MiSTer FPGA Forum

If you connect a display to your MiSTer via HDMI, here's some good summary info and advice.

Apparently, the DE-10 Nano's HDMI port is not electrically buffered or isolated from the rest of the main board in the ways you (and other consumer electronics devices) would normally expect. For example, apparently its 5V line is tied directly to the main board's 5V rail, such that if 5V is present on that pin of the HDMI port, then some portions of MiSTer's hardware can appear to "turn on" even while the MiSTer itself is not powered up or even connected to power! As such, MiSTer can exhibit a myriad of weird issues while hooked up to a display system via HDMI, especially if you are using long cables or running through an HDMI switch, AV receiver, HDMI capture device, or any kind of standard HDMI adapter/converter:

• decreased SDRAM stability

• display device randomly losing sync/signal from the MiSTer or taking an excessively long time to lock in sync after a resolution mode change

• MiSTer's cooling fan spinning, or its lights lighting up, even while MiSTer is powered completely off

• HDMI switch refusing to change inputs and/or randomly changing inputs while you're not using MiSTer and it is powered off

• USB dongles for wifi and/or Bluetooth not coming up and working reliably on every boot

• ...and possibly even other weird symptoms

What I have found is that the following inexpensive HDMI booster gadget works to effectively isolate/buffer MiSTer's HDMI output:

https://www.amazon.com/gp/product/B07PWYN3L4/

And it seems to work without supplying it any external power over its microSD port (i.e. the power produced by MiSTer's HDMI port appears to be enough to make it work). By installing one of these between my MiSTer's HDMI output and the HDMI switch I used in my home theater system, I was able to cure a combination of the above issues that I'd been beating my head against for months.

So, if you're using HDMI at all, it's probably just a really good idea to get one of these little booster devices and install it, to electrically isolate things. And if you're putting MiSTer inside of a larger case of some kind, it's probably an even better idea to install one of these inside of the case, so that the HDMI port exposed on the outside of the case is already electrically isolated.

Any idea whether that repeater somehow buffers/isolates pin 13 vs. passing it through, for those of us with setups that encounter CEC/ARC problems?

ExCyber wrote: ↑Mon Jun 01, 2020 10:55 pm Any idea whether that repeater somehow buffers/isolates pin 13 vs. passing it through, for those of us with setups that encounter CEC/ARC problems?

No clue. How could I check?

EDIT: Also, what are these CEC/ARC problems you are having?

Broken link above, please edit.

Thanks for the info. Most of your points seem logic or possible to me, depending on the output devices you use with your MiSTer.

What I don't get is, why SDRAM stability can be decreased. Can you elaborate a little more?

lomdar67 wrote: ↑Wed Jun 03, 2020 8:02 am What I don't get is, why SDRAM stability can be decreased. Can you elaborate a little more?

Nope, because I don’t understand why in any more specific detail than what I’ve already described. It is just demonstrably related.

Vic20-Ian wrote: ↑Wed Jun 03, 2020 8:00 am Broken link above, please edit.

Fixed.

keith.f.kelly wrote: ↑Wed Jun 03, 2020 8:33 am

lomdar67 wrote: ↑Wed Jun 03, 2020 8:02 am What I don't get is, why SDRAM stability can be decreased. Can you elaborate a little more?

Nope, because I don’t understand why in any more specific detail than what I’ve already described. It is just demonstrably related.

Okay, thanks, but where comes this information from and how could I test the SDRAM stability effect? Have you experienced it for your self? If so how exactly would I know that I'm effected?

keith.f.kelly wrote: ↑Wed Jun 03, 2020 7:51 am

ExCyber wrote: ↑Mon Jun 01, 2020 10:55 pm Any idea whether that repeater somehow buffers/isolates pin 13 vs. passing it through, for those of us with setups that encounter CEC/ARC problems?

No clue. How could I check?

EDIT: Also, what are these CEC/ARC problems you are having?

It's a series of issues that have been reported when connecting DE-10 Nano to a CEC-capable TV. In my case, audio (from all sources, not just the DE-10 Nano) is played through the TV speakers instead of the receiver. This is apparently because pin 13 (CEC) on the DE-10 Nano HDMI connector is connected directly to ground, and CEC is used to negotiate the ARC connection. People have reported success by connecting the DE-10 Nano through certain HDMI switches, physically removing pin 13 from an HDMI cable, or using an adapter whose explicit purpose is to not pass CEC signals. I thought that a repeater might "boost" the CEC signal in such a way as to prevent the problem.

edit: Another method reported successful with a different device having the same design issue is to use an HDMI-to-DVI adapter (or cable) followed by a DVI-to-HDMI adapter (or cable). This takes advantage of the fact that the CEC signal isn't present on DVI and the fact that HDMI audio is sent over the RGB pairs.

If you have a multimeter and it shows continuity between pin 13 on both sides, that would indicate that the repeater is just passing the pin through and wouldn't do anything to help with this problem. Other than that, I don't think there's a good way to check without being able to reproduce the problem.

There was a thread on the old forum about it, but the Wayback Machine apparently didn't get it.

There's also a Reddit post about the issue, and another Reddit post about the issue.

lomdar67 wrote: ↑Wed Jun 03, 2020 8:48 am Okay, thanks, but where comes this information from

Some from Sorgelig, some from my own first-hand experiences, some from corroborating experiences shared by others.

lomdar67 wrote: ↑Wed Jun 03, 2020 8:48 am and how could I test the SDRAM stability effect?

Using the memtest core in its default/auto mode. Let it run for 15 minutes and see what the fastest speed is that it considers stable.

lomdar67 wrote: ↑Wed Jun 03, 2020 8:48 am Have you experienced it for your self?

Yes.

lomdar67 wrote: ↑Wed Jun 03, 2020 8:48 am If so how exactly would I know that I'm effected?

If you've experienced any of the symptoms I've listed, and you use HDMI as your display connection, there's a decent chance this is what you're running into. It's worth checking and either ruling out or addressing.

ExCyber wrote: ↑Wed Jun 03, 2020 1:02 pm It's a series of issues that have been reported when connecting DE-10 Nano to a CEC-capable TV. In my case, audio (from all sources, not just the DE-10 Nano) is played through the TV speakers instead of the receiver. This is apparently because pin 13 (CEC) on the DE-10 Nano HDMI connector is connected directly to ground, and CEC is used to negotiate the ARC connection. People have reported success by connecting the DE-10 Nano through certain HDMI switches, physically removing pin 13 from an HDMI cable, or using an adapter whose explicit purpose is to not pass CEC signals. I thought that a repeater might "boost" the CEC signal in such a way as to prevent the problem.

...

There's also a Reddit post about the issue, and another Reddit post about the issue.

Interesting. At least one room in my house has such a setup (ARC from the TV feeds into the soundbar, and turning the TV on/off also turns the soundbar on/off and vice-versa). Are you saying that if I plug my MiSTer directly into some other HDMI port on that television, that should be sufficient to reproduce the issue, i.e. the sound will come out of the TV instead of the soundbar?

ExCyber wrote: ↑Wed Jun 03, 2020 1:02 pm If you have a multimeter and it shows continuity between pin 13 on both sides, that would indicate that the repeater is just passing the pin through and wouldn't do anything to help with this problem. Other than that, I don't think there's a good way to check without being able to reproduce the problem.

I do have a multimeter, but a simple continuity check wouldn't necessarily tell you whether it's electronically buffered or not. Plus, those HDMI pins are tiny little buggers... no way to get the tip of my multimeter probes inside and reliably touching only one pin.

keith.f.kelly wrote: ↑Wed Jun 03, 2020 11:23 pmInteresting. At least one room in my house has such a setup (ARC from the TV feeds into the soundbar, and turning the TV on/off also turns the soundbar on/off and vice-versa). Are you saying that if I plug my MiSTer directly into some other HDMI port on that television, that should be sufficient to reproduce the issue, i.e. the sound will come out of the TV instead of the soundbar?

I'm not sure how common that specific problem is. It seems that different TVs have differences in how they respond to this problem (basically, I think a TV can just wire pin 13 of all connectors together, but not all TVs necessarily do). But plugging the DE-10 Nano into a port and then power cycling everything should show some kind of oddity (e.g. audio not going to the soundbar, TV failing to turn the soundbar on and off, consoles failing to turn on the TV) if your TV is susceptible to the problem.

ExCyber wrote: ↑Thu Jun 04, 2020 3:22 am I'm not sure how common that specific problem is. It seems that different TVs have differences in how they respond to this problem (basically, I think a TV can just wire pin 13 of all connectors together, but not all TVs necessarily do). But plugging the DE-10 Nano into a port and then power cycling everything should show some kind of oddity (e.g. audio not going to the soundbar, TV failing to turn the soundbar on and off, consoles failing to turn on the TV) if your TV is susceptible to the problem.

Cool. I'll give it a try this weekend to see if I can repro. If I can, I'll let you know if this particular booster thingy resolves the repro.

I am unable to repro any CEC/ARC issues either without or with the HDMI repeater. Sorry I can’t provide an answer. However, the good news is that the repeater is inexpensive, so it won’t cost much for you to simply try it for yourself.

Fair enough. Thanks for trying. There's a possibiltiy that this only occurs on some manufacturing runs of the DE-10 Nano (the schematic has "DNI" on the components connected to pin 13), but any further investigation of that would be off-topic for this thread.

Does the HDMI repeater add lag?

SegaSnatcher wrote: ↑Mon Jun 08, 2020 9:27 pm Does the HDMI repeater add lag?

No.

ExCyber wrote: ↑Sun Jun 07, 2020 2:55 pm Fair enough. Thanks for trying. There's a possibiltiy that this only occurs on some manufacturing runs of the DE-10 Nano (the schematic has "DNI" on the components connected to pin 13), but any further investigation of that would be off-topic for this thread.

Hey! I was finally able to repro the CEC problems! Which means this repeater does NOT disconnect pin 13.

[ADDED] My configuration is as follows: MiSTer and Xbox360 plugged into the two HDMI inputs on my soundbar, and the soundbar then connected to the TV via an ARC-supporting HDMI port. This allows the remote for the soundbar to additionally power the TV on or off, or the remote for the TV to also power the sound bar on or off and control its volume, and it also allows the TV's own sound output (from its built-in video streaming apps) to be played through the soundbar. My repro is that sometimes when I power on the TV and/or soundbar and switch the soundbar to the MiSTer input, I'll either get no picture and no sound, or picture without sound. I have to yank the HDMI cable from the MiSTer and re-connect it to get both picture and sound working again.

Now I’m considering prying pin 13 out of one end of an HDMI cable (the one between the MiSTer and the soundbar) in order to solve it... but I suppose I can see if pin 13 can be broken/disconnected on the repeater.

Stay tuned.... more info to come.

I tried to pry out pin 13 from a regular HDMI cable today as a workaround, and I can tell you this: unless you have an army of intelligent Borg nano-robots to crawl down inside that tiny connector for you and rip the damn thing out, you may as well forget that approach. I used the tiniest tools I own, under the brightest lights in my house, with a magnifying lens, and I still couldn't do it. That cable got permanently mangled and is now in the trash.

[EDITED] I was able to dig through my TV's settings and find an option to disable AnyNet/CEC without disabling ARC, but further testing has revealed that doesn't resolve the issue. So now I'm trying to find a solution to this too.

So I haven't ordered one of those Lindy CEC-less adapters. Looks like they're vastly overpriced (especially here in the USA, because I guess they're manufactured in the UK and international shipping is a disaster right now thanks to COVID-19), and there are many angry Amazon reviews who all reported the same problem: early batches actually passed through pin 13 due to a manufacturing mistake, rendering that item's one job in life a complete failure! And because so many people order them from Amazon, discover they got a bad one, and then naturally return them to Amazon for refunds... and then clueless/lazy/evil Amazon refuses to take the bad ones out of circulation and just ships them right back out to other unsuspecting customers... it seems that most of the defective batch is still in circulation years after the fact. What a cluster!

And of course, no other company appears to make anything like that CEC-less adapter. Missed market opportunity, if anyone's looking for a get-rich-quick scheme. All you'd have to do is get some manufactured correctly by applying a tiny bit of quality assurance and attention to detail, then sell them like hotcakes for $15 USD apiece online with free domestic shipping, thus undercutting the only competition by at least 50%.

No, seriously... pretty please, will someone do that? Seems like it should be easy for all the folks who make and sell custom add-on boards for MiSTer. Surely someone in this community must be up to the challenge.

Look what I found...

Micro SATA Cables CEC-Less HDMI Cable

It has a confusing listing because the name of the brand is "Micro SATA Cables", but it looks like the item is actually a two-inch-long CEC-less male-to-male HDMI adapter cable. That should actually work great together with the repeater I linked to earlier, as it's a female-to-female coupler.

I'm going to order one so I can try it out, and I'll report back.

Wow. I bought that Lindy thing a couple months ago and it didnt solve my issue. I was stuck thinking I would never get this solved, only to find out that I probably had one of the bad batch plugs. So after reading these posts I managed to push in pin 13 and it actually did fix the arc problem. Now I can leave my Mister unpowered and with a hdmi cable plugged in and not causing any Arc issues. Perfect!

I wonder if this is what's causing some issues I've seen.

1. After long hours of playing, the refresh rate or something seems to get de-sync'd. How I see this is the screen seems to occasionally redraw the pixels noticeably from the top to bottom (like crt are drawn), but it's visible and a bit jarring.

2. Randomly on some soft reboots or switching from one core to another (not counting menu core) the graphics can get all funky and in the case of the menu core then there is this heavily distorted version of the default static background with the wallpaper occasionally showing itself. The menu isn't affected or anything else. If I load a core without doing a cold reboot or manual shutdown and powering on again, I see issues sometimes in cores as well. It's very rare.

But these issues are so rare I haven't been worrying about them. I've been more occupied to fix the fact that the dynamic contrast setting on my tv being turned off doesn't seem to change its reducing the brightness of colors when things being displayed around the edges of the output are black. Walking into a house in Final Fantasy VI that doesn't light up the whole screen, for instance... Probably something I can't fix.

I received the two CEC-Less HDMI cables a couple weeks ago, but I've had other things going on so I haven't had a chance to try one with my MiSTer up in my guest bedroom again (where I have a setup that reproduces the ARC problem). I've finally got a little time this weekend, so I'll try it out and report back if it cures the ARC issues I was seeing.

aberu wrote: ↑Thu Jul 02, 2020 2:32 pm But these issues are so rare I haven't been worrying about them. I've been more occupied to fix the fact that the dynamic contrast setting on my tv being turned off doesn't seem to change its reducing the brightness of colors when things being displayed around the edges of the output are black. Walking into a house in Final Fantasy VI that doesn't light up the whole screen, for instance... Probably something I can't fix.

You should start a new forum thread to discuss that issue. I recently went through this issue with multiple flat panel LCD TVs in my home, and I was able to cure it for at least one of them (a more recent SAMSUNG model) by accessing its secret service menu and applying some non-obvious settings I found in a forum someplace. Fortunately, I took notes on what I did, in case I ever need to repeat the process. And I also managed to track down the super-hard-to-find information about how to access the secret service menus in my other two TVs, even though they ultimately didn't expose settings to change that behavior. That kind of information is priceless, as is compiling a list of which LCD HDMI televisions actually support the slightly-non-standard-sync-rate trick that MiSTer uses for its "butter-smooth" vsync mode.

I lucked out with my 70" Sharp Aquos TV in my main room, as it seems to support a range of vsync frequencies centered around both 50Hz and 60Hz. But if I feed anything other than an exact 60Hz signal to my newer-model SAMSUNG in my guest bedroom, it will achieve sync and display the image, but the image will exhibit bizarre stuttering behavior (every so often, it will momentarily show a frame that is from 2-3 frames ago, such that any smooth motion appears to suddenly lurch backward for a frame). These are the kinds of unofficial compatibility that make or break using an HDMI display device, but unfortunately it's not part of any published specs... so the only real way to know if it will work before buying is to get verified information from other owners of the same model who have already tried it, or to take your MiSTer to a store, hook it up, set if for "butter smooth" sync mode, and then play it a while and see how the display handles it. I'm not sure how open to such a thing most stores or retail salespeople generally would be, and given the current COVID pandemic happening here in the USA, buying electronics physically in person at retailers isn't really a viable option right now.

keith.f.kelly wrote: ↑Fri Jul 10, 2020 10:59 pm

aberu wrote: ↑Thu Jul 02, 2020 2:32 pm But these issues are so rare I haven't been worrying about them. I've been more occupied to fix the fact that the dynamic contrast setting on my tv being turned off doesn't seem to change its reducing the brightness of colors when things being displayed around the edges of the output are black. Walking into a house in Final Fantasy VI that doesn't light up the whole screen, for instance... Probably something I can't fix.

You should start a new forum thread to discuss that issue. I recently went through this issue with multiple flat panel LCD TVs in my home, and I was able to cure it for at least one of them (a more recent SAMSUNG model) by accessing its secret service menu and applying some non-obvious settings I found in a forum someplace. Fortunately, I took notes on what I did, in case I ever need to repeat the process. And I also managed to track down the super-hard-to-find information about how to access the secret service menus in my other two TVs, even though they ultimately didn't expose settings to change that behavior. That kind of information is priceless, as is compiling a list of which LCD HDMI televisions actually support the slightly-non-standard-sync-rate trick that MiSTer uses for its "butter-smooth" vsync mode.

I lucked out with my 70" Sharp Aquos TV in my main room, as it seems to support a range of vsync frequencies centered around both 50Hz and 60Hz. But if I feed anything other than an exact 60Hz signal to my newer-model SAMSUNG in my guest bedroom, it will achieve sync and display the image, but the image will exhibit bizarre stuttering behavior (every so often, it will momentarily show a frame that is from 2-3 frames ago, such that any smooth motion appears to suddenly lurch backward for a frame). These are the kinds of unofficial compatibility that make or break using an HDMI display device, but unfortunately it's not part of any published specs... so the only real way to know if it will work before buying is to get verified information from other owners of the same model who have already tried it, or to take your MiSTer to a store, hook it up, set if for "butter smooth" sync mode, and then play it a while and see how the display handles it. I'm not sure how open to such a thing most stores or retail salespeople generally would be, and given the current COVID pandemic happening here in the USA, buying electronics physically in person at retailers isn't really a viable option right now.

