gscartsw / gscartsw_lite / gcompsw switches support thread

[superg]

You said in july you emailed me. You just made up an email address and put it in. Then this week you said this: \

I never mentioned July. Yes I've send the initial email on October 15 to chris@windygamingchicago.com which turned out to be a fake address, I explained that. Usually when people post emails they are trying to obfuscate it to protect from automatic collection and spam. "chris @ windy gaming" seemed like a legit email.

Thats fiction. The email you did send was sent on monday as a forward from the initial bad email you sent me. you did not send anything a couple of times last week. Thats just lying. So maybe 'less than honest' was me being generous. I'm straight calling you a liar now. Now if you know how email works, you cant forward an undelivered mail. Once we found out about the email situation, Yes we emailed twice.

(how I forwarded "undelivered" email response from gmail)


(how you acknowledged the fact that you've actually got that "undelivered email" in your Junk folder)

That last one you put up?

you know the only one in that chain that doesnt have any email info on it (time stamp and whatnot)? yeah. you never sent that. you just added that to the chain for the sake of your response post. You never replied to my last email. I'm sure you'll have time to "type up" a time signature since you arent using screen caps though.

Like i said. I'll drop screen caps when i get home. rather than a mass copypasta . You need to learn how email works dude. Honestly I'm have a hard time running an honest business while being dragged through the mud by a straight up liar.

And finally the whole email conversation:


I've just posted all the conversations I've had with you. I may seem unfriendly at times, but I'm definitely not a liar.
You've exposed yourself, the community will make the conclusions. Your reputation have been affected and you made it even worse by accusing me of something I haven't done. While I always say that modded consoles are unsupported I tried to resolve the issue with your arcade conversion (and for the benefit of the retro community), you became defensive right from the start, leaning back and forth, explaining the miscommunication issue and correcting everybody that it's not "Windy City" but "Windy Gaming" instead of getting straight to the point. I offered to send you the unit and the screen caps are here, you haven't responded. Then you called me a liar and said that I made up that last email which is totally untrue.
I politely ask you to leave my thread and I wish your business to improve. This have been over the top and I'm not going to continue this discussion with you.

[DirkSwizzler]

cdamm / Windy gaming:

You said you haven't logged in for a long time. You should be aware that you're entering a community where SuperG has a very long track record of being an upstanding member of the community. Doing great work.

Even if the problem were on the switch side your tone is that of someone who doesn't handle negative feedback very well.

While I'm not an EE. I have been software systems engineer for over a decade. Discussing what the problems might be is no place for such heated emotion and mud slinging.

For both:

Anecdotal "it works with everything else" generally isn't an acceptable argument on it's own for pointing fingers. It's really only useful to narrow down "why is this different?" quickly. And it sounds like that's been achieved. But cdamm, testing against other switches is not useful if the problem has already been narrowed down.


From my non-EE external read of the problem. It sounds like Windy Gaming device was not specifically designed for SCART. But instead has that as an option. While the gscartsw was most definitely designed for SCART. And it sounds like people are converting the device to a SCART connection without bringing the signal into spec.

If that's a correct read. Then neither device is directly at fault, and the users are for using the wrong settings on the supergun or the wrong cable. Fine enough as a baseline of how to take action.

At this point I have to admit that I'm biased toward thinking the gscartsw is an awesome and well designed product. So the rest of this subjective data is probably not as useful, but maybe there's something useful in there.

I believe SuperG has been clear enough in the past that his device only supports the SCART standard. And the frequent topic of CSYNC being out of spec is explicitly not supported. So I think the overall messaging is clear there. If there were to be any change, I would ask that something be printed directly on it's shell to remind users that modifications can be harmful to their equipment and they should triple check that everything going into the switch is well designed.

It would be nice if it had protections, but I believe SuperG has also been very clear about cutting costs wherever possible. So the logic seems sane for omitting them as well.

I believe the Windy supergun is probably providing the features users need to bring the output within the range of the SCART specifications. But easily allows users to output something that is out of spec. So the direct fault in that case would be on the users. But having looked at the device itself. I could easily see myself leaving it in the wrong configuration. Especially since there isn't a label near the switch itself. Would it be easy to move the attenuation switch to the front of the box and label the positions? Or possibly make a separate SCART output on the back? Users sometimes need someone to save them from themselves.

EDIT: I also personally believe that adding as much protection to the gscartsw products would be advantageous to my setup. I know SuperG has stated he's not doing any more design this year. But I'd definitely be interested in a more expensive "save me from my own dumb decisions" gscartsw

[clintkolodziej]

cdamm / Windy gaming:

You said you haven't logged in for a long time. You should be aware that you're entering a community where SuperG has a very long track record of being an upstanding member of the community. Doing great work.

Even if the problem were on the switch side your tone is that of someone who doesn't handle negative feedback very well.

While I'm not an EE. I have been software systems engineer for over a decade. Discussing what the problems might be is no place for such heated emotion and mud slinging.

For both:

Anecdotal "it works with everything else" generally isn't an acceptable argument on it's own for pointing fingers. It's really only useful to narrow down "why is this different?" quickly. And it sounds like that's been achieved. But cdamm, testing against other switches is not useful if the problem has already been narrowed down.


From my non-EE external read of the problem. It sounds like Windy Gaming device was not specifically designed for SCART. But instead has that as an option. While the gscartsw was most definitely designed for SCART. And it sounds like people are converting the device to a SCART connection without bringing the signal into spec.

If that's a correct read. Then neither device is directly at fault, and the users are for using the wrong settings on the supergun or the wrong cable. Fine enough as a baseline of how to take action.

At this point I have to admit that I'm biased toward thinking the gscartsw is an awesome and well designed product. So the rest of this subjective data is probably not as useful, but maybe there's something useful in there.

I believe SuperG has been clear enough in the past that his device only supports the SCART standard. And the frequent topic of CSYNC being out of spec is explicitly not supported. So I think the overall messaging is clear there. If there were to be any change, I would ask that something be printed directly on it's shell to remind users that modifications can be harmful to their equipment and they should triple check that everything going into the switch is well designed.

It would be nice if it had protections, but I believe SuperG has also been very clear about cutting costs wherever possible. So the logic seems sane for omitting them as well.

I believe the Windy supergun is probably providing the features users need to bring the output within the range of the SCART specifications. But easily allows users to output something that is out of spec. So the direct fault in that case would be on the users. But having looked at the device itself. I could easily see myself leaving it in the wrong configuration. Especially since there isn't a label near the switch itself. Would it be easy to move the attenuation switch to the front of the box and label the positions? Or possibly make a separate SCART output on the back? Users sometimes need someone to save them from themselves.

Yeah, sounds like this was an error made by a user (or users) who either bumped the switch on back, didn't take the time to understand what it was for (and what the implications would be of having it ON/OFF), or knew what it was for and was trying to extract amplified audio after the switch from the SCART cable to power arcade speakers, which is out of the SCART spec and damaging to the switch. In any case it doesn't sound like it was a direct issue with any hardware, more of a user-friendly issue. I'm a programmer as well and users never read instructions, so the quote "Users sometimes need someone to save them from themselves" rings true to me for sure.

At this point, the argument on who said/did what isn't going to help future users from the same issues, but since it sounds like the issue was narrowed down to high voltage audio through SCART the productive things would be to:

A) Better document (stickers on back, updates to manuals, websites, etc) each device and how they interact, and clearly mention what is or is not supported, as was mentioned before

B) Perhaps in future revisions of each product add enhancements to avoid future issues and headaches such as this (more protection on gscartsw for out-of-spec voltages, limit SCART output either via included cable or directly on supergun as DirkSwizzler mentioned)

C) Work together, or with community members (like Voultar) who were trying to assist how to remedy this for users of current hardware (mod the SCART cable with a resistor or some other workaround)

I get how everyone is getting pissed off, this is causing work that no one really wants or has time to do, and people's products and reputations are at stake. Really, as an end-user I'd just prefer that issues are productively worked out, as I don't really care about all of that, I'd be more concerned to know what I need to do to prevent my hardware from having issues. We've gone through this time and time again as DirkSwizzler mentioned with out-of-spec CSYNC, TTL vs 75Ohm, wonky sync signals, etc, but in the end we just want our stuff to work.

Anyway, sh#t was said back and forth already, but hopefully we can turn the page on this and move forward to help out the affected users, and help others that have the same or similar hardware know how to avoid this.

[tusecsy]

Windy City Gaming.....wow you're an asshole bro. And that says a lot coming from me.

