PM3050 Phillips Scope: 1x/10x problem on channel a

[Mounty2010]

hello. first of all i've to say that i'm german and i'm trying my best to write in english.
i have a problem with my Philips PM 3050 scope and maybe someone can help me
or point me the right way to fix it.

second, i'm not new to using scopes, i also own a cheap rigol 1052 dso and i'm
some kind of familiar in measuring NF signals, doing it for years now. but i have
no idea what kind of problem my PM 3050 has - i'm trying to explain now:

- i'm using 2 probes (1x/10x, set up to 1x) connected to A and B of the scope
- i connect both of them to the CAL output of the scope
- CAL output delivers an ~ 1.86khz, 1.2Vpp square wave
- i press the "auto" button and now something strange happens:

- the scope sets Channel A to 5.0V / DIV
- the scope sets Channel B to 0.5V / DIV

both signals looks the same, they perfectly align BUT channel A and B have different
volt/div settings ! I found out, that channel A is the problem, because B reads about
2.4 divs * 0.5V = 1.2Vpp (which is okay) but channel A is measuring wrong by factor 10,
like the probe was set wrong - but it isnt. when i switch/change the probes, the problem
allways remains on channel A - so its not probe related. channel A thinks, there is an input
signal of 12Vpp but we all know, that it is only 1.2Vpp (like channel B says and rigol 1052)

So whats up with this scope ? i have attached a photo so you can see what im talking about.

i hope someone can tell me whats wrong with this scope. i really need your help and i know
this is the right place for finding well versed electrician or electronic engineers here at eevblog.

[Shock]

I have checked my Philips service manual (similar model to yours) and you should be able to rule out most of the attenuator PCB being a problem by triggering some high and low frequency signals on each channel with the other probe disconnected when not in use, then test varied voltage signals.  Try also removing the power from the oscilloscope a few times. See if you can isolate further it this way, you would expect the problem channel to always read incorrect no matter what.

The most likely cause I think is the probe indicator circuit, there is an IC that measures the resistance of the probe identification pin for each channel to ground, and then scales the LCD V/DIV accordingly. There is a capacitor on each of those inputs as well, compare their values and ESR.

Get the service manual and read through that section. That circuit should be located on the Front Unit PCB and is static sensitive, take precautions.

[HighVoltage]

I would think it is a simple setting of the scope.

1. Make sure you can repeat the problem by exchanging the probes.

2. This scope has two turning knobs to set the vertical amplification seamlessly.
Make sure that both knobs are set to the CAL position.

[voltz]

Have to agree with Shock there.
As any attenuation problems like probes or internal faults would show the range going lower not higher to display the signal. Its almost as if the signals is being made bigger, not smaller. Of course thats not really whats happening here though.
Yes, also looking at the diagram, I see the probe indicator pins (right next to the BNC sockets of A and B) detect a resistance to indicate what range to display. This seems right, because your display looks fine and yet the LCD is showing exactly x10 higher than it should on channel A.
The indicator pins are meant to be used with probe types PHILIPS PM 8926/59. These have a special outer ring that contacts the sens pin when connected. Inside the probe there is a resistance to indicate what attenuation the probe is.
The chip in the 'front unit' is marked as D7004. This chip reads the sens resistance and changes scale.
It is very likely your scope has a fault here.

[Mounty2010]

thanks for your answers guys, i got the service manual and took a look at it.

The most likely cause I think is the probe indicator circuit, there is an IC that measures the resistance of the probe identification pin for each channel to ground, and then scales the LCD V/DIV accordingly. There is a capacitor on each of those inputs as well, compare their values and ESR.

The indicator pins are meant to be used with probe types PHILIPS PM 8926/59. These have a special outer ring
that contacts the sens pin when connected. Inside the probe there is a resistance to indicate what attenuation
the probe is. The chip in the 'front unit' is marked as D7004. This chip reads the sens resistance and changes scale.
It is very likely your scope has a fault here.

