Hydra Scart Switcher (and OSSC tests)

I have lots of systems to connect via RGB scart, at least 15 at the last count. I wanted a single display format (HDMI), so I have everything going through an OSSC Scaler; I needed a multi-way scart switch in front of it in order to handle all of the systems I wanted to connect up.

I chose to go with the Wide Hydra from Lotharek, which at first glance looks to be an excellent solution.

However, I've since found that it is extremely sensitive to both the type of scart cable you use, as well as the type of signal being sent over the cable; specifically the voltage of the video sync line (scart pin 20).

The device is designed closely to the scart specification which lists a voltage of 0-1v for most of the lines… however lots of systems don't appear to be anywhere near this specification unless you make some modifications to the video signal they're outputting.

As a result, virtually none of my systems work as-is with the Hydra, when they used to work perfectly directly to a Sony PVM monitor, or work perfectly when connected directly to an OSSC scaler.

I'm going to try and summarise my findings here, I'll also include the findings of other kit connected through the OSSC, but not dependent on the Hydra - VGA and Component inputs, for example:

Conclusion

Although it seems like a great solution, until I can work out how to actually get all of these systems working with standard scart cables, and not some mega-expensive versions, I really cannot recommend the Hydra. It seems that it is far too picky about signal type for its own good.


As I test each system, either connected to the Hydra, or bypassing it and connecting directly to OSSC, I will update the totals below.

  • Number of systems working to Hydra: 3
  • Amiga 500, Atari STe, Gamecube
  • Number of systems not working to Hydra: 9
  • Amiga 1200, Atari Jaguar, BBC, Master System, Megadrive, MSX2+, Saturn, SNES, ZX Spectrum
  • Number of systems working direct to OSSC: 12
  • Amiga 500, Atari Jaguar, Atari STe, BBC, Gamecube, Master System, Megadrive, MSX2+, PC-9821, Saturn, SNES, ZX Spectrum
  • Number of systems not working direct to OSSC: 1
  • Amiga 1200
  • Number of systems that remain to be tested: 4
  • Amstrad CPC, Neo Geo AES, PC-Engine, Turbo Duo

Amiga 500

  • Sync type:
  • Sync voltage level:
  • Result with Hydra: Working (with 1 of 2 scart cables); Port activates, signal passed to OSSC correctly
  • Result direct to OSSC: Working.

The Amiga 1.3 kickstart screen displays clearly, albeit with very faint ghosting.

Testing a game (Shadow of the Beast III) the result is less clear cut - both 312p and 625i output is shown nicely:

But then it goes out of sync:

I'm not overly concerned with the above test results - this A500 has sat unused for years, so there could be anything wrong with it (or the floppies I was using). Regardless of the game, it sat happily at the kickstart screen without issue for a good half hour, so the cabling, Hydra and OSSC seem okay with that signal, at least.

Amiga 1200

  • Sync type:
  • Sync voltage level:
  • Result with Hydra: Not working
  • Result direct to OSSC: Not working (problem with Amiga itself?)

I have two different-looking Amiga scart cables. I have a suspicion that it isn't a difference between the systems that allows one to work with the Hydra and one to not, but is actually a difference in construction of the cables. One has a plastic shell over the 23-pin connector, the other is metal. I still need to measure the voltages and have a look at the construction of the cable itself to find any internal differences.


  • Sync type:
  • Sync voltage level:
  • Result with Hydra:
  • Result direct to OSSC:

Untested


  • Sync type:
  • Sync voltage level:
  • Result with Hydra: Working
  • Result direct to OSSC: Working

Syncs perfectly to Hydra or direct to OSSC. Output is good and works fine in line 5x mode on the OSSC. I haven't adjusted any OSSC timings - it seems stable at the defaults.

However, the STe itself exhibits ghosting on the output signal, regardless of whether you feed it through the Hydra, direct to the OSSC, or plug the Scart cable straight into a CRT, so is not an artifact of either the scart switch or the scaler. This problem is easiest to see in the image of the Desk menubar, as well as the close up of the Atari TOS license text.

A possible fix for shadows or ghost images is detailed here: http://exxoshost.co.uk/atari/last/VIDEOFIX/index.htm


  • Sync type:
  • Sync voltage level:
  • Result with Hydra: Not working
  • Result direct to OSSC: Working; some instability in picture - wobbly image.

Not detected by the Hydra. Works in a fashion with the OSSC but the image is very unstable - text is wobbling up and down.


  • Sync type:
  • Sync voltage level:
  • Result with Hydra: Working
  • Result direct to OSSC: Working

A cheap aftermarket rgb scart cable is working perfectly with both the Hydra and directly to the OSSC. Ideally I'd be using component or HDMI output, rather than interlaced RGB, but I don't have those options available right now.


  • Sync type:
  • Sync voltage level:
  • Result with Hydra: Not working
  • Result direct to OSSC: Working

Requires some fiddling with the settings on the OSSC to get a nice picture (can lose sync at higher OSSC vertical line output settings; i.e. linemode 3X works okay, but 4X and 5X drop out), is never detected by the Hydra at all.


  • Sync type: C-Sync
  • Sync voltage level: 1.13v (cable unplugged from Hydra)
  • Result with Hydra: Not working
  • Result direct to OSSC: Working, some banding.

Has a 470 Ohm resistor and a 220uF capacitor on the Scart sync pin, so this should be working to bring the voltage down to standard levels, but doesn't appear to play well with the Hydra. Works with the OSSC, though there is some rolling interference pattern cycling up the screen every second or two.


  • Sync type: C-Sync
  • Sync voltage level: 3.22v (unplugged from Hydra)
  • Result with Hydra: Not working
  • Result direct to OSSC: Working

Has a 470 Ohm resistor and a 220uF capacitor on the Scart sync pin, so this should be working to bring the voltage down to standard levels, but doesn't appear to play well with the Hydra.


  • Sync type:
  • Sync voltage level:
  • Result with Hydra: Not working
  • Result direct to OSSC: Working

Perfect with the OSSC; really crystal clear. Not detected by the Hydra.


  • Sync type:
  • Sync voltage level:
  • Result with Hydra:
  • Result direct to OSSC:

Untested


The PC-98 has a VGA type RGB+HV output (640×400@56Hz in 16 or 256 colour), although it's not VGA compliant as it syncs at 24KHz. With the optimal timings and output of the OSSC set to 1920×1200 along with light scanlines it very closely mirrors a period CRT display.

  • Sync type:
  • Sync voltage level:
  • Result with Hydra:
  • Result direct to OSSC:

Untested; I don't currently have the system modded for RGB out - considering either a RGB amp board mounted in a Tennokoe Bank unit, or the slightly dodgy Super SD System 3 backup unit from Terraonion.


  • Sync type:
  • Sync voltage level:
  • Result with Hydra: Not working
  • Result direct to OSSC: Working

Two different cables tested - an original SEGA branded RGB scart cable, and an aftermarket one. Both work fine with the OSSC, but neither are detected by the Hydra.


  • Sync type: Luma Sync
  • Sync voltage level: 2.02v (cable unplugged from Hydra)
  • Result with Hydra: Not working
  • Result direct to OSSC: Working

Wired for sync-on-luma (an original C-Sync cable didn't work with either the Hydra or the OSSC). Works crystal clear on the OSSC, but does not sync with the Hydra, hence no signal passed through.


  • Sync type:
  • Sync voltage level:
  • Result with Hydra:
  • Result direct to OSSC:

Untested; my RGB scart connection for this system recently broke when transferring from the old office to new.


  • Sync type:
  • Sync voltage level:
  • Result with Hydra: Not working
  • Result direct to OSSC: Working

Works perfectly through OSSC (best Spectrum image I've ever seen!), not detected by the Hydra.

  • blog/retro_scart.txt
  • Last modified: 2019/08/19 09:09
  • by john