Following on from my post of a couple of days ago, looking for a schematic set for a JVC with a dead radio section, not having found one, I had a little search through my extensive stock of JVC manuals, and came up with one for a similar(ish) model of the same vintage, which used the same radio tuner sub-board.
Using this, I was able to determine that all rails were established and correct, that the PLL chip was idling and unlocked, that the FM subsystem IC was idling, and that the VFD for the radio section had the correct supplies and drives, but was being 'muted' by the drive IC, which is also the system control processor. No I2C data was being sent to the PLL IC, hence the reason that it never ramped the tune voltage to the RF module. The I2C bus was however, correctly active for other control functions, such as handling the CD section. About the only conclusion that I could come to, was that either the microcontroller was in some subtle way faulty or, more likely I felt, its software was screwed.
Last week, we had a couple of claps of thunder locally, so I began to think that this might be a 'storm damage' job. Either way, in view of its age, it was not going to be a practical repair, so I informed the store that it came to me from, and put it back together. It was at this point that I noticed the PAT test label wrapped around its power lead, and this indicated that the test had been carried out only a week or so before it became faulty. At this point, I began wondering if the fact that it had had the tests performed on it, probably by someone with an automatic tester and little understanding of what it is doing, and where it is appropriate to run which tests, had resulted in damage to the microcontroller. I have seen other equipment on a number of occasions over the years, with similar 'odd' failures, and sometimes power supply failures, shortly after being PAT tested.
I've never been a great believer in the appropriateness of applying these tests to double insulated and transformer based equipment, particularly given that large DC spikes are applied, which with some testers are actually twice the nominal line voltage. Given that in much microcontroller based electronic equipment, digital grounds are not directly bonded to other system grounds, it seems to me that having big voltage spikes flashing around between the primary side of the power supply, and cabinet metalwork, which is not grounded to any line power earth, but may well be AC common to internal DC grounds via low puff ( and sometimes not-so-low ) caps and high value resistors, is asking for trouble of the same nature as you might expect from static damage, or pulse damage from nearby lightning strikes.
I would be interested in knowing if I'm on my own on this one, or if anybody else involved professionally in service work - or indeed anyone who carries out PAT testing - has any similar experience, or opinions on this.
Arfa