DB9 Vs USB game control tests

[onaryc29]

AFAIK, the snac board is exactly that, a level shifter (5v to 3v). I think antonio include directly an equivalent to the snac board and put a db9 connector instead of a usb one (some board has also hdmi connector to use blissbox cable).

Personally, i think i will buy a snac board at some point, to be able to use native controller (i also like my PS4, xbox one, ... controllers connected in wireless or cable).

IMHO, a lot of mister users use it, instead of software emulators, for accuracy and very low input lag (i think the ability to use a CRT is also a big plus). I've used a lot of emulation solutions in the past 25 years and you do not need a lot of frames the feel the input lag and be bothered by it (especially the lack of homogeneity in this delay). That's why i switch to real hardware and was waiting for good FPGA solution (and here we are )

[benitoss]

AFAIK, the snac board is exactly that, a level shifter (5v to 3v). I think antonio include directly an equivalent to the snac board and put a db9 connector instead of a usb one (some board has also hdmi connector to use blissbox cable).

Personally, i think i will buy a snac board at some point, to be able to use native controller (i also like my PS4, xbox one, ... controllers connected in wireless or cable).

IMHO, a lot of mister users use it, instead of software emulators, for accuracy and very low input lag (i think the ability to use a CRT is also a big plus). I've used a lot of emulation solutions in the past 25 years and you do not need a lot of frames the feel the input lag and be bothered by it (especially the lack of homogeneity in this delay). That's why i switch to real hardware and was waiting for good FPGA solution (and here we are )

The Antonio board is exactly that , a level shifter (5v to 3v). But the big diferent is the electronic part.
The SNAC board requires of additional extensions so, finally you have to connect externally to the board through the HDMI connector (used in different way that a standard HDMI, that it creates confusion) another board in a complete solution very complex

The Antonio's DB9 is simpler, only level shifter (5v to 3v) to a one standard DB9 connector compatible directly with Atari standard Joysticks and MegaDrive -Genesis - MegaCD 3 & 6 buttons Joy-pads. Actually the Megadrive (Genesys) and MegaCD cores support them

Later through an simpler adapter without any electronic parts (only connectors to native ports) will support NES and SNES controllers

Why do you complicate with boards and extension when with only the DB9 simple board you can connect the Atari and Genesys controller directly?

[Newsdee]

What I'd like to know if is this is exactly SNAC or a separate implementation using different pins. Because those pins were reserved officially for that, and deviating from the spec will cause incompatibilities with newer hardware.

[onaryc29]

Well i prefer official hardware (well at least official support). You need specific version of the cores for DB9 support which is a pity (and for what reason, you already have snac support or can add it in the official repo). Personally, i prefer the versatility of the current official version (i do not know if there is a use of the direct to fpga usb3 port without level shifter but if there is or will be, the antonio board would not be able to do it)

Imho, the level shifter should be added on a future version of the official io board (and you can select to activate it or not, but perhaps it is too much complexity) a DB9 connector could be put instead of the usb for those who want that. But there you have 2 close implementations of the same thing with some efforts that can be put elsewhere. But, i know it is your time and you do want you want with it.

[the80scomputers]

I think it's perfect, everyone will have their preferences and can choose freely

The truth is than DB9 with its input lag 0 absolute (like the original systems) is already a fact because there is already hardware and a repository of adapted cores, and of course, all open source and its much cheaper hardware than the other options (5 euros it cost me to adapt my MiSTer
to DB9).

Let everyone take the path they like best.

[the80scomputers]

We have recently made a significant expansion of the number of cores adapted to DB9 which we keep continually updated. They are not all absolutely but almost.

I leave here the link for those who want to enjoy it:

https://github.com/antoniovillena/MiSTe ... tter_DB9MD

[dshadoff]

I see that TurboGrafx is on that list, but if you are using the user port (open-collector), the rise time is too slow with 10K ohm pull-up resistors: This means that there will be problems with 6-button joysticks, multi-taps, mouse, and pachinko controllers. In short, anything other than a single 2-button controller.

Are you using the "user port" as I believe that you are ? And if so, what is the value of your pull-up resistor ?

[Sinclair]

You have everything in that link, including the answer to those questions because you also have access to the source code from there.

[dshadoff]

You have everything in that link, including the answer to those questions because you also have access to the source code from there.

I followed the link. Lots of binaries, no source.
Up a level, tons of code, no hardware. Not useful.

So I will assume that you don't know either, and you haven't tested for any of the issues I mentioned (which almost certainly exist on this board as well).

EDIT:
Looking at the code for this core, it scans the user port asynchronously from the CPU at 195KHz - which may still be too fast.
So, while multi-tap and 6-button controller might work, pachinko pad, mouse, memory base 128 will not work, because the protocols are not properly observed.

And it reduces the functionality of multi-tap (if working) to only 2 joypads.

So, I'm not sold that this is better than SNAC in any way.

[Sinclair]

See that they are in that link you can find in addition to the binaries the source code of it.

I cannot answer your technical questions but you can find your answers because EVERYTHING is there, just as you have been told.
Therefore, if you are an expert you will find there the answers easily.

And yes I have tried it and it work fine, there is no lag and I can use the native controls, but if you want to continue using the USB controls you can also continue using them from these cores.

[redsteakraw]

Is there any progress on this getting mainlined this looks interesting?

[Sorgelig]

Is there any progress on this getting mainlined this looks interesting?

Use USB joysticks. With 1ms polling it also has zero lag. You won't need any adapters. You will be able to use high quality modern gaming gamepads.

[redsteakraw]

Is there any progress on this getting mainlined this looks interesting?

Use USB joysticks. With 1ms polling it also has zero lag. You won't need any adapters. You will be able to use high quality modern gaming gamepads.

Yeah I am going to make some gamepads, I might mod a broken genesis 3 button to USB using deamonbite. There is just something nostolgic and compelling about being able to plug in original controllers. Serious question about 1ms is it default yet if not what is holding that back?

[Threepwood]

Serious question about 1ms is it default yet if not what is holding that back?

1000Hz polling is in the kernel (not default), you can enable it via script.

[redsteakraw]

Serious question about 1ms is it default yet if not what is holding that back?

1000Hz polling is in the kernel (not default), you can enable it via script.

I know of the script but I was wondering if there was any technical reason for it not being default as it is touted as the future of MiSTer input. Are there any bugs or incompatibilities that 1ms polling introduces that would prevent it from being the default?

[Threepwood]

I know of the script but I was wondering if there was any technical reason for it not being default as it is touted as the future of MiSTer input. Are there any bugs or incompatibilities that 1ms polling introduces that would prevent it from being the default?

Apparently some gamepads have issues with the setting.

[reflex]

I know of the script but I was wondering if there was any technical reason for it not being default as it is touted as the future of MiSTer input. Are there any bugs or incompatibilities that 1ms polling introduces that would prevent it from being the default?

Apparently some gamepads have issues with the setting.

I actually had overall instability when I had this enabled. I haven't had time to go and try it again.

[Sorgelig]

It's not exactly must be 1ms. It can be set to 2ms.
Also some gamepads have polling rate 1ms by default and not require it to be enabled.
Arduino Micro based converters in MiSTer repository by default are 1ms, so tweaks aren't required.