AB763 Deluxe Reverb & Vibroverb Questions

[RazzleDazzle]

Hello TAG...

I want to build a Blackface Deluxe Reverb (current project) and a Blackface Vibroverb (next project) AB763. I want the amps without the vibrato on the reverb channel and a mid control added on both channels tho, thx to Rob from robrobinette.com that was quite easy!

I also upped the voltage and capacitance rating of the bias board filter electrolytic capacitor to 50uF @ 100V. After my understanding the 50V voltage rating of the original one gets pretty much maxed out and the 25uF extra filter capacitance are a good idea to combat hum from the bias circuit.

On the Deluxe Reverb the rating of the first two "Dog House" filter electrolytic capacitors was also upped to 20uF so the the loose low end of the Deluxe Reverb gets firmed up a little.

I got a couple questions I cannot answer myself tho, since my knowledge is limited enough so the answers are not obvious to me...

1. I modified the regular main board layout a little bit, spread everything apart more to accommodate for the extra space and added the 68K grid stop resistors to the main board as well (like it was done in Tweed style amps). Does all of that look ok?

2. Do I need the two 100R heater resistors on a 2x6L6 or 4x6L6 amp? Basically all "new" layouts for the Deluxe Reverb show them, but the 2x6L6 amps with a GZ34 tube rectifier (Super Reverb & Vibroverb) and the 4x6L6 amps with a solid state rectifier (Twin Reverb) do not show them.

3. Can a 500V voltage rating capacitor be used for the A/C shunt capacitor, instead of the original 200V voltage rating capacitor, without it causing any issues in the circuit? SoZo only makes 500V "Blue Molded" capacitors and I would like to use a SoZo "Blue Molded" there because the original ones used a "Blue Molded" capacitor in that spot as well.

4. Is it ok to experiment with the NFB tail resistor value on a Blackface Vibroverb style amp? I think with a Weber Copper Cap solid state rectifier, adjusted bias and a 100R NFB tail resistor the Vibroverb will have a little more clean headroom compared to a stock one with a 47R NFB tail resistor and a GZ34 tube rectifier, which I would like. I'm not completely sure if it's fine to change that value in that particular amp tho so I'd rather ask around before I do something stupid!

My last questions needs a little more info added...

I know this debate is all over the internet but I do think that carbon composition resistor add some "mojo" to the amps they are used in. IMO it has more to do with the second harmonic distortion that carbon composition resistor add, rather than the almost mystical sound qualities some people believe they have...

Carbon composition resistors only make sense in a couple strategic spots tho, spots that have high voltage across the resistor and spots that have large signal swings across the resistor. Only under those conditions will they give the amp enough second harmonic distortion to benefit from. Everywhere else they only add noise because those resistors do not have enough high voltage and large signal swings across them for them to produce an appropriate amount of second harmonic distortion that is worth the extra hiss. If introduced in the early stages of the amp the hiss will also be amplified by the later stages, which will add up, so metal film resistors are definitely a much better choice in that case.

That means the PI plate load resistors (100K & 82K) in the last gain stage before the global feedback loop, or the phase inverter if the amp has no global feedback loop, are the perfect choice.

They are also a good choice for the grid stoppers (1.5K and 1.5K) across the power tube sockets. A low inductance resistor, carbon composition resistors have very low inductance, is perfect for that purpose. They tend to fail short and not burst up into flames and fail open, like metal film resistors could, as well, which is a good idea in that spot too.

I heard they are also good as power tube grid leak resistors (220K & 220K) in the output stage as well. I haven't found much information about that sadly tho, which leads me to my last question...

5. Do the power tube grid leak resistors (220K & 220K) in the output stage have enough high voltage and large signal swings across them for carbon composition resistors to make sense in that spot?

6. Are there any other good spots to add carbon composition resistors? Maybe the phase inverter grid leak resistors?

My layouts also show an extra bias filter electrolytic capacitor and a reverb noise filter capacitor to reduce hum and noise. I'm not sure if I will install them but I think I will try the amps without them first and after they are finished clip them in with alligator clips to see if I like the change or not.

[dorrisant]

1 - That layouts should be fine. Spreading it out a little bit shouldn't hurt anything, condensing it might. Low gain though, so less worry.

