Super SD System 3

[ASDR]

Did you consider to just add a flying resistor to ground on your existing PCBs to get both correct attenuation and impedance? That's not exactly pretty, but I'd personally take ugly on the inside over ugly on the output any day I guess with an initial batch of 100-200 units that would be an afternoon of soldering for two people. I'd also imagine you'd fix this properly on the 2nd batch PCBs, would be kinda awkward to have two kinds of SSDS3s out there requiring slightly different cables.

In other news, My Life in Gaming just posted a video on the state of RGB output for the 1-CHIP SNES consoles:

https://www.youtube.com/watch?v=gvliOHNfCAA

The unmodified 1-CHIP and its early RGB bypass mods are well-known to be too bright. They show some great pictures of games and the 240p testsuite to illustrate the difference between the vanilla RGB / early bypass mods and the correct boards available today. I had great success with the input gain setting on the OSSC in correcting consoles that are way too low, but there isn't a lot of headroom to turn down an overly bright image.

[Ripthorn]

The Genesis 2 conector was used to avoid the user needing to buy custom cables for their systems, that was just for user conveinence.
We are trying our best to get this sorted out, but keep in mind that the boards are already produced, so we can only make small changes like component values, or bypassing the capacitor with a 0ohm resistor.

You could have used the SEGA Saturn mini din connector.
Easy to find on Aliexpress, same for the cables. So no need for custom cables.
The good part is that it has no components inside the cables, no need to take into consideration caps and resistors.

I hope in the next batch/version you decide to switch to the mini din 10 connector with the signal properly attenuated.

[ASDR]

You could have used the SEGA Saturn mini din connector.
Easy to find on Aliexpress, same for the cables. So no need for custom cables.
The good part is that it has no components inside the cables, no need to take into consideration caps and resistors.

I hope in the next batch/version you decide to switch to the mini din 10 connector with the signal properly attenuated.

A Saturn cable wired for csync needs the same 470ohm/220uf components on the sync line. And then you'd have the potential for the same kind of issues, like people frying their scaler/TV with TTL level sync because they bought a poorly made cable with missing components. In any case, it doesn't matter which cable / connector is used. Nintendo AV, Mega Drive 2, Saturn, etc. All of them are readily available and all of them are capable of delivering great picture quality if you design your device properly for them. The issue here is that the elevated output levels of the PC Engine and the actual, correct output specifications of the MD2 were seemingly not considered until last minute.

[donluca]

I think people are really overreacting over the 0.8vpp vs 0.7vpp.

Keep in mind some TV sets don't have proper termination on their inputs, so what about those? They are gonna have a brighter image with ALL your consoles and you've probably never even noticed it.

This is a moot point and, IMHO, I'd rather not put a 21Ω resistor in series instead of the 0Ω because it would lead to the risk of an impedance mismatch which I'd personally avoid.

Between the two values there's not such difference, NeoSD made a video about it and unless you make a back to back comparison, you would hardly notice it, especially while gaming.

On the other side, if your concept of gaming is staying in front of your TV watching the various 240p test suite patterns then you might have a point, but I'd re-consider some of your priorities in life

[Syntax]

It's not over reacting when it's easy enough to do it correctly.

Neodev, I suggest you look at Voultars PCE RGB bypass board and copy that, it's open source, and he knows what he's doing when it comes to video.

[DirkSwizzler]

It's not over reacting when it's easy enough to do it correctly.

This.

Neodev, I suggest you look at Voultars PCE RGB bypass board and copy that, it's open source, and he knows what he's doing when it comes to video.

The cable that's for is a different connector and spec though. I *think* it's entirely passive, which is what they were designing for in the first place.

[strygo]

Is 0.7vpp an RGB/video standard or just a SCART standard?

If it is the former, then including the 21ohm resistors makes sense to me. If it is the latter, it feels like a problem better solved in the cable.

[Syntax]

It's not over reacting when it's easy enough to do it correctly.

This.

Neodev, I suggest you look at Voultars PCE RGB bypass board and copy that, it's open source, and he knows what he's doing when it comes to video.

The cable that's for is a different connector and spec though. I *think* it's entirely passive, which is what they were designing for in the first place.

So follow Voultars design and use a Saturn cable.

.7vpp is a TV standard, it's what every set in the world is expecting.

[DirkSwizzler]

So follow Voultars design and use a Saturn cable.

I'm just trying to be specific to hopefully avoid oversight is all.

[Syntax]

Voultars board is not designed for any cable really. He even states that wiring his amp is your problem to figure out.

I wasn't aware a Saturn cable was the same as a md2 cable without components untill ASDR mentioned it, I haven't checked if he is correct but it seems like a no brainer to use it with Voultars design.

Edit *** are the Saturn and MD2 Dins different?
If so we should stop focusing on the Saturn cable, we can't expect any hardware changes past smd components.

Both major cable makers offer custom cables without components, so it's not hard for someone without any soldering skills to aquire one.

[Voultar]

This is all pretty simple.

If you want to drive video directly through Genesis II cables and avoid all of this attenuation headache. Take care of business right off of the 6260 outputs and couple the video into your 6dB driver. Keep the impedance low. The outputs will take care of themselves. Drive the outputs into whatever cable you want.

[Syntax]

Thankyou for coming to the show Lord Voultar

Show em how it's done!

[Voultar]

Or, more specifically.

Spoiler

I guess I don't understand why this is such a problem. You can shape this out so that you don't have to worry about attenuating anything on the outputs, Therefore, Genesis 2 cables will do the job without any modification.

Oversaturating the video (.8vPP) is bad. It's not the end of the world if you go too dark as you can recover the video with your TV/device.

Going too bright, you can't go back. That video content is gone.

[Neodev]

This is all pretty simple.

If you want to drive video directly through Genesis II cables and avoid all of this attenuation headache. Take care of business right off of the 6260 outputs and couple the video into your 6dB driver. Keep the impedance low. The outputs will take care of themselves. Drive the outputs into whatever cable you want.

Yes, that was another option I was thinking, attenuating before the amplifier, but that can't be done on the current design because the boards are already produced, so I was looking for a solution with the current board.

[gojira54]

What values have you put on the amp input currently? Also is the LPF disabled or enabled pls?

[Neodev]

What values have you put on the amp input currently? Also is the LPF disabled or enabled pls?

0.1uF on input, and 3M6 ohm to Vcc . Bypass is ON, so LPF is disabled.

[gojira54]

Thank you =]

[viletim]

Or, more specifically.

Spoiler

I guess I don't understand why this is such a problem. You can shape this out so that you don't have to worry about attenuating anything on the outputs, Therefore, Genesis 2 cables will do the job without any modification.

Oversaturating the video (.8vPP) is bad. It's not the end of the world if you go too dark as you can recover the video with your TV/device.

Going too bright, you can't go back. That video content is gone.

The PC engine's video chip has a PNP + high impedance output. An resistor attenuator connected to it would load the signal. It's better to attenuate on the amplifier output.


Neodev,

I recommend you have your contract manufacturer replace each 75 ohm series resistor on your amplifier with a 22 ohm resistor, if possible. This will solve both your problems (cable compatibility and amplifier attenuation) at once. It does increase the total output impedance to 96 ohms (75 in cable + 22 on board), but that shouldn't have an adverse effect on the video quality.

[viletim]

I'm by no means an expert, but:

- I would say it is extremely unlikely they did that. I think they would have mentioned it by now. Also if they were trying to get proper levels, they probably would have realized earlier that their cable had extra resistors in it.

- IMHO, it looks too bright and washed out with 75 ohm resistors. You can wait till you get yours and see how you feel about it. I know at least RCA makes custom cables. Ive never bought from RGC though so im not sure about them. They might just not have any 91 ohm resistors sitting around to do it as its not really a common value, but i'm sure they could buy some if someone wanted a cable with them in it.

Thanks! I'm trying to understand your suggested cable construction. My knowledge here is incredibly basic, so not critiquing anything, simply trying to understand it for myself.

You suggest turning the in series 75 ohm resistor into a voltage divider, 86 ohm in series, 586 ohm to ground, like this

0.8*586/(86+586) ~ 0.69Vpp

Very close to ideal. But we want to use standard values resistors, so it'll be

0.8*470/(91+470) ~ 0.67Vpp

Still close, probably better to undershoot in any case.

What confuses me, why do we use a simple in series 470 ohm resistor to attenuate TTL csync into video level sync, but need a voltage divider to attenuate our video levels? Is that because the sync pulses are of a known/fixed voltage & current draw?

The attenuator for the Mega Drive sync signal is series resistor (470 ohms) and capacitor (220u). The values are not critical, but both components are essential. What's going on here is the sync output form the Sega video chip is open collector (can only drive low) with a pull up resistor (maybe 2.2k - I can't remember exactly). It's connected to the video encoder chip and the AV out jack. The trick is to deliver a sync signal to the TV (> 0.3 Vpp) without attenuating the signal to the video encoder (TTL input). There's a significant potential across the cap always, when the sync pulse comes it drives negative. It works like a charge pump circuit.

[gojira54]

Thanks for the input here =]
So what would be the correct components to put on the sync line? A 470 ohm resistor only?

[ASDR]

Thanks for your explanations Tim & Voultar! Hopefully the Terraonion guys can implement one of your suggestions with their existing PCBs.

[Voultar]

The PC engine's video chip has a PNP + high impedance output. An resistor attenuator connected to it would load the signal. It's better to attenuate on the amplifier output.

You're right, Tim. That's my mistake. Ste of Rectalvision did a bunch of research on this a while back and got some data out of the DAC. I know that I put 8 or 9 different loads on those and did some observations on the actual behavior, from 1K all the way to 13K.

If you load the outputs with around 10K, it's perfectly fine.

[ASDR]

Manual:

http://www.neosdstore.com/news/wp-conte ... MANUAL.pdf

Output details:

[strygo]

It looks like they revised it slightly: http://www.neosdstore.com/news/wp-conte ... UAL_v2.pdf

Based on the details you pasted, it would seem they went the 0 ohm resistor route. I suppose that implies that you should use 91 ohm in the cable. Does this sound right?

[ASDR]

They said they'd definitely remove the output components and target a standard MD2 cable, so no surprise there. I was hoping they would correctly attenuate on the input side like Voultar suggests, even if it meant adding flying resistors to their 1st revision board. Or at least attenuate correctly like Tim suggested, even if it leaves the impedance a bit off. Seems like they decide to not do any of that and just have the output slightly too bright.

From what everybody has been saying, and as far as I can verify it with my clumsy EE-math skills, a cable with 91 ohm in series & 470 ohm to ground should get us within spec.

It's not the end of the world, but it just seems like such an avoidable issue. 5min talking to the various community members here that have worked on PC Engine video devices, properly looking up the specs of MD2 output/cables and measuring PCE expansion / SSDS3 output levels with a scope or just looking at the color/greyscale ramps of the 240p test suite should've shown these problems right away.

Congrats to all the people smarter than us who waited for the second batch

[mickcris]

They said they'd definitely remove the output components and target a standard MD2 cable, so no surprise there. I was hoping they would correctly attenuate on the input side like Voultar suggests, even if it meant adding flying resistors to their 1st revision board. Or at least attenuate correctly like Tim suggested, even if it leaves the impedance a bit off. Seems like they decide to not do any of that and just have the output slightly too bright.

From what everybody has been saying, and as far as I can verify it with my clumsy EE-math skills, a cable with 91 ohm in series & 470 ohm to ground should get us within spec.

It's not the end of the world, but it just seems like such an avoidable issue. 5min talking to the various community members here that have worked on PC Engine video devices, properly looking up the specs of MD2 output/cables and measuring PCE expansion / SSDS3 output levels with a scope or just looking at the color/greyscale ramps of the 240p test suite should've shown these problems right away.

Congrats to all the people smarter than us who waited for the second batch

If you want to make it easier and do what Viletim suggested, just use a 100 ohm resistor in your cable instead of 75 ohm.

[ASDR]

That's true. I'll give it a try with my cable once I have the system.

Talking about something more fun, what's on everybody's SD card? I've spend the past days watching every Top 25 & Hidden Gem video I can find to make a list of games I'd like to play. I prefer shooters, JRPGs, puzzle games, racing & turn based strategy, not super big on platformers, head-to-head fighting, sports & beat'em'ups. Here:

HuCard

Code: Select all

1943 Kai (Japan).pce					Magical Chase (USA).pace
240p Test Suite (PD) v1.02.pce				Mega Man 2 (PD) v87a Stereo pcedev.wordpress.com.pce
Aero Blasters (USA).pce					Mesopotamia (Japan).pce
After Burner II (Japan).pce				Military Madness (USA).pce
Air Zonk (USA).pce					Necromancer (J) T+Eng v0.01b pikachumanson.pce
Alien Crush (USA).pce					Neutopia (USA).pce
Atomic Robo-Kid Special (Japan).pce			Neutopia II (USA).pce
Blazing Lazers (USA).pce				New Adventure Island (USA).pce
Bloody Wolf (USA).pce					Out Live (J) T+Eng v1.0 Nebulous .pce
Bonk III - Bonk's Big Adventure (USA).pce		Out Run (Japan).pce
Bonk's Adventure (USA).pce				Puzzle Boy (J) T+Eng v1 KingMike.pce
Bonk's Revenge (USA).pce
Boxyboy (USA).pce					R-Type (USA).pce
Cadash (USA).pce					Raiden (USA).pce
Castlevania (PD) v0.44 pcedev.wordpress.com.pce		Salamander (Japan).pce
Contra (PD) v0.35 pcedev.wordpress.com.pce		Silent Debuggers (USA).pce
Darius Plus (Japan).pce					Sinistron (USA).pce
Devil's Crush (USA).pce					Soldier Blade (USA).pce
Dungeon Explorer (USA).pce				Special Criminal Investigation (Japan).pce
F1 Circus (Japan).pce					Splatterhouse (Japan).pce
Final Lap Twin (USA).pce				Street Fighter II' - Champion Edition (Japan).pce
Final Soldier (Japan).pce				Super Mario Bros (PD) v0.21 pcedev.wordpress.com.pce
Galaga '90 (USA).pce					Super Star Soldier (USA).pce
Gomola Speed (Japan).pce				Time Cruise (USA).pce
Gradius (Japan).pce					Victory Run (USA).pace
Legendary Axe II, The (USA).pce				World Court Tennis (USA).pce
Legendary Axe, The (USA).pce

