HP Yig Oscillator tear down. Fits 8562/8563A SA

[jlmoon]

Greetings, 
It has been some time since I have contributed by posting some interesting information.  I have been working with my 8563A spectrum analyzer and have finally determined the local oscillator (YTO or YIG) has failed.  So the journey has begun, and I figured I would take a closer look at just what is up with this one.  I hope these pictures will shed some interesting perspectives into the black magic within one of these really cool devices.  If anyone can see the failure by my pics please let me know your opinion, maybe it will help others and me as well.  It is such a treat to see the inner workings up so close.  Enjoy!

Hope my pictures are detailed enough but not too large in size to make out what is going on.

[jlmoon]

One more, this one the FM coil looks like it has been hot.  Not sure!

[0culus]

VERY neat pics, thanks for sharing. YIG oscillators are amazing bits of high precision engineering.  Hope you are able to get that 8563 fixed.

[wilfred]

One more, this one the FM coil looks like it has been hot.  Not sure!

What is going on here? looks broken but can't be sure.

[jlmoon]

Hello,
That image (screenshot) isn’t my device is it?  If not, then yes looks like way to much fm coil current was applied to that one.  Magic blue smoke was set free!
I think wht you’re seeing is the limited field of view with my microscope. 
I looked at my device and can see the very small wires going to the term points.  What I am wondering is if that coil has over heated at sometime in the past. 

[wilfred]

It's just a little piece I cut from yours. I don't know the names of the parts so that seemed the best way to point to it.

[jlmoon]

Wilfred,
Had to go check out those wires, the area you're seeing is where they are reflecting light and twisting together creating an optical illusion. 
Just for grins, I measured the resistance of both windings, the main coil read 41 ohms which should be okay, while the fm coil read around 2 ohms. 
Thank you for pointing that out. 

Jon

[SoundTech-LG]

what looks like a burned coil...  but why on just one part of it? the current is the same everywhere in the coil. An even burn seems more likely.
It also appears possibly as corrosion, and also the magnet in the middle seems corroded in spots as well.

[hendorog]

I have read that HP YIGs in general often fail in the microcircuits.

The ones from the 856x series are not particularly robust apparently:
https://www.qsl.net/ct1dmk/wbond_ex.html

[jlmoon]

Hendorog,
I concur!  I have seen a few of these failures as well.  I have one pic that appears to show a substrate crack below a semiconductor.  Not sure until I turn up the magnification a bit more.  img_2298.  A very small hairline crack going East to West!  Possibly a manufacturer inspection scratch.

[jlmoon]

Hello all,

Thought I would give a status update on the YIG / HP8563A trouble.  She sings again!!!
Now, I am going to venture into the pains and pleasure of Class 100 clean room etiquette and attempt to wirebond a new output semi / oscillator transistor.  All this with my future time allotment.  Thank you for everybody's input, it was Great!

Jon

[0culus]

Very nice! Please do share photos of the wirebonding repair if you can.

[jlmoon]

Will do!  I have only found one wirebondable part so far.  Die size 15 x 15 mil.  Jury is still out on the actual part that is fused to the substrate.  GaSFet, bipolar or what.  To be determined.
SiGe  www.sige.com now SkyWorks is the only source so far that has anything remotely possible to use.  They have a 30Ghz "LPT16ED" bipolar that looks promising, could be a good fit.  Not sure yet! 

[Samogon]

So at the end it wasn't coil? Fine tune coil is really small and 1-2 Ohms should be ok?

[DL9SDL]

Hi experts,
I am quite new in this forum.
My situation: I´ve got error messages on the HP8563A, #301, 317  and sometime 334.
I have bought a used one in UK and then replaced the existing one.
I have passed the routine of YIG alignment in the service manual, but I don´t get the lower frequency position with 3.02 GHz adjusted with the spindle poti.
So I put the jumper on A14 back to it´s original position and I tried (try and error) to get at the lower band a sweep from 10 kHz up to 2.9 GHz, all the time monitoring the LO output with another SA. Frequency flatness is quite good and also the measured LO-level of apr. +14dBm. No error message is present.
By pressing full span, I get 301 and 317 error again. what are the correct values in the frequency analysis menu ?
Has there any calibration code to be typed in after having changed the YIG ?
I ask for good guidlines from people, having had the same problem .

Kind regards from "cold" germany

[PA0PLY]

Hi JP,
Can you help to fix my yig, or support for getting new transistor?
Thanks in advance,
Jan info@pa0ply.nl