AB763 Stunflower

[Ten Over]

It turns out that the AES parts are XVive. In fact I can't seem to locate anyone (that seems reputable) selling a vactec vtl5c1. In what way are the newer parts deficient for use in a trem-o-nator?

Vactrol @1mA
18.5k Average Resistance
44k High
4.4k Low
Sample Size: 20 Units


Chinese @ 1mA
4.6k Average Resistance
12k High
0.6k Low
Sample Size: 10 Units


Vactrol @ 0.1mA
3.9M Average Resistance
6.6M High
1.0M Low
Sample Size: 10 Units


Chinese @ 0.1mA
8.7k Average Resistance
14k High
2.4k Low
Sample Size: 10 Units


Chinese @ 0.01mA
34.9k Average Resistance
52k High
21.5k Low
Sample Size: 10 Units


So the Amazon Chinese ones have a far lower resistance range than the Vactrol ones that I bought from AES a number of years ago. The Chinese resistance range squats the 12AX7 down too much and makes it sound like the neon roach tremolo. It also makes the overall volume drop when heavy tremolo is engaged.

Three of the Chinese units never got below 11k no matter how much current I gave them. These would be useless in a trem-o-nator because there isn't enough variation in resistance.

[romberg]

Thank you so much for posting these numbers. They will help alot in knowing what I'm looking for the part to do. My guess is that I'll not luck out and have useable chinese vtl5c1. I may have to search for an alternate opto-coupler to use that is still in production. Maybe something like a nsl-32 would work? I'll have to start looking at data sheets...

Mike

[sluckey]

Don't give up on the xvive yet. Some people have had no issues with them.

[romberg]

My bag on repro vtl5c1s appeared today. So, I started to test one out. Right out of the gate, I discovered that the plastic case is *NOT* light proof. With no connection at all to the LED I could get wildly differing resistances just by waving my hand around near the device. And the max resistance (with no current) was very low.

So, I slipped one into some heat shrink tubing and clamped off each end (leaving some room for the leads to poke out each side). This gave more promising results. The photocell now read over 40m (max on my meter) when no current was fed through the LED.

Next I hooked the led side up to a power supply, meters and a pot to measure some data points. I got this:

cvtl5c1.png

Looks similar to the Xvive data. I'm assuming the useful range for the oscillator tube is in the neighborhood of 0-2mA But there is not much variation (relative to 50k) after about 80uA. Since the LED also has to bias the oscillator tube, I don't see a way to feed it less current and get it working in the steeper part of it's curve.

I'll spot check the rest of my grab bag. But I'm not very hopeful. Unless the oscillator dips down to cut off or very close to it. I guess I'm not exactly sure what the current range would be in operation.

Mike

[pdf64]

Given that the optocoupler coating seems to be translucent (FFS! ), did you take those readings with the optocoupler wrapped in something opaque, eg black insulation tape?

[romberg]

Given that the optocoupler coating seems to be translucent (FFS! ), did you take those readings with the optocoupler wrapped in something opaque, eg black insulation tape?

Yes. The part was inserted into some heat shrink tubing. Then each end of the tube was clamped with a forceps with a length of component lead hangin outside. This made the largest difference in bringing up the resistance values.

But then I also draped the entire setup with a heavy cloth and then covered that with a cardboard box. These two steps were probably overkill as the readings on my resistance meter did not change much after doing it.

So, I'm reasonably sure that the vast majority of photons hitting the photoresistor came from the LED. The part does produce a nice linear change in resistance to current. But the line is not very steep in the region of interest for biasing a tube.

Mike

[Ten Over]

My bag on repro vtl5c1s appeared today. So, I started to test one out. Right out of the gate, I discovered that the plastic case is *NOT* light proof. With no connection at all to the LED I could get wildly differing resistances just by waving my hand around near the device. And the max resistance (with no current) was very low.

So, I slipped one into some heat shrink tubing and clamped off each end (leaving some room for the leads to poke out each side). This gave more promising results. The photocell now read over 40m (max on my meter) when no current was fed through the LED.

Next I hooked the led side up to a power supply, meters and a pot to measure some data points. I got this:

cvtl5c1.png

Looks similar to the Xvive data. I'm assuming the useful range for the oscillator tube is in the neighborhood of 0-2mA But there is not much variation (relative to 50k) after about 80uA. Since the LED also has to bias the oscillator tube, I don't see a way to feed it less current and get it working in the steeper part of it's curve.

I'll spot check the rest of my grab bag. But I'm not very hopeful. Unless the oscillator dips down to cut off or very close to it. I guess I'm not exactly sure what the current range would be in operation.

Mike

So I shined a white LED at the Chinese units and the resistance dropped just like you noticed. External light had no effect on the Vactrol units.

Another thing about the Chinese units is that they idle at 1.8V while the Vactrols idle at 1.6V (idle = 1mA).

I drew in a Vactrol curve on your graph for reference. Xvive and Vactrol have the same curves, but somebody at Xvive didn't understand logarithmic graph paper.

[romberg]

So, I tested my remaining 9 opto-couplers all at 1ma and they were all pretty consistently near 540 ohms. That is pretty much spot on for what I measured for the part I made the curve for. So, I did not test them any further. They are pretty consistant. So, I don't think they are rejects from a line making other parts. They just are not parts that match the specs for a vtl5c1.

