Garnet Stinger…taming the beast

[TubeGeek]

Hello all,


I have in front of me a Garnet LB190D Pro amplifier with the "stinger" fuzz circuit.


So far I have installed a new grounded AC cord and a full electrolytic cap job (left the original cap cans installed but disconnected) and installed discrete caps inside the chassis.


I powered up this morning and measured voltages.  This is where I am in the process.  It is working but I feel like there's some gremlins to chase out.


Some voltages are a bit higher than expected and it turns out the PT isn't original. It has been replaced previously. Will look into this more...


Couple microphonic and rattling tubes but they are functioning.


I have worked on a handful of these old Garnets over the years and this time I want to dig into the stinger a bit more.


It's too much as is…just way too noisy and such a big volume boost when engaged, it almost seems unusable.


Are there any tips or techniques of taming this circuit to be a bit more useable?

[TubeGeek]

pics


Feel free to roast me on the cap job. I think it worked out ok for as much work as it was to do.

[TubeGeek]

more pics.


Stinger close-up pic...

[TIMBO]

Hey mate, You could try a voltage divider before the Pentode input to reduce the signal and/or disconnecting the 25uf bypass cap

[TubeGeek]

Hey mate, You could try a voltage divider before the Pentode input to reduce the signal and/or disconnecting the 25uf bypass cap

I wish it were that easy. The stinger uses a 6AN8 triode/pentode.


I’m still wrapping my head around the design and what’s going on.


There is no 25 µF bypass cap here.

[sluckey]

The schematic you posted only shows the preamp.

[TIMBO]

OK, looking at the wrong schematic

[WimWalther]

The re-cap job looks totally ship-shape to me, I'm not sure what you're concerned about.

Now I'm assuming you drilled holes to mount the terminal strips, and that's always best to avoid if at all possible - but either way, the job looks very neat and workmanlike.

ETA: Here's a tip for replacing can caps.. Disconnect the old cap and dismount any parts, resistors, etc. Cut off flush all but one cap terminal, and bend it over 90°.

Now solder the mounting tab of a terminal strip to the bent terminal and use the strip to mount the new caps, parts etc. You may not need to add any new holes.

[HotBluePlates]

I have in front of me a Garnet LB190D Pro amplifier with the "stinger" fuzz circuit.

Hey mate, You could try a voltage divider before the Pentode input to reduce the signal and/or disconnecting the 25uf bypass cap

I wish it were that easy. The stinger uses a 6AN8 triode/pentode.
...
There is no 25 µF bypass cap here.

A different schematic (hosted here) shows the 6JW8 tube for the Stinger circuit, and a different bias method.

Are there any tips or techniques of taming this circuit to be a bit more useable?

I’m still wrapping my head around the design and what’s going on.

Both schematics have a couple things in common: apply a signal to the pentode section of a pentode-triode tube, and rig that pentode for the highest voltage-gain it can manage.

Despite the tube types being different, both use a 4.7MΩ resistor to the screen, and a 1.5MΩ plate load.
  - The large screen dropping resistor lowers screen voltage a lot.
  - The low screen voltage makes plate current very small.
  - Very-small-plate-current means the tube is running near cut-off.
  - The very-small-plate-current is passing through a very-large-value plate load resistor.
  - The very-large-value plate load resistor means even a very-small variational-plate-current (AC) creates a large voltage-output.

  - The small plate current also implies a small grid-bias voltage (especially when cathode-bias is used).
  - Every tube distorts hard when the peak signal driving the tube exceeds the tube's bias.

  - Small-grid-bias-voltage means the distortion noted above happens on the positive-peak of the driving AC.
  - Operating near cut-off means the tube also distorts rapidly on the negative-peak of the driving AC.
  - This Stinger circuit is rigged to deliver lots of odd-harmonic distortion (output clipped on both sides of the waveform).

The differences:

  - The Hoffman-hosted schematic has:
     - 6JW8 pentode.
     - Cathode bias: 5.6kΩ cathode resistor bypassed by 25µF.
     - Takes signal after the 1st gain stage.
     - Has a simpler coupling-cap setup feeding the Stinger's "Vol" and "Color" (Tone) controls.

