Reverb JFET Tails Switching - no switching

[drdn0]

Building a Riptide, and having a couple of issues getting it working well.

Biggest issue I'm having is the JFET switching for tails (always-on preamp) - one pole of the switch controls this, the other controls the LED. The pedal is always on, but LED switches on/off. I did use Fairlchild MMBFJ201's for Q1/Q2, and I was sure I'd read somewhere that a J201 would absolutely work in this position - is this my issue?

JFET switching is witchcraft to me

[Rob Strand]

It's not clear what is going on from the schematic.  We need know where wire at the top to C21 and the JFET goes.    The circuit appears to be using both single supply and dual supply circuits.

For the circuit to work the drain/C21 side of the JFET needs to be at a DC voltage of about 4.5V.   If it's at 0V the JFET will stay on.

Assuming the circuit is correct from that perspective, the thing to check would be the pinout of the JFET.   Next it might be worth testing the JFET as you could have a fake.

[drdn0]

It's not clear what is going on from the schematic.  We need know where wire at the top to C21 and the JFET goes.    The circuit appears to be using both single supply and dual supply circuits.

For the circuit to work the drain/C21 side of the JFET needs to be at a DC voltage of about 4.5V.   If it's at 0V the JFET will stay on.

Assuming the circuit is correct from that perspective, the thing to check would be the pinout of the JFET.   Next it might be worth testing the JFET as you could have a fake.

https://scientificguitarist.wixsite.com/home/riptide-spring-reverb

Power supply is bipolar, tested and working - +8.8/-8.7v.

JFETs appear to be legitimate - purchased from Mouser, and all tested within the average specs I'd expect from a J201. Pinouts match the PCB, which matches the schematic.

[Rob Strand]

https://scientificguitarist.wixsite.com/home/riptide-spring-reverb

Power supply is bipolar, tested and working - +8.8/-8.7v.

JFETs appear to be legitimate - purchased from Mouser, and all tested within the average specs I'd expect from a J201. Pinouts match the PCB, which matches the schematic

The JFET should be good from Mouser.

The problem is the circuit.  I can't see how it could ever work, no matter what the JFET, as the JFET is always on.

The JFET drain gets its bias from R17 and IC1B.  Since the unit has dual supplies the voltage on R17 will be 0V.  So that means the JFET cannot be turned off.  At best you will get some distortion in the off position.

To turn it off:  the switch contact going to the point SW1c needs to be wired to -9V instead of ground.  Should be OK to leave the SW1a contact switching to +9V for the on position.  The J201 JFET has a 40V gate breakdown so it shouldn't have any problem with the +9V and -9V.

FWIW, the spec'd J112 would be better than a J201 in that position but it still won't work in the original circuit.
Here's a simulation which turns the JFET off and on.  Signal is a 1kHz sinewave which you can't see the details of.  Switch is on for 100ms then off for 100ms.

Not working, 0V and +9V


Working, -9V and +9V

[drdn0]

https://scientificguitarist.wixsite.com/home/riptide-spring-reverb

Power supply is bipolar, tested and working - +8.8/-8.7v.

JFETs appear to be legitimate - purchased from Mouser, and all tested within the average specs I'd expect from a J201. Pinouts match the PCB, which matches the schematic

The JFET should be good from Mouser.

The problem is the circuit.  I can't see how it could ever work, no matter what the JFET, as the JFET is always on.

The JFET drain gets its bias from R17 and IC1B.  Since the unit has dual supplies the voltage on R17 will be 0V.  So that means the JFET cannot be turned off.  At best you will get some distortion in the off position.

To turn it off:  the switch contact going to the point SW1c needs to be wired to -9V instead of ground.  Should be OK to leave the SW1a contact switching to +9V for the on position.  The J201 JFET has a 40V gate breakdown so it shouldn't have any problem with the +9V and -9V.

FWIW, the spec'd J112 would be better than a J201 in that position but it still won't work in the original circuit.
Here's a simulation which turns the JFET off and on.  Signal is a 1kHz sinewave which you can't see the details of.  Switch is on for 100ms then off for 100ms.

Not working, 0V and +9V


Working, -9V and +9V

Legend - it absolutely works! I've sent the feedback to SG and hopefully he can make the schematic/PCB changes easily.

There are some other issues (feedback, etc) but I'm guessing the fact I got sent 600 ohm tanks instead of 150 ohm tanks probably isn't helping.

[Rob Strand]

Legend - it absolutely works! I've sent the feedback to SG and hopefully he can make the schematic/PCB changes easily.

Cool.  It's a bit weird you are the only one to notice the switch didn't work.

There are some other issues (feedback, etc) but I'm guessing the fact I got sent 600 ohm tanks instead of 150 ohm tanks probably isn't helping.

Yes, it's going to create some problems.    A 600 ohm tank on a 150 ohm circuit is going to drive the tank at twice current it should be.  A simple mod would be to change R4 to 330 ohm and perhaps R7 to 6.8k.  If the drive current is high the reverb level is louder and that will cause more feedback.

Another source of feedback is the drive amp and recovery amp let through a lot of low frequencies.   To help feedback you could decrease C6 to say 4.7uF, perhaps even down to 2.2uF.

