Simpson 464D multimeter LEDs flickering

[TimeBandit]

I recently noticed that a few of the LED segments in the display for my Simpson 464D multimeter have begun to flicker. I replaced the electrolytic caps about a year ago, and just checked the voltage going to the LED drivers: around 6.5V vs. the 5.3V in the schematic.

I'm thinking about replacing the 2N3638 drivers. Thoughts?

[CaptDon]

Highly doubt the drivers are the problem!!! Your supply voltage is out of spec by over 20%, doesn't that indicate the problem is in the power supply. Maybe shorted regulator / pass transistor with tons of A.C. ripple? Why do you think the driver transistors would cause this problem?

[floobydust]

I recently noticed that a few of the LED segments in the display for my Simpson 464D multimeter have begun to flicker. I replaced the electrolytic caps about a year ago, and just checked the voltage going to the LED drivers: around 6.5V vs. the 5.3V in the schematic.

I'm thinking about replacing the 2N3638 drivers. Thoughts?

If you know a rail is way out of spec, that needs to be fixed first before guessing what is wrong.
The +5.3V rail is unregulated off the transformer. High mains voltage which is common nowadays, might be the problem. I'd check the +15V rail is not too high either, it's not regulated as well.

IC7 DM7447N max. is 5.25V absolute max. 7V, IC5 ADC LD110 max is 6V...

[TimeBandit]

I rechecked the rail voltage using a different multimeter. It's 5.3V as it should be.

A few solder joints, including at pin 16 of IC7 and emitter of Q6, looked suspect, so I reflowed them. No flickering so far.

[floobydust]

Hopefully that was it. I like LED display bench meters so I have a few oldies I keep running.

[TimeBandit]

No luck. It's acting up again. Very randomly though. In the latest instance, a few of the segments disappeared altogether. 

[floobydust]

When the segments disappear, is it something valid in the 7447 datasheet fonts?
Just to see if it's bad codes coming out of the LD110 or if the 7447 is misbehaving. There is a Lamp Test at pin 3 if you ground it. Note the DM is a power pig 60-100mA should run warm.
I have a lot of PMOS get senile with age, the ion implantation seems to fade out and the IC's go bad. I don't like PMOS to TTL interfaces, they are no fun. I would scope the 7447 BCD input pins and see what that looks like, should be TTL levels.
The transistors are strictly doing digit work, so they would cause a digit to stay off or get stuck on.

[TimeBandit]

Turns out that a DM5447 is what's actually installed, which makes sense I suppose since the max supply voltage is 5.5V vs. 5.2V for the 7447, and I'm measuring anywhere between 5.3 and 5.5V at pin 16.

I did the lamp test and all the segments lit up. I still need to check the BCD input pins with my scope.

[TimeBandit]

Ok, here's what I'm seeing at the pins for the B, C, and D inputs.

[floobydust]

What the V/div? Not sure if x10 probe or not.

Roughly, it should go below 0.8V for TTL logic 0, and over 2V for a logic 1 (at Vcc 5V) but a little higher OK for 5.3-5.5V VCC you have.
I think "D" looks a bit sick, unless you have it mixed up with "A" which has Q4 loading it as well. Is it going -2.5V? The LD110 power is OK, -12V and +5.0V?
Old slow switching PMOS usually gives ghosting but you don't see that? So an "8" or "9" could smear between digits (D) possibly.

Attaching schematic snip for those playing at home.

[TimeBandit]

Sorry for the lack of context. I attached some new shots. These are all at 0.2V/div, probe set at X1.

The supply voltages are 5.24V and -11.77V, which look fine.

[TimeBandit]

Ok, I'm thinking that replacing the LD110 (and LD111?) is worth a shot.

[floobydust]

I can't make sense of your scope traces, waveforms. One "D" I think looks bad. What's with the intensity modulation and where is 0V and at 0.2V/div I don't see a logic 1 level getting reached so it's a little confusing to me.

If it's old PMOS (which I suspect it is), heat can flush that out try warm up the LD110 and see if that causes the fault to appear. I think it will have trouble sinking current at the BCD outputs.

[coromonadalix]

i would check your 555 circuit,  i think its the scanning clock for the display, 
if you see the flicker, it may be possible the scanning frequency as drifted lower and your eyes perceive it, i would try to crank it up

check the parts around it for guessing its frequency or with the scope  ...


theses ld110 and ld111    do ressemble a bit like the ca3161 and ca3162 pair,   one is the a/d converter to bcd ouput and the other is the display driver