Author Topic: Re: Repair: PM3394 - "Can't communicate with UFO" (not quite solved)  (Read 18517 times)

0 Members and 1 Guest are viewing this topic.

Offline grumpydocTopic starter

  • Super Contributor
  • ***
  • Posts: 2905
  • Country: gb
So, just acquired a dead Phillips PM3394 combiscope.

On power up there is the normal clicking of attenuator relays but the front panel is unresponsive and there is a very dim display which says - as far as I can see (had to turn the lights out)

           CAN'T COMMUNICATE WITH UFO

at least that's what I think it says.

Haven't, as yet, started fault finding - has anyone with any experience of these 'scopes and, if so, do you know what the above means. I can't see any module on the block diagram which fits with the initials "UFO" although I wonder whether the problem is that the main CPU can't talk to the front pannel CPU.
« Last Edit: November 14, 2015, 03:53:11 pm by grumpydoc »
 

Offline kfitch42

  • Frequent Contributor
  • **
  • Posts: 300
  • Country: us
Re: PM3394 - Can't communicate with UFO
« Reply #1 on: December 10, 2014, 08:50:04 pm »
I suspect you are in the right neighborhood. I googled the model number and found another EEVBlog forum thread ( https://www.eevblog.com/forum/beginners/help-with-oscilloscope-purcase/ ) with a link to the user manual: http://assets.fluke.com/manuals/PM3370B_umeng0000.pdf

Page 6-14 makes a reference to UFO being related to the "frontpanel control software."

P.S. I didn't even try googling the exact error message... I suspect the results would actually decrease my IQ.

P.P.S. Here is a service manual in case anyone cares. Looks like its just scanned images, so no searching the document for UFO. http://www.download-service-manuals.com/en/manual.php?file=FlukePhilips-4778.pdf
« Last Edit: December 10, 2014, 08:59:51 pm by kfitch42 »
 

Online PA0PBZ

  • Super Contributor
  • ***
  • Posts: 5120
  • Country: nl
Re: PM3394 - Can't communicate with UFO
« Reply #2 on: December 10, 2014, 08:58:35 pm »
The dim display and the error message should make you check the voltages of course.
Keyboard error: Press F1 to continue.
 

Offline grumpydocTopic starter

  • Super Contributor
  • ***
  • Posts: 2905
  • Country: gb
Re: PM3394 - Can't communicate with UFO
« Reply #3 on: December 10, 2014, 09:26:42 pm »
I have the service manual - two copies in fact - one for the original series and one for the "A" variant - the link for the "B" variant might well come in useful.

Reading down the service manual further confirms that the 83C652 front pannel processor is, indeed, labelled "UFO" in the schematics and the error message means the main processor can't talk to it.

This is possibly something as simple as a dodgy ribbon cable - or else the front panel CPU is hosed which will mean waiting for a(nother) parts unit or sacrificing my 3382

Voltages and ripple will, of course, be checked in due course.

It might be a while before I can take it apart - for one thing the bench is piled high with crap errr, other repairs at the moment.
 

Offline tszaboo

  • Super Contributor
  • ***
  • Posts: 7302
  • Country: nl
  • Current job: ATEX product design
Re: PM3394 - Can't communicate with UFO
« Reply #4 on: December 10, 2014, 11:09:08 pm »
The UFO must be out of range. Try signaling them, maybe if they return.
 

Offline Alex Eisenhut

  • Super Contributor
  • ***
  • Posts: 3330
  • Country: ca
  • Place text here.
Re: PM3394 - Can't communicate with UFO
« Reply #5 on: December 11, 2014, 04:09:30 am »
Play this to the scope

Hoarder of 8-bit Commodore relics and 1960s Tektronix 500-series stuff. Unconventional interior decorator.
 

Offline oldway

  • Super Contributor
  • ***
  • !
  • Posts: 2172
Re: PM3394 - Can't communicate with UFO
« Reply #6 on: December 12, 2014, 01:38:43 pm »
These oscilloscopes are not easy to repair because the plugin boards are not accessible and we must have extension cards.
For the front end  board, it is even more difficult because it would be necessary to have a test bench to make measurements on both sides of the board.
It seems to me essential to have another oscilloscope (working) of the same type to test the cards one by one.
For those how are unaware about, it use a very expensive TDA1541 DAC sought by audiophiles.
 

Offline grumpydocTopic starter

  • Super Contributor
  • ***
  • Posts: 2905
  • Country: gb
Re: PM3394 - Can't communicate with UFO
« Reply #7 on: December 12, 2014, 04:26:01 pm »
Quote
These oscilloscopes are not easy to repair because the plugin boards are not accessible and we must have extension cards.
For the front end  board, it is even more difficult because it would be necessary to have a test bench to make measurements on both sides of the board.
It seems to me essential to have another oscilloscope (working) of the same type to test the cards one by one.
For those how are unaware about, it use a very expensive TDA1541 DAC sought by audiophiles.

Yes, they look quite nice and modular when you get it open - only to realise that the boards are rather difficult to actually get at! Phillips had an extender card but I suspect that it would be hard to find one these days.

I'll post some photos when I get the chance

Fortunately I have plenty of other 'scopes for debugging and a 3382 which I can swap boards from if necessary.

Initial investigations suggest I'll have to do some propper debugging.

 

Offline HighVoltage

  • Super Contributor
  • ***
  • Posts: 5453
  • Country: de
Re: PM3394 - Can't communicate with UFO
« Reply #8 on: December 12, 2014, 05:40:00 pm »
I have many of these Fluke Philips Combiscopes and they are usually pretty easy to repair.

Well, sometimes they are beyond repair...
I just bought a 2 channel version on ebay, the PM3380B and it arrived as junk.
Everything is broken in to pieces, most likely from shipping damage.
May be I can salvage some PCB's and the tube.

Here are some pictures, it shows the modular build up of the scope pretty well.
These scopes were state of the art in the 1990th.

 
There are 3 kinds of people in this world, those who can count and those who can not.
 

Offline grumpydocTopic starter

  • Super Contributor
  • ***
  • Posts: 2905
  • Country: gb
Re: PM3394 - Can't communicate with UFO
« Reply #9 on: December 12, 2014, 06:11:17 pm »
Quote
I just bought a 2 channel version on ebay, the PM3380B and it arrived as junk.
Everything is broken in to pieces, most likely from shipping damage.

Ouch, I hope that you get a refund.
 

Online Shock

  • Super Contributor
  • ***
  • Posts: 4200
  • Country: au
Re: PM3394 - Can't communicate with UFO
« Reply #10 on: December 13, 2014, 03:20:42 am »
Check if the power supply X2 filter cap and the other electrolytic filter caps have been replaced and if not do so. They can cause complete failure and intermittent problems.
Soldering/Rework: Pace ADS200, Pace MBT350
Multimeters: Fluke 189, 87V, 117, 112   >>> WANTED STUFF <<<
Oszilloskopen: Lecroy 9314, Phillips PM3065, Tektronix 2215a, 314
 

Offline grumpydocTopic starter

  • Super Contributor
  • ***
  • Posts: 2905
  • Country: gb
Re: PM3394 - Can't communicate with UFO
« Reply #11 on: December 13, 2014, 09:29:39 pm »
Quote
Check if the power supply X2 filter cap and the other electrolytic filter caps have been replaced and if not do so. They can cause complete failure and intermittent problems.
Check that.

However at the moment I'm focussing on the front panel CPU and its communications with the main CPU.

And it doesn't look promising for the front panel 83C652, it has 5V, the oscillator is running and there is a stream of reset pulses from the main processor at about 25ms intervals but no activity on any pin (mental note - double check ALE).

I think I will try swapping the front panel & CPU from the 3382 and see what happens tomorrow,

HighVoltage - you wouldn't be interested in selling the front panel board from that broken 'scope you have, would you?
 

Offline HighVoltage

  • Super Contributor
  • ***
  • Posts: 5453
  • Country: de
Re: PM3394 - Can't communicate with UFO
« Reply #12 on: December 14, 2014, 09:49:17 am »
HighVoltage - you wouldn't be interested in selling the front panel board from that broken 'scope you have, would you?

Well, I have to wait what the claim with the shipping company and the seller will bring.
I filed a claim and hope of a partial refund.
The seller claims he packed it well
The shipping company claimed we signed for the acceptance of the package and because of that we released all possible claims.

But.... lets say we will find a way ...
You have a 4 Channel scope and this one I showed is a 2 Channel scope
Yours is the PM3394 and the one I have is the "B" version (2 generations later)
I am not sure, if this panel would work at all.

Let me remove the front panel PCB from the frame and I will take some detailed pictures for you.
There are 3 kinds of people in this world, those who can count and those who can not.
 

Offline grumpydocTopic starter

  • Super Contributor
  • ***
  • Posts: 2905
  • Country: gb
Re: PM3394 - Can't communicate with UFO
« Reply #13 on: December 14, 2014, 11:10:48 am »
Quote
Well, I have to wait what the claim with the shipping company and the seller will bring.
I filed a claim and hope of a partial refund.
The seller claims he packed it well
The shipping company claimed we signed for the acceptance of the package and because of that we released all possible claims.
You certainly need to sort out any claim for the damage. Ebay might be of help but it's always possible they will insist on you sending it back to the vendor.

It is a pain that the courier is denying any responsibility, but you can't always inspect the contents of a package when it is delivered. Often one of my neighbours will take a package for me which means I don't get to check it until I arrive home from work.

Even if you can get around the "you accepted the package" argument you may well then hit the "it has a CRT" argument. Though most couriers will take CRTs (usually with a size limit) they will not insure them.

However I would argue that the seller didn't really pack it well. Unconstrained polystyrene chips and recycled Amazon crushed paper packaging are not really appropriate or adequate for an item the size and weight of this 'scope.

Quote
But.... lets say we will find a way ...
You have a 4 Channel scope and this one I showed is a 2 Channel scope
Yours is the PM3394 and the one I have is the "B" version (2 generations later)
I am not sure, if this panel would work at all.
Looking at the service manuals I can't see any differences in the front panel PCB between versions - except for the rotary encoders for channel 3+4 position and obviously the rubber membrane. No mention is made of checking a firmware version on the front panel processor.

I'm going to try swapping the front panel from the 3382 later - if the scope seems to work (no guarantee of that) I will probably then see if the PCBs can be swapped.

Quote
Let me remove the front panel PCB from the frame and I will take some detailed pictures for you.
OK, that will be useful. Thanks
 

Offline grumpydocTopic starter

  • Super Contributor
  • ***
  • Posts: 2905
  • Country: gb
Re: PM3394 - Can't communicate with UFO
« Reply #14 on: December 14, 2014, 01:19:57 pm »
OK, not the front panel board/CPU..... works perfectly in the 3382.

Time to swap the A3 main CPU board.
 

Offline HighVoltage

  • Super Contributor
  • ***
  • Posts: 5453
  • Country: de
Re: PM3394 - Can't communicate with UFO
« Reply #15 on: December 14, 2014, 02:19:31 pm »
However I would argue that the seller didn't really pack it well. Unconstrained polystyrene chips and recycled Amazon crushed paper packaging are not really appropriate or adequate for an item the size and weight of this 'scope.
Are you having magic powers or how did you know what the seller used for packing this scope.
(see picture)
Making the matter worse, on the side of the box it said "Hand Towel ZIGZAG FOLD"
So, I can not even blame the shipping company 100%
There are 3 kinds of people in this world, those who can count and those who can not.
 