[drdn0]

Legend - it absolutely works! I've sent the feedback to SG and hopefully he can make the schematic/PCB changes easily.

Cool.  It's a bit weird you are the only one to notice the switch didn't work.

There are some other issues (feedback, etc) but I'm guessing the fact I got sent 600 ohm tanks instead of 150 ohm tanks probably isn't helping.

Yes, it's going to create some problems.    A 600 ohm tank on a 150 ohm circuit is going to drive the tank at twice current it should be.  A simple mod would be to change R4 to 330 ohm and perhaps R7 to 6.8k.  If the drive current is high the reverb level is louder and that will cause more feedback.

Another source of feedback is the drive amp and recovery amp let through a lot of low frequencies.   To help feedback you could decrease C6 to say 4.7uF, perhaps even down to 2.2uF.

Going to 330/6k8/4u7 made some massive improvements - still not quite there and will runaway feedback with everything dimed, but it's a million times better. The tone control still really does nothing which is pretty odd, but it's actually usable now which is nice!

[Rob Strand]

Going to 330/6k8/4u7 made some massive improvements - still not quite there and will runaway feedback with everything dimed, but it's a million times better. The tone control still really does nothing which is pretty odd, but it's actually usable now which is nice!

Well that's a step in the right direction.

It's possible the circuit has too much gain in the recovery amplifier.  Clearly if there's too much gain at some point the circuit will take off.   Try reducing R11 from 22k to say 10k;  as a rough guess I'd start at 15k.

If you turn the reverb level (Dwell and Mix) up to full, normally the reverb and dry signals are about equal level when listen to individually.   If the reverb is clearly louder then that's a sign there's too much recovery gain and R11 should be reduced.

You can also keep reducing that cap from 4u7 to 2u2.  However if the recovery gain is way too high then it's solving the wrong problem.

[Rob Strand]

I had a closer look at the circuit and the signal levels in the circuit.

The way to avoid feedback is to drive the spring fairly hard and use the lowest gain possible.  On top of that you can roll-off some of the lows and mechanically/acoustically isolate the tank chassis from the enclosure.

Drive level:  For a 600 ohm coil, I think you can leave R4 at 150 ohm.   The extra drive is within the range of the 600 ohm coil and it will allow a lower recovery amp gain.

Recovery gain:  Perhaps knock R11 down to about 15k.   You could go down to 10k provided the reverb level is still OK.

Low-frequency roll-off:  leave C4 at the modded value of 4.7uF.

With that set-up:
- see if the reverb level is good.
- if you still get feedback see if you can isolate the tank from the enclosure using rubber feet.
  Some old amps had a padded bag.

If you still get problems then you might need to drop the C4 value a bit more.

[drdn0]

I had a closer look at the circuit and the signal levels in the circuit.

The way to avoid feedback is to drive the spring fairly hard and use the lowest gain possible.  On top of that you can roll-off some of the lows and mechanically/acoustically isolate the tank chassis from the enclosure.

Drive level:  For a 600 ohm coil, I think you can leave R4 at 150 ohm.  The extra drive is within the range of the 600 ohm coil and it will allow a lower recovery amp gain.

Recovery gain:  Perhaps knock R11 down to about 15k.  You could go down to 10k provided the reverb level is still OK.

Low-frequency roll-off:  leave C4 at the modded value of 4.7uF.

With that set-up:
- see if the reverb level is good.
- if you still get feedback see if you can isolate the tank from the enclosure using rubber feet.
  Some old amps had a padded bag.

If you still get problems then you might need to drop the C4 value a bit more.

Now we're getting somewhere.

• R4 - 150
• R11 - 12k
• R15 - 100k (to get some more range out of the tone controls)
• C22 - 100pf (to try and get rid of some of the noise)

It's not -perfect- (still fairly noisy, but manageable), and the tone control isn't what I'd want (it cuts a bit more bass up and is almost flat in the middle, but not really the range of control I'm used to), but given the fact it actually works, given a very washy classic reverb and even does the palm-muted rain drop thing incredibly well I'd call this one a success.

Thankyou so much for your help, absolutely appreciate it!

[Rob Strand]

Now we're getting somewhere.

• R4 - 150
• R11 - 12k
• R15 - 100k (to get some more range out of the tone controls)
• C22 - 100pf (to try and get rid of some of the noise)

[/list]

It's not -perfect- (still fairly noisy, but manageable), and the tone control isn't what I'd want (it cuts a bit more bass up and is almost flat in the middle, but not really the range of control I'm used to), but given the fact it actually works, given a very washy classic reverb and even does the palm-muted rain drop thing incredibly well I'd call this one a success.

Thankyou so much for your help, absolutely appreciate it!

That's good progress.

If you want to shave some noise without really affecting anything try changing C7 from 470pF to something around 1n8 to 2n2;   Any larger than that and you need to spend time tuning the value to suit the tone you like.

You could get rid of some noise by tweaking the circuit around Q1 but with C2 and C20 present it gets a bit messy to get rid of the noise and not change something.

[drdn0]

Turned out great!