Now, I was under the impression that all Chinese vtl5c1s were like other products over there and the same thing existed under hundreds of different names. But maybe not. Steve did mention that some folks have had good results with the XVive parts. I noticed that XVive does have it's own data sheet (not just a copy of the original).

https://www.tubesandmore.com/sites/defa ... 1_2019.pdf

It says it has 20k resistance at 1ma. Which is what the trem-o-nator is looking for. So, just maybe not all chinese vtl5c1s are alike. I've ordered one and will soon find out. While I was at it I picked up a 500kL dual ganged pot. If all goes well I may try the other oscillator.


Mike

[Ten Over]

I'm assuming the useful range for the oscillator tube is in the neighborhood of 0-2mA But there is not much variation (relative to 50k) after about 80uA. Since the LED also has to bias the oscillator tube, I don't see a way to feed it less current and get it working in the steeper part of it's curve.

At 400V with a 220k plate resistor, the range is about 0.1mA to 1.5mA. The oscillation goes from about -4.5Vgk to 0Vgk regardless of the DC bias point.

LFO 220k 400V.png

At 400V with a 470k plate resistor, the range is about 0.1mA to 0.78mA.

LFO 470k 400V.png

[romberg]

At 400V with a 220k plate resistor, the range is about 0.1mA to 1.5mA. The oscillation goes from about -4.5Vgk to 0Vgk regardless of the DC bias point.
LFO 220k 400V.png

At 400V with a 470k plate resistor, the range is about 0.1mA to 0.78mA.
LFO 470k 400V.png

Oh! The load line makes alot of sense. Thanks for pointing that out! It seems kinda obvious once you think about it. But for some reason (probably 'cause it is not a conventional gain stage) it never occurred to me. I guess the 0-4.5Vgk thing comes in because it varies from cutoff to saturation.

So, in theory, one could select a 4m plate resistor and get the 0Vg current down a bit under 0.1mA. This might get the chinese opto-couplers into the steep part of their curve. But this is way way down at the bottom of the charts for the 12ax7. I'm just guessing an oscillator with a 4m plate resistor would not work for reasons I don't even know at the moment.

Mike

[Ten Over]

Xvive's came in.

0.1mA
143.2k Average
155k High
129k Low
Sample size: 5


1.0mA
8.3k Average
9.8k High
7.3k Low
Sample size: 5


1.5mA
5.7k Average
6.7k High
5.0k Low
Sample size: 5


1.61 voltage drop across LED @0.5mA

Not sensitive to outside light sources.

Looks exactly like the old VTL5C1's.

[romberg]

My XVive vtl5c1 arrived today (since they cost 10x more than the others I just got one). My tests look similar to yours:

opto-couplers.png

This looks more promising. And now I understand why the 470k plate resistor. You wanna run this thing at low current to operate in the steep part of the range.

Once I finish fitting my chassis and cabinet together, I'm going to switch over to a trem-o-nator.

Mike

[romberg]

Finally got a bit of time to switch my tremolo over to slucky's trem-o-nator. I decided to keep the fender style footswitch setup and did not use a 10k cathode resistor. Everything else is per the trem-o-nator schematic. I've updated the schematic in the first post of this thread to match the amp as built (including the rem-o-nator).

This circuit made a huge differentce in the quality of the tremolo. Here is a scope shot of what the output looked like with the stock tremolo circuit (max intensity):

pre-tnator.png

I think those extra caps I pulled in from the UltraPhonix schematic which keep the thing from ticking really mangle the sinusoidal nature of the effect. For all I knw the McGiver nature of the DIY opto-isolator probably makes the effect vary quite a bit.

And here is a max intensity shot with the Trem-O-Nator:

post-tnator.png

Here the low frequency signal looks much more like a sin wave. Plus the oscillator covers a lower frequency band. It might be nice to tweak this to get a little bit more strength out of the intensity. The trem-o-nator is a bit weaker intensity wise than the stock bug. Not sure how to do this off the top of my head. I'll think about it. My guess is that the stock photo resistor gets to a lower resistance than the XVive vtl5c1.

Overall this is very much an upgrade worth doing for fender style amps with tremolo.

Mike

[Ten Over]

Finally got a bit of time to switch my tremolo over to slucky's trem-o-nator. I decided to keep the fender style footswitch setup and did not use a 10k cathode resistor. Everything else is per the trem-o-nator schematic. I've updated the schematic in the first post of this thread to match the amp as built (including the rem-o-nator).

This circuit made a huge differentce in the quality of the tremolo. Here is a scope shot of what the output looked like with the stock tremolo circuit (max intensity):

pre-tnator.png

I think those extra caps I pulled in from the UltraPhonix schematic which keep the thing from ticking really mangle the sinusoidal nature of the effect. For all I knw the McGiver nature of the DIY opto-isolator probably makes the effect vary quite a bit.

And here is a max intensity shot with the Trem-O-Nator:

post-tnator.png


Here the low frequency signal looks much more like a sin wave. Plus the oscillator covers a lower frequency band. It might be nice to tweak this to get a little bit more strength out of the intensity. The trem-o-nator is a bit weaker intensity wise than the stock bug. Not sure how to do this off the top of my head. I'll think about it. My guess is that the stock photo resistor gets to a lower resistance than the XVive vtl5c1.

Overall this is very much an upgrade worth doing for fender style amps with tremolo.

Mike

The saw-tooth pattern is typical for the roach tremolo.

If you have a 470k plate resistor on the LFO, then try 220k and see if that increases the intensity.

[sluckey]

Hey Mike, would you mind if I put a link to this topic on my Trem-O-Nator webpage? Another success story would be great, especially since you used Xvive.