restoration of a wavetek 3000 signal generator
by ralph klimek 2014
epilogue 2020
The wavetek 3000 is a late seventies advanced multi PLL signal generator with coverage from 10Mhz to 500Mhz and numerous options for coverage and modulation methods.
This unit had been recovered from a dumpster and was non functional.  Recently a poor quality manual has been found on the internet.  I have extracted the images from this source and cleaned them up using the GIMP.
The recovered manual was used to find the many faults that were present and still present with this unit. My advice to people  finding one these units at a flea market or dumpster is to leave it there. There are too many problems with this unit due to the serviceman un-friendly construction methods, most notably needing to remove and open up numerous tin cans to get at the circuits.  Mine has been brought back to life in as far as the unit produces a constant level frequency stable output , that is, alas only loosely related to the thumbwheel settings !
it also had numerous inherent implementation issues that were beyond tolerance or repair.
First Fault
You will find that the +/-  18Volt supply rails are at zero volts.  This is due to numerous tantalum capacitors inside the many tins  having turned into short cicuits.  Finding them was the motivation for my Kelvin Probe project. The dead bug construction method has not contributed to reliability either.  Wavetek did not believe in using PCBs !  It has all been done with point to point wiring, nicely done, but it must have cost them a fortune in expensive american labour. I redid the PSU with three terminal regulators. Lets face it, the designers of three terminal regulators were better designers than me !

Second fault.
The power supplys did not use 3 terminal regulators, which were available when this unit was designed. Instead, they made their own regulators using the impossible to replace LM135 "indestructible" transistor. Guess what ?

Third Fault.
This is the one that even after all my work, will consign this unit to landfill. Lack of consistency in the manual means I cannot really debug the electronics, the manual does not describe all the tin cans, only the basic default models units. My present fault condition renders the output frequency to be only loosely  related to the machines settings, but very stable !

Some  Praise is also  due.
The circuit design is highly advanced and I believe, that when this unit worked, it displayed unprecedented low phase noise and lack of spurious output. I have also read that this unit provided a Gold standard in phase stability and low noise for its time.   The circuit design is worthy of something that HP would have done but implemented correctly.

The tin cans are not sealed, can be easily removed just by unscrewing the lockdown screw. It is not require to unsolder all the connectors....I was laboring ,incorrectly, under this illusion. The teflon terminal posts are actually a plug and socket !  The maintenance method was to indentify the failing tin can and return to manufacturer. Some tins were utterly and completely beyond inspection and repair due to internal progressive assembly.

Utter and complete condemation.
see my epilogue below.  IC sockets in a place that is totally and utterly beyond reach and repair.
IC sockets should never ever be used for production equipment, only for test and development and eproms.

clicky clicky  here ! is the complete manual image archive !  I have downloaded the manual that has recently appeared (I have lost the reference but many many thanks to the kind person that scanned it and make it available.  I have doctored the scanned images  to remove crud and make them legible and printable  using the tools provided by the GIMP.
this is my manual archive  images are in ppm format
all these capacitors must be replaced. They will all become shorts or are shortsthe front panel pcb has been liberatedthe businesslike front panel
millions of tin cans. label the coaxs before removing them !  Not all the tins are described in the manual. Ouch.thumbwheel connectors. One of my thumbwheels has jammedoutput attenuator module
well worth saving from landfill
the backplane provides many usefull test pointsthese are not just terminal posts, they are sockets ! Dont unsolder them !most of the tin cans contain this style of construction. It is very amatuer dead bug style, good for one off construction but it must have cost a fortune in labor. Maybe  I got the "missing prototype" unit ?
blankthumb.imgp6957-wavetek-module-internal.jpgon the left, IC sockets inside all the tin cans.  Intermittants anyone ? Why were they not soldered in, why no PCB ?  adjustments not reachable from outside the tin. Was my instrument the prototype ?  This is a relatively clean and good example.Some ic sockets were inside fully soldered up tins. Many Adjustments could not be reached from outside the tins.

My wavetek has now completed its journey to landfill. I decided not to waste my precious time on this lemon, and that
is being unjust to lemons !   My journey with this abomination ended inside a tin that contained one of the divider chains.
Here I found inside a tin, inside a fully soldered up sealed tin another sealed tin containing the divider chain. The ics were on a pcb.
Thats good, that is what ics should be mounted on. Compare and constrast this where they used point to point wireing on the ics.
The ics on this board in a tin, in a tin, in a tin were in SOCKETS. Scream with impotent rage ! Those sockets are intermitant as are
most ics sockets from this era.  Rage and Fury emoticons.  (please supply your own, I am out of them !)
The only thing that I wound up salvaging was the very well made precision attenuator which has found another life in the lab.
My advice to "lucky" people finding one of these in a dumpster or flea market is to run away, far away, fast, very fast.

Fri Mar 21 19:05:16 EST 2014
Wed 22 Jul 2020 05:35:05 PM AEST   this is my last word on this unfortuneate topic.