Hi Adri
?
Here are the measurements connected to the first 2 converters as a reference.
?
For the second converter I get a voltage of -1.678V instead of -3.5V as written in the service manual, I would say that we have similar voltages for this particular point.
It seems far from the nominal point, but there may be an error in the manual.
?
You are close to the goal, and this module may be repairable if there is no micro circuit, I have never opened a second converter in this series, but it is a possibility.
?
When you open your second converter, send me pictures, I'm curious.
?
Yves
?
dear Yves, I didn't realize that you had made the measurements expressly to help me! I owe you then a double (even triple) thanks! Especially considering the fact that the front end of this instrument is really very dense, complicated and delicate!
After my last email, in addition to doing the (essentially mechanical) intervention on the SYTF, I continued to check the input signal through the various sections by alternately injecting 300MHz @-20dBm and 5GHz @0 dBm
The signals were found to be in specification up to the A3A9 bandpass (where the two paths join to form the intermediate frequency of 321.4MHz). At this point I could see the 5GHz signal (converted), present at that point at -47dBm (the specification says -43/-56dBm).
Instead, no trace of the 300MHz input signal
Investigatine?the second converter A3A4 in more depth
I checked?the signals at its input
On J1 there is 3.9214GHz at -46dBm? --->OK
On J3 there is 321.4MHz @ -37.7dBm? --->OK
On J4 I did not check 600MHz @-1dBm but the fact that the converter works in band 1 tells me that the oscillator is present and is in specification
The power supply is quite OK (5V are actually 4.75; 10V are 9.98 and the DC that controls the internal switch that should be +15V? (band 0) and -3.5V?(band 1) is actually +14.60V and -1.35V
This last value is rather far from the nominal and would have worried me BUT band 1 is exactly the one that works
Conclusion: it would seem that the culprit is the second converter and in particular (probably) the internal switch that has the sole task of letting the signal? (which is already at 321.4MHz and does not need to be converted)?pass from its input to the output
In the next few days I will test the converter that meanwhile?has been?disassembled?; In the meantime I found a cheap one on the bay that should arrive in about ten days
Yves, I have a question: on the block diagram A3A4 is represented as a block that has 5 ports of which one (J5) would seem to be the connector that carries the power and the DC that controls the switch and that connects to A7 Analog Interface through the W35 cable
Actually,?this connector on the real component, does not have a name
Instead,?the name J5?, stamped on the case , indicates an SMB connector that has the writing LO TEST and in my analyzer?it is the one that?conveys the 600MHz LO signal?
Now,?on the block diagram the connector that conveys the 600MHz signal has the name J4 (in my analyzer, the J4 connector was disconnected)
This confusion of names and roles made me doubt that years ago, when I had disassembled this section I had not reassembled it correctly
thanks again and best regards
----- Messaggio originale -----
Inviato: 06/09/2024 23:21:52
Oggetto: Re: [HP-Agilent-Keysight-equipment] HP 5086-7906 YTO
Hi Adri,
?
I had sent you the measurements of the RF input-output mixer signals verification, I had disassembled the High Frequency section of my 8595E to be able to provide you with this information.
If you have no more questions relating to this section, I will close my 8595E and do a flatness check to see if everything is correct.
?
Yves
?
?
thank you very much for all the suggestions and information with which I will continue my investigation
In the meantime I received the 2.5mm socket wrench and after?a little mod to adapt it?for the purpose (photo), I proceeded to align the spheres of the SYTF
I must say that the operation done under the microscope is quite simple
The nuts must be loosened (not too much) and with a small screwdriver that acts as a delicate lever, ?the supports of?the?spheres can be moved so that the coupling ring completely shadows them (i.e. the sphere is exactly in the center)
Removing the gold mesh was not necessary
The insertion loss has improved enormously:
at 5 GHz it is -6dBm at 6GHz it is -10dB (before the signal starting from this frequency was no longer visible) at 6.5 GHz it is only -3dB
I will install it when I have solved the problem of "blindness" in band 0 since the calibration constants that the instrument has in memory (in practice the tuning current data) are relative to the original SYTF and the "new" SYTW requires a? SYTF CAL is done
