6G3 Capacitor Question

[dwinstonwood]

Hi,
I've been looking at the 6G3 schematic, and I'm not clear on the purpose of the cap I've circled in the attached image. Given my limited knowledge, it looks like it's letting part of the signal bypass the pots and the second gain triode. If so, why? If not, what is it doing?

Thanks!
David

[sluckey]

It shunts some hi freq signal to ground thru the B+ filter caps. This makes the normal channel darker than the bright channel.

[dwinstonwood]

Thank you! If I was to build a single channel version of this amp, I guess I could add a switch to take it in or out of the circuit. That seems easier than moving the guitar cable to another jack.

[HotBluePlates]

Thank you! If I was to build a single channel version of this amp, I guess I could add a switch to take it in or out of the circuit. That seems easier than moving the guitar cable to another jack.

If you build a single-channel version, you might just try the amp with & without the cap.  My guess is you'll pick one and stick with it.

Until a few months ago, I had a '62 and '63 6G3 at the same time.  I mostly played in the Bright channel (no-cap).

[dwinstonwood]

If you build a single-channel version ...

Thanks HotBluePlates. I'm thinking about it. (My schematic uses the bright channel, and I probably attached sluckey's trem-o-nator in the wrong place.)

[sluckey]

I probably attached sluckey's trem-o-nator in the wrong place.

Yes, you did. Well, it's actually connected to a good point but there must be another cap between C12 and the LTP-PI circuit. Speed pot is wired wrong. And the optocoupler does not use a phototransistor. It uses a LDR.

[2deaf]

This type of tremolo works by applying a varying load on the 12AX7 causing a varying change in gain.  In order to do this, the intensity pot must be connected to the plate via a coupling capacitor.  Connecting the intensity pot to the 15K/100K junction barely changes the load on the 12AX7 and it is in the load range that doesn't have very much gain change with load change. 

The output impedance is around 14K as drawn.  It is going to take an LDR that is capable of a very low resistance to work against this low output impedance.

[dwinstonwood]

Thanks sluckey and 2deaf.

I couldn't find an LTSpice model for a VTL5C1 so I just stuck something in there that looked kind of similar to me.  

[66Strat]

Some suggestions

[dwinstonwood]

Thanks 66Strat. I think I fixed the things you pointed out.

Until I can find or teach myself how to model a VTL5C1 my LTSpice schematic won't run. It just freezes up. No big deal.

One thing I did discover, there's no way I'll get the same plate voltages on the first 12ATX7 shown on the original schematic using those plate load resistors.

[2deaf]

One thing I did discover, there's no way I'll get the same plate voltages on the first 12ATX7 shown on the original schematic using those plate load resistors.

That doesn't matter.  The preamp will operate just about the same on a wide range of voltages.

[66Strat]

You're overthinking things. Voltages in Fender schematics and layouts rarely match real life conditions. The plate resistors in the schematic are key in voicing and setting the gain structure of the amp. The 220K ohm resistor in combination with the 1 meg volume pot sets the gain of the first triode. The 15K ohm and 100K ohm resistors are a distributed load that sets the gain stage of the second triode.

As you only plan for a single channel amp, the resistance of the first triode cathode resistor should be double to maintain the same bias and the by pass capacitor should be halved to maintain the same frequency response of the stock circuit. However, this would be an area to experiment with to voice the amp to your taste. The 10uf capacitor in combination with the 3K ohm resistor has a 3db down turnover frequency of 5 hz compared to the stock 25/25 combination of 4 hz. Bypass capacitors of 4.7 uf and 2.2 uf in combination with the 3K ohm resistor would have turnover frequencies of 11 hz and 24 hz respectively.

[HotBluePlates]

And another thing to be aware of related to voicing:  Whatever speaker you pick, it will probably be darker than the original Oxford.

When I bought one of my 6G3 amps, the prior owner had "saved" the Oxford 12K5 and installed a Celestion alnico Cream speaker.  He wanted to know if I wanted to buy the amp with the new speaker, and my answer was "No!"  I wanted to hear the amp as it originally was.

I pretty much stopped building amps the past 8 years or so and went down a vintage rabbit-hole, mostly coming to find that while voltages matter (especially the imperfections of some power transformers), that speakers matter a whole lot more.

Another finding was that sometimes if you "fix the flaws" in some amp designs (or make the "more versatile") you can wind up squashing the amp's unique identity.

[dwinstonwood]

Wow! Thanks for all the input and guidance. Much appreciated.

I should state up front that I've never played through a 6G3. And, I think the examples on YouTube are not a good source because who knows what's been modded, or how the recording was done.

The truth is, I've been casting around for a second amp project and this one looked "not too complicated," yet pretty cool. Also, given the tweeds, browns, blondes and blacks that I have played through or heard others use, I've always liked the brown and blonde tolex era amps best. Not sure how empirical vs subjective that preference actually is.

HotBluePlates, I've felt that way about speakers for a long time. I plan to buy a 12" cabinet for this amp so I'll have the option of using a great speaker.

Having said all that I did choose a 58yr old design after all, so I would like it to sound as close - within practical reason - to the original configuration. I'm thinking now that I should build it as the two channel original. It's actually the same number of tubes either way.

The AC4 I just built sounds great. I love it. sluckey and others helped me a great deal in making that project a success. Their help made it possible. However, over the weekend I went back in and changed the .1u coupling cap à la AC15 to the .047u shown on the schematic. Man, the sparkle really came out. Noticeably brighter, but not shrill. I guess chime is the word Vox people use. So, there's an example of the original parts working better for me, at any rate.

So, back to the original schematic it is!

[sluckey]

So, back to the original schematic it is!

Hoffman has an eyelet board for $22.

[dwinstonwood]

So, back to the original schematic it is!

Hoffman has an eyelet board for $22.

Super! Thanks. His pricing and shipping are legit. Too tired to struggle with LTSpice modelling tonight. I'll get your Trem-O-Nator added to my drawing tomorrow. I have no knowledge of tremolo workings, but I like the idea of not wobbling the bias voltage, for some reason. BTW, I was trying to follow the rat's nest of a tremolo circuit in the 5G13 Vibrasonic. Looks like that was only used about once from what I can tell. Is that true pitch vibrato?

Added the other channel. FWIW, and I agree, I get too absorbed in LTSpice, there's only 10V difference between the first and second preamp tubes now - 160 vs 170. That brackets the 165 on the schematic!

[66Strat]

So, back to the original schematic it is!

Hoffman has an eyelet board for $22.

I have no knowledge of tremolo workings, but I like the idea of not wobbling the bias voltage, for some reason.

Why not? There is nothing wrong with the original tremolo circuit.

I've actually played through several vintage 6G3 amps. The first belonged to a buddy of mine that I played with in 1968 when I was a freshman in high school. I have played through more than one since that time. The only thing lacking in those amps was the Normal "Mud" channel. It certainly wasn't the tremolo.

I like Slukey's idea of using Hoffman's circuit board as a starting point. I would loose the .003 uf cap shunt across the Normal channel plate resistor to clean up the "Mud" channel. And, I would change the Bright channel coupling cap to .002 uf.

Modded to add Waylon just because.
https://www.youtube.com/watch?v=PO4CGuh9xR4

[dwinstonwood]

Modded to add Waylon just because.

Nice. I had to go listen to some Don Rich after that. That guy on the Electric XII with the mop cut and sideburns looks like he would be more at home in the Turtles than in Waylon's band.

Good point about the Fender tremolo. I think I get stuck in the cathode biased amp mentality. This would be the first fixed bias amp build for me (only my third tube amp build ever). So, would there be any need or real advantage to adding a bias pot here?

[66Strat]

My best friend growing up was a huge Don Rich fan. I enjoy Buck and Don, especially after consuming a few adult beverages. There's nuthin more funner than drinkin beer and playin Buck Owens songs. Gotta put flatwounds on that tele to capture that early Don Rich tone.
https://www.youtube.com/watch?v=UkNsqdGm0wo

It's not essential to have a bias pot, but there certainly would be no harm in adding one to the circuit.

