HP 8720C VNA - S2, S3 20GHZ Microwave Switch control voltages?

[John_ITIC]

Hello. I recently picked up a HP 8720C Network Analyzer on Ebay. It was listed as "dim screen" and "phase lock failures". I fixed the screen via some pot adjustments and I have tracked down the phase lock failure to two RF switches that are not letting through the YIG Oscillator output signal. I looked up the switches but I cannot find the exact model datasheet online. However, I found one on kratosmed.com that looks the same except that it is marked 2677 rather 1699 that my switches have printed on them. I suspect it is a compatible unit but i cannot be sure. I'm not sure anyone would know if I contacted them directly and asked. Perhaps its worth a try.

https://www.kratosmed.com/gmcatalog/microwave-switches/rf-switch-spdt-switch-model-2677

The above kratos RF Switch states that a port is turned on by setting its LGC control pin to 7V. It is turned off by setting the same pin to -1V.

I measured on my 8720 system and it is feeding -4.53V to the LGC pin for J1 and +13.03 to the LGC pin controlling J2.

Also, these control voltages don't change when I set the 8720 into either low or high band when in CW mode. They also stay the same when sweeping the full 0-20 GHz.

I hoping that someone here has some deep knowledge of these units and can confirm whether the control voltages are bad (and possible have burned my RF switches).

I have attached some relevant information.

Thanks,
/John.

[John_ITIC]

More attached files.

[John_ITIC]

Attached application notes explaining internal workings of these RF Switches.

[John_ITIC]

One more piece of information: The control voltages are coming from the interface control board as shows in the attached picture. I will remove the board and see where the vias lead to.

[John_ITIC]

Update: The RF switches are controlled via the two 1858-0008 DIP-14 ICs on the Test Set Interface Board. I have Googled and I cannot find any reference to these ICs. It seems Motorora made them custom for HP? Image attached.

Edit: Found a reference to this part here:
http://hparchive.com/Bench_Briefs/HP-Bench-Briefs-1991-01-03.pdf

It is a Motorola SPQ526X. FETs. I have yet to find a datasheet...

Edit: I found that an equivalent part number is MHQ6001. The ICs contain two NPN and two PNP transistors.

/John.

[John_ITIC]

Tonight I confirmed that the two RF switches are working properly. They are not at all sensitive to the actual voltages put onto the control pins, anything between a few volts up to 14V enabled the switch. Anything from -1V down to -5V turned the switch off. That's as far as I dared to test. When the control input was floating (0V), the switch let through the signal with some 35dB attenuation. An "off" switch attenuated some 70dB (down in the noise of my spectrum analyzer).

I also found out that the 08415-60057 number printed on the RF switches is the HP part number. Apparently, over the years, there have been multiple models of 18GHz and 20GHz switches, from multiple manufactures, that have been marked with this same HP part number. So, it is possible that the 1699 model (which I don't have a datasheet for) may actually require a higher span on the control voltages than the 2677 I found a datasheet for.

So, now I will continue checking the downstream modules (A68 attenuator and A53 mixer) per the previously attached block diagram. I'm now more confident that the control voltages of -4.62V and +13.18V may be correct (although not per the 2677 RF switch dataheet).