  - Tubegeek's example has:
     - 6AN8 pentode.
     - Grid-leak bias: 0.005µF cap into 10MΩ grid-leak.
     - Takes signal from Input jack before 1st gain stage.
     - Has a voltage-divider (220kΩ into 150kΩ) w/ 1MΩ "Vol" control across lower leg.
     - 0.01µF cap across upper leg of the divider to shave bass.
     - Different-value "Color" (Tone) pot that changes the turnover point of this control.
     - Use the triode side of the pentode-triode tube for additional voltage-gain.
     - 6AN8 triode has a crazy-high 1.5MΩ plate load & 200kΩ cathode resistor.
     - 6AN8 triode cathode resistor bypassed with 0.047µF: more super-high emphasis?

So how to tame?

... Garnet LB190D Pro amplifier with the "stinger" fuzz circuit.
...
It's too much as is…just way too noisy and such a big volume boost when engaged, it almost seems unusable.
...

6AN8 pentode plate is direct-coupled to the 6AN8 triode grid.  How about trying:
     - Break that direct connection from pentode output (Pin 6) to triode input (Pin 2).
     - Unsolder the 0.01µF from the 6AN8 triode plate (Pin 1).
     - Use a jumper wire to solder the free-end of this 0.01µF coupling cap to the 6AN8 pentode plate (Pin 6).

     - This should bypass the 6AN8 triode section, and avoid hiss imposed by the 6AN8 triode's 1.5MΩ plate load & 0.047µF cathode bypass cap.

Does the Stinger "Vol" control now have a better range towards the bottom of its sweep?

[TubeGeek]

The schematic you posted only shows the preamp.

Fixed. Power supply schematic uploaded on the same post.

[TubeGeek]

The re-cap job looks totally ship-shape to me, I'm not sure what you're concerned about.

Now I'm assuming you drilled holes to mount the terminal strips, and that's always best to avoid if at all possible - but either way, the job looks very neat and workmanlike.

ETA: Here's a tip for replacing can caps.. Disconnect the old cap and dismount any parts, resistors, etc. Cut off flush all but one cap terminal, and bend it over 90°.

Now solder the mounting tab of a terminal strip to the bent terminal and use the strip to mount the new caps, parts etc. You may not need to add any new holes.

It's just my sense of humour.  Thanks for having a look and your feedback.


I don't prefer to drill holes in the chassis and I always clear it with the owner first if I feel I need to.  This amp is something the owner has sentimental attachment to and he will not be selling it.  I also don't think a Garnet's value is being reduced significantly with a few new holes.  There were already a few other holes drilled out previously too.


I have used the technique you described before, it works good if new holes aren't an option.

[TubeGeek]

I have in front of me a Garnet LB190D Pro amplifier with the "stinger" fuzz circuit.

Hey mate, You could try a voltage divider before the Pentode input to reduce the signal and/or disconnecting the 25uf bypass cap

I wish it were that easy. The stinger uses a 6AN8 triode/pentode.
...
There is no 25 µF bypass cap here.

A different schematic (hosted here) shows the 6JW8 tube for the Stinger circuit, and a different bias method.

Are there any tips or techniques of taming this circuit to be a bit more useable?

I’m still wrapping my head around the design and what’s going on.

Both schematics have a couple things in common: apply a signal to the pentode section of a pentode-triode tube, and rig that pentode for the highest voltage-gain it can manage.

Despite the tube types being different, both use a 4.7MΩ resistor to the screen, and a 1.5MΩ plate load.
  - The large screen dropping resistor lowers screen voltage a lot.
  - The low screen voltage makes plate current very small.
  - Very-small-plate-current means the tube is running near cut-off.
  - The very-small-plate-current is passing through a very-large-value plate load resistor.
  - The very-large-value plate load resistor means even a very-small variational-plate-current (AC) creates a large voltage-output.

