HP8644A Signal Generator Fan drive circuit failure

[tkamiya]

I'm working on HP8644A RF generator.

Problem is, fan does not work.  If I drive the fan with an external power supply, it works.  If I watch the voltage on PCB marked "fan", open voltage is 28V.  With fan, it's like 3 volts.  According to circuit diagram, there is a "voltage regulator" reducing the voltage to 12 volts.

So it is obvious the problem is this regulator.

Problem is, circuit diagram does not go any further than saying it is a regulator and doesn't show anything other than a box or location of the component.  Since it is a multilayer board and trace goes under heatsink, I can't tell where it is going. 

Does anyone have a good circuit diagram, or have some experience?

[coromonadalix]

Yeah  i checked the service manual, it is an bloc section only

Well the regulator could be an 3 pins to-220  like an lm7812,  but i think in thoses years  it could easily be an power transistor with an 13.1 volt zenner a resistor and small capacitor  ...

I think youll have to check if you dont have cold joints solders, a bad capacitor, damaged regulator section, i dont think this puppy will be easy to service, tons of boards and stuff


photos ?

[tkamiya]

Well....  photo would rather be useless.  On the actual board, there is a connector and the trace goes immediately under a heat sink and under yet, another heatsink.  Tracing from the other side, I can't even see the trace.  It's probably in the middle layer.  For something this simple, it's really hard to trace.

It's quite upsetting.  A stupid 12volt regulator circuit is stopping me from completing this project.  I might just rig up a 7812 somewhere and call it a day.....   grrrrr.....  but it pains me to deviate from HP's design.

[coromonadalix]

I understand, and tearing down this unit may cause more troubles ?? 

[tkamiya]

I started to randomly poking around what looked like a collector terminal on power transistors.  BINGO!  HP 1826-0423!  Cross referencing, it's just an LM317K.

I think this is it.

[tkamiya]

LM317 K in TO3 is awfully expensive.  I'm going with one in TO220.  I think I got the case number right.

[coromonadalix]

lm317 reference design : http://www.mh-audio.nl/lm317.asp

[tkamiya]

Thank you.

The part is already on order.  Once it is here, I'll take that portion apart and transplant the new chip.

When I do repairs, I prefer to approach it surgically.  I really dislike, say, replace all caps, kind of approach.  PC board after 20+ years are already pretty fragile.  It would be a shame to damage excellent working HP classic for sake of repairing silly fan circuit. 

I wonder why HP didn't use 7812 type chip?  The same thing can be done with less part count and a bit cheaper.  Hum....

[TERRA Operative]

Could be that at their volume pricing at that time, it was cheaper this way, or maybe it was just how the design engineer liked to do things.

[edpalmer42]

Does the block drawing show a 12V output from the regulator?  Even though it's a 12V fan, maybe they decided to run it at a different voltage for noise, longevity, or some other reason.  If so, a 7812 wouldn't do the job.

Have you checked for bad solder joints on the LM317 and/or the resistors that set the output voltage?

[coromonadalix]

In some cases an lm317 could be used as a current regulator like any fixed 3 pin to-220 regulator,   maybe they have used an lm317 due to parts availability they had their reasons for sure.

[tkamiya]

In diagram, it clearly shows input coming from 20V rectifier circuit into a box labeled +12V regulator, and goes to another box, FAN.

No details on +12V regulator circuit.  Circuit trace is somewhat oddly routed and not entirely clear.  I'm considering leaving this alone and simply build another regulator circuit elsewhere.  FAN is the only thing this circuit feeds.

No, I didn't reheat joints but looked very closely.  They are intact.

[tkamiya]

More I look at this, more I think HP engineers over complicated this.

From transformer, it goes to rectifier and filter circuit.  Then to fuse.  At this point, it's +/-20V.  It goes to another transistor and into a regulator circuit to get +/-15V.  Right after fuse, +20V is tapped off and into a 12V regulator into a fan.

One could easily used 120V fan as source it right there, or 24V fan out of +20V, or tap off -10V to drive a 12V fan.  From what I can tell, fan bearing is drying.  Driving this fan with external power supply shows power consumption doubles and becomes slower and noisier in a minute or so.  I bet it burned out a regulator. 

I'm more inclined to just run it off 120V.... or 24V and skip the dedicated regulator.

[tkamiya]

One thing I forgot to take into consideration....

This RF sig-gen keeps some circuits alive while the machine itself is turned off.  Sure enough, some of "Regulator" circuit is connected to "On/off control" line. 

Odd thing is, this fan line is not connected to it.  But the fan is not always on.  However, it's clearly turned off when front panel switch is OFF.  In other words, functionality does not match the circuit.  Weird.

[tkamiya]

I just went to a local lab and did spare parts bin jumping.  Found a replacement power supply board and a fan.  Paid $20 for both.  My HP8644A is functioning as it should.

I'm a bit disappointed in HP's design and documentation.  Fan does not come on unless power switch is ON.  However, some internal circuits are kept alive.  Rest remains in powered down state unless the switch is ON. 

Circuit diagram shows partially how this is done, but this provision is not connected to the fan circuit.  Plus, just to drive a fan, the circuit is really over complicated.  There may be a good reason for it, but I really don't see it. It's really nice for HP to make most of these documentation available though. 

Anyway, my repair project is done.