8640B Bites the Dust

[nq5t]

Flipped on my (always loyal and trusty) 8640B tonight, and apparently it has suffered a stroke.  It powers up, but the display is filled with nothing but zeros and there's no output.  Without having taken it out of the case yet, my bet's on a power supply board — but that's a wild guess and I'm rarely right.

Anyone know of a place that repairs/rehabs these things (for less than a king's ransom) or if there's a parts and spare board source beyond the vagaries of eBay?

As much as I'd love a fancy dancy even used synthesized generator, the prices I've seen are a bit beyond what my "accountant" of 30 years is willing to allow in the midst of dumping cash into house renovations.

Grant NQ5T

[KO6YB]

Hello NQ5T, The power supply regulators are on plugin pcbs. They are under the top cover plate, near back toward the rear of the chassis, close to the fan. They have leds to indicate that each voltage is there. Also test points for a voltmeter. Both for each voltage. A quick check would not hurt. There is also an extender board so you can elevate these boards, that extender is somewhere near the power transformer IIRC. These regulator boards are almost always listed on the bay individually. It could be a cheap fix. At least check it out.

Stan

[MikeKE0ZUinkcmo]

Get a manual and dig in.  I suspect your guess is a good one.

Another thought is that many synthesizers are in the same price as an 8640 even if you don't spend too much time looking.   I found a 8657a, the 990 meg version, for fifty bucks in perfect working order.   Took a few weeks, but if you don't beat the bushes, you won't find the deals.

Good luck on your project.

[nq5t]

I have an original manual.  Probably had the 8640B for — I don't know — maybe 15 years or more.  It's just "worked", and I've never even had the covers off.  Guess I'm in for interesting times this week.  No matter what happens, I can't say I haven't gotten my money's worth out of it

[Opcom]

I realize this is a non-answer, but if you find trouble elsewhere and it is weird or expensive, I have one with a beat up front panel that apparently works properly. It might be of value in that case.

[Jim/WA2MER]

Don't you just love it when you ask a question and get a bunch of non-answers?

A friend dealt with this guy, who he found quite helpful: http://electronicsrevisited.com/html/hp_8640b.html

73,
Jim

[w1vtp]

Grant

Check the TIME BASE  INT/EXT switch on the rear panel.  It's a toggle switch that is easily pushed up by lifting the 8640 when moving or repositioning the unit and reaching back with your left hand - fingers curling over the cover.  The fingers naturally touch that switch and may flip it up to the EXT position.  When that happens the digital display goes out and the RF out goes away.  Flip the switch down to INT and you should be good to go.  I lost count how many service calls I made when an engineer would make a panic call to me.  I used to (as a private joke) do a little tinkering twisting this flipping that and then switch the time base back to INT which restored the display and the RF out.

I know, perverse sense of humor but it helped make the day go by.

Good luck, Al

[KA2DZT]

Al,  good advice,  sounds like the most likely thing that happened to the generator.  Could even be the switch needs a little cleaning spray with WD-40.

I have a similar problem with the time base switch on the back of my HP 5245 counter, have to turn it a few times to get things back working.

Fred

[nq5t]

Grant

Check the TIME BASE  INT/EXT switch on the rear panel.  

Thanks.  That was the first thing I checked, and I flipped it several times for good measure.  If it had been flipped, it would have had to have been the Ghost of Christmas Past or something

Perhaps the internal timebase is no longer time-basing.  I need to wrap up the in-progress project on the service bench and then uncable everything so I can move the stack of stuff that will actually let me get down to the 8640B.  Too little space to be neatly spread out …

[w1vtp]

Grant

Check the TIME BASE  INT/EXT switch on the rear panel.  