[bmccowan]

On the bias pot issue - I have built both a 6G2 and a 6G3 - I put a bias pot in both and am glad I did - from my post in the 6G2  build thread (including a Monty Python reference):

My plate B+ is about 316, which is right on spec, but I found that with the standard -35V bias supply I could not set the power tube bias where I wanted (+/- 24ma). The highest I could get it was 13ma - pretty darn cold, so I changed the trim resistor from 27K to 22K and got a good range on the bias pot, and -28V.
Interesting to note that Rob Robinette's site states (for the 6G3 which uses the same trem circuit) that "Bias is non-adjustable because adjusting the bias will change the intensity of the tremolo which is the major weakness of bias wiggle tremolo." But I would counter that adding adjustable bias allows you to dial in the sweet spot. I found that if you bias the tubes too hot, the tremolo starts to make a flutter noise like the wing beat of an unladen African Swallow. But setting to about 60-70% of max range gets great drive and nice tremolo. So I suspect Fender biased these cool intentionally to avoid issues with the range of tubes that might be installed.

I suppose that an amp manufacturer wants to make sure that the amp will function ok with any set of power tubes - but we have the ability (ok, need) to tinker, eh? So once you decide on your power tubes, you can dial in the trim resister and set the bias to your own ears.

[dwinstonwood]

Great info bmccowan. Thanks!

As everyone here has probably always known, the big expense of amp building - which I'm now realizing - is in the tx's, chassis, tubes, cabinet and speaker(s). In other words, if I'm going to commit to investing $800 or more in an amp I need to be sure this is the amp I really want.

Another option - and, I'm leaning this way - is to not build this as a combo (but, just as a chassis) and buy a 12" speaker cabinet that could also be used with my AC4, etc. I'll never be using more than one amp at a time.

[bmccowan]

I built mine as a head for that reason and for the weight issue.

[66Strat]

I built mine as a head for that reason and for the weight issue.

Nice!

[dwinstonwood]

That looks great bmccowan!

[bmccowan]

Thanks - I bought the chassis and plates from Marsh. I built the cab from pine - I enjoy the woodwork, but hate gluing up Tolex.

[66Strat]

Just a thought. If you're searching for a new project, why not build a ReVibe. You could use this with your AC4 as well as any other amp that you may build or acquire in the future. I'm not much of a reverb fan, but I always preferred the 6G15 reverb unit to Fender amps with on-board reverb, and the ReVibe is an iteration of the Fender harmonic tremolo that arguably is one of the best trems ever.

https://el34world.com/Hoffman/files/Hoffman_ReVibe.pdf

[dwinstonwood]

Just a thought. If you're searching for a new project, why not build a ReVibe. You could use this with your AC4 as well as any other amp that you may build or acquire in the future. I'm not much of a reverb fan, but I always preferred the 6G15 reverb unit to Fender amps with on-board reverb, and the ReVibe is an iteration of the Fender harmonic tremolo that arguably is one of the best trems ever.

https://el34world.com/Hoffman/files/Hoffman_ReVibe.pdf

Now that's an interesting idea! I've been looking at that harmonic tremolo circuit in the "6-series" amp schematics, and I've heard it's really good. It would be a challenging project, too. I'm going to study up on that. I could definitely get a lot of use out of one. Thanks!

edited to add: I could build that, and buy a speaker cabinet for less than a full-blown combo project, too.

[dwinstonwood]

Thinking it over, I'm going to build the 6G3. Mojotone has a chassis/faceplate combo that's reasonably priced. That saves a lot of drilling and filing by hand. I'll use the Hoffman board. I can buy a head cabinet later, after it's built.

But, I'm finding these old Fender layouts difficult to follow in places.

