low frequency gain error on broadband preamplfiier HP70000 70620B

[coppercone2]

Does anyone know if there is a failure mode of a broadband RF preamplifier where the low frequencies (say under 1GHz) are severely attenuated for some reason?

I am getting expected behavior at high frequencies but when I try the low frequency, with the cal output, I am getting something like -70dBm on a -10dbm signal that is 300MHz.

If I connect it to a external signal generator and sweep it I get like what looks to be a high pass filter transfer function.

Does anyone know if the OPT001 has different address settings on the side bit? The traditional preamp is good from 1-26.5 GHz, but the OPT001 is supposed to bring it down to 100KHz, but it kinda looks like I am getting a high pass filter effect around 1GHz, or severe attenuation under 1GHz however you wanna put it. It's pretty linear when I disconnect the preamp using the menu option.

I don't know if its damage coincidence or if I fucked it up in software because of where it begins to function as intended.

holy dogshit this makes me wanna punch a baby with the money I paid for that fucking thing.

Fucked up pin diode control?

I hook it up to a signal sweeper and check 100mhz test points, very linear, as soon as I switch the preamp on i get a massive drop at like 500MHz and its kinda wavy on the lower frequencies.

[coppercone2]

Does anyone know what kinda extension cable this thing uses?

Otherwise I need to solder in a buncha taps and possibly drill holes to test the power rails.

its a ~60db loss on the lower frequencies. At least it looks like you can test the modules. God I hope its software or some kinda replaceable transfer switch, the 100khz-2.7ghz preamp is unobtainum. 

[radioactive]

Based on the high-pass filter type response you described, I would suspect 1) a low series capacitance on the preamp input feed line.  This could be broken coupling cap, bad solder joint, bad connector.   2)  A low shunt inductance to ground on the preamp feed line (seems less likely).

btw,  just watched the most recent video on SignalPathBlog where he has similar issues with two preamp units (different models than yours).  In his case, both units had issues with the RF input connectors  (low series capacitance) only passing higher frequencies.

[coppercone2]

No, I can switch the preamp out with a shorting relay, then its linear, so it is not the front panel connector, which is a 3.5mm apc.

I think what happens is a transfer switch is used with the sweep signal to move it from 1 amplifier t othe other amplifier during the sweep (RF AMP and MW AMP)

Do you think its possible the untouched RF connector inside of it went bad? Like on the RFAMP can?

I thought maybe the RF can is not being switched in, so its stuck on the MW amp, which has a smaller capacitor.

Do you think that it is possible the unobtanium RF AMPLIFIER is damaged? Is that something that can happen, with semiconductor damage, that it will give accurate response but severe attenuation?

I am REALLY hoping its what you say it is with a cap or switch. I tried wiggling the connections but I would need a extender cable. I think its about to have a fucking mess of kynar comming out of it with little labels 

[coppercone2]

[coppercone2]

K1 definatly works. A5 appears to work.

So I think it might be either a control signal to A4/A6, general damage to A4/A6, or A7 (nightmare scenario). If its A7 hopefully its a bias issue or some stupid shit

Any idea what it would look like if you had actual damage to A4/A6 rather then a bad control signal...?

The lack of extension cable is killing me. I absolutely hate having to put taps on daughter cards.

Does this work like I think, where the sweep signal that controls the YIG systems will enter a window comparator type circuit to toggle the A4/A6 switches based on what the yigs are doing? So I can probe the sweep and the A4/6 control to see a correlation ? or at least one deflection per sweep cycle (i suspect the tune scan might be digitized because the preamp does not have a analog input on the back)



There is also some kind of lookup table. That might have chowdered calibration adjustments maybe.

[coppercone2]

I am going to make some splices that lead to kynar wire lenghts.

I think its going to be a good thing to own in your lab home made adapter kit, it uses those single pin connectors that fit over headers that are crimped onto wires. I will solder together a bunch with some heat shrink and strain relief so I will have ready taps so you can splice into a very long header cable without a breakout PCB that would be hard to fit in the same annoying form factor.

I have previously made banana to header (female) so you can connect headers to a lab PSU, I keep a small kit of these around along with some tektronix micrograbbers of various sorts, and the banana male to shrouded banana female so I can connect DMM clips to a PSU.

This should be a welcome addition and fun to make. It should be perfect for testing power rails. I am thinking about making some that have some shunt resistors too so you can monitor current draw to something annoying like a can sealed inside of a enclosure. I have ancient ceramic resistors that have been waiting for an application for nearly a decade now...

To those interested, the RF cans are soldered shut.

[ogden]

You shall check cables around RF/MW switch assemblies and preamps. Two primary suspects are cables connecting A7 RF amp assembly.

Similar problem is solved here:

[coppercone2]

Oh boy I just made like 7 splices and a shunt that have 6 or so feet of kynar wire attached to them so I can probe the power rails at least.

I have seen fucking extensions go for more then the god damn spectrum analyzer system. It's insane. What a buncha fucking vultures on ebay. I have not even managed to find one, but for another spectrum analyzer I was working on the fucking criminal wanted like 900$ for some extender cards (listed as like 1200$), when i bought the high tech SA for like 400$. Yea I am gonna pay that for a few PCB's with no components on them 

serious profiteering. I hope the occupy your stock room till protons decay.

[coppercone2]

I got an old school soldering iron that is a copper block on a stick, that should work. The problem is focused between 400-700 MHz.