  - The small plate current also implies a small grid-bias voltage (especially when cathode-bias is used).
  - Every tube distorts hard when the peak signal driving the tube exceeds the tube's bias.

  - Small-grid-bias-voltage means the distortion noted above happens on the positive-peak of the driving AC.
  - Operating near cut-off means the tube also distorts rapidly on the negative-peak of the driving AC.
  - This Stinger circuit is rigged to deliver lots of odd-harmonic distortion (output clipped on both sides of the waveform).

The differences:

  - The Hoffman-hosted schematic has:
     - 6JW8 pentode.
     - Cathode bias: 5.6kΩ cathode resistor bypassed by 25µF.
     - Takes signal after the 1st gain stage.
     - Has a simpler coupling-cap setup feeding the Stinger's "Vol" and "Color" (Tone) controls.

  - Tubegeek's example has:
     - 6AN8 pentode.
     - Grid-leak bias: 0.005µF cap into 10MΩ grid-leak.
     - Takes signal from Input jack before 1st gain stage.
     - Has a voltage-divider (220kΩ into 150kΩ) w/ 1MΩ "Vol" control across lower leg.
     - 0.01µF cap across upper leg of the divider to shave bass.
     - Different-value "Color" (Tone) pot that changes the turnover point of this control.
     - Use the triode side of the pentode-triode tube for additional voltage-gain.
     - 6AN8 triode has a crazy-high 1.5MΩ plate load & 200kΩ cathode resistor.
     - 6AN8 triode cathode resistor bypassed with 0.047µF: more super-high emphasis?

So how to tame?

... Garnet LB190D Pro amplifier with the "stinger" fuzz circuit.
...
It's too much as is…just way too noisy and such a big volume boost when engaged, it almost seems unusable.
...

6AN8 pentode plate is direct-coupled to the 6AN8 triode grid.  How about trying:
     - Break that direct connection from pentode output (Pin 6) to triode input (Pin 2).
     - Unsolder the 0.01µF from the 6AN8 triode plate (Pin 1).
     - Use a jumper wire to solder the free-end of this 0.01µF coupling cap to the 6AN8 pentode plate (Pin 6).

     - This should bypass the 6AN8 triode section, and avoid hiss imposed by the 6AN8 triode's 1.5MΩ plate load & 0.047µF cathode bypass cap.

Does the Stinger "Vol" control now have a better range towards the bottom of its sweep?

Wow, thanks for taking the time to write all that out. I am going to give it a try.


I am going to chew on that and report back when I get back to the amp.  It'll be a few days before I am back to sitting at my amp workbench. 

[TubeGeek]

I measured voltages in the amp and wrote them down. I have attached a pic of that sheet.

I wanted to measure voltages before I make changes.  The original PT has been replaced so I am seeing higher voltages than expected.

Tips on lowering the B+ from the rectifier by 30 volts?

My first thought is to add a resistor between the diodes and first filter cap "A". Unsure of what value to start with...

After the B+ is lowered, I can dial in the voltages of the filter cap nodes accordingly.  Then I can get back to working on the stinger.

Power section schematic attached for reference.

[HotBluePlates]

Tips on lowering the B+ from the rectifier by 30 volts?

My first thought is to add a resistor between the diodes and first filter cap "A". Unsure of what value to start with...

After the B+ is lowered ...

I would add resistance between the PT winding & rectifier (so 2 individual resistors).

The 6AN8 voltages don't make sense:  The triode side has its cathode (Pin 3) voltage higher than its plate (Pin 1), and the pentode side looks like the circuit is disengaged (the cathode at Pin 9 must be 0v when the Stinger is On).  I think you'll need to have the Stinger On to get valid readings, even if you need to turn the Fuzz Vol all the way off.

Personally I would fix the Stinger before chasing voltages (+/-30v is "no big deal" in a B+ supply).

[TubeGeek]

What led me down the voltage rabbit hole was the voltages I am seeing at the filter caps.  They seem to be much higher than expected from point C to F. 


When troubleshooting the stinger, I noticed a little smoke coming from the power tubes or the filter cap at point C (hard to pinpoint).  I haven't fully let the magic smoke out yet, but something's up.  I wanted to get that sorted out before I damage something.