It's pretty obvious at this point whats going on here (your product is shit). Why haven't you just posted a video of the oscilloscope as requested from the beginning? Seems like an easy fix no? Or are you afraid you'll be exposed for the fraud you are? Fear the whole house of cards is about to collapse, douchendance style?

but on the flipside...

@superg, still no answer on voultars question about the diodes, i'm guessing your product doesn't have them, so why doesn't it?....your dodginess isn't helping your case here.

Does the Windy City Gaming S.G. have protection diodes or any kind of protection on there? The same question can be asked for the GSCART switch, too.

[DirkSwizzler]

@superg, still no answer on voultars question about the diodes, i'm guessing your product doesn't have them, so why doesn't it?....your dodginess isn't helping your case here.

Can't speak for SuperG. But I'm assuming it's because it would need 8 sets of them and he's always been about keeping costs at a minimum. That's just my guess though. I know I personally would pay more for them if that were the only reason.

[Ruprit]

@superg, still no answer on voultars question about the diodes, i'm guessing your product doesn't have them, so why doesn't it?....your dodginess isn't helping your case here.

I think Superg already answered that question.

Does the Windy City Gaming S.G. have protection diodes or any kind of protection on there? The same question can be asked for the GSCART switch, too.

There is nothing before multiplexer, different multiplexers are used in gscartsw v3.4 and gscartsw_lite, v3.4 ones clamp -2.5V..+2.5V range and lite -5V..+5V range. Lite has sync 600OHm resistor after the multiplexer for additional protection. I had both v3.4 and lite switches for repair from the same Windy City supergun.

[clintkolodziej]

Windy City Gaming.....wow you're an asshole bro. And that says a lot coming from me.

It's pretty obvious at this point whats going on here (your product is shit). Why haven't you just posted a video of the oscilloscope as requested from the beginning? Seems like an easy fix no? Or are you afraid you'll be exposed for the fraud you are? Fear the whole house of cards is about to collapse, douchendance style?

but on the flipside...

@superg, still no answer on voultars question about the diodes, i'm guessing your product doesn't have them, so why doesn't it?....your dodginess isn't helping your case here.

Does the Windy City Gaming S.G. have protection diodes or any kind of protection on there? The same question can be asked for the GSCART switch, too.

LOL, don't fan the flames dude!

If you want your answers this needs to calm down a bit so we can be productive and work together to solve it. I agree some scope captures would help see what the output is from the supergun, and superg or someone like Voultar that knows how to read them would need to take a look to see if it can be made more foolproof (if the switches are misconfigured) for use on the gscartsw (or other devices), as well as confirm the hypothesis that the cause is the audio attenuation switch.

Regarding the doujindance comment, I get the joke about his old work, but he took the feedback from Voultar and other members of the community and has really improved his work. I'd say he is a perfect case for how to take criticism and improve your product.

[superg]

@superg, still no answer on voultars question about the diodes, i'm guessing your product doesn't have them, so why doesn't it?....your dodginess isn't helping your case here.

I think I answered that here:

Does the Windy City Gaming S.G. have protection diodes or any kind of protection on there? The same question can be asked for the GSCART switch, too.

There is nothing before multiplexer, different multiplexers are used in gscartsw v3.4 and gscartsw_lite, v3.4 ones clamp -2.5V..+2.5V range and lite -5V..+5V range. Lite has sync 600OHm resistor after the multiplexer for additional protection. I had both v3.4 and lite switches for repair from the same Windy City supergun.

To summarize there is currently nothing that will prevent switch from dying if amplified audio will be connected.
Why it doesn't have such protection? - Because first of all it goes against SCART specifications. Currently I have a polarity protection and overvoltage protection (in a sane range obviously) for power supply.
Second, even if I would implement such audio protection that will have to be added before the multiplexer for both channels of audio which means 3 (resistor capacitor and diode) * 2 (stereo) * 8 (inputs) = 48 new board components. That will affect components price and assembly price. Obviously I don't want every owner to pay $10 extra for something that can hardly be called a feature. Moreover this kind of protection is just to protect the components, this won't fix the audio and you will hear garbage. That's what I think about it at this point. Yet I might consider it if I will be making some other changes to lite design, have to calculate the impact anyway.

