DIY Component --> VGA converter (Component transcoder)

[beun]

The DIY VGA to component transcoder is a very successful thread that is still going on in finishing the PCI version.

In the meantime I have received a number of requests for a Component to VGA transcoder. I think that we could start this design ending in the same results:

1) A freely available design database plus full description including PCB's for all forum members.
2) An assembled converter with warranty available from me.

At this time I would like to get some idea's on the options that people are interested in and I need to investigate how the design should look.

Please feel free to comment

[ChrisW6ATV]

This sounds like a great project idea. I have built and tinkered with circuits since the 1970's. I would be happy to help with testing of such a device. The 'options' that first come to my mind are actually 'features':
-Able to handle video from 15 kHz through 1920x1080
-Able to be powered by the VGA device, if there is power available on any of the HD15 pins

Thanks-

[beun]

Chris,

The first should be doable, the second I need to investigate. The VGA port is an output, so I doubt that the input on the other side has any power on it.

[Wireless]

One of the key things for HDTV compatibility is proper tri-level to bi-level sync conversion.

[beun]

yep, I agree. This is actually not too difficult, the most interesting problem is horizontal picture shift.


Kim

[stjr]

In addition to proper tri-level to bi-level sync conversion, the transcoder should be able to remove all traces of sync from the RGB output, or at least have the option available.

I would personally like to have a separate converter for converting tri-level sync to bi-level sync only, because my RPTV seems operates at its best using YPrPb+HV (5 inputs) with bi-level sync.

[beun]

steve,

Stripping all the sync from the RGB output is exactly the plan, now I just have to figure out how to do it.

When we have the sync stripped and converted it should not be too hard to make it available including a YPrPb pass through.


Kim

[steve]

(1) I would like to see two inputs- one that is compoent to be transcoded to rgbhv, and a second input that can be configured either as a second component to be transcoded, or as a rgbhv pass-thru. Switching between inputs would either be via ir or rs-232.

(2) Able to handle 480i, 480p, 720p, 1080i and 1080p. In this regard, please note that I currently have a transcoder made by Current Design that transcodes well, but has an annoying quirk. I receive hdtv signals via an OTA stb in all formats, which is fed to a front projector. When the input signal switches from 1080i to 720p, the projector will not sync up, and I get an image on the left half of the screen and an identical image on the right side of the screen. Going from 720p to 1080i does not exhibit this problem. I know the quirk is with the transcoder, because when i connect the rgbhv output of the stb to the projector, this problem does not occur.

(3) selectable tri-level or bi-level output.

(4) ability to strip out 7.5 ire when a digital version of an ntsc signal is inputted.

(5) ability to strip out macrovision.

(6) compatibility with the jvc dvhs recorders.

Steve

[beun]

Is there a website on that Current design transcoder so I can look it up, I can only find kayaks under that name.

This is an interesting feature list that you would like to see and at first glance it will not be in the $100 range anymore.


Kim

[steve]

Current Design is an off-shoot of Progressive Labs, the maker of the CA-1 color analyzer. They never had a website, AFAIK.

Steve

[steve]

Progressive Labs no. is 201-342-1235. I think I may have contact info (e.g., who designed it) on Current design at home. I will try to find it.

Steve

[MrHifi]

This Colorimiter Walks on Water AFAI Concerned.

[Briands]

Originally posted by steve
(1) I would like to see two inputs- one that is compoent to be transcoded to rgbhv, and a second input that can be configured either as a second component to be transcoded, or as a rgbhv pass-thru. Switching between inputs would either be via ir or rs-232.

...

(Steve

I second this part with a bit more constraint...

It would be great if the RGBHV in and out were BNC and if the switch bandwidth were high enough not to kill the benifits of an MP1 moded card.

Would it be possible to have it auto switch with selectable priority and override.

By the way is Mike involved in this project at all? I know he discussed a transcoder design previously as well.

[beun]

Briands,


I don't know if Mike is reading these posts. BNC inputs is definately an option and bandwidth should end up around 160MHz or so.

Let me ask another important thing. How much should this thing cost in the end. Do we want to make the mother of all transcoders, just a simple one or one that can be expanded with different options?

[MrHifi]

I've spoken to several HT owner who have the same problem. Most of us own a Front Projector TV with RGBHV input. Several pass through a processor/doubler/transcanner that only proceses 480i input information. All of these units have a passthrough RGBHV directly to the projector. This passthrough is a selectable input via a remote control. In this case, it is the only means for providing HDTV info to a data grade monitor. Several of us have RGBHV switchers to select between different RGBHV HDTV inputs. The switcher's RGBHV output goes to the RGBHV input/passthrough of the Transcanner. In order to accomodate newer STB's, PVR's and Progressive scan DVD Players, a remotely switchable transcoder to process YPbPr to RGBHV is required. The switch would select between the RGBHV input to the unit for passthrough to its RGBHV output or would select the YPrPb input for transcoding to its RGBHV output. For upgrade, the unit could be offered with 1,2,or3 YPbPr inputs and the RGBHV passthrough. Please keep it simple. Plug it in and select. Try to keep dip switch setting to a minimum. Thankyou!! I'll offer to buy or test the first unit if you need an old physicist whose played with this stuff for 50 years.

[ChrisW6ATV]

beun-

I would rather see a simple, small, low-cost device. Keep any multi-input or auto-switching circuits as a separate project.

