Warren,
I was assuming the Si5351 gave a consistent square wave from clk1 into the bridge.
However, the Si5351 CMOS drivers can be programmed for a max current of 2,4,6 or 8 ma,
and at 8ma it can just barely drive a 3v pk-pk square wave centered on ground into
the 177 ohm load presented by the bridge as per the last paragraph of post 1257.
Reducing the max current by a factor of four down to 2ma also reduces that
square wave voltage by a factor of four, making for a 10*log10(4*4) = 12 dB loss in power.
So that mechanism would be quite effective in reducing signal levels into the bridge.
I understand the disruption when doing measurements that scan through 300mhz,
where the Si5351 is switched from using the fundamental to using the 3'rd harmonic..
How that 300mhz disruption could have caused some sort of minor spike
in any measurement while scanning through 600 mhz still evades me.
Of course, there's lots of things that evade me like that.
Jerry
toggle quoted message
Show quoted text
On Mon, Aug 26, 2019 at 10:18 AM, Warren Allgyer wrote:
Jerry
There was no gain adjustment on the Rigol.
The yellow sweep begins at Marker 1 at 250 MHz. When it reaches 301 MHz the
fundamental drops back to Marker 2 at 100 MHz and the 5351 level is boosted by
more than 10 dB. This is what brings the third harmonic up to a level at 301
MHz that is comparable to the fundamental level at 300 MHz.
So there is a lot of stuff going on with the 5351 when it transitions the 300
MHz boundary. The level is boosted and the dividers are reprogrammed to drop
back to 100MHz. It is this complex transition that is my #1 suspect as the
spike culprit.