[clintkolodziej]

@superg, still no answer on voultars question about the diodes, i'm guessing your product doesn't have them, so why doesn't it?....your dodginess isn't helping your case here.

Can't speak for SuperG. But I'm assuming it's because it would need 8 sets of them and he's always been about keeping costs at a minimum. That's just my guess though. I know I personally would pay more for them if that were the only reason.

I'd pay a few bucks extra to make sure my dumb @$$ doesn't connect something without thinking and fry my almost 200$ investment. That surely would save time and headache down the road from this same thing happening again.

[clintkolodziej]

I'd pay a few bucks extra to make sure my dumb @$$ doesn't connect something without thinking and fry my almost 200$ investment. That surely would save time and headache down the road from this same thing happening again.

With that said I still plan on picking up a _lite from the next batch, as long as my finances allow and my wife doesn't kill me for buying more retro gear.

[clintkolodziej]

Moreover this kind of protection is just to protect the components, this won't fix the audio and you will hear garbage.

Garbage in, garbage out. I think at least in this particular case someone would hear a bunch of garbage and get an idea that something isn't set up correctly so they can debug it, rather than just blowing out the switch. Anyway, it's up to you, I think I'd pay for the peace of mind but I can only speak for myself here.

[superg]

While I'm not an EE. I have been software systems engineer for over a decade. Discussing what the problems might be is no place for such heated emotion and mud slinging.

I'm a software engineer for over a decade as well. I'm not considering myself an electronic engineer, if not expert help of my dad I wouldn't be able to complete such a project myself. That said I always appreciate viletim and voultar feedback here (and calpis on ASG forums) and I know so much more now comparing to what I knew when I started.

Anecdotal "it works with everything else" generally isn't an acceptable argument on it's own for pointing fingers. It's really only useful to narrow down "why is this different?" quickly. And it sounds like that's been achieved. But cdamm, testing against other switches is not useful if the problem has already been narrowed down.

It reminds me "works on my machine!"

[DirkSwizzler]

Anecdotal "it works with everything else" generally isn't an acceptable argument on it's own for pointing fingers. It's really only useful to narrow down "why is this different?" quickly. And it sounds like that's been achieved. But cdamm, testing against other switches is not useful if the problem has already been narrowed down.

It reminds me "works on my machine!"

For real.

[clintkolodziej]

Anecdotal "it works with everything else" generally isn't an acceptable argument on it's own for pointing fingers. It's really only useful to narrow down "why is this different?" quickly. And it sounds like that's been achieved. But cdamm, testing against other switches is not useful if the problem has already been narrowed down.

It reminds me "works on my machine!"

For real.

Gotta clear your cookies and try again. Sorry, web developer humor.

[DirkSwizzler]

It reminds me "works on my machine!"

For real.

Gotta clear your cookies and try again. Sorry, web developer humor.

Guys, I've done a lot of very thorough analysis. The problem is that the gscartsw doesn't have an on/off switch.

[chromium]

Update: On a bright note the current lite batch is finally dispatched and will be here on Monday!

Great does this mean gcompsw is coming soon?

Checking this thread every day for news on it, is there a better way?

[superg]

Update: On a bright note the current lite batch is finally dispatched and will be here on Monday!

Great does this mean gcompsw is coming soon?

Checking this thread every day for news on it, is there a better way?

Let me put lites on sale first and ship some units, then I will be planning gcompsw.
I have to come up with some sort of mailing list really, I realize it's annoying for everybody to go through all this drama in topic, I'll think about it.

[tusecsy]

Update: On a bright note the current lite batch is finally dispatched and will be here on Monday!

Great does this mean gcompsw is coming soon?

Checking this thread every day for news on it, is there a better way?

Let me put lites on sale first and ship some units, then I will be planning gcompsw.
I have to come up with some sort of mailing list really, I realize it's annoying for everybody to go through all this drama in topic, I'll think about it.

thanks for answering the diode question i'm not well versed in these things so didn't realize you answered it. just didn't want another krikzz situation on my hands where i have to sell off and rebuy things down the road. i'll shut up on the subject now.

as for the mailing list, twitter would do it. i know it's largely garbage but a simple account we could follow just so we get notified when things go up would be great.

i'd also pay an extra 10-20$ for piece of mind, i think most people willing to pay 200$ for a scart switch wouldn't mind the added cost. but that said it's not the end of the world and i'm def in on the next order.

