Stanford Research SRS DS345 Function Generator Repair.

[MaxFrister]

I just finished repairing a new-to-me function generator.  I’ve been intrigued by this company given the rave reviews in The Art Of Electronics textbooks.  This is an older model but is still available new.

The machine was sold as not-working-for-parts from an infamous auction site.  It arrived dirty and completely dead.

After checking the fuse and transformer I brought the power up.  While the display was dead the internals showed signs of life.  I spent some time checking power rails (there are 10 of them).

While tracing the power-on circuit I found that power made it as far as the main mechanical power switch but no further.  There was a bad solder joint on the switch.

Here is where chance helped with the repair.  With the switch repaired, it powered up and displayed all 8’s on the display and no output.   I decided to probe around the digital circuits and disconnected the output from my analog scope to switch to a digital one.   I heard some relay clicking and noticed that the machine was booting.  I connect the output to the scope and I heard a spark and it went back to the all 8’s behavior.  This is the first time I’ve found a problem by *not* testing it.

The outputs of these machine are supposed to be floating but somehow were sitting at precisely  -5V relative to earth.   I spent an embarrassingly long time looking for the problem.  I eventually tracked it to a missing insulating washer; without it a +5V rail was directly connected to earth.


Observations:

* Stanford research system is not a nice company for those of us that like older equipment.  While they provide the manual online, the schematics have been stripped out.  They might sell you a new one for $100.  Hint: google is your friend.

* They will not sell parts for these machines.  They will not even service it unless you are the original owner.

* While not impossible to service these are difficult machines  to work on with crowded interiors and multilayer PCBs with tiny through-hole pads.   

[LapTop006]

* They will not sell parts for these machines.  They will not even service it unless you are the original owner.

At least from my attempts they simply ignored me when I tried to purchase cosmetic parts (turn some rack mount units back into benchtop). Their kit can be very nice, but good luck if it breaks.

[Zenith]

@MaxFrister,

This fits exactly with everything I've heard about Stanford Research as used instruments. Very high performance, but if they go wrong, and they do, they almost certainly can't be fixed.  So steer clear.

I didn't know they were so uncooperative as a company.

Kudos to you for fixing it though.

[DaJMasta]

Don't know what the hate is over SRS on saying they're unrepairable - they don't provide schematics but because the parts used are generally so old, it doesn't take advanced tools to figure out what's going on and parts are very often available, though not from the manufacturer - you even get the complete BOM in the manual.

Their current instruments are basically designs from the 90s or earlier (maybe the very latest instruments are from the early 2000s), mostly through hole stuff, and they are generally reliable and high performance.  Their company's policy with schematics and repair is very in line with Keysight's, Tek's, and others in the 21st century, it's just that instruments that look comparable from the 80s and 90s from other manufacturers often had their schematics shared.  It's not helpful and they shouldn't be praised for their closedness, but it's typical and expected in my book for current test and measurement equipment.


It's certainly easier to source parts for a current SRS instrument vs. a current instrument from many other companies.

[MaxFrister]

This fits exactly with everything I've heard about Stanford Research as used instruments. Very high performance, but if they go wrong, and they do, they almost certainly can't be fixed.  So steer clear.

I’m not ready to completely write off Stanford research systems (SRS) equipment yet.  Repair difficulty is relative.  Except for the missing insulator, this would have been a relatively easy repair. 

You have to compare it to others.  For example the similar vintage HP 33120 is SMD and HP can’t even remember their own name, let alone sell you spare parts.

Nice features of the DS345:

Red LED display
Most functions have a physical button
Lots of features
Beautiful design
Cost-is-no-object parts
Some thought given to service (top PCB folds out for probing while connected).
Most components are off-the-shelf

Not nice features:

Complicated design
Few test-points or labels on the pcbs
Dense through-hole boards with tiny annular rings
A few components are SRS only and they will not sell replacements.


I found an interesting marketing blurb that compares the arbitrary waveform generators that were available at the time:

http://www.hysen.com/uploaded/fckeditor/625_ARB_comparison_chart.pdf

[MaxFrister]

After a few months of use, I really like this function generator.  It seems really well designed and constructed.

Like any complex piece of test equipment, it has its quirks.  For example frequency sweeps are not always started at the same place (i.e., they don’t trigger at the same point in the waveform). This is both good and bad, overall probably good and an intentional design decision.

I appreciate that all of the features are a direct key or at most a shift key away.  The controls are, for the most part, well designed.  I also like the clear and bright led display.

I liked it well enough that I recently bought a second broken one and repaired it. The two can be slaved together for various options.

Hints for the next person repairing one:  Beware of leaking electrolytic caps.  On one of them, there was some damage to the pcb.  The power switch takes a significant push to engage and it is pcb mounted.  On both machines, the solder connection between the main board and switch was broken and the switch contacts were dirty.