Hmmmm....
Before I watched your video my suspicion was that maybe you had inadvertently set the meter ballistics to "slow"...
But after seeing what's going on, man...I don't know, and I'm not all that familiar with Amek desks...
Boswell ? Kurt Foster ?
And didn't we have an RO Member who was rebuilding an Amek console a few years back? Was it Joel?
@Asaf Smilan ...
Is this happening every time you stop signal? Or is it intermittent?
Has this been happening for a while, or did it just start?
Are you noticing any other issues... Noise? Hum? Any sudden increase or decrease in signal?
Assuming you've made no recent mods or other repairs to the desk...
Just thinking out loud....

Seems like an internal power problem. Some capacitors may need to be changed..

Double check the phantom power is working properly, problems with the two matched resistors can mean there is a DC imbalance and the input meters are seeing voltage, which eventually discharges via the capacitor and the meters return to normal. Unlikely to come and go, but if there is a switch to turn it off try it. Had this on a soundtracks desk in the 90's. You can get the same effect when you connect a sub mixer that has phantom accidentally turned on.

There are several significant observations to make about the behaviour of the metering shown in the video.

Firstly, all the channels are affected, so that, for example, it's unlikely to be a phantom power problem on any one channel.

Secondly, the fault gives similar, but not the same, magnitude of effect on every channel, and corresponding decay times.

Thirdly, there are not many places in a complex analogue system such as the Amek Angela that have time constants of order 1 minute.

The initial suspect area is the power supply rails, as Paul suggested. I've not been able to find any Angela schematics on line, so a guess would be that the design of the metering section is not proofed against variations in the internal power rail voltages.

A cause of the problem could be leaky reservoir capacitors, or high-impedance contacts at the rectifier diodes or at the voltage regulators. If I were sitting next to the console with its lid removed, I would put a digital voltmeter on the power rails feeding the metering board and see how much variation in rail voltage could be invoked by feeding a signal generator or some loud music through many channels at once.

If you accidentally apply power to a Soundcraft mixer (the one I was using was a Live 8-32) input - I had it's output plugged into channels 1 and 2 of a Yamaha, with phantom turned on accidentally - and all the Soundcraft meters, including the groups all lit up to the same level. sticking power up the output of a mixer I can imagine causing problems in the output section, but it managed to get to every channel. The cause of this fault will be a strange one. The trouble with phantom power is that any fault will be applied to one point on the desk, but once it's there, it can spread because all the channels share so much - the axes and groups, PFL lines and stuff like that. A fault in the master section could leak back through the grouping on old style desks. I had an old Mackie desk years ago that when you turned the power supply on - every meter kicked and fell gently back. Much defy common sense.

Watching the video also shows that while music is playing leds stay on longer then they should between the individual channel peaks....if you have music only one channel does any other channel leds turn on?

I've already replaced all the electrolytes on the meters PSU and it doesn't made any change.
Since this problem is happening in all channels / meters can we say that the fault happens before the meter circuits?
Attached the schematic:
https://i.ibb.co/jgD34qy/angela-meter-schematic.jpg
At this moment the console is installed in a way that preventing me from access to the meters PCBs ant to the power distribution board.
Next time I'll have access to those parts I'll start by checking the diodes and resistors on the power distribution board?

The meter circuit you posted is not capable by itself of creating the sort of fault you report. That means you have to look a bit further afield. Do you have the schematics of the remainder of the mixer and its power supply?

I have the schematics of the input and master channels, but I don't have the meters PSU schematic.
When I've measured the voltage coming from the meters PSU it seems to be right.
I've also noticed that when I'm turning off the meters PSU the positive voltage LED indicator on the PSU panel turns off immediately while the negative voltage LED indicator turns off after 10 seconds.
When I'm disconnecting the meters PSU from the console, both LEDs turns off together after a minute or so.

Asaf Smilan, post: 460849, member: 51171 wrote: When I'm disconnecting the meters PSU from the console, both LEDs turns off together after a minute or so.

I was following you up to this point. Could you explain what action you are describing here?

Also confused...Are you suggesting that more current is coming from the positive supply to the Meter bridge than from the negative supply?

I hope this short video will make it clearer

Could be normal.

Most circuitry in the meter bar uses the positive rail, only a small part uses the negative rail.

That's supposing the PSU supplies the same amount of current on both rails...

In the first part of your video when you had the control cable connected to the console, was the console powered or unpowered? The behaviour in the second part (the meter circuit unplugged from the console) I would consider completely normal.

The implication is that the positive rail of the meter PSU is back-powering some of the circuitry in the console. Without access to more schematics, it may be difficult to take this any further, including forming an opinion on whether it is related to your original problem of the meters not returning to zero in a reasonable time.

The LEDs behaves the same way with the the console powered / unpowered.

I'll try to find the meters PSU schematics.

I've marked in the the actual voltages I've measured on the 'Power Distribution Board'

What should be the voltage on the O/L REF point (pin 4 after the trimmer)?
I can set the voltage to anything between +1.5V to +16.3V.

That adjustment sets the signal level above which it's deemed to be overload. It's arbitary, but I would set it about +12V.

Thanks, I've set it for about +12V.
Now what other schematics you would like to see that can help you to understand this problem (the too long decay time of the LEDs)?
I have the schematics for the input channels and for the master channel.
Would it help?

Here's the schematic of the meters PSU.
*It's the same design but without the +48V section.

Your mixer doesn't have phantom?

The phantom comes from the main PSU.
This PSU producing +17.5V / -17.5V only for the meters.

The PSU is definitely asymmetrical.