Offline HighVoltage

  • Super Contributor
  • ***
  • Posts: 5453
  • Country: de
Re: PM3394 - Can't communicate with UFO
« Reply #16 on: December 14, 2014, 02:32:26 pm »
Although you said it is not the Front Panel PCB, here are the pictures of the PCB as it just came out of this horribly broken PM3380B.

It seems the only thing missing to make this a 4 Channel PCB are the two missing encoders.

There are 3 kinds of people in this world, those who can count and those who can not.
 

Offline grumpydocTopic starter

  • Super Contributor
  • ***
  • Posts: 2905
  • Country: gb
Re: PM3394 - Can't communicate with UFO
« Reply #17 on: December 14, 2014, 03:27:16 pm »
Quote
Are you having magic powers or how did you know what the seller used for packing this scope.
It was obvious from your first picture.

Quote
It seems the only thing missing to make this a 4 Channel PCB are the two missing encoders.
Typically the case for equipment with several options is to have as much as possible of the design common to evry variant with some unused/unpopulated.

In fact I've just taken the opportunity to give the front panel of the 3382 a good clean and I note that the rubber switch membrane was obviously common as well - switches in blanked off positions have been simply cut from the membrane.
 

Offline grumpydocTopic starter

  • Super Contributor
  • ***
  • Posts: 2905
  • Country: gb
Re: PM3394 - Can't communicate with UFO
« Reply #18 on: December 14, 2014, 08:55:39 pm »
OK, that was enlightening. A bit.

Swapping the 3394 CPU board to the 3382 moves the error with it.

This is good and bad news - the good news is that the problem must be on that board and there are only four signals forming the link between the two boards, as well as the data bus.

I'm assuming the bus is OK or nothing would work.

That leaves two strobes (a read and write) which are generated respectively from a 74LS32 and a 74LS138. It will probably be easier just to replace these than to try to debug the lines. I don't keep 74LS in surface mount so will have to order some. I do, for some odd reason, have some 74ACT138's so I might try one of those tomorrow. The LS32 OR gate shows signs of having been replaced already anyway.

The not so good (AKA bad) news is that if it's not one of the strobes the "data available" lines are all that's left and they go to the CPU (an 80C196). If swapping the LS138 and/or 32 don't do the trick I might try the processor - they can still (just) be found on ebay but only one or two sellers seem to have the QFP version, most having the PLCC. In fact I do have one kicking around but it's a PLCC, not QFP :(

After that I don't think I can be bothered to hack on the CPU card and will just keep an eye out for a parts 'scope.

Swapping the good CPU card to the 3394 gets a good bright trace - yay!  Sadly however there's no readout text - not even the very faint text with the original CPU card.

The text is generated from the acquisition card so I'm not sure why it has disappeared - except that getting the CPU card out is no picnic.

The manual says remove the retaining screw, remove the screw holding the 9-way RS232 connector to the back panel and the ribbon cable down to the A1 analogue board and the CPU card will lift up.

The manual lies!

You can see the problem in HighVoltage's photos - there are a set of coax cables which loop over the CPU board RS-232 connector to plug into the acquisition board. These make it impossible to remove the board. In the end I unplugged them from the analogue board to get some leeway but not before trying to get the required slack by unplugging the acquisition board - which promptly got wedged and I needed to manhandle it rather roughly to get it out. Or rather not get it out as there are yet more cables plugged which become obvious as you try to lift it up. So I unplugged the other end pr the cables needed to free the CPU card, however I'm now wondering whether I damaged the A8 card.

Hopefully it's OK and it's just a Z-drive issue but I need to sit down with the schematics and trace out the signals.
« Last Edit: December 14, 2014, 09:00:20 pm by grumpydoc »
 

Offline grumpydocTopic starter

  • Super Contributor
  • ***
  • Posts: 2905
  • Country: gb
Re: Repair: PM3394 - "Can't communicate with UFO"
« Reply #19 on: December 15, 2014, 11:14:42 pm »
OK, as promised - some photos to complement those from HighVoltage, and the start of repair efforts.

