FLUKE PM6680B teardown photos and upcoming repair.

[H.O]

Hi,

So I got hold of a FLUKE PM6680B timer/counter for, what I thought, a fair price. I paid $300USD for it untested but "most likely working just fine".
It arrived the other day and what I found so far (after about 10 minutes of playing with it) is that channel B seems to be working fine while channel A has some issues. It seems to count fine up to around 130kHz above that it either does what looks like aliasing or just plain stops counting. (No the 100kHz input filter is not enabled).

Manuals, service manuals and schematics are available online so as long as it's not an unobtainable part that's failed it should be repairable (it does contain two custom ASICs...).
At the moment I don't want to allocate the bench space and time to attempt troubleshooting and repair but since I couldn't find much info about this instrument on the EEVBlog forum I thought I'd take it apart, have a look around, and take some photos.

First a general overview of the front and back panels:




Taking the instrument apart is really easy, two screws at the back and two at the bottom then the whole thing slides out of the case. The front panel is a fairly thick diecast aluminium piece, the back is piece of sheet metal. Holding the front and back panels together are two pieces T-slotted extrusion to which the PCB is also mounted. A really nice and accessible design.




In one of the back corners we have the GPIB board. Made in Sweden, like me :-)


In the other corner we have the power supply:


The backside auxillary BNC connectors soldered directly to the board:


To be continued....

[H.O]

Here's a closeup of the backside of the board. The schematic mentions a delay line, this must be it:


Moving on to the input sections hiding under the shielding can. Note the two ferrite rings directly on the PCB:




This is where I found the first evidence of damage, a 47ohm 1206 resistor looking quite bad:


Initially I thought this was the switchable 50ohm input termination resistor but after looking at the schematic it turns out it's not, the termination is the two thru hole 100 ohm resistors in parallel, pretty obvious once you actually look at it..... Instead it's in series with the input, it's burnt open yet channel A seems to sort of work at low frequencies....

Further up on channel A is evidence of either a previous repair (or attempt to repair) or a bodge. There also seems to be more flux residue around the transistors on channel A compared to channel B so I think perhaps someone has been in here before me. The instrument did arrive with intact cal stickers though (for whatever that's worth).


After the input section comes the reference oscillator and main chipset, a pair of custom ASICs. This instrument did not come with any options (except the GPIB) but adding a better oscillator should (if needed) be fairly straight forward, the pinheader for it (BU105) is hiding just under the GPIB interface PCB.


The digital section with CPU, memory and logic. Ribbon cable to the front panel obviously:


I did create a stitched together ~8800x4600 pixel image of the instrument, turned out quite nice, but it's a little too big to upload (26MB).

Anyway, once I get to further troubleshooting and (hopefully) reparing the unit I'll report back. In the meantime, if anyone has experience with this instrument please let me know.

[PTR_1275]

How did you go with this repair? Find anything more or have you run into a dead end?

[H.O]

I haven't actually got around to it yet :-(
Guess it goes to show how much I actually needed a new frequency counter doesn't it?
But now that you've reminded me I'll see if I can at least replace that 47ohm resistor and see if that's all it is.

[HighVoltage]

I always wanted one of these but never bought one. And now, that I have my Agilent counters, I will probably never buy one.
But I am really looking forward to your repair.

[H.O]

Thank you guys for prompting me to open this puppy up again!
I used my WXMT tweezers to remove the damaged 47ohm resistor and the WXMP to solder a new (47.5ohm because that's what I had at hand) one in its place and now look at it:



Channel A now counts properly and gives the same readings as channel B. Both slightly low compared to what I dial in on my Rigol DG4162 but I don't know much about its reference so I can't say if it's the Rigol or the FLUKE (or both) that's out. Need a better 10MHz reference.... Tried both 1V and 50mV (RMS) amplitude and various frequencies between 1kHz and 150MHz and it seems to work fine as far as the previously found problem is concerned. That's not to say there can't be other issues with it of course.

All in all the fix took less time that it took to write this post.....

I need to play around with it and have a read of the manual.

[Gyro]

Both slightly low compared to what I dial in on my Rigol DG4162 but I don't know much about its reference so I can't say if it's the Rigol or the FLUKE (or both) that's out. Need a better 10MHz reference....

Hard to say, Philips/Fluke were/are rather good at oscillators but the board looks configured to use the basic on-board one (non-TCXO, non-oven) which has aging of <5e-7 /month, <5e-6 /year. It would be worth trying to search out and fit one of the higher stability oscillator options.

Edit: Quite surprising given the resolution, it's always possible that someone removed the high stability osc before you got your hands on it... the rear panel label will tell you: PM6680B/_2_ is TCXO, /_4_ and /_5_ are OCXOs (/_1_ is the basic on-board one).

[dom0]

The PM6680 and it's little brother PM6685 (twins) superseded the PM6652 / PM6654 as Philip's laboratory counters (they all have a normal, non-precision, un-ovenized, un-compensated quarz oscillators as the standard time base, ovens where an option), which were most likely connected to a local lab reference anyway.

Edit: If memory serves me right there was however a variant of the 6680 with a built-in rubidium standard.

[Samogon]

I know that it is really old topic, but short question:
I you apply power does it boot or stays in standby mode? There is switch JP16 which is controls power on behavour, but mine can be in two modes always on and always off, no logic controlled boot.

Thank you

[H.O]

When power is applied to mine it powers up completely, display on etc.

[Samogon]

Oh thanks, i though it is some sort of failure.