CD-ROM

Code: Select all

Akumajou Dracula X - Chi no Rondo (Japan) (FABT, FACT)		R-Type Complete CD (Japan) (Rev 1)
Buster Bros. (USA)						Rally Championship (Japan) (FABT)
Cotton_-_Fantastic_Night_Dreams_(NTSC-J)_[HCD3043]		Rayxanber II (Japan) (Rev 3)
Dungeon Explorer II (USA)					Rayxanber_3_(NTSC-J)_[DWCD2002]
Faussete_Amour_(NTSC-J)_[NXCD2017]				Seirei Senshi Spriggan (Japan)
Forgotten Worlds (Japan)					Spriggan_Mark_2_-_Re-Terraform_Project_(NTSC-J)_[NXCD2008]
Gate of Thunder (Japan) (FACT, FADT)				Star Parodier (Japan) (Rev 2)
Ginga_Fukei_Densetsu_-_Sapphire_(NTSC-J)_[HCD5080]		Super Darius (Japan)
Gradius II - Gofer no Yabou (Japan)				Super_Darius_2_(NTSC-J)_[NAPR-1031]
Hyper_Dyne_SideArms_Special_(NTSC-J)_[HACD9002]			Super_Raiden_(NTSC-J)_[HCD2023]
Image_Fight_2_-_Operation_Deepstriker_(NTSC-J)_[ICCD2002]	Vasteel (Japan) (Rev 10)
Last Alert [U][CD][TGXCD1007][Telenet Japan][1990][PCE]		Ys Book I & II (USA) (Rev 2)
Lords of Thunder (USA)						Ys III - Wanderers from Ys (USA) (Rev 1)
Neo Nectaris (Japan) (Rev 3)					Zero_Wing_(NTSC-J)_[NXCD1003]

This all fits on a 16GB SD card with nearly 900MB left for save memory, firmware updates and the game database files. So probably no need to go broke buying a 256GB card.

[Neodev]

They said they'd definitely remove the output components and target a standard MD2 cable, so no surprise there. I was hoping they would correctly attenuate on the input side like Voultar suggests, even if it meant adding flying resistors to their 1st revision board. Or at least attenuate correctly like Tim suggested, even if it leaves the impedance a bit off. Seems like they decide to not do any of that and just have the output slightly too bright.

From what everybody has been saying, and as far as I can verify it with my clumsy EE-math skills, a cable with 91 ohm in series & 470 ohm to ground should get us within spec.

It's not the end of the world, but it just seems like such an avoidable issue. 5min talking to the various community members here that have worked on PC Engine video devices, properly looking up the specs of MD2 output/cables and measuring PCE expansion / SSDS3 output levels with a scope or just looking at the color/greyscale ramps of the 240p test suite should've shown these problems right away.

Congrats to all the people smarter than us who waited for the second batch

Well, there is an error there in the manual. There is a 22 ohm (or 21, I can't remenber) resistor to attenuate the output on a proper MD2 cable.

[Fudoh]

Is there a chance of getting an audio mixing option in the future ? To manually set the volume balance between chip audio, CD audio and ADPCM audio ?

When using an old PCE along with a interface unit one could get chip audio from the PC and raw CD audio from the CD drive, which allowed manual external mixing (minus the ADPCM tracks though which were only output through the interface's RCA ports).

[Neodev]

Is there a chance of getting an audio mixing option in the future ? To manually set the volume balance between chip audio, CD audio and ADPCM audio ?

When using an old PCE along with a interface unit one could get chip audio from the PC and raw CD audio from the CD drive, which allowed manual external mixing (minus the ADPCM tracks though which were only output through the interface's RCA ports).

Right now, the audio is balanced so it sounds like an actual cd unit (or as close as possible). The pce audio volume can't be changed without altering the amplifier resistors, but the cd & adpcm ones could be changed, as they are digitally generated, and we could apply a factor to the values before sending to the DAC, it is already done for the fadein/out anyways. The current firmware doesn't allow changing that but it's something that can be done in the future without hardware changes.