All the images are linked to a full-size version.so just click on the image if you want a more detailed view.

First of all this is what the inside should look like when it hasn't been mangled by poor packaging and being used as a football by the courier. This is my 3382 rather than the 3394 but they are essentially identical.



The lowest card in the picture above is the acquisition and data processing card A8 - basically the DSO part of the 'scope plus triggering and text generation. It has several custom chips, two 200Ms/s ADC's and a TMS 320C25 DSP. I'll probably need to have a more detailed look at this card to figure out why I'm not getting any readout text from the 3394.

The next card up is the main CPU A3, then the power supply A6 and at the top, just visible the final X, Y and Z amp board A2.

Underneath is the A1 board carrying the analogue front end and signal paths.



The white coax carries differential signals from each of the four channels to the digital board. If you look at the first photo you can see how these loop up and over the RS232 connector on the CPU card, making it almost impossible to remove, in the end I just unplugged them - I don't need the digital section to work to debug the CPU to UFO link. They are neatly colour coded though - if you look carefully at the full sized version you can see that each of the connectors on the board has a coloured ring which matches the one on the cable.

The black coax brings some signals back - mostly related to TV triggering. Just about every chip on the A1 board apart from the TEA1017 serial to parallel converter is a custom Phillips part so if these die the only option is a swap from another 'scope. Even the TEA1017 is obsolete so not all that easy to come by.



The front panel connector is in a spectacularly awkward place. I've ended up cursing this several times. Worryingly there appears to be provision for a further card in the chassis - I don't know what variant of the 'scope would have that but it would be impossible (rather than merely very difficult) to connect and disconnect the front panel ribbon cable with it in place.



So if you remember the problem was that ET couldn't phone home.


A total of 4 signals provide the handshake between the main CPU and the front panel processor. Looking at the circuit diagram for the CPU card 2 of these signals are generated from address decode logic and two connect directly to the CPU. This is difficult to see on the reduced size version so do look at the original if you want to see what's going on..



CPU card processor side - there's really very little going on on this card. The main processor is an 80C196 in qfp package which you can see just above/to the right of the three capacitors. There's a GAL used for decode (DIL chip with label) and two flash ROMS. To the left of the board there is the (unpopulated) GPIB option.



The rear of the CPU card is slightly busier. Over on the left the largest chip is a 32kx8 SRAM ffor the main CPU, various discreet logic in the middle and more unpopulated areas for the GPIB option. We're interested in the 74LS32 and 74LS138 just above/to the right of the SRAM.



Actually debugging the signals is not all that feasible. For one thing the getting a probe to the rear of the card would be impossible. For another we're probably trying to capture just one event as the CPU card resets the front processor and tries to establish comms with it so the easiest way of debugging the fault is just to swap suspect parts and see what happens. I'll do the 74LS138 first as a) I happen to have a replacement handy and b) when I first got the CPU card out it was obvious that the 74LS32 has already been replaced.

So, out with the hot air gun on the rework station and here are the pads cleaned and tidied for the replacement.



The spare turned out to be an 74AHCT138. I really can't remember why I bought these. Still it's just as well that I did :)

Looks like I cooked the conformal a bit too much - oh well, I'm still getting used to working and re-working SMD parts. At least I didn't lift any pads this time unlike my FE5680A :(



Turn the 'scope back on and - TA-DA



So, there you are - we're up an running on the CPU board at least. I hope you all enjoyed the photos  and going through a bit of logical hardware debugging (plus a smidgen of luck).

Hopefully I can get the text readout sorted as well. However I think the first thing is to swap the CPU cards back into their respective 'scopes and probably close up the 3382 and put it back on the bench as a 'scope, not a test-bed. I doubt swapping the 3394 A8 card into the 3382 will give me any useful information and swapping my 3382 A8 card into the 3394 won't quite work correctly anyway as the 100MHz and 200MHz A8 cards are slightly different (and will probably be detected as a wrong hardware version by the CPU).