[DirkSwizzler]

I have to come up with some sort of mailing list really, I realize it's annoying for everybody to go through all this drama in topic, I'll think about it.

I hate to admit it. But I really liked Matt Buxton's use of Telegram for the OSSC updates. Basically the same amount of signal as following someone on twitter, but much less noise.

[clintkolodziej]

Update: On a bright note the current lite batch is finally dispatched and will be here on Monday!

Great does this mean gcompsw is coming soon?

Checking this thread every day for news on it, is there a better way?

Let me put lites on sale first and ship some units, then I will be planning gcompsw.
I have to come up with some sort of mailing list really, I realize it's annoying for everybody to go through all this drama in topic, I'll think about it.

As Tusecsy mentioned a twitter account might be easiest, but that may risk people tweeting issues at you there instead of here. I guess you could be clear that it is just for broadcast messages on inventory going for sale though and not respond to messages there.

A mailing list would be an upgrade over twitter, but many out there either aren't free, or only allow a small number of users on the list before pushing you to a paid tier, so if you want free then twitter might be best.

[chromium]

As Tusecsy mentioned a twitter account might be easiest, but that may risk people tweeting issues at you there instead of here. I guess you could be clear that it is just for broadcast messages on inventory going for sale though and not respond to messages there.

A mailing list would be an upgrade over twitter, but many out there either aren't free, or only allow a small number of users on the list before pushing you to a paid tier, so if you want free then twitter might be best.

The GDEMU guy uses wordpress and gives updates. I just subscribe to the wordpress page and get emails whenever they post. Seems to work

[Voultar]

Nobody needs to get bent out of shape about this. It's a simple flaw that can easily be remedied with a few disclaimers and customer awareness.

Jamming +12V audio through the same connector though might not be the best approach. Someone could easily flip that switch, and there she goes banging off the rails. Maybe dedicate a port that's exclusively used to drive speakers, directly?

At any rate, it's easily solvable. It's going to be OK!

[DirkSwizzler]

Maybe dedicate a port that's exclusively used to drive speakers, directly.

Yeah. That seems the sanest the more I think about it.

This is just an opinion from looking at the thing for 2 minutes. I think if I were to own one of these, I'd want BNC connectors for the outputs instead of mini-din. BNC seems like a good baseline for adapting to whatever your A/V wants.

And if it already had BNC (or BNC+RCA), it'd be trivial to have separated audio outputs with big scary labels on them for driving speakers directly.

[Ryoandr]

To summarize there is currently nothing that will prevent switch from dying if amplified audio will be connected.
Why it doesn't have such protection? - Because first of all it goes against SCART specifications. Currently I have a polarity protection and overvoltage protection (in a sane range obviously) for power supply.
Second, even if I would implement such audio protection that will have to be added before the multiplexer for both channels of audio which means 3 (resistor capacitor and diode) * 2 (stereo) * 8 (inputs) = 48 new board components. That will affect components price and assembly price. Obviously I don't want every owner to pay $10 extra for something that can hardly be called a feature. Moreover this kind of protection is just to protect the components, this won't fix the audio and you will hear garbage. That's what I think about it at this point. Yet I might consider it if I will be making some other changes to lite design, have to calculate the impact anyway.

First, the following is not blaming on SuperG. I don't have one of his products, but given the praise and feedback, I can only have much respect for him.

Now, about input protection, nearly every commercial device has it, even if it's planned to be used only at a max level, because, quite simply, "shit happens", even more with sound since their levels can easily go up. You can go in the RGB modding thread and see various TVs schematics, everyone has some input protection, being a diode or a zener. The JVC add-in cards for monitors also have input protection despite being a professional device for a (normally) controlled environment.
The cost and complexity concern is also perfectly valid, even more with such low production numbers.

So, while protection on all inputs is understandably out of question, how about this :
have 1 input fitted with input protection, with a LED after the zener that would light if overdriving occurs, and have clear instructions (a label maybe) to first test your devices on that port. You might need some components to have that LED clamp for a visible duration, but I'm pretty sure it would be doable.

[FBX]

Jamming +12V audio through the same connector though might not be the best approach. Someone could easily flip that switch, and there she goes banging off the rails. Maybe dedicate a port that's exclusively used to drive speakers, directly?

