Simple 1-channel RGB amplifier circuit.

Go To Last Post
14 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

A friend of mine collects old video game consoles, and has a NEC Duo with only a composite output. He wanted a RGB output to plug into his SCART switch, and bought a kit from some French company before he talked me into wiring it up. The kit is a small board and a couple of caps plus a piece of double sided tape, and comes with instructions on an amateur website that I know for a fact is at least partially wrong, for the small sum of around 90$ with shipping (which is ridiculous).  I did wire it up as per instructions, but the board quality was horrible (and he paid a small fortune for it!). Every exposed pad for input and output is SMT, copper is super thin, and every pad I tried soldering to ended up lifting in seconds... I did succeed in wiring it by soldering to the onboard through hole caps, and got everything wired nicely inside the casing, but the final result, while it does work and colors do look fine, is full of ghosting to the point where you can't read text on screen. I checked my connections, and added some caps on the output, redid the wiring with twisted pair, but the result is always the same. At this point I think the dubious board is the problem, and I am determined to fix him up a good working replacement. Now I am vaguely aware of what RGB signals look like, but I am wondering if there is any kind of tried and tested RGB amplifier circuit you guys could suggest. TI used to make nice little chips like the LM1203, but they are discontinued and they list no replacement part. A search on digikey for other RGB amplifiers turned out empty. I will be grateful for any leads you could give me... Note that this is not for high-resolution (the console generated 320x240 256 colors AFAIK), but it does need to be able to drive a high resolution LCD. It does not need to be perfect (let's face it it's a 25 years old console), but it needs to be cheap to build, either from discrete components or a chip, small and work properly.

Last Edited: Fri. Jul 10, 2015 - 05:44 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

How about the TI THS7316DR  (HDTV VIDEO AMP 3CH 8-SOIC).  

  

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

You know I did a little digging and found another version of this chip, the THS7314DR, in SOIC8... Not sure what the difference is between them, but I think the 7314 has less bandwidth (advertised for SDTV only, not HDTV like above), and in my case that is fine since the source is low res. The schematic above doesn't look quite right though (no DC blocking or biasing of the inputs? and no DC blocking on the output? Not sure about YPbPr but pretty sure you can't have DC on the output signal for RGB, or am I mistaken?). This chip actually looks great, I'll tinker around a schematic and post it here a little later.

Last Edited: Fri. Jul 10, 2015 - 05:42 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I haven't used the THS7316.  I ran across it a while back when looking for a wide band amp.

The datasheet states that it can be used in both DC & AC coupled input configurations.

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

"Ghosting" can be from several causes:

   1. Mismatched delay between the color channels - seems unlikely in this case

   2. Ringing and overshoot - pretty likely

   3. Poor power bypassing - I don't see any in the schematic

   4. Poor grounding, especially loops  involving the nonexistent bypass cap and the output. Each output sees a load of 150 ohms so there can be transients of several 10s of mA and thats enough to make problems.

 

Jim

Jim Wagner Oregon Research Electronics, Consulting Div. Tangent, OR, USA http://www.orelectronics.net

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Thanks Jim and David...

 

David: The console is actually a DUO, so that mod is not the same (this one doesn't have an expansion port like the FAMICOM or basic PC Engine; stuck taking RGB signals directly off the 6260. In any case the mod is really all done, the only problem is the amp itself.

 

Jim: The schematic up there is not the actual board I am using. This one has all the necessary decoupling. Ringing and overshoot seems pretty likely though... This board he bought for 60$ + shipping has a chip on it, but I can't identify it (no markings). It seems to be missing a few caps and resistors too, but I can't really know for sure. The new one (which will cost less than 10$) has internal filters and overshoot protection. So it should be an order of magnitude better.

 

EDIT: This should work nicely... I used the output stage component values off another RGB amp appnote, and from the few I checked it seems to always be pretty much the same for a 5V VCC. For the input stage, the caps don't really matter (just DC blocking) and the pull ups are values I saw on a reference design for the chip.

 

Last Edited: Fri. Jul 10, 2015 - 05:38 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

By the way this is how the mod looks like when completed, with the bad amp:

 

http://ddata.over-blog.com/2/11/04/71/OTAKUS-STORE.NET/TUTO/NEC/global.jpg

 

Before you ask, yes I have no choice about the wire length, the RGB signals are off a chip on the backside of the board. Wiring is not the issue I think, I redid this mod 3 times with various placements and wires (my first guess was cross talk or EMI), and the ghosting was pretty much identical every time. There is also a limited amount of space in the casing once reassembled, so I cannot put the board anywhere I want.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I would use shielded cable for the RGB - ideally run via triple co-ax http://www.cablesforless.com/Belden-7794A-High-Quality-RG59-Three-Coaxial-Conductor-Cable-By-The-Foot-P6833.aspx.  BTW the second YouTube link is the DUO but references the mod in the first.

 

<edit> the co-ax should eliminate impedance mismatch which causes reflections, i.e. ghosting.

 

David 

Last Edited: Sat. Jul 11, 2015 - 07:59 AM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Like I said I tried many different tactics, from simple wires to a GND interleaved ribbon to three twisted pairs, and I tried moving the board and wires around the mainboard, but it did zero difference. I don't think the problem is the wire at all... The other issue with the wire is that it has to be thin and flexible, I am soldering directly off TSSOP pins on the other side of the board, anything too rigid will stress the solder joints too much.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Have you looked at the signals on a scope?  That'll give you an idea of what is causing the ghosting if it is as bad as you said it could be you are overdriving the inputs to the amps.

 

David 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

No unfortunately I did the work at his place, didn't have my proper tools.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Well, it turns out the problem came from his SCART cable... Cheap cable used unshielded wires for video inside, instead of coax.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Glad to hear it's sorted.

 

David