I do not know what the parts for the RGB-to-component project cost, but I'd like to see this one be less than US$50 not including case or power supply, but including the circuit board (and maybe connectors). The one 'option' that I'd like is separate versions/boards/layouts for multi-BNC outputs versus HD15 ("VGA connector") output. Thanks!

[ChrisW6ATV]

Update-

I had only glanced at the RGB-to-component discussion at its beginning and I had seen a simple schematic with a few TTL chips and op amps. Now, I have seen the Crescendo product and schematic-it is obviously a much more complex device than in the beginning. So, the $50 price is probably not likely, but I am interested anyway.

[Wireless]

Mike has his own transcoder that he is putting together. I should have one to test on Saturday. Based on Mike's previous work, I expect it will be a higher end unit.

[beun]

When Mike already has a transcoder that does everything there may not be a need for a new one.

I will however see if can make some time today to draw up a simple block diagram that captures the ideas from your recent postings. It will consist of a basic transcoder that does the sync-stripping and the color matrix ( both of them) with optional input and output switching. I will try to avoid a microcontroller (programming issues) but if anyone would like to see remote control that is probably unavoidable.

[mp20748]

Originally posted by beun

When Mike already has a transcoder that does everything there may not be a need for a new one.

Yes there could still be a need for what you're doing. I'm doing something a little different. I'm putting emphasis on performance, so I've not considered some of the request that are listed on this thread. My unit will only have two component inputs, but all in's and outs will be true 75 ohm BNC's. And the inputs will be selected from a front panel switch.

I'm not into bells and whistles. So I've spent my time tweaking and tuning for performance, so my unit will be very straight forward.

However, I really like what you've done with the other transcoder. And it's a blessing to have someone like you around to make things like this available to many. And of course I really like how you've made it available to those who could not assemble one themselves.

Keep up the good work !

[DoyleS]

Here is a site with some schematics that show a couple different versions of YUV to RGB conversion.
http://elm-chan.org/works/yuv2rgb/report.html
Some more complex than others. There are also a lot of new high bandwidth op amps out there designed for these video applications with built in sync strippers.
http://www.intersil.com/data/an/an9513.pdf
From my perspective it would be important to have the transcoder either calibrated or to have the ability to calibrate the RGB output properly. In the end, the reason we are doing this is to get the black levels correct. My STBs output RGB and Component, if indeed the component is the most accurate of the OTA and Satellite HD formats, then accuracy of that conversion to RGBHV is critical.
..Doyle

[DoyleS]

Hmmm.....

Seems this thread went dead. Are we still thinking this is a good idea? I for one am up for a kit or at least PCB of a good Component to RGBHV transcoder. I can't do the > $200 dollar deal.

..Doyle

[MrHifi]

I hope it is not dead. I have put off buying a ready made transcoder expecting the gentleman who offered to build one and sell it to do so. I'll pay up to $300 for remotely switchable Component to RGBHV with RGBHV passthrough. as I discussed in my previous posting.

[beun]

No the thread is not dead, it is just that I have a lot to do and so little time to do it.

What I would be proposing is the following.

1) make a modular system, with the component->VGA transcoder as the core. This one will do the sync-stripping and the color conversion. There will be dipswitches or so to select the color matrix. This one can be had separately.

2) Create a front-end switcher. Depending on the requests it can be relatively simple but high quality using relays, or it can be a fancy (and expensive) cross-point switch that will have slightly reduced bandwidth (200MHz for the simple one vs 100MHz for the cross-point).

3) Create a backend switcher for those that want to do fancy things with the recovered sync.

4) Anything fancy requires a microcontroller as the IC's to use need to be serially programmed. This microcontroller can than also be used to receive remote control.

This design is going to be a fairly lengthy process. The core transcoder probably needs 3 months of design and first round printed circuit boards. I myself am very bad in programming microcontrollers so I would need some help there .

[MrHifi]

This is rather disheartening. I guess those readily available transcoders for $199 aren't such a bad deal.

[Chuchuf]

John,

So youare getting MP's transcoder this weekend?? Can't wait to see it!!

Terry

[beun]

Art,

The transcoder core which will be comparable to the AA or KD transcoder and put in its own case can be designed and made for a price that will be very competitive. It is all the switching functions that drive up the costs.

[gn2]

I own an AA transcoder, you do not want one that performs so poorly. Snych stripping and accurate color output should be the core, so true HD black levels can be maintained. Its sooo disheartening to have an HD source and not be able to use it because there is no affordable transcoder that can adequately display the signal.
My vote, both in words and with real money, is for a very basic transcoder that outputs an RGB signal with all of the correct levels of synch(ie. none on the RGB) and correct color signals in order to result in a fully detailed HD image on my front projectors. There are TONS of good switchers already available on ebay for very little, we need a good transcoder ONLY.
If MP can sell his for $300. or less, I'll buy one of his. But I would also support this effort, as it doesn't sound like MP is heading anywhere near that price point.

[Wireless]

I've had the MP transcoder for a few days and my initial impression is very very positive, much better than the inexpensive units.

[gn2]

That is GREAT news, but only if he'll be able to sell it for an affordable amount. Like a Faroudja scaler, it doesn't really matter if its the best if I'll never be able to afford one. Any word on pricing or availability yet ?