Seems you both were right - i checked the "inidcator pins" and shorted them to the bnc's ground and that happens:

when shortening the indicator pin of channel A to the BNC's ground -> nothing happens
when shortening the indicator pin of channel B to the BNC's ground -> the display switches from 1V/DIV to 10V/DIV

So for me it looks like channel A thinks there is a 10x probe connected, even when there is no probe connected.
Shortening the indicator pin of channel A doesnt do anything as it is already shortened somewhere else maybe.

Now the really problem begins: the attenuator board was not meant to be removed outside the scope easily
i read the service manual and there is no way described, how to remove the attenuator board. im stuck and dont
know what to do now.

while waiting for your suggestions i'm going to measure the resistance between indication pin and bnc's ground
to find out, if there is any difference between channel A and B (because i think, channel A is shortened somewhere).

edit:

i measured resistance between the "indicatior pin" and the BNC's ground (scope turned off) and thats what i got:

channel A: 1.820 Ohm (1.8k)
channel B: 4.150 Ohm (4.1k)

so this seems to be the problem and it must be on the attenuation board - which i cant remove, because it is the
only board not meant to be removed easily

[voltz]

when shortening the indicator pin of channel A to the BNC's ground -> nothing happens
when shortening the indicator pin of channel B to the BNC's ground -> the display switches from 1V/DIV to 10V/DIV

Correct and backs up the fault again. Its possible other resistances will give different ranges on B as stated in the manual.

channel A: 1.820 Ohm (1.8k)
channel B: 4.150 Ohm (4.1k)

I would expect these values to be equal (looking at the diagram). It does indicate again there could be fault on A sens pin / circuit

Dont forgot to check the capacitors as 'Shock' mentioned. Possibly not the chip at all. The usual process of elimination needed here.
Isn't the chip located on the 'front unit' as it says? Is that not on the front panel and easy to get to?

Anyway, good luck.

[Mounty2010]

I took a look into the service manual and found chapter 10.1.7 "probe indicator" (page 109).
It says, pin 3 (A) and pin 16 (B) are the probe inputs on IC D7004 (Philips 0Q044). So i went
to chapter 10.3 (page 110) and found out, that probe A is PIN33, probe B PIN31 on the X7001
connector of the front unit. I removed the front unit and measured resistance between:

- PIN33 (probe A) and PIN11 (GND): 1.7 kOhm
- PIN31 (probe B) and PIN11 (GND): 3.8 kOhm

For me that means, the failure should be (as some of you guessed) definitely on the front panel.

On the schematic diagramm (page 111) in the right upper corner is the D7004 probe indicator ic.
It shows that both probe inputs have 10nF capacitors (C7004 & C7006), nothing more, thats all.
In other words does that mean for me, at least one of these capacitors has reached its lifetime
and must be faulty - am i right ? what is the right replacement - 10nF ceramic but what voltage ?

[voltz]

It could be the caps or the chip, thats the point.

Remove both 10nF capacitors (C7004 & C7006).

reconnect everything and switch on, test if channel A works, if so, you found the fault and renew both of them.

If not, then the chip is at fault and needs replacing.

[Mounty2010]

I already did that when you was writing your post i removed the two capacitors but nothing changed.
The inputs 3 and 16 of the IC D7004 have different resistance (1.7k and about 3.8k) to ground, so it is
the IC that has to be replaced ... *damn*. the capacitors where damaged when de-soldering (thought
they were faulty, so i didnt care about when removing them) so i have to replace them - what voltage
should they have ? (50, 63, 100V, 400V or these 2kV ones ?)

[voltz]

Yes, it does look like the chip. I hope you can find one, thats an odd looking number.
The cap voltage should be 63v i think. It does not need to be high voltage as its just for the sens pin with a low voltage on it. The voltage value probably is not critical as long as its not too low. The type is ceramic, and yes they do crumble and break sometimes, especially if old. Just replace both, and the chip.

[Mounty2010]

you were right again - i didnt find anything using the numbers "00Q0044" or "41215MC" 

in that moment i read your last comment i got the point with the voltage of the capacitors.
its just a low-voltage IC so i wont expect high level signals on these inputs. thanks again.

