6G3 Build Blocking Distortion?

Howdy Folks,


I am finishing up a build of a 6G3 brownface deluxe, but am having an issue that is vexing me. So I humbly ask the gurus of TDPRI for some input. When using the tremolo, I am hearing some faint distortion that almost sounds as if it’s “behind” the main guitar sound. The more I increase the intensity, the more prevalent it is. I believe this to be blocking distortion of the power tubes.


I think the issue is rooted in the common problem that seems to plague almost every brown deluxe: all the voltages are way over written spec. The specified -26V bias voltage was measuring -36V off the bat. I was having some tremolo low frequency thump (the other common 6G3 problem) until I put a diode across the intensity pot. This made the thump 90% better, but it had the side effect of bumping the bias voltage to -42V.


I replaced the stock 22K bias resistor with a 20K fixed resistor and a 10K pot. This allows me to dial the grid voltage down to about -38V with proper 6V6 bias, but with the tremolo fully engaged, it can swing to about -62V.



Some troubleshooting steps I’ve done:

• Adjusted the bias
  
  
• Used a variac to lower input voltage down to 110V.
  
  
• Removed the diode on the intensity pot. A slight reduction in distortion, but obviously the thump returns.Added grid stopper resistors. I tried various values from 1.5K up to 5K to no avail
  
  
• Reduced grid leak resistors down to 100K
  
  
• Changed the 100k resistor in the bias supply (tried various values from 100-150K). I’m not sure if this is typically done, but it was a wild idea that I had. With 150K, I could get the grid voltage to -28V. It seems to be improving the distortion some, but it throws the bias of the power tubes off enough that I’m unable to bring them into the correct range.
  
  
• Because my plate voltages were about 445V with a GZ34, I’ve been experimenting with 5U4GB and 5Y3 rectifier tubes in an attempt to get the bias set correctly.
  
  
• Changed speakers and cabinet just for a sanity check
  

Am I on the right track? Am I way off base in thinking that the wild voltage swings of the bias wiggle tremolo are to blame for the issue? What else can be done about it? I can post some pictures or measurements if need be.

 

@Jon Snell Thanks so much for the input. That makes a lot of sense now!

With the decreased bias voltage, the issue for me becomes getting the plate dissipation down to an acceptable level. Any advice on also reducing the plate current? Is it just modifying the values of the bias resistors or should I be setting my sights elsewhere?

 

I’m using a Classictone 40-18016. It’s a 330V-0-330V (like the schematic calls for), but in retrospect I could see how a 315 or even a 300 would be a better fit.

I have a vintage 5Y3 in there at the moment to help get the plate voltage down, but the plate current can get a little out of control.

 

I think @ThermionicScott hit the nail on the head with the power transformer assessment. Another model does seem like the best solution.

I’m not sure how clear I was about the issue. When setting the bias voltage to a more appropriate level (-30V), the plate current increases so much that even with the drop in plate voltage, my total plate dissipation is too high.

I guess I was wondering if there were any “tricks” to help burn some B+ power. A crazy idea I had would be to add some cathode resistors and make it a mixed bias. Not sure how well it would work, though.

Thanks everyone for the input and the advice, by the way.

 

Sorry for all the confusion. I was indeed measuring plate voltage, but since that is unaffected by the changes to the bias section, I was not reporting those numbers and just stating the total plate dissipation. Perhaps the actual numbers will help shed some light on the issue.


These numbers are with JJ 6V6S tubes, so I’m using 14W as the max. Measurements listed are with the bias pot in the same position and are measured on the same tube in all instances (the other one is flowing about 2mA less current).



GZ34

Bias Voltage: -38V

Plate Current: 19mA

Plate Voltage: 447V

Percentage: 60%



5U4GB

Bias Voltage: -38V

Plate Current: 21.6mA

Plate Voltage: 429V

Percentage: 66%


5Y3 (and changing 100k resistor to 120k in bias section):

Bias Voltage: -35V

Plate Current: 23mA

Plate Voltage: 402V

Percentage: 67%



With any of these settings, the amp is problem free with no/minimal tremolo. Unpleasant distortion gets introduced between 4 and 6 on the intensity dial.


I was indeed trying to do what @_Steve had mentioned. The suggestion was made to get the bias voltage to -30V. When I lowered the voltage down to -30V, this would increase my plate current to 40mA and I would get PDs in excess of 90%. So my question was “how do I get the plate current and/or plate voltage down with the new bias setting?”. Hence the experiment with different rectifier tubes.


If I’m now understanding it correctly, the bias voltage itself isn’t the answer. The answer is a lower plate dissipation by whatever means necessary (within acceptable limits). I’ll take another whack at it and report back the results with biases less than 60%. Looking at it now, the GZ34 does seem like a good option. But I was a little scared of the 447V because the spec sheet max screen voltage for those tubes is 450V and there aren’t any screen resistors.


And again, I really appreciate everyone’s time and input. I’m an enthusiastic amateur in the DIY amp game, but I’m fully immersing myself in this world and trying to soak up all the knowledge that I possibly can.

 

I don't want to beat a dead horse here, but I thought I'd provide an interesting update: I’ve been playing with it for a while now. I adjusted the bias all the way down to 30% and the distortion was still. In fact, as the bias got hotter, the problem decreased (presumably because the tremolo weakened).

I found a post on another forum where someone had the same issue. He tried all the standard and logical things that were suggested to no avail. Eventually, he realized that the output transformer he had in there was not right for the job (it had a 27:1 turns ratio instead of 30:1). Swapping the OT instantly fixed the issue.

A light bulb went off in my head. The 4 Ohm tap on the OT was not working, but I had just dismissed it as a wiring problem and had not considered that the OT could be the issue. I have a ClassicTone 40-18038 OT, which has 4 and 8 Ohm taps, but otherwise is supposed to be correct to vintage specs. Measuring the winding ratio revealed that instead of 30:1, it was actually almost 33:1. In other words, my 6K6 Ohm primary impedance is actually 8K5 Ohms!


With the sudden demise of Classictone, that makes getting a replacement impossible. I'll have to order another transformer, but I may keep this one for an eventual Princeton Reverb build (if only the mounting patterns lined up!)

 

That’s exactly the power transformer that I have!