Yeah I think that's what I need to do, to find an actual hidden menu is probably the only way, since it's same Make as the one you did. I suspect it's not dissimilar from the mess that is mfg's not complying with EDID standards.

Thanks a TON for taking the time to make this post!

My MiSTer completely breaks HDMI-CEC on my LG UJ6200 TV for all other devices connected to the TV. Meaning, my Nvidia Shield can no longer automatically turn on and off the TV via using the Shield remote, which is SUPER annoying.

I wonder if there's a cheaper version of the Lindy out there?

Confirmed — the CEC-Less cable cures the ARC / CEC problems.

So my setup now consists of the CEC-Less cable plugged straight into MiSTer, then plugged into the CableMatters repeater’s “input”, then an HDMI cable connected from the repeater’s “output” into one of my soundbar’s inputs.

Nytron wrote: ↑Sun Jul 12, 2020 7:43 pm I wonder if there's a cheaper version of the Lindy out there?

Yes — get the CEC-Less cable I linked to earlier.

keith.f.kelly wrote: ↑Mon Jun 01, 2020 10:13 pm

If you connect a display to your MiSTer via HDMI, here's some good summary info and advice.

Apparently, the DE-10 Nano's HDMI port is not electrically buffered or isolated from the rest of the main board in the ways you (and other consumer electronics devices) would normally expect. For example, apparently its 5V line is tied directly to the main board's 5V rail, such that if 5V is present on that pin of the HDMI port, then some portions of MiSTer's hardware can appear to "turn on" even while the MiSTer itself is not powered up or even connected to power! As such, MiSTer can exhibit a myriad of weird issues while hooked up to a display system via HDMI, especially if you are using long cables or running through an HDMI switch, AV receiver, HDMI capture device, or any kind of standard HDMI adapter/converter:

• decreased SDRAM stability

• display device randomly losing sync/signal from the MiSTer or taking an excessively long time to lock in sync after a resolution mode change

• MiSTer's cooling fan spinning, or its lights lighting up, even while MiSTer is powered completely off

• HDMI switch refusing to change inputs and/or randomly changing inputs while you're not using MiSTer and it is powered off

• USB dongles for wifi and/or Bluetooth not coming up and working reliably on every boot

• ...and possibly even other weird symptoms

What I have found is that the following inexpensive HDMI booster gadget works to effectively isolate/buffer MiSTer's HDMI output:

https://www.amazon.com/gp/product/B07PWYN3L4/

And it seems to work without supplying it any external power over its microSD port (i.e. the power produced by MiSTer's HDMI port appears to be enough to make it work). By installing one of these between my MiSTer's HDMI output and the HDMI switch I used in my home theater system, I was able to cure a combination of the above issues that I'd been beating my head against for months.

So, if you're using HDMI at all, it's probably just a really good idea to get one of these little booster devices and install it, to electrically isolate things. And if you're putting MiSTer inside of a larger case of some kind, it's probably an even better idea to install one of these inside of the case, so that the HDMI port exposed on the outside of the case is already electrically isolated.

Oh my God man,
i have to give you a big thank!
i didn't belive, but since my mister setup was degrading day after day, i tried this teeny tiny device.
AND IT DID THE MAGIC!
it cured these symthoms:
wifi magically started to work again (more to follow)
keyboard issues (sometime kaystroke missing, sometime keystroke repeated until another key is pressed)
joystick issues (player stuck in one direction until you move again in the stuck direction)
rtc! yes! exactly, rtc! i supposed the battery was dead since the clock wasn't even found. after installing this device, the rtc magically started to work again!
some syncing problems. i notice some cores displayed signal for some seconds then "unsupported resolution" appeared and the image disappeared. no more "unsupported resolution" error!
since my first concern was the wifi dongle that stopped to work one day, i made some tries with various windows machines. the wifi was ever working on windows 7 and windows 10 machines. with this hdmi booster now it started to work again (hope is not a coincidence) however, when installed on windows, the wifi led is blinking, but when installed on mister the wifi is rock solid!

When hooked up to my main gaming TV, I run through a ViewHD VHD-UHD4X1A switch. I struggle to recall any issues when running through it (wifi, 4x wireless controllers). So, it makes me wonder if it has a similar effect of electrically (isolating|buffering). I don't really have the equipment/training to tell for sure, though, so it's pure speculation on my part.