[Shock]

I would first reflow that area and associated circuit with a some new flux and leaded solder. Then when the caps arrive put them in and measure the supply voltages to that IC (see schematic), As well as ensure the IC is grounded properly.

After that I would desolder the IC (the reflow should make it easier to remove) and install a socket if there is clearance, then I'd modify another socket to piggyback and pin lift the IC to further check any conditions which make the input resistances change.

Looks like there are more caps on those input lines possibly around the Pre Amp PCB so make sure you're testing both sides of that circuit. Trace the connectors in your schematic and map out the whole circuit then check the resistances to ground with that side isolated (if you haven't already) and also check any caps and any other components that could affect your inputs.

So after ensuring whats coming into 33 and 31 is acceptable I'd probably end up swapping the IC inputs over if it didn't look like it was going to cause any issues.

[Mounty2010]

I measured the voltages on the D7004 IC and this is what i got (see image please).
Seems like the 5V and -12V are okay, but the IC doesnt change output level for A
when connecting different probes (or shortening the probe indicator to BNC's GND).
As you can see, the IC works fine on channel B (switches the output signal correctly).

I think the D7004 is faulty and this is where the repair stops, because i dont see any
possibility to get that suspect Philips OQ0044 IC. On page 187 of the service manual
there is a list of components but, guess what - the D7004 is not listed. There is D7001,
D7002, D7003, D7006. I think this chip could be a custom made one for this device.

That means, i have to give up, i've no other choice  its not worth buying another
(maybe cheap and faulty) PM 3050 to steal the D7004 from it. I think its best to sell
this device and let the buyer know, that on channel A only 10x probes are suitable.

Thanks to everyone who helped, really. without your comments i wouldn't be able to
trace the problem and finally reach the D7004 IC. thanks for all your help and support.

[voltz]

I spent quite some time searching for the chip. No good, came to a dead end.
But i have emailed NXP, who took over from Philips semiconductor. They may or may not get back with data. But it does look unlikely.

Assuming the chip is definitely not available and no equivalent, then you still have the possibility to carefully remove the chip taking great care not the pull up tracks on the upper side of the PCB, and then use links or low ohm resistors to pre-set the logic state on pins 7-12 and 6-13.
Basically, it means the probe detect will not work anymore but at least the scope will read the correct scales as x1 probes.
Its a work around, sure, and not great. But without the chip, what other options are there?.
Maybe NXP get back with news would give it a few days...
 

[Mounty2010]

hey voltz, thanks for your effort (mailing nxp) and your idea to find another solution 

i like the idea to put a switch onto the scope (maybe right beside the bnc connectors).
i just have to emulate the digital 1 (+4V) on Pin 8 of the D7006 IC and that would fix
the 1x probe thing ... thats a good idea, i will think about it

[Shock]

Check further down in the service manual it's probably listed but yes it may be a custom IC.

You can still find old parts for these scopes though. The PM30xx and the PM33xx (check service manuals) had many similar parts, you may find someone who has a wrecked beyond repair scope or even multiples laying about. These were very popular in Europe.

It's not a bad problem either just a slight cosmetic one.

[voltz]

Final update, NXP came back negative. No part listed in their database.

[Mounty2010]

hey voltz, thanks for your effort and this information. now its up to me to find a solution, i'm thinking about the switch 

[oldway]

I would choose that solution : let the indication of the two channels in x1 fixed position as it is the case in many oscilloscopes that do not have such probe selector x1 x10.

To do this, remove the integrated circuit D7004 and connect pins 6,8,13 and 15 of D7006 to 0 or + 5V respecting the original logic levels.

[Mounty2010]

to finish this thread, here is what i did. i attached a switch to the front panel and wired it
according to the schematic below. i had to cut and remove (desoldering) the pins 6 + 7 of
the D7004 IC. then i used the holes in the circuit board to solder the wires of my switch.

by cutting and desoldering pin 6 + 7 of D7004 the input pins 6 + 15 of D7006 were open,
so i had to use a 4.5 kohm pull-down resistor to get a clear "zero" (0V) on these inputs.

now everything works fine using the manual switch. left = 0V (x10), right = 5V (x1)