In the detail below, the line with "+270V" beside it is clearly connecting B+ to the two plates of the first 12AX7 via the 220k load resistors. So, I guess the X's inside the boxes must indicate that those two points are connected under the board, since the other X leads to the second 12AX7 plate via the 15k and 100k resistors. The little circle with the dotted line makes no real sense, unless it's trying to imply that the connection is made under the board. I assume that the grounded line with "Black" beside it simply means that B+ gets grounded inside the chassis, as opposed to being grounded at the doghouse?

It's not a big deal, I can follow the schematic to see where things are connected. I need to Google photos of actual chassis wiring. Then, the odd symbols in the Fender layout will make sense. I just need a visual reference to verify things.

[dwinstonwood]

OK, this photo clears up the drawing on the layout and my assumption about the black grounding wire...

But, why run the yellow wire through the hole and connect it to B+ under the board? What's the reason for that?

[shooter]

But, why

1st thought, safer;
as the "B+" wire bounces to the beat n saws into the eCap it's riding on
keeps it close to chassis to possibly help with PS noise
looks cooler, but that's cosmetic 

[dwinstonwood]

Thanks shooter.

I also see that if soldered on top it would create a junction of five connections in one place.

[66Strat]

Thinking it over, I'm going to build the 6G3. Mojotone has a chassis/faceplate combo that's reasonably priced. That saves a lot of drilling and filing by hand. I'll use the Hoffman board. I can buy a head cabinet later, after it's built.

But, I'm finding these old Fender layouts difficult to follow in places.

In the detail below, the line with "+270V" beside it is clearly connecting B+ to the two plates of the first 12AX7 via the 220k load resistors. So, I guess the X's inside the boxes must indicate that those two points are connected under the board, since the other X leads to the second 12AX7 plate via the 15k and 100k resistors. The little circle with the dotted line makes no real sense, unless it's trying to imply that the connection is made under the board. I assume that the grounded line with "Black" beside it simply means that B+ gets grounded inside the chassis, as opposed to being grounded at the doghouse?

It's not a big deal, I can follow the schematic to see where things are connected. I need to Google photos of actual chassis wiring. Then, the odd symbols in the Fender layout will make sense. I just need a visual reference to verify things.

Good plan.

IMO, your biggest challenge will be finding a power transformer that will give you a target B+ voltage 370V DC with a GZ34 rectifier, that will also fit the chassis. IMO, the transformer that will get you closest voltage wise is the Hammond 272DX, but it will not fit the Mojo chassis.

[66Strat]

The Hammond 290BX is the best option IMO that will fit the chassis, but the B+ voltage will come in around 400V DC with a 5U4 rectifier. That wouldn't be a bad option, but you will need an adjustable bias pot to optimize the bias.

[dwinstonwood]

The Hammond 290BX is the best option IMO that will fit the chassis, but the B+ voltage will come in around 400V DC with a 5U4 rectifier. That wouldn't be a bad option, but you will need an adjustable bias pot to optimize the bias.

Yeah, that's the one Hammond lists as a drop-in replacement for the 125P2A.

Classictone has one with multiple secondary windings. This datasheet shows 375V with a 5U4GB:

http://www.classictone.net/40-18028.pdf

[66Strat]

That 375V is at a 120 milliamp current draw. The voltage at 120 ma would be pretty much full tilt boogy for the 6G3. It will be around 400 volts at idle. The 600 volt taps will get you closer.

[dwinstonwood]

Thanks.

So, my math skills are nonexistent. Using Tubenit's current draw pdf I get the following range for tube current draw in the 6G3:

Preamp 12AX7 - 1-2ma * 2 = 2-4ma
PI 12AX7 - 5-10ma * 1 = 5-10ma
6V6 - 70ma/pair = 70ma
Totals = 77ma - 84ma

So, how do I calculate the PT's voltage drop with loads of 77ma to 84ma with 120V wall AC?

Then, I guess I would multiply that figure by the GZ34's voltage increase factor of 1.3 to get plate voltage?

I'd like to be able to calculate rough estimates when shopping for transformers.