2 - If your PT doesn't have a center tap for the filaments, you must use two resistors for a virtual center tap... No matter what power tubes are used or how many.

3 - Yes, going up in voltage rating is fine. I wouldn't use a lower voltage rating.

4 - Yes... put it on a pot with a low limit value resistor in series with the pot to experiment or modify on the fly.

CCs - I am not a huge fan of CC resistors. I won't answer this directly but will tell you to look at what Dumble did and maybe follow his advise.

5 - They are supplying the negative bias vltage for the power tubes. I would suggest using metal films here just for stability.

6 - Refer to comment above.

Additional ? - Try it without, add them if you think it needs it.


And welcome!!

[pdf64]

… Can a 500V voltage rating capacitor be used for the A/C shunt capacitor …

Which particular component is being referred to by ‘A/C shunt capacitor’?

… I think with a Weber Copper Cap solid state rectifier, adjusted bias and a 100R NFB tail resistor the Vibroverb will have a little more clean headroom compared to a stock one with a 47R NFB tail resistor and a GZ34 tube rectifier, …

Weber make several types of copper cap rectifier, which one are you referring to?

… the PI plate load resistors (100K & 82K) in the last gain stage before the global feedback loop, or the phase inverter if the amp has no global feedback loop, are the perfect choice…

I can’t make sense of the above; the phase splitter is inside the global feedback loop.

…They tend to fail short and not burst up into flames and fail open, like metal film resistors could …

Why do you think CC resistors tend to fail short?
Why do you think metal film resistors tend to fail open?
With regard to bursting into flames, my guess is that a chunk of carbon presents a bigger and more readily ignitable fuel source than the ceramic substrate and metal spiral of an MF resistor.

[RazzleDazzle]

1 - That layouts should be fine. Spreading it out a little bit shouldn't hurt anything, condensing it might. Low gain though, so less worry.

2 - If your PT doesn't have a center tap for the filaments, you must use two resistors for a virtual center tap... No matter what power tubes are used or how many.

3 - Yes, going up in voltage rating is fine. I wouldn't use a lower voltage rating.

4 - Yes... put it on a pot with a low limit value resistor in series with the pot to experiment or modify on the fly.

CCs - I am not a huge fan of CC resistors. I won't answer this directly but will tell you to look at what Dumble did and maybe follow his advise.

5 - They are supplying the negative bias vltage for the power tubes. I would suggest using metal films here just for stability.

6 - Refer to comment above.

Additional ? - Try it without, add them if you think it needs it.


And welcome!!

Thank's a lot dorrisant, that was very helpful...

Is the filament center tap the green/yellow wire on this Mercury Magnetics PT for the Deluxe Reverb? If it is and if I understood you correctly then I think my Deluxe Reverb build will not the heater resistors either, because I plan on using that PT.

I know the CC resistor thing is a highly debated topic but even tho I agree with you and Dumble, in that metal film resistors are a better choice in general, I think with vintage amp designs the second harmonic distortion is a part of what made them sound the way they did tho. While using all CC resistors is not a good choice, they will add a lot of hiss that could be avoided with metal film resistors, using all metal film resistors is perfectly fine. IMO there is no need to use CC resistors for an amp to sound great, but on vintage amp designs some might be needed tho to get the same sound.

[RazzleDazzle]

… Can a 500V voltage rating capacitor be used for the A/C shunt capacitor …

Which particular component is being referred to by ‘A/C shunt capacitor’?

… I think with a Weber Copper Cap solid state rectifier, adjusted bias and a 100R NFB tail resistor the Vibroverb will have a little more clean headroom compared to a stock one with a 47R NFB tail resistor and a GZ34 tube rectifier, …

Weber make several types of copper cap rectifier, which one are you referring to?

… the PI plate load resistors (100K & 82K) in the last gain stage before the global feedback loop, or the phase inverter if the amp has no global feedback loop, are the perfect choice…

I can’t make sense of the above; the phase splitter is inside the global feedback loop.

…They tend to fail short and not burst up into flames and fail open, like metal film resistors could …

Why do you think CC resistors tend to fail short?
Why do you think metal film resistors tend to fail open?
With regard to bursting into flames, my guess is that a chunk of carbon presents a bigger and more readily ignitable fuel source than the ceramic substrate and metal spiral of an MF resistor.

Which particular component is being referred to by ‘A/C shunt capacitor’?

I attached a layout from robrobinette.com that shows the capacitor I am referring to, along with the name.

"Weber make several types of copper cap rectifier, which one are you referring to?"

Either the WS1 (without sag) or the WZ34 (with GZ34 style sag), I'm not sure tho which one I will use.

Is there anything wrong with using either of them?

https://www.tedweber.com/ws1/

https://www.tedweber.com/wz34/

"I can’t make sense of the above; the phase splitter is inside the global feedback loop."

Yeah, you are right! I have misread/misunderstood something, thanks for pointing that out to me!

I took most of my information on the topic from this website: http://www.geofex.com/article_folders/c ... oncomp.htm

"The place to use CC's is where there's big signal - plate resistors, and ideally the stage just before the phase inverter. The phase inverter would otherwise be ideal, with plate resistors carrying the highest signal voltage in the amp, but phase inverters are often enclosed in a feedback loop. The feedback minimizes the distortion the resistor generates. "

Could you perhaps point me in the right direction of what resistors I should have to replace that meet the criteria that R.G. Keen outlined on his website?

I might have chosen the wrong resistors, I'm only sure that they are a good choice for the grid stopper resistors across the power tube sockets because of the low inductance that CC resistors have.

"Why do you think CC resistors tend to fail short?
Why do you think metal film resistors tend to fail open?
With regard to bursting into flames, my guess is that a chunk of carbon presents a bigger and more readily ignitable fuel source than the ceramic substrate and metal spiral of an MF resistor."

Its something I read while researching.

I'm only sure that CC resistors are a good choice for the grid stoppers across the power tube sockets because of the low inductance that CC resistors have.

If the failing short part, and not being likely to burst into flames because of them being more likely to fail short, is true or not I do not know for sure.

[martin manning]

4. Is it ok to experiment with the NFB tail resistor value on a Blackface Vibroverb style amp? I think with a Weber Copper Cap solid state rectifier, adjusted bias and a 100R NFB tail resistor the Vibroverb will have a little more clean headroom compared to a stock one with a 47R NFB tail resistor and a GZ34 tube rectifier, which I would like. I'm not completely sure if it's fine to change that value in that particular amp tho so I'd rather ask around before I do something stupid!

Most people insert a pot to vary the series resistance when experimenting with GNFB level. It's hard to do anything stupid. You can play with GNFB and change rectifiers to your heart's content. Just make sure your PT has sufficient current rating if you try a 5U4, and reset the bias for each rectifier type so you have a valid comparison.

The Deluxe Reverb seems to have become fashionable lately. My recent build is here: https://ampgarage.com/forum/viewtopic.p ... 61#p428861

I know this debate is all over the internet but I do think that carbon composition resistor add some "mojo" to the amps they are used in. IMO it has more to do with the second harmonic distortion that carbon composition resistor add, rather than the almost mystical sound qualities some people believe they have...

R.G.'s CC resistor manifesto is internet lore by now, and most here are familiar with the theory. Here is some some recent discussion where I tried to produce the voltage coefficient of resistance effect by tracing CC and MF resistors by applying voltage pulses up to 350V: https://ampgarage.com/forum/viewtopic.p ... 98#p444398 It didn't appear.

[pdf64]

...
Either the WS1 (without sag) or the WZ34 (with GZ34 style sag), I'm not sure tho which one I will use.

Is there anything wrong with using either of them?

My WZ34 has a series pair of 39ohm internal resistors to simulate sag. That drops far more voltage than a real GZ34. In an amp, it drops the power output perhaps 25%.
A GZ34 is better simulated by a 22ohm resistor.

Guitarists use the term headroom to refer to both sytem power output and system gain. The latter is somewhat nonesensical. Altering the 47 or 100ohm feedback resistor only affects system gain.
Fender used 100ohm for amps intended for a 4 or 2ohm load, 47ohms for amps intended for an 8ohm load.
Of the ideas you suggested, only changing to a solid state rectifier eg WS1 will increase the amp's power output and hence its headroom.

Could you perhaps point me in the right direction of what resistors I should have to replace that meet the criteria that R.G. Keen outlined on his website?

I think RG is suggesting the the V4 pin6 anode load resistor is the best choice for CC.
Personally I've been using CC for all anode load resistors, my experince is that the signal level at the input stage anode is plenty high enough to make it theoretically worthwhile. I've have not encountered excessive hiss, even with highish gain overdrive preamps.

I might have chosen the wrong resistors, I'm only sure that they are a good choice for the grid stopper resistors across the power tube sockets because of the low inductance that CC resistors have.

I think that's making an issue when none exists, resistor inductance is negligible for our application. rather it is far more important to use good lead dress and to mount the grid stopper with its body as close as is practicable to the socket terminal.

Its something I read while researching.

I'm only sure that CC resistors are a good choice for the grid stoppers across the power tube sockets because of the low inductance that CC resistors have.

If the failing short part, and not being likely to burst into flames because of them being more likely to fail short, is true or not I do not know for sure.

I suspect that either your research has led your to bad info, or you've completely misunderstood it.
As they age, low wattage CC tend to drift higher in value. That is a bad characteristic for the grid circuit of output valves, due to it having a max resistance limit, which Fender typically exceeded anyway.
If a CC resistor's dissipation is significantly exceeded, the carbon block that is its resistive element is a fuel load that can set alight. It is very common for old Fender amps to exhibit wiring damage from the screen grid resistors doing this, typically in response to an output valve shorting.
Whereas metal film or oxide resistors are far more stable and can be obtained with a flame retardent coating. They are very unlikely to set alight when they blow.

Also, just to note that the black fibreboard traditionally used for Fenders is something of a bodge, I don't think it's insulative properties make it a suitable substrate for high voltage, high impedance circuits.
Adjacent eyelets with a high voltage between them can short, stray voltages can appear where none should be, proving that slightly conductive paths must exist.
Many builders seem to get away with it, but my experience is that problem do emerge and there's so much work in a circuit board that the risk of those problems make it just not worth using it.

[RazzleDazzle]

My WZ34 has a series pair of 39ohm internal resistors to simulate sag. That drops far more voltage than a real GZ34. In an amp, it drops the power output perhaps 25%.
A GZ34 is better simulated by a 22ohm resistor.

Guitarists use the term headroom to refer to both sytem power output and system gain. The latter is somewhat nonesensical. Altering the 47 or 100ohm feedback resistor only affects system gain.
Fender used 100ohm for amps intended for a 4 or 2ohm load, 47ohms for amps intended for an 8ohm load.
Of the ideas you suggested, only changing to a solid state rectifier eg WS1 will increase the amp's power output and hence its headroom.

Thanks a lot, that makes perfect sense!

I will leave the 47R resistor as is in that case because I plan to use a 8R OT secondary load on both amps.

I think RG is suggesting the the V4 pin6 anode load resistor is the best choice for CC.
Personally I've been using CC for all anode load resistors, my experince is that the signal level at the input stage anode is plenty high enough to make it theoretically worthwhile. I've have not encountered excessive hiss, even with highish gain overdrive preamps.

I misunderstood the placement of the CC resistors in that case, this makes way more sense!

Does that mean the two V1 plate load 100K resistors should also be replaced with CC resistors to add some second harmonic distortion to the normal channel?

I hope I understood this correctly now but please correct me if I'm wrong again...

V4 is the 3rd preamp of the reverb channel and because it's best to add the CC resistors at the latest point they go there, they could also be added to V2 tho (which I will probably do). The normal channel only has two preamps so they are added at V1...

I think that's making an issue when none exists, resistor inductance is negligible for our application. rather it is far more important to use good lead dress and to mount the grid stopper with its body as close as is practicable to the socket terminal.

Ok, I will try to mount them as close as possible.

Lead dressing is another topic I have a couple of questions about, I will do so in a new post tho for simplicity's sake tho...

I suspect that either your research has led your to bad info, or you've completely misunderstood it.
As they age, low wattage CC tend to drift higher in value. That is a bad characteristic for the grid circuit of output valves, due to it having a max resistance limit, which Fender typically exceeded anyway.
If a CC resistor's dissipation is significantly exceeded, the carbon block that is its resistive element is a fuel load that can set alight. It is very common for old Fender amps to exhibit wiring damage from the screen grid resistors doing this, typically in response to an output valve shorting.
Whereas metal film or oxide resistors are far more stable and can be obtained with a flame retardent coating. They are very unlikely to set alight when they blow.

You're right, I will use metal film resistors for the grid stoppers!

Also, just to note that the black fibreboard traditionally used for Fenders is something of a bodge, I don't think it's insulative properties make it a suitable substrate for high voltage, high impedance circuits.
Adjacent eyelets with a high voltage between them can short, stray voltages can appear where none should be, proving that slightly conductive paths must exist.
Many builders seem to get away with it, but my experience is that problem do emerge and there's so much work in a circuit board that the risk of those problems make it just not worth using it.

I plan on using 3mm thick G10 FR4 boards with either eylets or turrets. I never liked the floppy black fibreboards that Fender used back in the day!

I always disliked them because I think turret boards look cleaner. What you just told me made me dislike them even more, with better reasoning this time I suppose tho...

[RazzleDazzle]

Most people insert a pot to vary the series resistance when experimenting with GNFB level. It's hard to do anything stupid. You can play with GNFB and change rectifiers to your heart's content. Just make sure your PT has sufficient current rating if you try a 5U4, and reset the bias for each rectifier type so you have a valid comparison.

I will definitely try a Weber WS1 Copper Cap solid state rectifier in my Vibroverb build and a 5U4, like you did, in my Deluxe Reverb build.

The Deluxe Reverb seems to have become fashionable lately. My recent build is here: https://ampgarage.com/forum/viewtopic.p ... 61#p428861

...

The ground scheme has been completely redone to separate the preamp and PI grounds from the power section.

...

Very nice and clean looking build, I hope I get mine to that level as well...

I like the idea of separating the preamp and PI grounds a lot, I will do that on my builds as well!

I do not understand a couple of things tho and I'm not sure everything translated correctly because I removed the vibrato on the reverb channel...

1. Why is there a jumper between pin 2 and 3 on the 10K bias pot?

2. Why was there a 10K resistor added to connect the diode on the bias board to the bias pot? Does this need to be a 1W resistor?

3. Why did you use a 15K max bias resistor instead of the original 10K one?

4. Why was the 470R bias board resistor changed to 1.2K?

5. Where will the V3 cathode resistor and the bypass electrolytic capacitor connect to ground?

6. Where will pin 8 of the second 6V6 power tube socket connect to ground?

Sorry but the answers to those questions are not at all obvious to me...

Could you maybe check the layout I attached to see if I interpreted your modded layout correctly? I'm especially not sure if I translated the speaker jack wiring correctly since the original Fender layout you modified does not show the names for the speaker jack terminals.

I marked all the parts that I am unsure of in pink to make them easier to spot.

R.G.'s CC resistor manifesto is internet lore by now, and most here are familiar with the theory. Here is some some recent discussion where I tried to produce the voltage coefficient of resistance effect by tracing CC and MF resistors by applying voltage pulses up to 350V: https://ampgarage.com/forum/viewtopic.p ... 98#p444398 It didn't appear.

I have not seen that thread before, very interesting!

[martin manning]

1. Why is there a jumper between pin 2 and 3 on the 10K bias pot?
2. Why was there a 10K resistor added to connect the diode on the bias board to the bias pot? Does this need to be a 1W resistor?
3. Why did you use a 15K max bias resistor instead of the original 10K one?
4. Why was the 470R bias board resistor changed to 1.2K?

A second RC filter (10k & 47u) has been added to the bias supply, and the other values have been changed to get the desired range. The pot is wired as a variable resistor so that a wiper contact failure results in minimum negative voltage going to the grids. Note all components are on the bias supply board, except for the resistor going to the reservoir ground point, and the pot case is not used as a ground.

5. Where will the V3 cathode resistor and the bypass electrolytic capacitor connect to ground?

You've missed the separation of the B and C filter grounds from the A (preamp) filter ground. Break the link between the BC and D negative terminals on the filter board and run a new BC negative terminal lead to the pink eyelet, and from there a lead to a ground point near the 6V6's.

6. Where will pin 8 of the second 6V6 power tube socket connect to ground?

The 6V6's are grounded through 1 ohm current sensing resistors located on the test jacks.

I'm especially not sure if I translated the speaker jack wiring correctly since the original Fender layout you modified does not show the names for the speaker jack terminals.

The Fender layout shows the lug positions for Switchcraft jacks. It's easy to determine which is which from the feedback connection.