Thoughts of RetroUSB's AVS?

This is an archive of a topic from NESdev BBS, taken in mid-October 2019 before a server upgrade.
View original topic
Thoughts of RetroUSB's AVS?
by on (#182580)
Anyone got it? What do you think about it?
Re: Thoughts of RetroUSB's AVS?
by on (#182582)
Very happy with it as a gaming platform. Performance (in terms of lag) on my TV was noticeably better than using a regular NES with the composite input. Never noticed any glaring inaccuracies or issues with actual games.

The scoreboard feature is pretty fun. I currently hold the number 1 score on the Dr. Mario and Tetris 2 leaderboards. :D

The only compatibility issue I encountered was that it did not support the PowerPak NSF player, although that seems to have been fixed in the latest beta firmware. I haven't been able to test it since my powerpak has been returned for repairs and
Re: Thoughts of RetroUSB's AVS?
by on (#182588)
Ordered mine late last week (not here yet), though I have no actual display that supports HDMI to test it on (will get one soon; baby steps).
Re: Thoughts of RetroUSB's AVS?
by on (#182596)
I had the chance to check one out at the Syracuse retro game convention.

I liked the palettes.
Scanlines looked well done, all locked to NES lines properly (no free-run business here)
Menu is kind of nifty, I like the scoreboard idea.
I wasn't able to figure out whether or not it had 3.3V <-> 5.0V buffers, since traditional NES carts are 5V devices.
Thought the case shape was wonky, but well made, and the Famicom slot is well integrated.
I liked the clear plastic case, a bit bummed those aren't just available.
Interesting that expansion audio is handled with an ADC, instead of re-implementing the way Kevtris did. Advantage: Not having to turn it on and off per-game.
Re: Thoughts of RetroUSB's AVS?
by on (#182629)
Quote:
I wasn't able to figure out whether or not it had 3.3V <-> 5.0V buffers, since traditional NES carts are 5V devices.


with smd/smt it's not easy to make out whether a regulator is 3.3v or 5v just by looking:

http://www.hobbytronics.co.uk/image/cache/data/misc/sot223-500x500.jpg

However, i think it should be non-critical.
Re: Thoughts of RetroUSB's AVS?
by on (#182643)
mikejmoffitt wrote:
I wasn't able to figure out whether or not it had 3.3V <-> 5.0V buffers, since traditional NES carts are 5V devices.

https://www.youtube.com/watch?v=kxrYjL6f_wQ
Kevtris did a teardown / analysis video not long ago, I'm pretty sure he covers it. I'm at work now and can't be bothered to watch the hour long video to check though.
If I remember correctly he had some concerns with the AVS' design; I remember the grounding of the USB and HDMI shields being a concern which I thought was kind of interesting. Had me wondering if bunnyboy was deliberately trying to isolate the FPGA's digital ground from chassis ground. I know it comes up in telecom hardware design a bunch (that's the industry I work in) depending on the need to isolate a given system from noise, but I don't know HDMI and USB standards at all.

Wow that was a tangent, sorry. :lol:
Re: Thoughts of RetroUSB's AVS?
by on (#182648)
This is the time mark: https://youtu.be/kxrYjL6f_wQ?t=2140

5 volts, "i have to guess". if he doesn't revise the statement later on from there.
If you have a voltmeter it's U68 according to the vid
Re: Thoughts of RetroUSB's AVS?
by on (#182652)
I'd really like to get one, but the price is already high enough without the international shipping and the 60% of import taxes our government demands. It could easily end up costing me twice what it costs to someone in North America to get an AVS. :cry:
Re: Thoughts of RetroUSB's AVS?
by on (#182686)
WheelInventor wrote:
Quote:
I wasn't able to figure out whether or not it had 3.3V <-> 5.0V buffers, since traditional NES carts are 5V devices.


with smd/smt it's not easy to make out whether a regulator is 3.3v or 5v just by looking:

http://www.hobbytronics.co.uk/image/cache/data/misc/sot223-500x500.jpg

However, i think it should be non-critical.

