Philips / Fluke PM6654C High Resolution Counter - Teardown

[CaptnYellowShirt]

Here are some pictures from my most recent acquisition -- a Philips PM6654C. Sadly, the previous owner seems to have been a life-long smoker. And as such, he left this very nice piece of lab equipment covered with a thick coating of yellow-tar-gunk. In the process of alcohol washing the entire thing, I thought I'd take a few pictures and share them with you all.

In a related thought, why did they allow smoking in the MCC...



... yeah, thumbs up to lung cancer, buddy.

Anyway, this is the base model PM6654C. It doesn't come with any of the exotic time-base options, just a single bare quartz crystal can folded right over onto to the FR4... doesn't inspire a lot of confidence in a device that can display 9 1/2 digits and claims to have a resolution of 2ns.

No worries though... my lab has a Rubidium standard that pipes a solid 10MHz right into the back of this device. Seriously though, I'd just as well put a piece of duct tape over about half of those digits if I had to rely on the non-compensated crystal. That being said, the ol' counter did alright for itself right out of the box: Reporting 1.000000185e3 on a 1kHz (0.1ppm) signal divided down off of the Rubidium clock and 1.000000656e6 on a 1MHz signal -( 0.6ppm)

[CaptnYellowShirt]

First up is option plate for this thing.... a sad state of affairs....

The next picture is the first inside view (from the rear) after you open the device up. In the back left, you have the GPIB interface board. Directly in front of that is the location for the high frequency counter option (>100Mhz). You'll notice a small SMB-style coax just dangling off into nothing there. To the left of that, you'll see a PCB labeled area that reads "Remove Crystal KT101 when optional oscillator in installed" -- its not, so nothing's there. We'll see the 'high tech' crystal in a bit... its dark secrets are too much to be reviled all at once.

All the caps look good... no leakage there... but what's all this gunk? That's right, 20+ years of cigarette smoke. You can't see it in the final picture, but that swab has a yellow tinge to it. This thing was in such bad shape that your fingers would get sticky after handling it. You could even tell what position the knobs on the front had been placed in for a long time because the knurling on one side contained a yellow gook that you couldn't get out with windex.

[CaptnYellowShirt]

Onward and downward....

Pulled the GPIB interface out and now I'm faced with this nice RF box. Could this be where the time base is hidden? Wanna, wanna.... nope. However, there are some interesting looking inductors (?) on the right-hand side -- the square spirals printed on the PCB. There's a blank spot to the right of each spiral has an identical spiral printed on the reverse side of the PCB. Anybody know what these are?

Ok, well if the time base isn't on that board... its gotta be below this section right? Where else could it be?

[CaptnYellowShirt]

Picture #4 (attached) is like a game of Where's Waldo? Trust me, the time base is in there, but if you can't find it have a look at the 2nd attachment...



I know, I know... when you first look at the front of this thing and see the 9 1/2 digits, you're thinking....



Sadly, if you don't get a timer upgrade option, this digit-driven-dream suddenly turns into....

[CaptnYellowShirt]

The last picture here is showing the device after its been cleaned out (yuck), hooked up to the Rubidium Standard, and placed in service.

Ahhh... the sweet sweet taste of 9 digits of perfection....

[Vgkid]

Thanks for the pictures. I figured they could have at least gone with a tcxo as the base oscillator.

[grumpydoc]

Thanks for the pictures. I figured they could have at least gone with a tcxo as the base oscillator.

I think the presumption was that a frequency standard would be used.

[burkm]

Well, not the youngest of threads... 
But just want to point to a blog, where an OCXO for this Fluke/Philips PM66xx family of counters is been done as a DIY project with source for PCB and listing of parts:
http://syncchannel.blogspot.de/2016/04/flukephilips-pm66xx-frequency-counter.html

So there is still hope...

[Fraser]

I own one of these units.

I must warn that the capacitors in mine had failed exuding a strang sweet small. Any owner of this unit should consider fitting new PSU capacitors.

I fitted a OCXO to my unit. It was easy to do as the service manual us available and the OCXO was designed to just replace the simple crystal oscillator an thee is space inside the case for it.

Fraser

[chickenHeadKnob]

To Fraser and Capt'n.
Uhg! Those blue ribbon cable female connectors give me bad bad flash-backs. In the early 80's  we used them all over the place on multiple articles. Started getting field returns traced to intermittent contact on the the pin sockets at the ends of the connector. They would begin failing after only 3-4 insertion remove cycles as the leaf spring contacts would get bent out of shape from people pulling them off at an angle, like numpties do. I am tempted to say they were T&B/ Ansley brand but can't remember for sure. We switched to samtec back then and never looked back, those bastards cost us so much time and grief I wouldn't use a blue connector ever again, even if I was assured the design had been improved. If you have intermittent problems after cleaning and generally 'dicking' with your unit look to those first.

[SaabFAN]

I just repaired my unit.
I'm not sure if the OCXO I fitted had anything to do with it (I now have it sitting in a little box, acting as an external frequency standard), but the 8741 Microcontroller that controls the display and the keyboard apparently had lost its memory or got damaged.

I managed to program a new one and get the counter back to life with Software I found online.
So it probably isn't a bad idead to read all the chips in there and refresh the program inside of them.
Unfortunately, the Willem Programmer for the parallel port or a self-made programmer is required to program the 8741 Micro.
The Firmware available online is the R.4 Version btw. And the Firmware inside the Micro and the EPROMs for the CPU have to be matched, or the counter won't work at all.

What I found surprising btw. is the fact that it can be powered from a 17 to 24V DC-Supply.

[ZigmundRat]

Just giving this a bump since I saw mention that the PM665x service manual was attainable. I've done a good bit of searching but only come up with operator manuals. Any link to the service manual would be much appreciated.

[ZigmundRat]

Just for posterity, I'm updating this thread. The PM6652 and PM6654 service manuals are available on the KO4BB archive site - http://www.ko4bb.com