Electronics > Repair
Wavetek 2520a 0.2-2,200MHz RF sig gen repair
(1/22) > >>
dazz1:
Hi
I am trying to repair an old Wavetek 2520a sig gen.
I have narrowed the fault to the main board.  This board drags the +/-18V supplies to near zero.  There is an almost dead short on the board.
The +18V regulator  on the power supply is hotter than the -18V regulator.  The regulators are based on LM395 so I can run the power supply into a dead short. 

The +8V supply feeds the +5V regulators on the main board.  That is OK.

The +/-18V supply feeds the +/- 15V regulators on the main board.  The input to the 15V regulators is near zero, but the regulators are not even warm.   The outputs are less than 0,5V.   None of the ICs I can touch are warm (there are some under a shield).  There are no burn marks or signs of the release of the magic smoke. 


I have a scan of the service manual but it appears that two schematic drawings are missing.  There are parts on the board that have no purpose.

I don't have a thermal camera to trace the current flow.
I have a Fluke 77 so not ideal to track the voltage drop along tracks on the double sided PCB. 

Does anyone have a scan of the missing schematics?
Does anyone have suggestions on fault finding this board?

Dazz

wn1fju:
Your description confuses me.  You say that the input to the regulators is near zero.  So why would you expect the output of the regulators to be anything other than near zero?  That doesn't necessarily imply a short on the main board.  If you had said the input to the 15V regulators was the normal 18V and the output was zero, then I would agree with you.

You should first unload the power supply by disconnecting the various boards and make sure all the power supply voltages are correct.  On my similar Wavetek 2510 unit, the Molex connector from the output of the main transformer to the rest of the power supply had corroded and the power supply didn't work correctly.  (I ended up removing the connector and soldering the wires since no amount of connector cleaning seemed to cure the problem).

There very well may be a short on the main board (or other boards).  Invariably it will be a capacitor hanging off the power rails.  You can start lifting one lead of the capacitors while watching an ohmmeter on the rails.  When the ohmmeter jumps away from zero, you've found your short.
dazz1:
Hi
I started by isolating the power supply.  With an open output, the voltages are all good.  The LM395 transistors on the power supply limit current to about 2.2A.  I can safely connect the power supply to a dead short.  Good for testing purposes.   The power supply is fully operational. 

I then isolated the fault to the main board.  When the main board is connected, the +/-18V supplies are near zero.  As a result the outputs of the +/-15V regulators on the main board are also near zero.    Something on the main board is taking about 2A.  None of the chips have a high enough supply voltage to operate, so there is no signal to trace and no function to check.

The 5V supply on the main board is OK so all of the digital logic devices are also OK.   None are exposed to the +/-18V supply.

Temperature measurements on the power supply shows that the +18V regulator is running hotter than the -18V regulator.    This could simply be because of the design loading on the +18V circuits is more than the -18V circuits.

There are several hundred parts on the main board so randomly lifting parts would be the option of last resort. 
I have started by marking the components supplied from the +/-18V including the +/-15V regulators and attached parts.  Without the 2 missing schematics, this is difficult to finish.

Ideally I would have a thermal imaging camera to follow the current along the heated tracks to the faulty component, but I don't have one. I am considering converting an old webcam to infrared.

Today I am going to try and trace voltage gradients along the PCB tracks, but I don't have the PCB drawing or a really sensitive volt meter.  I will pay particular attention to capacitors.  There are no electrolytics on the main board.

Dazz
wn1fju:
You're right, I apologize.  I just quickly looked at the mainboard schematic (at least the pages that aren't missing!) and saw the 15V regulators.  It didn't occur to me that the +/-18V was also propagated throughout the board. 

There may be no electrolytics, but it looks like there are quite a few (1 uF, 10 uF) tantalums which are notorious for going (near) short.  That would be my first guess.
dazz1:
Hi
The regulated +/-18V from the power supply is applied to 7815/7915 regulators on the main PCB.   There are two separate +18V regulators on the power supply, presumable because someone really wanted to use the LM395 transistors, and one couldn't supply enough current.   The power supplies are double regulated.  Definitely over engineered.

I am having a go at hacking a old webcam to make an infra-red camera.  It that works, I should be able to "see" where the current is going.

Dazz
Navigation
Message Index
Next page
There was an error while thanking
Thanking...

Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod