Bumble bee ain’t swellin. Resolved

[jimilee]

Like benny said the proper VGS of Q2 is critical to make the effect work. But, I don't think using one out of spec for the build will throw off voltages for the 2n3904 at all.

Worse comes to worst, I might be able to dig up a proper spec'd 2n5457 for you if your new batch doesn't pan out.

I feel like R17 and the resistor below it are swapped the build doc, am I wrong? The schematic and your numbering of the boardhas r17  4.7k below the 1n cap the silkscreen has the 39k there.

[Cybercow]

I feel like R17 and the resistor below it are swapped the build doc, am I wrong? The schematic and your numbering of the boardhas r17  4.7k below the 1n cap the silkscreen has the 39k there.

JimiLee - Wow! Great catch! That does look like they are swapped between the schematic, the parts ID image and the silkscreen. R17 (4K7 in the schematic & BOM) does indeed go to the emitter of Q5 (2N3904) - according to the red & blue trace image. And R20 (39K) is supposed to feed the base of Q5 (2N3904). But the silkscreen has those two resistors R17 (4K7) and R20 (39K) interchanged. That would certainly foul the swell. If there's a 39K resistor on the emitter of Q5 and a 4K7 resistor on the collector of Q5, the whole detection circuit is out of balance.

The 39K resistor should be above the 4K7 resistor where the part ID image says R17 & R20. Still, the silkscreen on the PCB is correct - with the 4K7 resistor going to the emitter. So I would not put that much stock in the part ID image. I think the Part ID image may off for those two resistors.

What value do you have on top position (referring to those two resistors)?

[Aentons]

Brian covered this in the setup part of the build doc but there is a slight discrepancy in the patreon bundle notes that I though I might point out just in case.

The vfe bundle notes state:
"do not install the JFET trimpot. Instead, install a jumper between the left-middle pin of that trimpot and its bottom pin. Use the TWEAK trimpot to fine tune the swell response"

So, it looks like you put a jfet trim pot there, and in your pic it looks cranked all the way over to pin 3, which I think is all the way up. Make sure its all the way down to start.

Also, not sure this matters either, just throwin it out there... the 3362 datasheet says:
-----------------------------
Absolute Minimum Resistance
.................................... 1 % or 2 ohms
(whichever is greater)
-----------------------------
https://www.mouser.com/datasheet/2/54/3362-776956.pdf

So, you are not starting at zero, you are starting at 200 ohms minimum.

[benny_profane]

Did you try swapping the JFET for one in the range indicated? I may have missed it if you did.

[jimilee]

Brian covered this in the setup part of the build doc but there is a slight discrepancy in the patreon bundle notes that I though I might point out just in case.

The vfe bundle notes state:
"do not install the JFET trimpot. Instead, install a jumper between the left-middle pin of that trimpot and its bottom pin. Use the TWEAK trimpot to fine tune the swell response"

So, it looks like you put a jfet trim pot there, and in your pic it looks cranked all the way over to pin 3, which I think is all the way up. Make sure its all the way down to start.

Also, not sure this matters either, just throwin it out there... the 3362 datasheet says:
-----------------------------
Absolute Minimum Resistance
.................................... 1 % or 2 ohms
(whichever is greater)
-----------------------------
https://www.mouser.com/datasheet/2/54/3362-776956.pdf

So, you are not starting at zero, you are starting at 200 ohms minimum.

Nope, I missed the part where he says not to use the trimpot, use a jumper. On page 7, he says to use the trimpot. Where did you say that was again?

I feel like R17 and the resistor below it are swapped the build doc, am I wrong? The schematic and your numbering of the boardhas r17  4.7k below the 1n cap the silkscreen has the 39k there.

JimiLee - Wow! Great catch! That does look like they are swapped between the schematic, the parts ID image and the silkscreen. R17 (4K7 in the schematic & BOM) does indeed go to the emitter of Q5 (2N3904) - according to the red & blue trace image. And R20 (39K) is supposed to feed the base of Q5 (2N3904). But the silkscreen has those two resistors R17 (4K7) and R20 (39K) interchanged. That would certainly foul the swell. If there's a 39K resistor on the emitter of Q5 and a 4K7 resistor on the collector of Q5, the whole detection circuit is out of balance.

The 39K resistor should be above the 4K7 resistor where the part ID image says R17 & R20. Still, the silkscreen on the PCB is correct - with the 4K7 resistor going to the emitter. So I would not put that much stock in the part ID image. I think the Part ID image may off for those two resistors.

What value do you have on top position (referring to those two resistors)?

I have a 39k in the r17 position instead of a 4.7k

Did you try swapping the JFET for one in the range indicated? I may have missed it if you did.

Not yet, I was working on getting my voltage right. The Jfets that came on Saturday were all below Spec. I figured I'd try to get the voltages right before I put too much more into the jfets. I've been working on some other stuff in the mean time.

[Aentons]

Brian's build doc (page 7) has this:
-----------------------
3. Peter recommended not installing the 20k trim and using a fixed 22k resistor in its place. Don't do this. Use the trimmer! I built two Bumblebees, one with the resistor and one with the trimmer. I could not get the swell to work using the fixed resistor method (and that was with a pre-tested 2n5457). IMO, the 20k trimmer is essential to getting this to work optimally. It may be he used JFETs with a very narrow cut-off range and was able to get away with the fixed resistor. But, my experience was different.

4. To set up the swell effect do the following: set the switch in its middle position, the MIX knob all the way counter-clockwise and all other pots to their middle position. Leave TWEAK in the center and the 20k trimmer all the way down. Now, adjust the 20k trimmer clockwise until you hear notes start to swell up in volume. Continue adjusting it until the volume of the swell matches your input. From there, you can adjust the TWEAK trimmer to fine-tune. In my build, the 20k is set at about 11 o'clock and the TWEAK trimmer pretty much in the middle.
-----------------------



In the notes that came with VFE board bundle that was sold on the VFE website and Patreon,  Peter says:
-----------------------
"do not install the JFET trimpot. Instead, install a jumper between the left-middle pin of that trimpot and its bottom pin. Use the TWEAK trimpot to fine tune the swell response"
-----------------------



Brian said his first one didn't work with a 22k fixed resistor and that's why he recommend using the trimpot, but he may have just misread where Peter says to use a jumper

[jimilee]

Brian's build doc (page 7) has this:
-----------------------
3. Peter recommended not installing the 20k trim and using a fixed 22k resistor in its place. Don't do this. Use the trimmer! I built two Bumblebees, one with the resistor and one with the trimmer. I could not get the swell to work using the fixed resistor method (and that was with a pre-tested 2n5457). IMO, the 20k trimmer is essential to getting this to work optimally. It may be he used JFETs with a very narrow cut-off range and was able to get away with the fixed resistor. But, my experience was different.

4. To set up the swell effect do the following: set the switch in its middle position, the MIX knob all the way counter-clockwise and all other pots to their middle position. Leave TWEAK in the center and the 20k trimmer all the way down. Now, adjust the 20k trimmer clockwise until you hear notes start to swell up in volume. Continue adjusting it until the volume of the swell matches your input. From there, you can adjust the TWEAK trimmer to fine-tune. In my build, the 20k is set at about 11 o'clock and the TWEAK trimmer pretty much in the middle.
-----------------------



In the notes that came with VFE board bundle that was sold on the VFE website and Patreon,  Peter says:
-----------------------
"do not install the JFET trimpot. Instead, install a jumper between the left-middle pin of that trimpot and its bottom pin. Use the TWEAK trimpot to fine tune the swell response"
-----------------------



Brian said his first one didn't work with a 22k fixed resistor and that's why he recommend using the trimpot, but he may have just misread where Peter says to use a jumper

Ah, thank you for that, I bet I've read this doc 100 times and missed it.


Sent from my iPhone using Tapatalk

[jimilee]

Like benny said the proper VGS of Q2 is critical to make the effect work. But, I don't think using one out of spec for the build will throw off voltages for the 2n3904 at all.

Worse comes to worst, I might be able to dig up a proper spec'd 2n5457 for you if your new batch doesn't pan out.

Can I buy a couple of those trannies from you?

[Aentons]

Like benny said the proper VGS of Q2 is critical to make the effect work. But, I don't think using one out of spec for the build will throw off voltages for the 2n3904 at all.

Worse comes to worst, I might be able to dig up a proper spec'd 2n5457 for you if your new batch doesn't pan out.

Can I buy a couple of those trannies from you?

I'm doing a bunch of the VFE builds all at once right now too, including the BB and tractor beam. I'm gonna try to solve some of these jfet woes by getting 100 J113's from Mouser for $13
https://www.mouser.com/ProductDetail/ON-Semiconductor-Fairchild/J113?qs=ljbEvF4DwONIvgQL%2FiOeMA%3D%3D

[benny_profane]

Do you have access to a reliable instrument to measure JFET characteristics or a test circuit? I had no issue finding an in-spec 2n5457 from the SBE devices. I also used the trim pot rather than the fixed resistor.

[jimilee]

Do you have access to a reliable instrument to measure JFET characteristics or a test circuit? I had no issue finding an in-spec 2n5457 from the SBE devices. I also used the trim pot rather than the fixed resistor.

Yes I do. I ordered some, but they were all crap. I have a bounced off matches ones, but they're around 3.0.  What are sbe devices?


Sent from my iPhone using Tapatalk

[jimilee]

Like benny said the proper VGS of Q2 is critical to make the effect work. But, I don't think using one out of spec for the build will throw off voltages for the 2n3904 at all.

Worse comes to worst, I might be able to dig up a proper spec'd 2n5457 for you if your new batch doesn't pan out.

Can I buy a couple of those trannies from you?

I'm doing a bunch of the VFE builds all at once right now too, including the BB and tractor beam. I'm gonna try to solve some of these jfet woes by getting 100 J113's from Mouser for $13
https://www.mouser.com/ProductDetail/ON-Semiconductor-Fairchild/J113?qs=ljbEvF4DwONIvgQL%2FiOeMA%3D%3D

Will j113s work?


Sent from my iPhone using Tapatalk

[Aentons]

I asked about the J113 for phasers here:
https://www.madbeanpedals.com/forum/index.php?topic=31196.msg300928#msg300928

Well, I don't know for sure if it will work in the BB, but I am going on the assumption that all that matters is we want an N-channel jfet with a Vgs(off) value between 1.4-1.5V. And since the drain and source pins are interchangeable on these, all you have to do is turn them around for the right pinout.


From the respective datasheets:
Vgs(off) Min/Max
J113 -0.5V / -3.0V
2N5457 -0.5V / -6.0V
2n5952 -1.3V / -3.5V

So, its seems like the J113 is not only usable for the BB, but it will also be easier to match them for phaser usage.

Someone please correct me if I am wrong

[jimilee]

I got the right 5457, it's all about the 5457. Carry on.