Optocoupler woes

[Mr_bibbles]

Hey everyone,

Got an Orange Matamp clone (one off, no schematic) on the bench, symptoms are low Clean Channel volume. Distortion channel sounds nice and loud. Guessing the amp is a MKII clone, but the schematic may not help considering the issue....

This is a custom one-off channel switching amplifier, I was able to trace the signal all the way through the preamp past the master volume but the signal is reduced dramatically after an Optocoupler. There are two optocouplers tied to a g5v relay, each one feeding the respective channe'ls master volume into the PI.

After bypassing the optocoupler completely, the signal returns, so we know that either the optocoupler or something connected to it (relay...) is bad. My question istwofold;

1. Is there an easy / recommended way to divide and conquer from here? The issue seems obvious (or at least isolated to this area) but I have never sourced optocouplers that aren't used for tremelos.

2. Without a schematic, is there a way to measure and source an equivalent optocoupler?

Thanks in advance for all you lovely folks.

[g1]

What number is on the opto? Half of it should be the optical section, and you need to know if that side is getting it's appropriate 'command'.

[Enzo]

An opto is a light shining on a photocell. BY measuring voltage across the light you can tell if it is lighting or not. The other side is a photocell, and you can measure resistance across that while turning the light off and on.

[J M Fahey]

Agree and add: check that the LED is shining or not.
Maybe there is nothing *broken* , simply Opto is doing its job, if LED does not light, CDS resistor will stay open, so no sound through.

To avoid cutting opto open to check whether Led lights or not, and to verify it´s alive, :
1) confirm that there is voltage across its legs
2) that voltage should be about 1.9V

Post clear readable Opto picture, any labels on it?

[Mick Bailey]

The most frequent failure mode with vactrols is the LDR side failing and it stays high resistance when on. The LED side is usually very reliable (except in some Mesa amps where the LED was run with excessive current from the switching matrix).

When replacing LDRs in tube amps it's important to consider the voltage rating of the LDR side. If you've established the original is faulty and you have a part number, take a look at the rating compared to the signal level. Sometimes it can be very close and reduce reliability, causing eventual damage to the LDR. Maybe take a look at the VTL5C2 as a replacement.

[Mr_bibbles]

Thanks y'all, this is incredibly helpful from all corners. FIrst, the optocouplers are all (obviously) sealed up with heat shrink so I'm avoiding taking it off until I have measurements. Here's what I've found.

1. BOTH LED's in this pair of optocouplers shows 4V across their legs when each respective channel is activated. This means at least the relay is working correctly, feeding each LED their voltages, but it does not ensure that the LEDs are turning on, right? It could theoretically still be passing the voltage and not lighting up? Perhaps I'm way off though and this is enough to know that the LDR side is faulty.

2. The offending optocoupler does NOT change resistance when the channel is selected, while the other channels shows a clear change from it's <30Mohms resistance when "unsignaled".

So this seems like everything is pointing towards the common failure mode of LDR staying at high resistance. At this point, I'll remove the heat shrink and examine to see if the LED is lighting and what the number code is on the LDR. This was incredibly helpful, but if anyone has anything to add or correct I'm happy to hear it

[Mick Bailey]

Commercial vactrols are sealed units - no heatshrink. This points to them being simply a regular LED and LDR pairing, maybe DIY. The way I double-check vactrols is to use two pairs of hook probes connected up with the amp switched off and discharged.. One pair goes the the LED side, one to the LDR side. Using two DMMs, set the LDR one for resistance and the LED one for diode test. When the diode is forward biased by the meter, there's enough current to light the LED and you should see the LDR resistance drop to around 50-100 ohms or so. In the absence of two DMMS use a 9v battery with a series 1K Ohm resistor on the LED side.

Spare a though for anyone working on Mesa Tremoverb amps. There are 28 vactrols, many stacked in pairs or threes and most early amps have numerous vactrol failures.

[g1]

Commercial vactrols are sealed units - no heatshrink. This points to them being simply a regular LED and LDR pairing, maybe DIY.

As he mentioned 4V across the opto, I wonder if there is a led & resistor in there.

[J M Fahey]

If heatshrinked then they are definitely homemade (I use nothing else) so feel free to cut heat shrink open, worst case you replace it.
Led lighting is self evident, any resistor will be visible, CDS cell must react to ambient light so testing is straightforward.

[Mr_bibbles]

As he mentioned 4V across the opto, I wonder if there is a led & resistor in there.

It was 4V across the Diode, no resistor in sight, but this unit has been working for a year like this. It once I was able to get a good idea of what was happening inside via the ohmage measurements, it was no problem to slice it open and see a perfectly fine Varistor and an unlit LED. Swapped a known good one in, shrunk it back up to the varistor and everything is sounding good again and switching as it should!