Thanks,
David

[66Strat]

Unfortunately Classictone doesn't give much information to work from. It would be nice if they provided no-load voltages or winding resistance specifications to go along with the voltage at the current rating. That is one of the reasons why I am not a fan of Classictone power transformers. Another is that there are inconsistencies/errors in the information that they do provide. Take for example the voltages that they provided. They show a 5U4GB dropping more voltage than a 5Y3GT at the same current load.

Using the voltage at current rating that they provide you can however back into an estimate for the idle voltage using graphic measurement of the rectifier tube operating characteristics graph. No math is needed. With 330 volts AC applied to the plates a 5AR4/GZ34 will provide about 385 volts DC at 120 ma. If we draw a line parallel with the AC voltage lines on the chart it leads us back to about 410 volts DC at 75 ma. These tube data charts are useful in selecting a transformer. For example pick a desired idle voltage at a desired current and mark that point on the chart. You can then estimate the amount of rectifier sag by drawing a line from the idle current point to the maximum signal current point on the graph.

[dwinstonwood]

That's really cool 66Strat!

So, I can see that if I want about 375V DC output from the rectifier at 75ma ( 355@120ma) I would need just under 300V AC input to the rectifier.

Using the Classic Tone datasheet, that would require the R/W wires and a 5Y3GT giving 354VDC at 120ma.

Thanks for that lesson.  Very useful knowledge.

[66Strat]

You can't rely on the Classictone numbers. The 5Y3 numbers are way off. The numbers are more in line with what one would expect from a 5U4. The red/white (600 volt) wires with either a 5AR4 or a 5U4GB will get you closer to your target voltage and will not sag like a 5Y3.

[dwinstonwood]

OK, got it. Thanks.

[66Strat]

The 5AR4 will come in around 380 volts at 75ma. A 5U4 will come in around 365 volts at 75 ma. Either would work well.

Verbiage added:

These numbers reflect how one would expect the rectifiers to perform. But we still have no idea as to the DC resistance of the transformer. The transformer AC voltage at the specified load of 120 ma. (300 volts) will be higher at the 75 ma. load. For that reason, I would opt for the 5U4GB.

[dwinstonwood]

Sounds good. Thanks for calculating that! I'd rather be on the low side anyway.

I'll also add that the PT doesn't have to fit the chassis cutout. I'm not planning on making a physical replica. Hopefully, a sonic replica, but who knows what these things sounded like 60yrs ago.

So, something like this might work, too:
https://www.edcorusa.com/xpwr014
I used Edcors in my EL34 stereo amp. They are affordable and of very good quality. But, they are built to order and take a while to get.

[66Strat]

Sounds good. Thanks for calculating that! I'd rather be on the low side anyway.

I'll also add that the PT doesn't have to fit the chassis cutout. I'm not planning on making a physical replica. Hopefully, a sonic replica, but who knows what these things sounded like 60yrs ago.

So, something like this might work, too:
https://www.edcorusa.com/xpwr014
I used Edcors in my EL34 stereo amp. They are affordable and of very good quality. But, they are built to order and take a while to get.

Edcor does make nice transformers. If it doesn't need to fit the chassis, the Hammond 272DX would be a good option that is readily available from several online retailers for about $80.

The 6G3 sounded real nice 52 years ago.


https://www.hammfg.com/files/parts/pdf/272DX.pdf

[dwinstonwood]

Cool. Would you still recommend a 5U4 with the 272DX?

I was in the first grade 52yrs ago. But, I was watching American Bandstand with my older sister!
There were these brothers in the neighborhood who had a - what seemed back then like a huge - Kustom amp with the vinyl glitter upholstery. Though not a tube amp, it must have been the first guitar amp I ever saw. They would put it in the driveway and blast the neighbors.

[66Strat]

I would use either a 5U4GB with the Hammond 272DX or a 5AR4 with the Hammond 270EX.

Attached are the 5AR4/GZ34 operating characteristic graph with an approximation of the Hammond 270EX AC voltage supply superimposed and the 5U4GB operating characteristic graph with an approximation of the Hammond 272DX voltage supply superimposed. For either tube and transformer combination, I would use a lower estimate for 6V6 tube current for the tremolo to operate properly (about 70% of maximum plate dissipation). To find an estimate of the DC voltage under load, extend a vertical line from estimated current under load to it's intersection with the curved blue line. Extend a horizontal line to the left from this intersection to find the DC voltage output on the graph.

I have also attached a copy of the 6G3 Schematic for reference. In real life, one would expect the DC voltage to be considerably higher than 375V DC with 333V AC under load applied to the plates. The schematic power supply numbers are more reflective of what would be expected with a 5Y3 rectifier rather than a 5AR4/GZ34. In comparison, the AB763 Deluxe schematic reflects 420 V DC under load with 325 V AC applied to the plates of a GZ34.

https://el34world.com/charts/Schematics/files/Fender/Fender_deluxe_ab763_schem.pdf

[dwinstonwood]

Thanks again 66Strat; I'll need to take some time to read through that and study the graphs.

...I would use a lower estimate for 6V6 tube current for the tremolo to operate properly (about 70% of maximum plate dissipation).

Not sure what this means. Are saying to alter the bias voltage (by using a bias pot) to lower the plate dissipation to 70%? If I used sluckey's trem-o-nator would that alleviate the need for that?

In the mean time, here's a good quality schematic that I cleaned up in Photoshop.

[bmccowan]

Unfortunately Classictone doesn't give much information to work from. It would be nice if they provided no-load voltages or winding resistance specifications to go along with the voltage at the current rating. That is one of the reasons why I am not a fan of Classictone power transformers. Another is that there are inconsistencies/errors in the information that they do provide. Take for example the voltages that they provided. They show a 5U4GB dropping more voltage than a 5Y3GT at the same current load.

66, I have used quite a few Classictone transformers - good quality at good prices. But I have been annoyed by the exact same thing. Its a real head scratcher that they publish so much misinformation. You are prompting me to send them an email.
And dwinston - my experience with that amp is if the dissipation climbs much above 70%, the trem start to produce a bit of a thump. Nothing wrong with 70% though. Its a good sounding trem - as is Sluckey's by all accounts. I'm going to try his soon in a blackface concert chassis I picked up.

[shooter]

If I used sluckey's trem-o-nator

don't believe it works in your configuration, the ToN wants to wiggle the signal not bias as your version does.

here's how I stuffed it into my last build, compare schematics so you have an understanding the difference.

[66Strat]

Thanks again 66Strat; I'll need to take some time to read through that and study the graphs.

...I would use a lower estimate for 6V6 tube current for the tremolo to operate properly (about 70% of maximum plate dissipation).

Not sure what this means. Are saying to alter the bias voltage (by using a bias pot) to lower the plate dissipation to 70%? If I used sluckey's trem-o-nator would that alleviate the need for that?

In the mean time, here's a good quality schematic that I cleaned up in Photoshop.

No. It appears that the 35 ma number was referenced from the 6V6GT data sheet. The 35 ma figure is valid for only one set of AB1 operating points. At a plate voltage of 285 volts DC, with 35 ma of plate current, plate dissipation is 9.975 watts or 71.25% of 14 watts at this operating point. As plate voltage is increased, plate current needs to be decreased, to remain in class AB1, and for the tremolo to operate properly. The intent of my post was to illustrate the voltage sag associated with transformer and rectifier IR losses and to provide you with some predictive data so that you could select your own operating points. A bias pot, while not a necessity, would be a good idea to help dial the amp in. I would also add 1 ohm 1% cathode resistors to each 6V6 cathode to measure tube current.

[dwinstonwood]

don't believe it works in your configuration, the ToN wants to wiggle the signal not bias as your version does.

Thanks shooter. But, that's what I was trying to say: if I use a trem that doesn't wobble the bias voltage would that alleviate the problem of pushing the 6V6's above 70%. Or, am I just completely confused about all of this?