That's pretty much what I was getting at. Having high voltage audio become active through the same din port with a flick of a toggle switch seems too risky to me. It should be routed separately and clearly labeled as such.

[thesteve]

a few things just to sum up
it appears the sound is the only possible culprit
the chips are clamped against over voltage
some of the chips actually exploded indicating high current
there is a switch to attenuate that is likely related




so now a solution on the side of the supergun
add a resistor to each sound pin
the sound channel is high impedance so even a 1k resistor should work fine

[superg]

Now, about input protection, nearly every commercial device has it, even if it's planned to be used only at a max level, because, quite simply, "shit happens", even more with sound since their levels can easily go up. You can go in the RGB modding thread and see various TVs schematics, everyone has some input protection, being a diode or a zener. The JVC add-in cards for monitors also have input protection despite being a professional device for a (normally) controlled environment.
The cost and complexity concern is also perfectly valid, even more with such low production numbers.

Of course mass produced consumer electronics have all kinds of protections. I'm not on that scale and will never be. A niche product for a small group of people. The protection will benefit me first of all because I will have less to repair, but it's charging people more and wasting my time and one more switch iteration and version bump.

So, while protection on all inputs is understandably out of question, how about this :
have 1 input fitted with input protection, with a LED after the zener that would light if overdriving occurs, and have clear instructions (a label maybe) to first test your devices on that port. You might need some components to have that LED clamp for a visible duration, but I'm pretty sure it would be doable.

I fully agree with the fact that protection is useful.
I can hardly believe that owner who toggles all kind of switches without understanding what they do will follow the instructions. What you propose is not as simple as you think.

I'm a bit tired of the situation right now.
Long story short - nothing will be done with the switch revision at this point, it's not a gscartsw fault and I won't be complicating the design just because of some conversion mishandling of one person. There will be a big red warning on the order page regarding the situation with Windy City SuperGun conversion. Enough is enough.

[thesteve]

you have input protection (diode clamping), so all that needs to happen is some resistance
should be done in the cable but would be an easy hack on the switch as well
doesnt require any change to the supergun and a small change to the cable

[tusecsy]

Now, about input protection, nearly every commercial device has it, even if it's planned to be used only at a max level, because, quite simply, "shit happens", even more with sound since their levels can easily go up. You can go in the RGB modding thread and see various TVs schematics, everyone has some input protection, being a diode or a zener. The JVC add-in cards for monitors also have input protection despite being a professional device for a (normally) controlled environment.
The cost and complexity concern is also perfectly valid, even more with such low production numbers.

Of course mass produced consumer electronics have all kinds of protections. I'm not on that scale and will never be. A niche product for a small group of people. The protection will benefit me first of all because I will have less to repair, but it's charging people more and wasting my time and one more switch iteration and version bump.

So, while protection on all inputs is understandably out of question, how about this :
have 1 input fitted with input protection, with a LED after the zener that would light if overdriving occurs, and have clear instructions (a label maybe) to first test your devices on that port. You might need some components to have that LED clamp for a visible duration, but I'm pretty sure it would be doable.

I fully agree with the fact that protection is useful.
I can hardly believe that owner who toggles all kind of switches without understanding what they do will follow the instructions. What you propose is not as simple as you think.

I'm a bit tired of the situation right now.
Long story short - nothing will be done with the switch revision at this point, it's not a gscartsw fault and I won't be complicating the design just because of some conversion mishandling of one person. There will be a big red warning on the order page regarding the situation with Windy City SuperGun conversion. Enough is enough.

Well, the gscartsw is technically at least part at fault. If you had the diodes it wouldn't be an issue. You also wouldn't have had to go through all of this drama and repairs, and warnings, etc. I really don't think saving people 10-20$ is a justifiable reason. Nobody would complain about the bump for safety.

If it comes down to laziness or unwillingness that's totally fine, but understand a lot of us in this community are perfectionists, and little things like this do irk us, even if "technically" it will never matter. Just look at the recent krikzz controversy, not a single person even had their everdrives or systems damaged, just the potential set everyone off.

[superg]

Just look at the recent krikzz controversy, not a single person even had their everdrives or systems damaged, just the potential set everyone off.

What is that krikzz controversy you keep mentioning about? Can you PM me some links, I'm genuinely curious, never had a single problem and I own almost all of his stuff.