Latest news are:
1. Power supply at the first sight doesn't show me any problem with voltage levels. It seems quite stable, however I haven't had good equipment to measure it.
2. Connecting DVD drive and HD to external power supply doesn't help.
3. I managed my "service" cable to work with the machine! It seems that no level conversions are required at all. So, I just connected my laptop RS232 directly to pins on machine board (I had to construct the cable anyway, because "mouse cables" or similar don't fit). That was really good experience. My observations are:
- there is Philips definitely inside HX717. It works with nucleus codes and especially answer for one of them is:
DS:> 903
090300:
BE version = 43.02.19. PHILIPS ,VAD8043 ,43021900,5VC0537341288
0200,
Optical unit version = 25.28.79.84.00
Test OK @
- I have found that this machine is not exacly as HDRW720, because some of nucleus codes doesn't work (however most of them work), and by blind typing I found some that are not documented for HDRW720 though.
- I have found next points to consider when repairing the machine - two tests failed:
Busy executing NUC115 5-27
Not all DMAs worked.
FAIL 011502
and
DS:> 2108
210802: Performance failure
Minimum access time = 204 msec
Maximum access time = 204 msec
Average access time = 204 msec
Number of commands below 160 msec = 0
Number of commands between 160 and 200 msec = 0
Number of commands between 200 and 250 msec = 1000
Number of commands above 250 = 0
Error @
The latter one is definitely Hard Disk failure. I am not sure about first one, but DMA may also be HDD dependant.
So, next step is to replace HDD, but first I have to found similar machine with working HDD or confirm what exactly Philips machine it is and have "Master DVD" for it.
What is interesting when playing with nucleus codes is for example - you may change region code of this machine

/jayu