Watch this space.
« Last Edit: December 15, 2014, 11:34:01 pm by grumpydoc »
 

Offline kripton2035

  • Super Contributor
  • ***
  • Posts: 2572
  • Country: fr
    • kripton2035 schematics repository
Re: Repair: PM3394 - "Can't communicate with UFO"
« Reply #20 on: December 16, 2014, 07:13:30 am »
couldnt be explained better. thanks for this.
 

Online tautech

  • Super Contributor
  • ***
  • Posts: 28136
  • Country: nz
  • Taupaki Technologies Ltd. Siglent Distributor NZ.
    • Taupaki Technologies Ltd.
Re: Repair: PM3394 - "Can't communicate with UFO"
« Reply #21 on: December 16, 2014, 07:53:27 am »
Quote
So, there you are - we're up an running on the CPU board at least. I hope you all enjoyed the photos  and going through a bit of logical hardware debugging (plus a smidgen of luck).
Sure did enjoy that, thanks for sharing.  :-+
I can see you smile from here.  ;)
Avid Rabid Hobbyist
Siglent Youtube channel: https://www.youtube.com/@SiglentVideo/videos
 

Offline grumpydocTopic starter

  • Super Contributor
  • ***
  • Posts: 2905
  • Country: gb
Re: Repair: PM3394 - "Can't communicate with UFO"
« Reply #22 on: December 16, 2014, 01:07:22 pm »
Quote
I can see you smile from here.   ;)
Yes, it's always satisfying to replace a component, switch the equipment on and find it is now working  :-/O
« Last Edit: December 16, 2014, 01:54:02 pm by grumpydoc »
 

Offline HighVoltage

  • Super Contributor
  • ***
  • Posts: 5453
  • Country: de
Re: Repair: PM3394 - "Can't communicate with UFO"
« Reply #23 on: December 16, 2014, 05:34:23 pm »
Great success on your repair.

The beauty is, that you can get all the schematics service manuals for these scopes.
The top of the line is the PM3394B with all options installed and extended memory.
Even in today's market with so many digital scopes, I do prefer for some measurements these Philips/Fluke scopes.

If you do a calibration, you need the access code and it is the same for all of these scopes: "32415"
With that you can adjust almost everything in the service menu.

One thing I have been looking for is an extender to have the boards exposed, while the scope is running.
It is a little difficult to take measurements on the installed boards.
Philips had these extenders developed for the service labs and I have seen them, but never found one that
was available.



There are 3 kinds of people in this world, those who can count and those who can not.
 

Offline grumpydocTopic starter

  • Super Contributor
  • ***
  • Posts: 2905
  • Country: gb
Re: Repair: PM3394 - "Can't communicate with UFO"
« Reply #24 on: December 16, 2014, 06:55:55 pm »
Quote
Great success on your repair.
Thanks - it looks like I have at least one more problem to fix before the 3394 is properly back on its feet. And, of course, fixing one problem in hardware of this age often only enables you to discover the next.

Quote
The beauty is, that you can get all the schematics service manuals for these scopes.
The top of the line is the PM3394B with all options installed and extended memory.
Even in today's market with so many digital scopes, I do prefer for some measurements these Philips/Fluke scopes.
Not just the schematics but have a fighting chance of actually debugging and repairing the hardware.

Quote
If you do a calibration, you need the access code and it is the same for all of these scopes: "32415"
With that you can adjust almost everything in the service menu.
Thanks, I think that code is in the service manual

Quote
One thing I have been looking for is an extender to have the boards exposed, while the scope is running.
It is a little difficult to take measurements on the installed boards.
Philips had these extenders developed for the service labs and I have seen them, but never found one that
was available.
Yes - I presume the connectors are DIN41612 so it should be possible to construct an extender.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf