F18A MK2

[+acadiel]

Also, could people please not PM me with complaints and insults for raising this issue. It's not my fault this has happened. I just alerted Matthew to the potential of risk because he has a right to know what he was getting into. If I knew and said nothing, I would have been partially to blame if they came after him. Even if it is only a one in 20 chance, I think any reasonable person here wouldn't take that risk with their home, retirement, etc. But at least they'd like to make an informed decision - which you can't do unless someone in the know gives you the information.

I want Matthew to succeed, and not lose a great deal in the process.

 

So please, stop blaming me. I'm not HDMI Licensing. If I was, I'd be better funded.

 

 

FYI, it's not me doing this to MrPix. Just want that out in the open.

 

However, like said before, still find it funny one can buy the connectors and the chips to do this and put something together and be completely oblivious to the fact that licensing is required. They could have handled it better, something like - "You want to buy this HDMI chip? You need to verify that you're an adopter, and then we can sell them to you."

 

Makes one wonder about projects such as the Amiga 600 Vampire card and others - there are like almost 0 discussions on any forum about this, nor are they offering up any "we're licensed" on their site. I'd also love to find out whether stuff like this is even licensed or not.

 

I do know for a fact that if I made some prototypes in my basement, they're not getting a licensing fee for me. Or even half a dozen - if the parts are publicly available, I see no need to tell them I'm building something for my use and maybe a handful of friends. The F18 is a much larger distribution, so I can see Matthew's reluctance and wanting to take the high road. Still doesn't change my viewpoint about the sucky licensing.

[+OLD CS1]

Which is my point. What Matt is doing with the F18A is commercial, not hobbyist, under strict definition. The problem is they see all commercial endeavors, no matter how small, as the same. It probably keeps them from dealing with licensing abuses.

 

EDIT: Does not make things any less frustrating.

[Omega-TI]

Mr. Pix,

It's sad when some want to 'shoot the messenger' simply for doing someone a huge favor and alerting them to the possibility of a major legal hassle. Rest assured, and take comfort in the fact that you can sleep at night with a clear conscious.

 

I love the idea of an HDMI F18A MK2 too, but not at the expense of a member of our community. Yes, the situation SUCKS, yes it's depressing, but a few people getting a card for a hobby computer is NOT worth a man risking his entire life's work and savings just to please a few people... that is just asking too much.

[RXB]

We can make due with whatever he decides, but what works and is best for price is the choice for me.

 

For the first time I will buy one hoping for less then $300.00 cost.

[matthew180]

@RXB: $300.00?? You do know the original F18A was only $86 on average, right? The MK2 will be no where near $300. Once the video interface issue is worked out, I will have a better idea of the final cost, but I'm pretty sure it will be less than $100.

[+OLD CS1]

@RXB: $300.00?? You do know the original F18A was only $86 on average, right? The MK2 will be no where near $300. Once the video interface issue is worked out, I will have a better idea of the final cost, but I'm pretty sure it will be less than $100.

 

He is talking about including licensing fees

[+acadiel]

 

He is talking about including licensing fees

 

Now that is a fan base. :) Hoo-rah!

[RXB]

Well either you are intrested no matter the price or not, but it has to have what you want.

 

Now that I have a SAMS 32K Loader/Saver/Switcher in RXB next is better graphics and waited years for this to happen.

 

Mind you RXB already had a 8K Loader/Saver/Switcher for lower 8K assembly since 2001.

[matthew180]

@RXB: I would argue that the F18A does offer better graphics. Maybe not strictly in terms of resolution, however more color can have a huge impact on graphic quality; and the F18A certainly does offer more colors. The F18A also has a little more resolution with the 30-row mode (256x240), which allows it to have the same (and better) colors and resolution as the NES. Couple that with smooth pixel scrolling, multi-color sprites, etc. and you can really do a lot with what you have.

 

I would be more than happy to get an F18A into your hands (if you don't already have one) if you would be willing to add F18A support into RXB.

[matthew180]

Update time.

 

So, I spent the last few days searching for the "free" DisplayPort 1.1a specifications. For something that is free, it sure as hell was hard to find! Seems to be a common theme with anything related to DisplayPort. After a quick review of the spec, it is very clear that DisplayPort is *very* different from HDMI, and a *lot* more complicated from an implementation standpoint. It also appears that DisplayPort requires, as a minimum, a single 1.62Gbps channel, and that requires a 162MHz clock (not too much of a problem) with a 10X serializer (a much bigger problem).

 

The only way (in an FPGA that I can afford) to do 1.6Gbps is with dedicated Gigabit Transceiver (GTS) blocks in the FPGA. The Spartan-6 has versions with GTS blocks, but not in the small package I need to fit the 600-mil DIP package of the original VDP. The 7-series FPGAs, specifically the Artix-7 (16T) has the GTS blocks in a 15mm package that I *might* be able to get to work (the FPGA I'm using now is a 13mm package), but it is not commonly stocked (i.e. hard to get) and therefore costs more when it is available. Ironically, the next size up, the Artix-7 25T, is very well stocked and is only $5 more than the FPGA I'm currently using (and has a lot more resources to boot), but the common package is the 17mm version, which is just too wide (it runs into the PCB pins). The only way to get into a package that wide is to go with a double-PCB solution and complete rework of the design (more time, money, and frustration).

 

See where this is all going? And it does not get any easier. Even if I could get an FPGA with the needed GTS blocks, implementing the DisplayPort core is a pretty major task. There is one open-source example done by Mike Field (Hamster) back in 2015 that I could maybe adapt, but even that would be a lot of work (I don't think he incorporated audio in his core).

 

There also seem to be very few, if any, single-chip DisplayPort solutions that are available and affordable. For a standard that is trying to take over HDMI, they are really making it hard to implement (from my perspective anyway).

 

In a last-ditch HDMI effort, I wrote a letter to the HDMI.org group asking if they would consider a very low-volume license that increased the per-unit royalty in exchange for no (or drastically reduced) annual fee, and a lifetime production of less than 5000 units. We'll see what they say.

 

So I'm still kicking around possibilities, and any other ideas people have would be great to hear. Again, to avoid a lot of rework and delay, I need something that can work in 9mm x 9mm of PCB area.

 

DVI with audio (dual-link I believe it is called) would be an option if I had a way to get the signals off of the MK2 PCB and over to a daughter board with the *really big* DVI connector on it. Of course that means people would have to modify their case to mount that giant connector. I can't find anything that works with this solution that would cost less than $20 or so. The solution requires a PCB connector on the MK2-side, a cable assembly to connect the MK2 to the daughter-board, and the daughter-board with the receiving-cable and DVI connectors.

 

VGA, ah good old VGA, how simple you are. I don't know if this is an option either. People would be upset for sure, and there is still the audio problem that the MK2 is trying to solve by using a digital video output interface. However, if someone could come up with a VGA connector solution, I might be swayed. It has to be small on the MK2 side, and a pre-made cable assembly (I'm not making video cable assemblies!), but if there was a viable VGA cable / header solution then I could probably add an audio output of some sort and call it good.

 

At this point I think lasers that just beam the video directly into your eyes would be easier. Bah! Video sucks.

[RXB]

@RXB: I would argue that the F18A does offer better graphics. Maybe not strictly in terms of resolution, however more color can have a huge impact on graphic quality; and the F18A certainly does offer more colors. The F18A also has a little more resolution with the 30-row mode (256x240), which allows it to have the same (and better) colors and resolution as the NES. Couple that with smooth pixel scrolling, multi-color sprites, etc. and you can really do a lot with what you have.

 

I would be more than happy to get an F18A into your hands (if you don't already have one) if you would be willing to add F18A support into RXB.

Ok how much?

[RXB]

Crap just go ribbon cable to VGA plug mounted on TI and plug this into it......

 

https://www.gearbest.com/cables-connectors/pp_1096148.html?wid=1433363&currency=USD&vip=14441751&gclid=Cj0KCQjwvezZBRDkARIsADKQyPlsW2RsIwNbX_DXOA7GhLnk-aTXm8_rjlBC_ljmvCUyZglyp8WBx3AaAvHaEALw_wcB

[Asmusr]

I know next to nothing about electronics, but isn't there some kind of standard small-form-factor connector cable that can be used to carry the signal (in whatever format) onto the daughter board? I read about "displayport over usb type-c". Or perhaps you could use a passive mini-displayport to DVI adapter cable and generate the DVI signal on the MK2?

 

Edit: Another idea would be to keep the micro HDMI connector and use a HDMI cable to carry a non-HDMI signal to a daughterboard. Would that require a license?

Edited July 5, 2018 by Asmusr

[TheMole]

IF it is indeed correct that the only real issue regarding licensing is the HDMI connector/form factor, and not the actual signalling, you could consider this: https://www.hdmi.org/manufacturer/hdmialtmodeusbtypec.aspx

[+Ksarul]

Looking at all of the options, the option to use DVI (with the necessary increase in cost for the connector board) would probably be the most optimal way to get the desired HDMI signal quality without the hassles of HDMI compliance requirements. Anybody that really needed to connect to HDMI at that point could use one of the many DVI to HDMI cables that are already out there--and the cable would be a user responsibility (and if sold in the US by reputable sellers, it is likely to be compliant in any case).

 

I would definitely pay the additional cost to go this route, as the capability this board provides is totally worth it to me. Even at $125 per board, it is still about $100 less than any of the other available video enhancement options out there--and it has the double bonus of being currently available too!

 

Once again, thank you @matthew180 for taking the time to work on this! I'll need at least three of them if you ever get to a release point (and I totally understand if the hassles of getting there are more than you want to deal with force you end up dropping the project--stressing yourself to meet inflated user expectations is neither necessary nor asked for here). I am just glad for all you've already done!

[jjh76]

Ok, from what I gather the HDMI licensing is for the design of the connector only, and not for the signals you send through it right? Well then, would it be possible to wire up an ethernet jack to the video out and use one of these:

 

https://www.newegg.com/Product/Product.aspx?Item=9SIAB1Z7GH3721&cm_re=hdmi_to_ethernet_adapter-_-2XW-0005-00041-_-Product

[Opry99er]

I'm seeing the same thing... Looks like the connector is the enforced patent...

 

Interesting.

[matthew180]

Ok how much?

 

I'll send you one of my F18A boards, and if you add support for the F18A into RXB, consider the F18A yours to keep. If you decide you don't want to add F18A support, then you can send me the F18A back or just pay the $86 price and keep it. How long to you need to get something into RXB, 6 months, a year? If you are interested, send me a Private Message with your address.

[matthew180]

I know next to nothing about electronics, but isn't there some kind of standard small-form-factor connector cable that can be used to carry the signal (in whatever format) onto the daughter board? I read about "displayport over usb type-c". Or perhaps you could use a passive mini-displayport to DVI adapter cable and generate the DVI signal on the MK2?

 

Edit: Another idea would be to keep the micro HDMI connector and use a HDMI cable to carry a non-HDMI signal to a daughterboard. Would that require a license?

 

Moving the DVI signals off of the MK2 board requires controlled cables since the high-speed serial lines are balanced pairs and all that. It would be just like the video cables themselves. Also, DVI to DP cannot be passive, they are completely different from an electrical and protocol standpoint. And, I can't generate DP signals from the FPGA I'm using (see previous post for the details about not being able to change the FPGA without a *complete* redesign and starting over.)

 

 

I'm pretty sure HDMI covers the connectors and some other aspects of the protocol, whether that includes any part of the electrical signaling I'm not sure; but it must cover some part of the data format. There are multiple big companies involved and multiple patents from what I understand (I have not really dug into the details of what is covered / protected by the HDMI license.)

 

I am very tempted to use a USB-C connector, which is only slightly larger than the micro-HDMI, and generate DVI signaling. But that relies on the USB-C to HDMI cables and HDMI monitors to accept DVI signaling. Most do from what I have read, but it is no guarantee. It also does not solve the audio problem, because DVI does not carry digital-audio data, and I might be infringing on the HDMI spec if I include the audio data. That is probably a grey (or gray) area. Going with a USB-C connector has an advantage too, it offers a path to USB updates for the firmware... with the inclusion of yet another chip to fit into the PCB.

