Keithley 2000: some symbols too bright

[exe]

Hi guys,

I think I screwed replacement of VFD on my Keithley 2000. One symbols is too bright and ERR is displayed. Could you please advice?

What could go wrong: I damaged wiring, at least two traces got broken. I repaired them and checked all contacts on VFD anyhow connected to drivers.

But now I'm puzzled: how is it possible one symbols is brighter than others? And why error is displayed? May be there is a short and multiple segments shorted somehow so they are always on (= no multiplexing) and that causes brightness problems and wrong segment ("err") to be displayed?

[Cubdriver]

I've no idea exactly how that's set up, but my guess would be that whatever is supposed to select that digit during the multiplexing scan is shorted, or somehow on all the time, as you suggested.  Perhaps the common line for that digit is shorted to ground?  Can you get any pictures of the traces that were damaged?

In the meantime, I'd try to limit how long it's on to reduce the risk of damaging the overly bright segments/parts of the display.

-Pat

[TheSteve]

Check the resistance between all of the pins - it doesn't take much to mess it up. Unless the driver is damaged I'm sure you'll find some lower resistance between the pins, removing it will fix the issue.

[exe]

Perhaps the common line for that digit is shorted to ground?  Can you get any pictures of the traces that were damaged?

Unfortunately, traces are under the display, I'd need to desolder it first. But I don't have desoldering gun. Desoldering braid proved to be useless on this board.

I checked all adjacent pins on display and drivers, no shorts . I also quickly checked waveforms on pins, no obvious problems (but I'm yet to read driver's datasheet).

Another idea is that one trace is broken. I checked that all pins are connected to drivers, but some pins I think connect to two drivers. This is not what I can check. Quick googling didn't get me exact schematic of the front panel.

I'm going to check for actual errors reported by the unit, may be this "ERR" segment is actually telling something.

Can I exclude the possibility of having a damaged VFD? I have no idea how to check it, especially in-circuit.

[Kleinstein]

It is unlikely that this due to a broken VFD - a broken driver chip is more likely.

In principle a VFD is kind of nice for checking in circuit, as the display input are essentially open.

The fault might be more like a line connected to the supply / GND, so that some segments are always on.

Pins connecting to two drivers are suspect. 

[exe]

I ran display test, looks like there is a short somewhere because offensive segments lights up simultaneously with other segments. To my understanding, this means there is no short to Vcc or GND. Now I need to figure out what pins got shorted...

Update: I don't think it's a short, I'm too good for making shorts on through-hole components More likely I destroyed a trace between two drivers. I'm trying to figure out which one. Schematic would help a lot, but I don't see any. I'll try to find a photo of the board...

Here is the video with the problem: https://youtu.be/q1GPIojpKFA .

[exe]

Huh, I thoroughly checked traces once again, no evidence shorts / disconnects (it doesn't mean they are not present).

Now I'm experimenting with the old display to see if I can reproduce the issue or localize the problem.

[saturnin]

If it helps, here is a photo of K2000 display board without VFD (sorry, it is not the best quality).

[exe]

If it helps, here is a photo of K2000 display board without VFD (sorry, it is not the best quality).

This helps a lot, thank you!

So, now I see each pin is driven by only one driver, good. I also checked connections once again, all looks good.

Okay, now I have only two ideas. It's either broken driver (but why?) or... flux residue conducts a little bit somewhere. These VFDs seems to be very sensitive, they light up even if I touch pins by hands. So, I'm removing residue first...

[exe]

TL:DR: Bottom driver is broken.

I cleaned the board from flux, no result. So I decided to swap drivers to confirm that one of them is broken. And guess what! The display died . Looking closely I found one pin didn't make contact after resoldering. I connected it with a jump wire and problem... inverted. Now I have one digit never show up. Damn... But! Found another pin that didn't make contact after resoldering. I fixed it and...

Now one row is always enabled. Using old display I traced the pin in charge of this row. Then I traced what this pin used to do before swapping drivers. Turned out it's the same pin created problems before! So I think I need to replace bottom IC.

Thank you guys for help.

[exe]

I just replaced the driver and now it works (though it took me three attempts and one extra IC to solder it right).

[Mr. Scram]

I just replaced the driver and now it works (though it took me three attempts and one extra IC to solder it right).

Hooray!

[Cubdriver]

Excellent news - I'm happy to hear.   

-Pat