Vibrolux 5f11 build from integrated amp

Hey folks - novice here with my first build. I had done some reading on converting old integrated tube amps into guitar amps when the tube complement and transformers made sense for a conversion. Found this 1959 Craftsman "Conterto" CA-11 amp on eBay a while back for $70 and grabbed it for the tube complement which I thought might match some of the old Fender circuits. Did some research and was originally going to build a Tweed Deluxe 5e3, but then figured the tube complement exactly matched a Vibrolux 5F11, so I'm going with that.

However: BIG PROBLEM. After wiring up all the PT leads, I realized there was 1 tap I don't have. It's marked "brown" in the layout illustration below, and supplies the -31v bias current to the power tubes through the tremolo circuit. Oops.

So, for the moment I'm all dressed up with nowhere to go. I'm wondering if anyone has any suggestions on how I might get that necessary 31v supply? A knowledgeable friend suggested maybe adding a small 24v doorbell transformer, which would end up supplying about 33v, give or take. Thoughts?

BTW - I am really a novice here. I know enough to follow a schematic and understand a bit of the theory, but am clueless beyond that. No ability to design or modify a design by myself.

Thanks.


Original front panel
http://files.skaye.fastmail.fm/img/vb-ca-11-front.png

Front panel parts installed:
http://files.skaye.fastmail.fm/img/vb-front.png

Layout showing missing "brown" tap
http://files.skaye.fastmail.fm/img/vb-layout.png

Schematic showing -31v route from "brown" tap
http://files.skaye.fastmail.fm/img/vb-schematic.png

Original shot from rear
http://files.skaye.fastmail.fm/img/vb-ca-11-rear.png

Guts of integrated amp before stripping out
http://files.skaye.fastmail.fm/img/vb-ca-11-guts.png

Project with new sockets installed and filiment runs in place
http://files.skaye.fastmail.fm/img/vb-bottom.png

Shot from side showing power and output transformers
http://files.skaye.fastmail.fm/img/vb-side.png

Look above the power supply

http://mhuss.com/Plexi6V6/Plexi6V6c.gif

The 220k resistor knocks down the high voltage. Don't forget the diode is backwards as the bias voltage is negative. You may have to adjust the resistor value to get the voltage you want.

Look above the power supply... The 220k resistor knocks down the high voltage. Don't forget the diode is backwards as the bias voltage is negative. You may have to adjust the resistor value to get the voltage you want.

Thanks printer2. So, looking back at the 5f11 schematic, you're saying I should tie into the PT high voltage tap at one of the 5Y3GT plates (pin 4 or 6) and run it through a little bias-supply network like the one shown with the 220k resistor and two 47uF/100v caps, adjusting the 220k resistor to get the desired 31v?

Thanks printer2. So, looking back at the 5f11 schematic, you're saying I should tie into the PT high voltage tap at one of the 5Y3GT plates (pin 4 or 6) and run it through a little bias-supply network like the one shown with the 220k resistor and two 47uF/100v caps, adjusting the 220k resistor to get the desired 31v?

You got it. Mind you the circuit shown has a pot to adjust your bias. When I say adjust the 220k resistor, that is to put the center of the pot roughly about your desired bias voltage for the tubes. You can also adjust the range by changing the pot value as compared to the other resistor values but I think it is easiest to do the 220k resistor.

When I say adjust the 220k resistor, that is to put the center of the pot roughly about your desired bias voltage for the tubes. You can also adjust the range by changing the pot value as compared to the other resistor values but I think it is easiest to do the 220k resistor.

I showed your idea to my friend and he thought it was a great solution - very elegant- so thanks again. He suggested tweaking the 15k resistor as opposed to the 220k/3w or the 10k pot to adjust the circuit to the target 31v. His thought was it's easier and cheaper to try a bunch of plentiful 1/2 watt resistors values in the 15k position rather than trying to locate the right 3w resisitor or pot value.

I showed your idea to my friend and he thought it was a great solution - very elegant- so thanks again. He suggested tweaking the 15k resistor as opposed to the 220k/3w or the 10k pot to adjust the circuit to the target 31v. His thought was it's easier and cheaper to try a bunch of plentiful 1/2 watt resistors values in the 15k position rather than trying to locate the right 3w resisitor or pot value.

Good idea. Not sure if a 3W is needed though. I did a rough approximation using 300V and the 220k which gives 1.4mA. 1.4mA X 300V = 0.4W so a 1W - 2W should work depending on how the amp is built (heat wise). Ohm's Law will tell you what you need, just at least double the wattage to get the resistor wattage.

Making slow progress, but having fun at it. Part of the challenge here is there's no room for a proper layout on a card or turret board, so I have to re-interpret the layout diagram and "fly" the components old school! I did wire up the power supply on a couple of terminal strips though. I'm trying to route all the connections properly to reduce hum and will be using a multi-star grounding scheme. A friend suggested covering the component leads with housing stripped from other wire which seems like a cool idea, so I've been doing that.

