Solartron/Schlumberger 7150 plus: "FAIL 1" ?

[Blogs14]

I went to use my Solartron 7150 multimeter yesterday, but when I turned it on I was greeted with "FAIL 1" on the display and no relay sounds like it would usually make on power up.

I found a copy of the service manual online, and opened the meter up. Poking around with the 'scope as suggested in the manual revealed that there are no signals on the TX and RX lines which link the floating and eartly logic boards, and furthermore there is no waveform on pin 40 of IC103.

IC103 has power, and the reset line was high (although power cycling a few times the reset line was sometimes high, sometimes low), which is where I got confused. Looking through the datasheet for IC103, it seems to imply that there should always be a waveform on pin 40 - so the clock isn't running?

Is there anyone else with a 7150 who could clarify how the waveform output on pin 40 should work? I really hope the chip isn't dead because I guess it'd be difficult to find a replacement at a justifiable price...

[Andy Watson]

it seems to imply that there should always be a waveform on pin 40 - so the clock isn't running?

Is the clock being generated on pins 2 and 3 of IC103?

[jkf1000]

Not familiar with that particular instrument but generaly the reset line should be held low for a few (maybe upto a couple of hundred) milli seconds to enable correct startup process. Probe VCC and reset lines at the same time while powering up. Trigger on channel 1 connected to VCC with channel 2 on the reset line, and you should see the delay before the reset then goes to high. Set your timebase to around 20-50ms and use single shot mode.. Karl

[Blogs14]

it seems to imply that there should always be a waveform on pin 40 - so the clock isn't running?

Is the clock being generated on pins 2 and 3 of IC103?

No, they're just sitting around DC 1.5 V.

Not familiar with that particular instrument but generaly the reset line should be held low for a few (maybe upto a couple of hundred) milli seconds to enable correct startup process. Probe VCC and reset lines at the same time while powering up. Trigger on channel 1 connected to VCC with channel 2 on the reset line, and you should see the delay before the reset then goes to high. Set your timebase to around 20-50ms and use single shot mode.. Karl

Seems like sometimes reset behaves as it should, sometimes it doesn't. Seems random whether it works or not having power cycled a few times.

(Waveforms attached)

[Andy Watson]

it seems to imply that there should always be a waveform on pin 40 - so the clock isn't running?

Is the clock being generated on pins 2 and 3 of IC103?

No, they're just sitting around DC 1.5 V.

If I'm reading the datasheet correctly, the clock runs irrespective of the reset signal (in fact, the clock is required for the MCU to complete a reset). You need to find out why the clock is not running. I would suspect low crystal activity.

Sanity check - does IC103 have +5V and 0V supplies - measured on its pins.?

[Blogs14]

I can measure 5.08 V between pins 1 and 7, so the MCU does have power. Since there are only few external components involved in the clock generation (2 caps and the crystal), it seems like there isn't much to go wrong, unless the fault is internal to the MCU.

[Blogs14]

Scratch that, I was just poking around C101 and C102 with the multimeter and the clock started running!

A capacitor problem then maybe?

[Andy Watson]

A capacitor problem then maybe?

Poking the pins with your probe may have been sufficient to jolt the oscillator into operation. Does it power-up? Examine the solder joints - possibly re-heat them. I would be inclined to replace the crystal.

[Blogs14]

Yes, it seems to power up pretty consistently - wasn't expecting that. I re-soldered the caps and the crystal; next time I'm ordering components I'll replace them.

Thanks for the help, hopefully this is fixed now.

[Kleinstein]

Normally the small caps at the crystal are rather reliable, especially the TH version that does not get mechanical stress. So failure is more likely the solder joint's or maybe the socket of the µC.

[Blogs14]

I would have said exactly the same if I was giving someone else advice, but I've already re-soldered all the joints, and I reseated the MCU before I posted.

So all I can say is what I've observed: when the clock doesn't start, I tap on one of the capacitors with the multimeter probe, and then it works. Weird huh? Perhaps it's a problem with the crystal, and the vibrations from tapping the capacitor are enough to get the oscillator going, it's electromechanical after all...

[Ordinaryman1971]

I have the same meter with exactly the same problem and according to what I've read it could be one of the optocouplers...
Don't have an access to the service manual at the moment but will try to link it a bit later.
It took me so long to find the manual that the meter got buried under other "more important" projects.

[Blogs14]

Yes, a fault with the optocouplers would result in the same error message going on what's written in the service manual.

Manuals available here: http://www.ko4bb.com/getsimple/index.php?id=manuals&dir=06_Misc_Test_Equipment/Solartron/Solartron_7150

[we]

I replaced C301 (2200uF/16V) to fix Fail 1