EEVblog Electronics Community Forum
Products => Test Equipment => Topic started by: pquadrat on February 05, 2018, 08:49:54 pm
-
Hi, this is my first post here, and I tried to find help searching around, but did not find anything useful.
This unit has a sensitivity problem on channel C (Option 030), which is specified up to 1.3GHz at -9,5dbm.
Sensitivity below 1GHz is much better.
Sometimes it reaches this sensitivity, even some db better, and sometimes it is about 10 to 15db worse.
I bought the service manual for it (Artek), and I found some tips on the Youtube Channes "The Signal Path", where Shahriar explains, that
the sensitivity on high frequencies is supported by some sort of oscillation because of a feedback loop around the divider.
The adjustment procedure for channel C is indeed to disable the peak detector and to adjust two resistors so that
the decade divider oscillates at ca. 1.15GHz, and the counter count this. Then You adjust the peak detector to disable the counting
as long as no good signal is detected.
Sometimes it does this (oscillating at 1.15GHz), and then the sensitivity is perfect up to 1.3GHz and even higher.
But sometimes it oscillates at 500Mhz or even lower, and the sensitivity above is bad.
I checked the input attenuator (skipping it), no help. I disconneted the divider fron the front end, no change,
sometimes it oscillates right, sometimes not. And I was not able to find anything that changes this behavior,
not cooling the circuit down, or heating it up, applying pressure, re-seating the chip in its socket, nothing changes.
My next try would be changing the decade divider IC, but I was not able to locate one.
Part is HP 1820-2382 / 1DC2A ECL decade divider. It is used in HP 5334A channel C, 5335A channel C, and und 5316B Option 003.
Any suggestions, or has anyone a working divider chip? On the third picture, it is the metal part (smaller than a TO-3) without the heatsink.
Thanks
-
Hi... is recommended a MC12080 prescaler by 10...in hp 5315 works fine... but if the divider (prescaler) is by 10.000 is possible make readings after 100ghz with the correct dividers...
a exemple is divider by 64 with MB506 and 2,5 x 3 with 3 74ls90 and by 10 with 74ls90 again ( 64x 2,5 x 2.5 x2.5 x10 ) reads 2.7 ghz ghz fine (3.5 ghz with upb1505 NEC)
bye
73
-
option C in hp 5315
-
U3 original prescaler by10 in hp5315A