 

 

Once again, thank you @matthew180 for taking the time to work on this! I'll need at least three of them if you ever get to a release point (and I totally understand if the hassles of getting there are more than you want to deal with force you end up dropping the project--stressing yourself to meet inflated user expectations is neither necessary nor asked for here). I am just glad for all you've already done!

 

Thanks for the kind words. I'm not willing to drop the project at this point, too much time and effort spent.

[jjh76]

Does this even really need to have vga or hdmi? Is the fpga you're using able to generate a component or rgb signal?

[+retroclouds]

Could you not have some kind of custom breakout box outside the console/computer that you could use for splitting the audio signal and which has the hdmi/dvi/vga or whatever is dediced for. That way you could have your own protocol into the breakout box and go from there. I know this does not solve all issues. Sure, it would increase costs but it would open some possibilities without having to do a major redesign. Could even have the possibility to add hdmi at a later stage.

[Asmusr]

Is this not an option?

 

STDP4028 LVDS or RGB to DisplayPort converter

http://www.megachips.com/products/displayport/STDP4028

[RXB]

 

I'll send you one of my F18A boards, and if you add support for the F18A into RXB, consider the F18A yours to keep. If you decide you don't want to add F18A support, then you can send me the F18A back or just pay the $86 price and keep it. How long to you need to get something into RXB, 6 months, a year? If you are interested, send me a Private Message with your address.

I have the money now for buying stuff, wife moved to Idaho so all that booze and drug money is in the bank vs her being wasted 10 hours out of 24.

 

Well announced 2018 released before end of year, so F18 support will be in next version 2019 is that doable?

 

2019 should be a RXB with SAMS/F18 support and would love a 4Meg version of SAMS and F18 MK2 by end of 2019 out....

[matthew180]

Does this even really need to have vga or hdmi? Is the fpga you're using able to generate a component or rgb signal?

 

When you say "RGB" I assume you are talking about analog red, green, and blue video signals with associated 15KHz horizontal and 60Hz vertical? That is the same as "VGA", but the horizontal is 31KHz for VGA, and yes the F!8A can generate that kind of signal (that is what the original F18A does). However, the 15KHz signals require a monitor that most people do not have, and defeats the main purpose of the F18A, i.e. to connect to modern monitors vs. legacy displays that are no longer available, breaking, and hard to get. Also, the original VDPs put out those kinds of signals already, so if you want that kind of video you don't have to mess with something like the F18A.

 

 

Could you not have some kind of custom breakout box outside the console/computer that you could use for splitting the audio signal and which has the hdmi/dvi/vga or whatever is dediced for.

 

I have looked for a solution like that for a long time. The problem is the data rate of the video. It is either wide (16 or more wires required) or very high-speed serial (LVDS, etc.) which requires cables designed for such signals, i.e. HDMI, USB3, SATA, DP, etc. and those are not small. Finding a connector + cable solution that is available and cost effective has so far eluded me. If anyone knows of a possibility, I'm eager to learn about it.

 

 

Is this not an option?

 

STDP4028 LVDS or RGB to DisplayPort converter

http://www.megachips.com/products/displayport/STDP4028

 

Yup, I have seen that chip many times; it comes up in just about every search I do related to this problem. The main issues with it are cost (is is almost $10), availability (it seems to only be available from DigiKey, and I like to have at least two suppliers for every component I use), and size (the chip is 164-pins and 12mm x 12mm). Unfortunately it tries to do too much. If they would just make a chip for LVDS to DP *or* RGB to DP, that would make it simpler, smaller, and cheaper.

[matthew180]

I have the money now for buying stuff, wife moved to Idaho so all that booze and drug money is in the bank vs her being wasted 10 hours out of 24.

 

Well announced 2018 released before end of year, so F18 support will be in next version 2019 is that doable?

 

2019 should be a RXB with SAMS/F18 support and would love a 4Meg version of SAMS and F18 MK2 by end of 2019 out....

 

Sometime in 2019 sounds good to me. PM me your address.