Opto cell power requirements

[Potato Cakes]

Hello!

I'm finishing an old Forsell Opto comp that I bought mostly completed from a forum member and after tracing down the signal to find out why it's not compressing, I get stuck when I get to the VTL5C1's. The one for the meter requires +24V and the one for audio does not, at least according to the schematic. I'm not so much asking about troubleshooting this board as it's pretty simple but trying to figure out what is going on with the Vactrol and why it wouldn't need power and the other does.

Thanks!

Paul

 

[Potato Cakes]

Also,  I believe that both are possibly bad as I can see voltages being affected by the front panel controls all the way up to the VTL5C1 in the audio path, which is why I am not getting any compression. I've attached a photo of them as I've never seen them look like this.

Thanks!

Paul

 

[shabtek]

The meter opto is waggling the needle by  modulating  current flow from +24...it is purely cosmetic

The other opto is squishing dynamic range by shunting signal to ground.

 

[Potato Cakes]

Thanks for the reply. I don't have any VTL5C1's left, but I have a couple of VTL5C4/2's. I'm wondering if I could just use one side of that just to test if it is the optos.

Thanks!

Paul

 

[Bo Deadly]

Looks like they're leaking goo? That would mean you're torturing them somehow.

From glacing at the datasheet on thonk it says power limit is 175mW which is not much (significantly less than a standard 1/4W resistor) so you have to be careful not to blast the photoresistors too much. Are you really using 24V rails?

There are no DC blocking caps in the schem so it might be important not to put DC through them. If your meter draws a lot of current or if there's a mistake in wiring / power supply causing an op amp output to drift, that could blow up the optos. Does the meter have a current limiting resistor? They both look equally tortured which is strange because they are performing very different functions.

Before you replace them, look at what voltages are across the photoresistors and LEDs with high level signal and while sweeping the controls to make sure the voltages are sane.

 

[Tubetec]

Ive had a wee look into busted vactrols before ,no  goo to be found in em ,most likely a drop of epoxy on one end to keep it together , in any dead ones Ive found so far its the drive side that tends to quit ,simple diode test should tell all .

 

[Potato Cakes]

There is a current limiting resistor in the meter.

And it turns out they are indeed dead. Or tortured, which I think I like better. I am getting about 24ohms between the + and - contacts, which would explain why everything seems to be fine right up to the opto cells. I double checked the contacts on a new VTL5c4/2 and I wasn't to get a reading, which is what I believe is to be the case.

I am running this at +24V.

I don't think a DC blocking cap will work as the signal is converted from AC to DC at D1 and D2, and continues from there to the +/- legs of the LED part of the Vactrol, so I think the DC is necessary.

With a 1kHz tone at 0dBu at the input, at most I will get about 1.8VDC right before it gets to VR5, then disappears as both +/- legs of the VTL5C1 were essentially shorting together and to ground. No matter what voltage I had pre VR5 the voltage on the LED legs were both 0VDC.

I've had this board for a while and always thought the opto cells on it looked funny. Now it seems that they were exploded when I got this PCB.

shabtek, thanks for the info on the optocell function. This new knowledge will prove vital when conquering the world, or at least another opto compressor issue.

Thanks!

Paul

 

[shabtek]

Looks like glue

Yeah pull 'em and test out of circuit... One side is a diode , other is ohms.


I think they have a yellow led in the capsule but whatever is handy could be tacked in to gauge flicker.

Could use other part#  if it has 2 optos in 1 package with suitable modifications

 

[Potato Cakes]

Looks like glue

Yeah pull 'em and test out of circuit... One side is a diode , other is ohms.


I think they have a yellow led in the capsule but whatever is handy could be tacked in to gauge flicker.

Could use other part#  if it has 2 optos in 1 package with suitable modifications

That's what I wound up doing. The LED legs both had the same approximately 20 ohms between them in and out of the circuit.

I think I'm going to just get some replacement VTL5C1's as the BOM specifies as I just want to get the compressor working as I'm cramming in some other circuits with this build before I can close the lid.

Thanks!

Paul

 

[shabtek]

Youd have to look at data sheet  to confirm response

But id just tack in the part you have. The leds in the schemo are in parallel : with the 2 in 1 opto the photoresistors are optically in parallel. The only thing you'd  give up is the optional matching trimmer

 

[Potato Cakes]

Copy that.

I'm going to at least try to get the GR meter to have some level of accuracy, so I'll need the trimmer. I found some new Xvive opto cells that should work for about $20 for the pair.

But this confirms what I was thinking in first place. Good to know that after all this time building stuff I'm also learning something.

Thanks!

Paul

 

[L´Andratté]

I don´t know if you´re a mastering engineer or something, but I have found that the trimmer on the meter is not needed unless you want super accurate metering, if you just want visual feedback on what it is doing and how much, untrimmed is ok (should be the same cell type though).

I´d get two new vtlc1, but check what went wrong before putting them in

There should be not more than 20mA through them a lot, or their life is shortened at some current higher than 50mA (very roughly from memory), they will burn out, what likely happened to your first optos.

EDITptos are different in dynamic range and drive current/resistance correlation. If you´re using a different type, you may have to make some adjustments in sidechain gain etc.

 

[gyraf]

(TL,DR) ...somebody modified those optos in your picture..

Jakob E.

 

[Tubetec]

I think Jacob is right ,the usual epoxy you find on vactrols is black ,which makes sense cause you dont need the possabillity of ambient light messing things up .

 

[Bo Deadly]

Actually that could not have leaked out because it's volume is too large compared to the size of the part. If that is epoxy I would be curious to know how conductive it is. Even a slight conductivity will limit off resistance. Flux is slightly conductive incidentally so you should take care to clean the flux off of the PCB around those cell leads.

Note that one has to be careful not to put too much current through the LED. For example, if you want to test one, putting a voltage across the LED will absolutely and instantly burn it out. You must have a current limiting resistor to limit current to at most 40mA but for testing purposes more like 10mA. Use ohms law to figure out what resistor value should be used for the supply voltage used to test (example: 9V - 1.5V drop / 0.01A = 750 ohms but 680 ohms would be ok).

 

[Potato Cakes]

Like I said, I bought the board from a guy here, so I do not know the history of the vactrols or what kind of mad scientist torturing they had to endure. I am sure the ones that were in there are fried, so there is no need to test them.

The solder pads were pretty clean before I started,  and they have generous amounts of space around them, so it would take a hefty amount of flux for there to potentially cause any sort of shorting. But I will pay closer attention to that when I install the new ones.

Thanks for the info on the LED testing procedure.

Thanks!

Paul

 

[L´Andratté]

Thing is, if blown, it is the led that´s wrong, the resistance should be up at Mohms, the low resistance actually measured points to something else, the epoxy being conductive, if it wasn´t maybe the things would still work...
Anyway best to use two new ones!

 

[rackmonkey]

Late again. To check a vactrol, use a VTVM or other high impedance ohm meter to measure the resistor side of the optocoupler (the side not labeled LED). With the LED off, you should get a very high resistance reading - in the gigohm range. Then wire a 1.5 volt battery and a 100 ohm resistor in series as follows: battery (+) —> resistor —> anode (+) connection of the LED side of the optocoupler, battery (-) —> Cathode of LED. Now measure resistor side again with the LED powered on. You should get a much smaller resistance. IIRC a 5C2 will have about 1000 ohms when measured like this.

Sorry I couldn’t get to this earlier.

BT