Whatever the regulator is feeding the cart, if it doesn't match the FPGA IO level then level shifting buffers would be best. This is just out of concern for longevity of both the console and the carts.
Re: Thoughts of RetroUSB's AVS?
by on (#182687)
I got my AVS today. This is probably the shortest "review" I've written on a product, but I've only put in about 30 minutes of test time.

- The first thing I did was update the F/W. v1.00 came with the unit, upgraded to v1.10 via the Scoreboard software (PC version).
- The Scoreboard software is incredibly simple, and I **love** the fact it's a standalone .exe vs. something you have to install.

- I forgot that my USB3HDCAP has HDMI capture capability, so I used that for testing the video and audio output.
- The video quality looks great. I haven't tested much other than SMB2 and Gyruss (FDS and US releases -- more on the FDS in a second), but they looked good. My capture device has all sorts of HDMI tweaking capability, which I didn't mess with.
- I didn't try games that are known to really use the NES NTSC palette to its best; like I said, ~30 minutes of test time.
- Audio output levels (for standard games and FDS games -- I haven't tried other expansion audio games yet) felt good/level.
- I didn't fool around with too many of the settings in the AVS (scanlines, etc.). I saw them, but didn't mess with them.

- Famicom and NES cartridge compatibility seems fine; everything I jammed into it worked. The Famicom slot tension feels great, while the NES slot tension feels tight but not excessively so (i.e. not as bad as some of those aftermarket NES console connector replacements -- the Blinking Light Win still is the best for those, BTW).
- FDS RAM adapter was tested -- saving the one issue (physical dimensions) for last -- and it worked fine. I used deadbody's fdsemu product without any oddities/problems.

- The physical size/shape of AVS isn't rectangular, instead it's more like an isosceles trapezoid. For some reason all the photos of the console are at an angle so it's hard to see; you can kinda tell from this one. I find this weird aesthetically, and it'll also make storage in square containers annoying. I'll get used to it though.
- The colour of the AVS is also a bit weird, but I believe it's intentional -- the AVS was the name of the "original" console Nintendo came up with before choosing the name NES, and I believe the prototypes used the colour scheme that the AVS has. It's very hard to describe -- it's like an "off-grey" with a slight tinge of brown or mauve in it.
- The NES controller connectors felt "just right" -- not excessively tight, not excessively loose -- when used with a couple different official NES controllers (original and dogbone). I didn't test the Famicom controller expansion port.

The two negatives I have at this point:

- I absolutely **hate** the lid/cartridge cover. It's too large, too flimsy, and feels extremely tacky/awkward. I'd rather the Famicom slot be covered (or removable? Possibly a push/slide-away plastic cover?) because it faces upwards/exposed to dust, and the rest left open.
- The Famicom FDS RAM adapter has some physical dimension incompatibilities with the "cartridge area". The FDS RAM adapter has a rubber cable that comes out of the back with a tension relief cover/sleeve. The cord/tension sleeve essentially rests on the AVS cartridge door (i.e. the cartridge area is slightly too short to fit the FDS RAM adapter), causing the RAM adapter's Famicom cart connector to be extremely "angled" (i.e. likely to bend pins more, or make a very bad connection). Had the "cartridge area" of the AVS been longer by maybe 1", things would have fit well. I ended up taking some pictures because it's hard to describe in text but easy to understand once you see it.

Overall I'm incredibly happy -- it means I don't have to have 3 separate devices on my table any more. Now just to do something about the lack of an HDMI TV... *chuckle*
Re: Thoughts of RetroUSB's AVS?
by on (#182976)
I bought the AVS on launch, and I've been doing a lot of beta testing on firmware updates. The latest beta firmware supports different palette selections, and issues with Xexyz, Gradius 2, and Pokemon Yellow (Lugia 2009's hack) have been fixed. Final Fantasy 7 reproduction carts (another hack) are still having issues due to CHR read timing issues, but bunnyboy has said he will correct that.

Overall, I have few complaints. The wire coming out of a RAM adapter is annoying, but I integrated an FDSstick into my ram adapter a while back and tossed the cable, so it works fine that way. I threw a lot of weird curveball carts at it without issue like a hacked Magic Dragon (R-Type) cart, Metal Slader Glory, Rockman 4 Minus Infinity, Legend of Link (MMC5 Zelda hack), etc.

I like the trapezoid shape it has, it's too bad the famicom carts stick straight up with no bracing except the door sitting on top, but nothing is perfect I guess.

It is 720p only, so expect your flatscreen to scale those scanlines wrong if you turn them on.
Re: Thoughts of RetroUSB's AVS?
by on (#184018)
I wouldn't mind having one but I really would want it to support Analog RGB AV output. I know not too many people would use it but supporting S-Video and Composite as well would have also been nice. This is the only current clone of the NES that doesn't suffer various incompatibilities or sound errors which is great but it's a shame that it's digital AV only. Perhaps a future version will address various issues or wishes people want.
Re: Thoughts of RetroUSB's AVS?
by on (#184022)
A lot of HDTVs will upscale 240p composite to a usable extent (modulo combing and lag) but choke on 240p RGB or 240p YPbPr. Would a simple line doubler that produces 480p analog component video, either RGB (aka "VGA") or YPbPr ("component"), be good enough? With 100% scanlines turned on, 480p component through a non-upscaling CRT should even work with the Zapper.
Re: Thoughts of RetroUSB's AVS?
by on (#184025)
tepples wrote:
A lot of HDTVs will upscale 240p composite to a usable extent (modulo combing and lag) but choke on 240p RGB or 240p YPbPr. Would a simple line doubler that produces 480p analog component video, either RGB (aka "VGA") or YPbPr ("component"), be good enough? With 100% scanlines turned on, 480p component through a non-upscaling CRT should even work with the Zapper.


That would be an interesting test for the zapper. A 480p CRT draws scanlines every 3.2 ms, while a 240p CRT draws scanlines every 6.3 ms. Of course, if you tell the 480p CRT to only draw half the lines, you may be able to fool the zapper's light detector into thinking it is seeing scanlines being drawn at the slower rate. You may want to increase the brightness, contrast and color to "bloom" out the drawn scanlines to lessen the obviousness of the undrawn scanlines.

I think the Open Source Scan Converter would really help out people who are looking for a 480p solution : http://retrogaming.hazard-city.de/ossc.html
Re: Thoughts of RetroUSB's AVS?
by on (#184052)
tepples wrote:
A lot of HDTVs will upscale 240p composite to a usable extent (modulo combing and lag) but choke on 240p RGB or 240p YPbPr. Would a simple line doubler that produces 480p analog component video, either RGB (aka "VGA") or YPbPr ("component"), be good enough? With 100% scanlines turned on, 480p component through a non-upscaling CRT should even work with the Zapper.


That's not quite right. Consider t to span from 0.0 to 1.0, where 0.0 is the start of one line (in 480p) and 1.0 is the end of one line (in 480p). The 480p image is line-doubling the 240p source in lockstep.

At 0.0, both the original 240p and 480p images are at the start of the line. At 1.0, both are at the end. However, at 0.5, the 480p display has already finished scanning an entire line, while the 240p source is halfway done. The 480p display then scans the same line a second time. If it matters, objects will be horizontally scaled in position. 100% scanlines or not doesn't really affect much; with 100% scanlines on (blanking the second line), the first half-equivalent of the 240p line is shown only to the zapper.
Re: Thoughts of RetroUSB's AVS?
by on (#184067)
Say the barrel is pointed four-fifths of the way from the left side of the screen. The light would come in at 0.8, 1.8, 2.8, 3.8, ... for the 240p monitor or 0.4, 1.4, 2.4, 3.4, ... for the 480p monitor showing 100% scanlines. Both would trigger a 15.7 kHz resonator.

And unlike with the Super Scope, Zapper games don't measure horizontal position, so games wouldn't be able to tell the difference between activation at 2.4 and activation at 2.8. In Zap Ruder's XY test, I tried to use an unrolled loop to find the start of light in 18-dot (6-cycle) units. It worked in some emulators, but on a real 240p monitor, the time of start of light was too noisy to be useful, darting back and forth by 1/3 of a scanline as I held the gun in place.