One bad fan and the other is wonky, so replacing both. Original fan did
come on at switch on and assumed to run faster on warming of amp as amp
never cut out.
Replacements do not come on at start other than a slight flick. One will
sometimes with the fans in a vertical plane , but amp used with them
horizontal. I tried heating the area of the PTC on the main heatsink,
with neither running at start and neither started up to the temp I was
prepared to take it.
The SM out there is for a different 1062 without SMPS and simple fan
circuit and fans in series. The PTC is 290R at 20deg C and it is PTC.
With the back now off again. The fan voltage, for 12V fans parallel in
this flavour of 1062 is 5.1V climbing to 5.5V at switch-on and first 10
seconds with no fan load.
The fan control circuit is off board, mixed in with complex SMPS not on
the PA panel and its heatsink. At the moment I intend putting 470R or so
in series with the PTC, if this fails to make fans start at powerup, ie
startup voltage of 8 or 9V say, where next? At the moment I don't know
if the control is proportional or stepped. When I've added the 470R I'll
jury rig and check the fan voltage response with hot air on the PTC
directly .
Are there specific types of fan ,and spec'd as such, that will start up
with progressively increased voltage and some that fail to start unless
given a good "kick" to start with?

Threshold and then proportional but as difficult to find what that
threshold is and what initial proportional voltage, looks like I'll have
to delve into the smps, at least only 6 sc hold it in place. As amp is
never used for "quiet" use no problem with elevating the startup voltage
to 8V or so, if I can determine the circuit enough to find where to
change a value, no preset seen

The fans that were in look as though they were 6V ones in paprallel , so
being overdriven when taken to 12V explains the knackered state of them.
No labels, but 3V starts the working one and takes .1A at 6V and at
physical size 60x60x10 is probably 6V,0.1amp rating. Not that they were
the originals as the fixing bolts were hacksawed to length and the
housings are for 20mm thick fans. What I thought was a fudge diode under
the collector pin of a TO92 was in fact a 6.6V zener, in the fan control
circuit on the SMPS pcb. That fudge diode has no marking on the overlay
but was of the same series as other zeners on that board and the TO92
uncropped long legs, to accommodate this zener , and all looked original
manufacturer's PbF soldering. So overall quite a mess. At least I now
know what zener sets the PTC sensing line, 2 of them in series one on
the SMPS rectifier heatsink and one on the PA . And which one sets the
12V as uses 28V supply it is quite critical. Now to find what this 6V6
zener function is, will change it to 3Vzener and suck it. So far the fan
control circuit has stuck resolutely to about 5.5V starting voltage via
the TO220 pass transistor.
The replacement fans I put in there are 12V
I suspect the pcb was designed for 2 fans in series,24V , 12V each,hence
28V supply, as in the earlier 1062 amps

Not as elegant as I would have liked but switch on fan voltage 8V
climbing to a top limit of 11V. As those fans if running at 12V, then 4
times the cooling power of what was in there before I got it, so will
leave it at that, as 11V/12V is not much difference in function.