> As you know, there have been a few problems recently with stability
> of the WFD in SM4 and occasionally in uVAX1. Last night, SM3 also
> started acting up and this morning *all* SM's started
> misbehaving. After a few hours, the problem appears to have corrected
> itself except that SM3 is still not working correctly. I am very
> suspicious that the problem is related to the clock distribution. Erik
> has looked at the clock distribution in several SM's and today (I just
> arrived) I had a first look. In SM3, the distributed clock has an
> amplitude of only about 100 mV. I don't know how the clock receiver
> works but that sounds like a pretty small amplitude to me. However for
Although I'm surprised it is that low, 100 mV should be OK. The signal
is immediately AC coupled and goes into a high speed logic buffer. The
input is biased to exactly the logic threshold (Vbb on the schematic- it
is an output provided by the MC10E116). So in fact, any small excursion
either way should cause the digital clock to switch. 100 mV is more than
enough.
> all of the SM's to go bad at once I think that the clock module itself
> must have been misbehaving too. I notice that for SM3 and SM4, the
> distributed signal (input to the fanout board) looks like a 600 mV
> (peak to peak) signal centered around about 2 V. I guess that the
> small amplitude is due to the slew rate of the driver chips? At any
> rate, I am suspicious that what happened last night and this morning
> could have been some drift in DC levels or sag in this amplitude that
> could have resulted in a marginal clock.
>
> Do you have any experience with this or suggestions?
Although the change in amplitude sounds fishy, it actually shouldn't
cause readout problems. Basically, the clock part of the circuit is
pretty robust... as long as a sine wave of a few 10s of mV reaches
each motherboard, they WFD should operate.
Statistically speaking, the commonest problems affecting the WFD system
have been:
general CAMAC problems (affecting STOP/START)
STOP master connections
mis-configured software (including camac lists)
mis-configured VIC-8250 network
SM-4 backplane
Ed