Based on the schematic I am following, there are some voltages listed on the left side of the schematic that I am assuming are what's expected at points A through F. It's not labelled super clearly to me.


I noticed that point C was of concern as it was measuring 453 VDC. It's a 47uF 450V cap.


When the amp came to me, the 10K resistor between points B and C was cooked.


This is what made me focus on that.  I did some experimenting with changing the value of that 10k.  I changed it to 20k and then settled on 18k for the moment.  This at least attenuated the voltage to 407.


I attached a sheet showing the different voltages I measured with the 10k, 18k, 20k dropping resistor.


If the voltages listed on the schematic are what I should be aiming for, then I still have some tweaking to do with the 27k's at those points.


I am not too concerned about the 30V on the B+, its more concerning to me down the line from C to F.  Or….should I ignore the voltages written on the schematic for now?


I'll try to sit at the bench tonight and turn the stinger on and take voltage measurements and report back.

[PRR]

The 120k for the bias supply is {drawn} connected wrong!!

Won't bias right neither.

[sluckey]

The 120k for the bias supply is connected wrong!

I looked at a lot of garnet schematics and that same error appears multiple times. I'm pretty sure it's just a draftsman error. Comparing to the correctly drawn schematics I can see how that can be an easy error. I doubt any of the amps are actually wired that way.

[TubeGeek]

The 120k for the bias supply is connected wrong!!.

Connected to the wall-outlet like that it can KILL YOU. 
Won't bias right neither.

Sluckey is right…just a drawing error. It's not actually wired that way.

[TubeGeek]

Tis the season for Garnets…I had another identical Pro amp dropped off last night for repair. It could be handy to have another for reference.

[sluckey]

Sometimes when it rains, it pours. (Saw that on the side of a Morton's Salt box many years ago.)

There's also another current Garnet thread.

[TubeGeek]

I am working on tracing everything out from the schematic to actual amp.  I'm about 80% done so far.


The "B" 0.005 cap on the schematic is not in the amplifier I have.  I have circled the cap in question on the attached image.



I'm thinking its drawing error.  It's not in the other amp chassis I have.

[AlNewman]

The .005 cap looks like a snubber cap to me.  Not sure why it's connected to the screen voltage, but I doubt you need it unless you're getting some nasty oscillations. 

[TubeGeek]

Opened up the second amp for comparison.


Through my tracing I did find the 6AN8 cathode resistor was not grounded properly.  That could be partly why the stinger was not working.  It's too late to verify this tonight…will do tomorrow.

[PRR]

> The "B" 0.005 cap on the schematic is not in the amplifier I have.

So forget it.

The usual excuse is a very cheap buzz reduction, by taking a little buzz at B and blending it at C. This actually rarely works except by accident (or hyper-critical and frequent adjustment).

[TubeGeek]

> The "B" 0.005 cap on the schematic is not in the amplifier I have.

So forget it.

The usual excuse is a very cheap buzz reduction, by taking a little buzz at B and blending it at C. This actually rarely works except by accident (or hyper-critical and frequent adjustment).

Forgotten. Moving on...

[HotBluePlates]

Through my tracing I did find the 6AN8 cathode resistor was not grounded properly. ...

The schematic says the method of engaging/dis-engaging the fuzz is by grounding/not-grounding the cathode components.  Both with a footswitch and the front-panel switch.

[TubeGeek]

Through my tracing I did find the 6AN8 cathode resistor was not grounded properly. ...

The schematic says the method of engaging/dis-engaging the fuzz is by grounding/not-grounding the cathode components.  Both with a footswitch and the front-panel switch.

The cathode resistor connects to one side of the on/off switch, which is grounded.  The other side of the on/off switch connects to the tip of the foot switch jack and pin 9 of the 6AN8. Follows the schematic.


It’s wired like it was when it came to me with a broken on/off switch.  I couldn't salvage the original 0.01 ceramic disk capacitor so I replaced it with a mallory 150M equivalent.


