" this stuff. The same goes for audio/video mods, accelerators, firmware boards, pc<->classic interfaces.. and overclocking.. and..

Edited July 13, 2017 by Keatah

[Wolf_]

I'm curious if Kevtris (or others) had any thoughts on this:

https://db-electronics.ca/2017/07/05/the-dangers-of-3-3v-flash-in-retro-consoles/

I was really saddened by Krikzz response. If you are out of spec you are out of spec. It isn't just a rumor if your regular input exceeds the TEMPORARY maximum voltage rating. This is strictly science and not something you can dispute.

 

That being said his point that old systems are incredibly resilient (namely anything Nintendo) is valid. While being out of spec is "damaging" these carts & systems I would estimate it at almost certainly being along the lines of an almost imperceptible difference in lifespan.

 

Edit: In the case of Krikzz carts anyways, some of that other stuff is super sketchy.

Edited July 13, 2017 by Wolf_

[phoenixdownita]

 

I agree with some of it but not all of it. I remember seeing that awhile back and was kind of horrified that they did it and that it "worked". Feeding 5V logic into a 3.3V only part that isn't 5V tolerant is pretty bad. I don't agree with the amount of current that will be sent through the flash chip pins, it definitely not is going to be as high as he says.

 

But it isn't a great idea to do it anyways. I highly doubt it will damage the system, but it can definitely damage the flash. One thing he doesn't really go over with the protection diode stuff is the 3.3V regulator can only supply voltage, and cannot sink it. So if the diodes forward bias, they will tend to pull the 3.3V *up* higher than 3.3V. He modeled it with a voltage supply/sink that will "pull down" anything higher than 3.3V and keep it there. This is not what would happen in real life. Regardless, if your 3.3V is getting pulled up near 5V, it is going to be bad news for the chips.

 

So yeah; it's a really bad idea to run your 3.3V flash with 5V IOs, but it's a lot more nuanced than is made out in that article. That's all IMO.

He measured something here:

https://db-electronics.ca/2017/07/13/ac-analysis-of-clamping-diode-effect-in-5v-to-3-3v-mismatch/

 

I'm not quite sure when he shows 3.9 at the flash end of the resistor what is the 3V3 voltage reg doing?

[Andromeda Stardust]

He measured something here:

https://db-electronics.ca/2017/07/13/ac-analysis-of-clamping-diode-effect-in-5v-to-3-3v-mismatch/

 

I'm not quite sure when he shows 3.9 at the flash end of the resistor what is the 3V3 voltage reg doing?

The internal protection diode in the flash chip dumps the remaining .6V to flash VCC. 3.9V - .6 = 3.3V so the actual VCC voltage is not exceeded. That said, it is bad design to rely on internal protection diodes in this manner, but it does not break anything. Increase in current consumption based on the entire console is 30-35mA higher than a standard cartridge, and the vast majority of that difference is in the current draw of the CLPD device.

 

A far cry from the 12.5mA per pin the db claims is being dissipated in the article. These measurements might be theoretical 1st approximation values assuming the console logic had an output impedance of zero ohms at 5V high logic, which is far from reality. Any stress on the actual hardware is likely negligible. The high logic signal is probably closer to 4.5V in real usage, and pulling it down by fractions of a volt due to current draw isn't going to burn out the logic chips.

[Keatah]

Don't forget duty-cycle and frequency. That have have a huge effect on how much gets dissipated.

[philyso]

I was really saddened by Krikzz response. If you are out of spec you are out of spec. It isn't just a rumor if your regular input exceeds the TEMPORARY maximum voltage rating. This is strictly science and not something you can dispute.

Except that, if you read the article carefully, Krikzz flashcarts theoretical voltage input (3,75V) was actually BELOW maximum voltage rating specified in flash datasheet (which is 3,9V and might go up to 5,3V for 20 ns during low to high transition). And this is with a theoretical 5V output on cartridge pins reduced to 3,75V after the serie resistor, but real measurements show that MD ouput voltage closer to 4,7V on these pins so input voltage will be closer to 3.6 V (with maybe a short peak up to 3,9-4V when transition occurs but that's sill within spec limits)

 

The rumor that was initiated by this article was also more that the console parts could be damaged and your console was going to slowly die because of excessive current flow and power dissipation , which has been proven wrong by real measurements. That was probably not db-elec's intention to freak people out but the autoritive 'avoid' verdicts, the poor cancer analogy and the overdramatic picture that introduced his article surely did not help.

Edited July 14, 2017 by philyso

[PateraSilk]

Excuse the noobish question, but how do you play .fds files? They're not recognized when I stick them in the NES folder.