The negative side only has the LM338K, the positive side is beefed up with an MJ2955, despite both primaries showing a 5A fuse. Bit weird and could explain why one LED stays on longer.

cyrano, post: 460903, member: 51139 wrote: The PSU is definitely asymmetrical.

The negative side only has the LM338K, the positive side is beefed up with an MJ2955, despite both primaries showing a 5A fuse. Bit weird and could explain why one LED stays on longer.

Can I make the PSU symmetrical by doing this tweak (eliminate the TIP32 / MJ2955 section on the positive voltage circuit)?

The schematic either is not the correct one for the meter power supply, or else Amek have used a old mixer supply for the meter unit, possibly not fully populated. The extra current capability on the positive rail is not relevant to the misbehaviour of the meters in normal operation.

The problem as described is that a short duration high level signal on one channel affects the meter display on other channels, coupled with the meters failing to return to a low level in the expected time. It would be easy to get diverted chasing a supposed problem to do with what happens on powerdown, or to go tinkering with the apparent asymmetry of the power regulators.

The schematic of the positive and negative power rails reflecting in 100% the meters PSU pcb.
This PSU unit came with this console from the factory.
I found some pictures of other Angela consoles with this type of PSU for the meters.
https://reverb.com/item/17155390-amek-light-and-meters-power-supply-black-for-angela-console-in-very-good-condition

http://vintageaudioworkshop.blogspot.com/2012/06/amek-angela-mixing-desk-part-2.html

https://www.njuskalo.hr/image-bigger/miksete/amek-angela-mki-slika-109976939.jpg

So it seems like Amek have used an old mixer supply for the meter unit, not fully populated (with out the +48V rail.

The problem with the too long decay time on the LEDs meters is happening either when I'm sending the signal to a single meter or when the signal being sent to more than one meter.
There's no situation where a signal on one channel affects the meter display on other channels.
Whenever a signal is being send to a meter (and the meter is illuminate for more than few seconds) it takes around one minute for the LED's to fade away.

Asaf Smilan, post: 460904, member: 51171 wrote: Can I make the PSU symmetrical by doing this tweak (eliminate the TIP32 / MJ2955 section on the positive voltage circuit)?

You could, but why would you?

I don't think the PSU itself is related to the problem. The LEDs on the PSU just show different current use from the meter bar, as I would expect.

If it was mine, I'd check the two tantalums on the output. Just to make sure. These fail after time and sometimes fail a bit strange. Again, don't think this is related to the meter problem.

Just guessing, I'd expect the meter problem itself be to leaking DC on the signal lines from somewhere. But without schematics, the only thing you can do, is open it up and check it.

cyrano, post: 460909, member: 51139 wrote: I'd check the two tantalums on the output. Just to make sure. These fail after time and sometimes fail a bit strange. Again, don't think this is related to the meter problem.

On my PSU C5 (and C10) have been replaced by the former owner to electrostatics.
Do you think tantalums would make a difference?

Just guessing, I'd expect the meter problem itself be to leaking DC on the signal lines from somewhere. But without schematics, the only thing you can do, is open it up and check it.

I have schematics for both input and master channels.
That would be helpful?

I've just tested the meters with two more PSUs, one from an Allen & Heath console (+17V / -17V at 6.5A) and a much smaller PSU from a small 8 channels mixer (+15V / -15V at 1.5A that couls only handle one group of meters) and it acts the same way with all 3 PSUs.
Since this problem occurs with three different PSUs there's must be something wrong with the 'Power Distribution Board' or with *ALL* the meter cards?
Any idea what else can I check?

Asaf Smilan, post: 460910, member: 51171 wrote: On my PSU C5 (and C10) have been replaced by the former owner to electrostatics.
Do you think tantalums would make a difference?

No, not in a PSU. Especially if that PSU is feeding a meter bar.

I have schematics for both input and master channels.
That would be helpful?

I think you either need schematics for the meter bar, or you need to take it apart. Photo's of one meter board and the power distribution board could be helpful...

This is the schematic of the meter bar and the power distribution board

I'm sorry, missed the schematics earlier. Like Boswell already said, there's nothing in there that would explain the symptoms you are seeing.

That leaves the suspicion that somewhere in the mixer, all DC blocking caps need to be replaced.

One thing you could do, if you have access to a scope, is have a look at the PSU's voltages, to see if there's some sort of RFI. But that's a longshot...

I'm still wondering if a loose ground somewhere in the mixer could cause this?

It seems like there's some RFI in the PSU's positive and negative rails.
I don't know if it's too much or it's reasonable?
I've took this measurements with the PSU and the meter bridge cases open so I guess it will be reduced when everything will be close.

I've pulled all the channels from the console in groups of four until I've left with only one channel and still no change.
So it seems like it's not a problem in a single channel...

I found the problem!
*ALL* the 4 pin molex connectors that feeding the signal to the meter card was plugged in a wrong way with pin 1 connected to pin 4, pin 2 connected to pin 3, pin 3 connected to pin 2 and pin 14 connected to pin 1.
According to the bands of the weirs it seems like it was connected this way for very long time...
As you can see here, now all meters working well!

Thank you everybody for your help!

Nice glad you solved it!

I'm pleased you found the problem, and also that no damage appears to have been caused by the mis-orientation of the plugs.

Something to check first...

I had a DDA a while back with a similar problem. Random dropouts on random channels. The owner had "checked" bus cables by pulling a few channels. Everything LOOKED like it was plugged in correctly.

I pulled all channels, ready to clean the mixer. No cleaning needed. In fact, the cleanest mixer internally I ever saw. After reseating all the channels, everything just worked.

When I was pulling cards, I was amazed how easy the connectors came out...

They shouldn't use connectors that make it possible to plug in the wrong way.