I spoke with the owner about the stinger and he mentioned it has never been useable.  It has always been crazy noisy.


Here is a short video of what the stinger is doing:


When the stinger is turned on, the tone pot change the pitch of the squeal and the vol pot just makes a ton of noise and sends the bias high on the power tubes.

i=y6Ory5jZV7VZxPpl

[TubeGeek]

Look at the differences between the two amps in the PI. The wiring of the 6SN7's is not the same.

As of now the stinger's of both amp I have are wired the same.


The new amp (amp#2) sounds good and works fine.


Geez, I don't know…amp #1 does correlate to the schematic I am looking at. Except for the caps on the plates and 10w resistor.  They are missing in this amp. Is the schematic incorrect here too? My head is spinning on this one.

[HotBluePlates]

Look at the differences between the two amps in the PI. The wiring of the 6SN7's is not the same.
...
The new amp (amp#2) sounds good and works fine.

Make Amp 1's phase inverter look like Amp 2's phase inverter?
(I haven't spent the time to decipher the electrical differences from the photos)

[TubeGeek]

I needed to take a break from the amp.  After a few days off I plugged a guitar into it and played it and it doesn't sound bad.  Without the stinger on, it sounds nice.  When I turn the stinger on I can get a nice fuzz with the vol pot at 2 and the tone at about 8.  It's quite useable there but as soon as I turn the fuzz vol up, it becomes too much.


Before I send the amp back to its owner I plan to try to bypass the triode section in the 6AN8 tube like you suggest.


I'm curious if this will make the fuzz more useable above 2 on the vol pot.


Other than that I think the amp is repaired and ready to go.


I'll probably get to this experiment this coming weekend. Just wanted to let you know I do plan on trying your idea out and that you haven't wasted time with helping me out.

[TubeGeek]

Well it has been a little bit. I put this amp to the side for a bit and finished some other repairs I had.


HBP…you are a genius! That's a great mod to this circuit. The stinger is much more useable. Thank you for your idea.


I think it works quite nicely.


The stinger can now be turned past 2 and it's not total chaos.  You could use it as a boost, overdrive or full out fuzz that has a characteristic I like.  It sounds like it has a noise gate or compressor on it. The notes do not sustain at all when palm muting the notes. It's like the signal envelope closes and mutes quickly.  It has a Jack white kind of thing going on. Very mid focused and will cut through any mix, I am sure.

[TubeGeek]

One more short video demo'ing the stinger mod:

[tubeswell]

Nice

[Platefire]

Another one bites the dust

[HotBluePlates]

HBP…you are a genius! That's a great mod to this circuit. The stinger is much more useable. Thank you for your idea.

Did you wind up bypassing the extra triode-stage?

Just wondering whether it was the "reduced-gain" or if you re-wired the Stinger to be cathode-biased like the 6JW8 examples.

[TubeGeek]

I kept it simple and just bypassed the triode stage.  That was good enough.


I have another identical amp up on the bench soon and will do the same mod to this amp if the owner wants.

[hiegdk]

For what it's worth I have Gar's book in front of me and the schematic for the LB190D references a 6JW8 in the Stinger section (pretty sure it's the same schematic as the one hosted on the Hoffman site. I don't know anything about these tubes other than that they were used in CRT TVs but maybe someone more knowledgable could indicate if there would be significant differences in operation between a 6JW8 and a 6AN8.

[HotBluePlates]

... indicate if there would be significant differences in operation between a 6JW8 and a 6AN8.

From Frank's:
   6AN8 data sheet
   6JW8 data sheet

The Triode section info appears to show the 6AN8 has a 12AU7-ish triode, while the 6JW8 has a 5751-ish triode (but with lower internal plate resistance & higher Gm).  The 6JW8 triode will definitely amplify more (but is not normally used in the Stinger circuit).

The 6JW8 pentode section doesn't pull as much current as the 6AN8's pentode section, but that doesn't matter too much in this application.  Both look like thy'd amplify fine, and the point here is to push a hobbled-pentode to clip fiercely.