[Tusecsy]

Except that, if you read the article carefully, Krikzz flashcarts theoretical voltage input (3,75V) was actually BELOW maximum voltage rating specified in flash datasheet (which is 3,9V and might go up to 5,3V for 20 ns during low to high transition). And this is with a theoretical 5V output on cartridge pins reduced to 3,75V after the serie resistor, but real measurements show that MD ouput voltage closer to 4,7V on these pins so input voltage will be closer to 3.6 V (with maybe a short peak up to 3,9-4V when transition occurs but that's sill within spec limits)

 

The rumor that was initiated by this article was also more that the console parts could be damaged and your console was going to slowly die because of excessive current flow and power dissipation , which has been proven wrong by real measurements. That was probably not db-elec's intention to freak people out but the autoritive 'avoid' verdicts, the poor cancer analogy and the overdramatic picture that introduced his article surely did not help.

I think rene has some kind of grudge with krikzz. That article comes off as a giant smear piece. You can't say something causes cancer and to avoid products because of "missing ground leads", which aren't missing, and still maintain your credibility.

 

Which is sad because the 5v to 3.3v problem is a problem that needs to be addressed. I sold my mega X7 and will either grab an X5 or wait for the revision.

 

If only the Super NT Mini were out I could just sell them all.

[NoKnownOrigin]

Excuse the noobish question, but how do you play .fds files? They're not recognized when I stick them in the NES folder.

 

 

FDS is not supported directly (yet?). Both the PowerPak and the Everdrive N8 have FDS support, but I would recommend getting the actual FDS RAM Adaptor and either the FDSStick or Magic Wild Card to load the disk images. Getting the real thing is going to be a lot cheaper, plus you get more authentic sound. Only reason to get one of the flash carts would be if you also want to play FDS games on an actual NES.

[Andromeda Stardust]

Somebody please pass this schematic on to Krikzz. I may post it on Krikzz forums later. Inline 100 ohm resistors works but it is a kludge.

 

 

 

A proper voltage divider setup places .5R inline between the 5V bus and the 3.3V bus, then places R between the input of the 3.3V bus and ground. In my example, I used a 2.4kohm for .5R and a 4.7kohm for R. This circuit also works for bidirectional data bus. In the event the 5V bus is receiving data and the 3.3V bus is transmitting, we have 3.3V across the 4.7kohm resistor to ground and the same 3.3V is passed through the 2.4kohm resistor to the 5V logic inputs of the vintage system. Generally a valid logic high is anything over 2.4V, so the 3.3V data from the flash chip output is no problem for a 5V input. 0V logic low is still below the .8V threshold for low logic, so again no issues here. In all cases, a 0V input will result in a 0V output. In the event of bus conflicts, it is also a non-issue. The cartridge side sees the output it is attempting to transmit, and the console side sees the output the CPU is attempting to transmit.

 

 

 

EDIT: This will not work per Kevtris. I did not consider the stray capacitance that exists on the bus. Nothing to see here, move along...

Edited July 19, 2017 by Kosmic Stardust

[kevtris]

Somebody please pass this schematic on to Krikzz. I may post it on Krikzz forums later. Inline 100 ohm resistors works but it is a kludge.

 

A proper voltage divider setup places .5R inline between the 5V bus and the 3.3V bus, then places R between the input of the 3.3V bus and ground. In my example, I used a 2.4kohm for .5R and a 4.7kohm for R. This circuit also works for bidirectional data bus. In the event the 5V bus is receiving data and the 3.3V bus is transmitting, we have 3.3V across the 4.7kohm resistor to ground and the same 3.3V is passed through the 2.4kohm resistor to the 5V logic inputs of the vintage system. Generally a valid logic high is anything over 2.4V, so the 3.3V data from the flash chip output is no problem for a 5V input. 0V logic low is still below the .8V threshold for low logic, so again no issues here. In all cases, a 0V input will result in a 0V output. In the event of bus conflicts, it is also a non-issue. The cartridge side sees the output it is attempting to transmit, and the console side sees the output the CPU is attempting to transmit.

 

voltage divider.png

That isn't going to work too hot unfortunately. Those resistors are huuuuuuuuge. The capacitance is going to kill your edge rate, and it probably won't work at all. Feeding signals back through will similar be a fail- there's probably even more capacitance on the bus on that side with maybe 6-12 chips or more, so you'll get a triangle waveform on the "output" side, as you charge and discharge the capacitance of all the chips + strays. For something low speed in the 100's of Hz it will work, but anything in the MHz range, not so much unfortunately.

 

The 100 ohm resistors were chosen so that they won't limit the slew rate so much and have enough oomph to overcome the capacitance, while still "limiting current".

 

The only good solution is level translators. There's no good shortcut to using them. Fortunately they aren't THAT expensive so there's no excuse for not using them IMO other than laziness and trying to shave cents from designs that aren't made in large volumes anyways.