Thanks.  That was the first thing I checked, and I flipped it several times for good measure.  If it had been flipped, it would have had to have been the Ghost of Christmas Past or something

Perhaps the internal timebase is no longer time-basing.  I need to wrap up the in-progress project on the service bench and then uncable everything so I can move the stack of stuff that will actually let me get down to the 8640B.  Too little space to be neatly spread out …

Got another sig gen with a 10 MHz timebase?  If so, run a BNC cable from it to the 8640 timebase in and flip the switch up for EXT and see if that works.  At any rate, I'd start with the timebase circuitry and see if the problem's there

GL, Al

[W4AMV]

The manual on the HP 8640 is excellent. They are not to bad to work on. The 8640 is NOT synthesized. The main master oscillator runs ~ 250-500 MHz and is divided down in octave steps. The phase noise is excellent. There is a neat bang-bang PLL circuit that is basically a one bit up down counter for locking. The counter shares its circuitry with the frequency read out counter. YES, look first at all power supply V's and be sure they are all there before digging to the next level. I recently fixed an HP Vector Impedance Meter where the entire digital display went out. I mean everything, dozens of LEDs and various annunciators all gone. Yes, it was a PS power, lost +5 V logic V. However, there was not a thing wrong with the PS components! Turns out a molex pin to socket interconnect from the motherboard to the rear panel power supply had simply OXIDIZED and the ground return was absent. One shot with contact cleaner and no longer a paper weight!

Alan

[N8ETQ]

   Is there an external input for just the counter?
HP uses that Goofy wide band amp chip that looks like a TO-3. 
Seems to me "No output" = nothing to count. He said it
shows "0"'s.

   I hope its not the PA. Might find one on Epay. My HP 8443A
Tracking Generator has a similar device but it only goes to 110Mc.
Pretty easy to find a sub.

GL  and please keep us up to speed on this.

73

/Dan

Flipped on my (always loyal and trusty) 8640B tonight, and apparently it has suffered a stroke.  It powers up, but the display is filled with nothing but zeros and there's no output.  

Grant NQ5T

[nq5t]

Looks like it's board A22 — -20VDC/+20VDC regulator.  The 8640B apparently used up its limited supply of +20 volts.  Needs a refill.

Just bought a board on eBay.  Of course, this may not be the end of it, but I won't know for sure until I get the new one.

[KO6YB]

Good going. Lacking +20V will stop the main oscillator module. Nothing to count nor output. The +20v (A22) voltage regulator board uses a 723 regulator IC. It will shut down the pass transistors if it senses overcurrent. You could have a short in the +20V distribution, like a filter capacitor gone bad (or whatever else). Current limit sensed between pins 1 & 10 of 723. So, that is a possibility. If it is in current limit, you will not have the ~+22V on the base of Q3 (pass transistor driver) board A22. There is also a fuse on the board (not like a 3AG, much smaller).

Stan

[nq5t]

I checked the fuse .. It's ok. Could still be the pass transistor.  I won't have the new board until next week.

While I was trying to get to the test points down in the middle of the board (no extenders), I managed to kill (most likely) the 723 on the -20 side, too. No attempt at a good deed EVER goes unpunished

Grant NQ5T

[Rob K2CU]

I had a similar problem with mine and found that a shorted bypass cap in the AM modulator unit had also blown the series RF choke in same power supply line. Two standard parts in lab stock and she was back to full health!

[nq5t]

Ahhhhh … Ohhhhh ... that gratifying feeling of popping in a new regulator board and having the generator come back to life

All is well.  

As a backup and a hedge, I also ordered a couple of new 2N3055 pass transistors (which won't be needed for now) and some LM723s in the TO-100 cans.  So I might repair the original board and just hang on to it as a spare for the future.  It's good to know that at least for now there is a reliable source of spare boards for these things.

[W2VW]

Now I'm going to check my 8640B which has not been lit in at least 2 years.

Send positive energy my way plz.

[KA2DZT]

Now I'm going to check my 8640B which has not been lit in at least 2 years.

Send positive energy my way plz.

I already did, earlier today.  It's in the form of rain.

 No rush, you can thank me later.

Fred