If anybody sees any problems or has any suggestions I'd be happy to hear from ya. Thanks.

http://files.skaye.fastmail.fm/img/vb-circuit-120620.png


Layout diagram showing completed wiring in orange:

http://files.skaye.fastmail.fm/img/vb-schem-120620.png

Hi folks - started this thread originally on the "Amp Central Station" forum, but just learned about "Shock Brother's DIY Amps" so moved it over here. Glad to find y'all and looking forward to following the forum!

That's a beautiful chassis for a conversion.
If you get tired of calling it a 5F11, you could also call it a 6G2. It's a little late to offer that comment since you've already started, but be aware that the schematics are nearly identical, however different triode sections are used in different places, which may possibly fit the layout of a conversion better. Then again, maybe not.

That's a beautiful chassis for a conversion.

Yeah, hoping it might work out well. I grabbed the amp for cheap on eBay a couple of years ago just after I got the Dave Hunter book to begin studying this stuff. I found these two articles about converting integrated amps and figured the tube complement and transformers might work well.

Converting Integrated/PA Tube Amp into Guitar Amps (http://www.geofex.com/article_folders/old2new.htm)

Notes on Converting PA Amps to Guitar Amps (http://www.rru.com/~meo/Guitar/Amps/PA2Guitar/pa2g-notes.html)

The limited space inside the chassis is a challenge though - no room for a circuit board so I'm wiring it point-to-point directly off the components. It looks like I'll have to hang half the circuit directly off the pots where I've got space since there's very little space under the chassis if I want to keep things relatively neat (see the hornets nest inside the original amp above!).

If you get tired of calling it a 5F11, you could also call it a 6G2.

The 5F11 seems like a pretty cool design and I liked the idea that it matched the components I already had from the amp. I'd be interested to know more though about what you like especially about the 6G2?

I've been studying up on grounding (i.e. "the other half of the circuit") and have come up with the star grounding scheme shown below. There are 18 ground points in the schematic and the different colors each represent a local star node. There are 5 nodes (blue, orange, brown, purple, yellow) and each will have a wire run to the star. Plus 4 individual ground points (9, 14, 16, 17 in green) which will each be connected directly to the star.

http://files.skaye.fastmail.fm/img/vb-star-v1.png

Any feedback (no pun intended!) very much appreciated.

I have not done much with a star configuration, more with a ground and the minor currents on one end and the major on the other, closest to the filter capacitor node. A couple of things you probably would want to do is not to ground one end of the heater supply but rather go across both terminals with the winding and then use a couple of 100 ohm resistor to make a virtual center tap if the transformer does not have one on this winding. If it does you would want to ground it. Even better would be not to ground the heaters but to elevate the windings on a dc voltage.

Also you have a two wire line circuit there, not something that is done any more for safety reasons. You would want a three wire grounded chassis to keep things safe.

You should probably wire your heaters the modern way. See a weber 5e3 for example. There are other things, but it seems lengthy to get into. Sounds like you have a friend irl (does that mean "in real life"?) to help with stuff. Grounding is a dividing subject, I like a split ground.
Good luck with the build!!
*edit* I see printer beat me to it :)*

...you probably would want to do is not to ground one end of the heater supply but rather go across both terminals with the winding and then use a couple of 100 ohm resistor to make a virtual center tap if the transformer does not have one on this winding. If it does you would want to ground it. Even better would be not to ground the heaters but to elevate the windings on a dc voltage.

Also you have a two wire line circuit there, not something that is done any more for safety reasons. You would want a three wire grounded chassis to keep things safe.

Thanks printer2. Right, #17 just represents the heater ground, which I've wired as a twisted pair and will ground through balanced 100 Ohm resistors (no center tap on the PT). One question I had was does it matter at what point I add the virtual center tap - e.g. at one end or the other? I'm thinking it makes no difference and can be put anywhere it's convenient. I was planning to tap it at the pilot light.

Right too - I'm using a grounded 3-wire cord.

re: "Even better would be not to ground the heaters but to elevate the windings on a dc voltage.". Can you explain what this means? That concept's a little beyond my learning to this point.

There are other things, but it seems lengthy to get into...

Would you be able to simply list the problem(s) you see - no lengthy explanation necessary. Would be really helpful. Thanks.

I've been studying up on grounding (i.e. "the other half of the circuit") and have come up with the star grounding scheme shown below. There are 18 ground points in the schematic and the different colors each represent a local star node. There are 5 nodes (blue, orange, brown, purple, yellow) and each will have a wire run to the star. Plus 4 individual ground points (9, 14, 16, 17 in green) which will each be connected directly to the star.

http://files.skaye.fastmail.fm/img/vb-star-v1.png

Any feedback (no pun intended!) very much appreciated.

What I'd do (just sayin')

Numbers 1-7, 10, 11 can ground at point 12 (seeing as how those parts of the circuit all get their power supply from that filter cap)

Numbers 8, 15 and 16 can ground at point number 14 (ditto). Number 13 can also attach to number 14.

Ditch the death cap on the PT primary side (so there is no need for number 9) and wire in a proper modern 3-chord mains cable with a dedicated separate grounded chassis wire bolt (fastened with a nylock-nut).

Number 18 can just be grounded at the speaker jack ground lug.

Ditch number 17. Wire each side of the heater winding to pins 2 and 7 respectively of the 6V6s and pins 9 and 4-5 respectively ofthe pre-amp tubes using twisted pair wiring. Then use a pair of 100R resistors (one from either side of the heater winding) as an artificial ground reference point that can either:

a) ground with number 14, or

b) wire up a 1/10th - 1/8th voltage divider across the reservoir filter cap and decouple the knee of that divider with about 100V 10-47uF cap, and then attach your '2 x 100R artificial ground point' to that knee. (that way the heaters with be elevated and it should make the amp run quieter, and you will also have a filter cap bleeder network - through the heater elevation voltage divider - to help discharge the filter caps quickly on switch-off)

Mostly just the grounding stuff, but I want to see how the star ground works, cause I'm curious.
Since you only have one input, I think you can do a 1meg to ground and 33k resistor to grid instead of the 68k to simulate the hi gain channel on amps. Someone else back me up :)

Yay tubeswell to the rescue!

Thanks so much tubeswell. Trying to learn here, so a few questions:

Numbers 1-7, 10, 11 can ground at point 12 (seeing as how those parts of the circuit all get their power supply from that filter cap)

Isn't it safer to bring #3,4 together and then directly to the star, given they make up the most sensitive segment of the signal chain?

Numbers #1,2,5 are involved with the 6V6 bias supply - wouldn't it be safer to gather them away from the signal gain stages and then wire directly to the star?

Numbers 8, 15 and 16 can ground at point number 14 (ditto). Number 13 can also attach to number 14.

Shouldn't #8 and #15 be gathered with #13 since #15 grounds the 6V6 cathodes which are supplied by #13, and #8 is the 6V6 bias supply?

Why combine #13 and #14 rather than wiring them separately to the star?

Ditch the death cap on the PT primary side (so there is no need for number 9)

Ahhhh, THAT's what the cap at #9 is for ("death cap" - funny!). Nice to be able to delete it.

Number 18 can just be grounded at the speaker jack ground lug.

I understood it's a good idea to ground the speaker jack at the injection stage when global negative feedback is used?

Ditch number 17. Wire each side of the heater winding to pins 2 and 7 respectively of the 6V6s and pins 9 and 4-5 respectively ofthe pre-amp tubes using twisted pair wiring. Then use a pair of 100R resistors (one from either side of the heater winding) as an artificial ground reference point...

Yup, that's the plan.

Isn't it safer to bring #3,4 together and then directly to the star, given they make up the most sensitive segment of the signal chain?

When I said ground #1-7 and and #10 and #11 with #12, its best to have all thee grounded at #3 (input jack ground) to make one ground return point. Not only are these the most sensitive ground returns - in terms of carrying the tinniest amounts of AC shunted to ground - but they all draw their power from the same filter cap (with the #12 ground return). So if they all ground back to the same ground return point as the #12 ground return point in the 1st instance, then you won't get affected by microwobble appearing within the ground return from other filter caps which shunt higher AC to ground. But if you start introducing different connection points for the more sensitive parts of the circuit, then you increase the risk that the small resistances between the different connection points will accentuate any microwobble that is sitting within the ground return. The key part of grounding is to tie each ground return point to the correct filter cap ground in the 1st instance. That you would use the input jack ground as the chassis connection point for all this is logical because you only want one ground return connection point for each filter cap ground return point (if you follow my meaning).

Now its best to have these ground return points (even the filter cap) all going by their own separate wires to the input jack ground lug - but seeing as how thats about 9 separate wires trying to squeeze into the input jack ground return lughole - you can do it by a running all the wires to a terminal strip, which is as close as possible to the input jack ground lug, and then run a short fat wire to the ground lug.

Numbers #1,2,5 are involved with the 6V6 bias supply - wouldn't it be safer to gather them away from the signal gain stages and then wire directly to the star?

The 6V6s bias supply is a separate supply, so I wouldn't ground #1 and #2 with it (for the reasons I explained above)

Shouldn't #8 and #15 be gathered with #13 since #15 grounds the 6V6 cathodes which are supplied by #13, and #8 is the 6V6 bias supply?

The filter cap at #13 supplies the 6V6 screens, which is part of the 6V6 tube current, so its better to ground that with the 6V6 plate current supply ground, and have the 6V6 cathode (#15) grounded to the same ground return node.

Why combine #13 and #14 rather than wiring them separately to the star?

They both supply parts of the same 6V6 tube current - see above

I understood it's a good idea to ground the speaker jack at the injection stage when global negative feedback is used?

The NFB signal is affecting the pre-amp signal, so if you cross ground it somewhere else, you'll start introducing more resistance points between the pre-amp ground and other grounds and multiple ground return points, which has more possibility to wreak havoc by accentuation of incompatible microwobble with the ground return.

YMMV

I'm thinking of doing a 5F11 also. I see now, I would have run into the same problem. Glad I read this thread. If I have to do a bias board anyway, I may just do a brownface Deluxe.

When I said ground #1-7 and #10 and #11 with #12, its best to have all thee grounded at #3 (input jack ground) to make one ground return point. Not only are these the most sensitive ground returns - in terms of carrying the tinniest amounts of AC shunted to ground - but they all draw their power from the same filter cap (with the #12 ground return). So if they all ground back to the same ground return point as the #12 ground return point in the 1st instance, then you won't get affected by microwobble appearing within the ground return from other filter caps which shunt higher AC to ground.


If we make a star ground chassis connection for all the pre-amp grounds at #3, and a second star ground chassis connection for all the other grounds, isn't that the scenario where resistance in the chassis between the two stars might wreak havoc, as opposed to collecting all grounds at a single star (point) so there are (virtually) no shared return paths which might 1) mix the ground currents and 2) introduce resistance variability leading to voltage noise between the now-mixed grounds? I'm guessing that may be academic and that two separate stars work fine in practice, but I'm trying to understand what's going on, so 2 stars can't actually improve things over a single star, can it?

..if you start introducing different connection points for the more sensitive parts of the circuit, then you increase the risk that the small resistances between the different connection points will accentuate any microwobble that is sitting within the ground return.

I thought that was the point of the star ground - i.e. to separate the ground returns (to individual components or to local nodes that collect components affiliated with the same stage) and collect them at just one point so there is no shared path where the different stage's potentially different ground voltages can mix?


The 6V6s bias supply is a separate supply, so I wouldn't ground #1 and #2 with it (for the reasons I explained above)


Right - the 6V6's bias supply ground is #8, which I was planning to ground with #13 and #15 - i.e. not with #1 and #2.


The filter cap at #13 supplies the 6V6 screens, which is part of the 6V6 tube current, so its better to ground that with the 6V6 plate current supply ground, and have the 6V6 cathode (#15) grounded to the same ground return node.


But still, why collect #13 and #14 together when they can have separate runs to the star? i.e. it wouldn't improve things, would it?


The NFB signal is affecting the pre-amp signal, so if you cross ground it somewhere else, you'll start introducing more resistance points between the pre-amp ground and other grounds and multiple ground return points, which has more possibility to wreak havoc by accentuation of incompatible microwobble with the ground return.

Is collecting 18 with 6,7,10 cross grounding it? I thought that would be grounding it with the pre-amp stage it's being injected to?

If we make a star ground chassis connection for all the pre-amp grounds at #3, and a second star ground chassis connection for all the other grounds, isn't that the scenario where resistance in the chassis between the two stars might wreak havoc, as opposed to collecting all grounds at a single star (point) so there are (virtually) no shared return paths which might 1) mix the ground currents and 2) introduce resistance variability leading to voltage noise between the now-mixed grounds? I'm guessing that may be academic and that two separate stars work fine in practice, but I'm trying to understand what's going on, so 2 stars can't actually improve things over a single star, can it?

The ground at the input is being decoupled through the filter cap at #12, and the chassis is acting as the connecting path to the other decoupled ground point(s). But you can run everything to a single star point instead, as long as you have separate wires from each ground return point to the star. See the attached article by R.G. Keen

But still, why collect #13 and #14 together when they can have separate runs to the star? i.e. it wouldn't improve things, would it?

What I mean is #13 and #14 can be grounded together (by their own separate wires to the common point)

Is collecting 18 with 6,7,10 cross grounding it? I thought that would be grounding it with the pre-amp stage it's being injected to?

Yes because the NFB is also part of the pre-amp circuit, which is more sensitive

Tonebrulee! Sorry I didn't see back on your heater picks, I see you are doing them the modern way (i.e. not grounding pin 9 on 12a_7 for example), sorry for not seeing that. I still wish someone would chime in on doing the input like a single HI style.

The ground at the input is being decoupled through the filter cap at #12, and the chassis is acting as the connecting path to the other decoupled ground point(s)...

Right, it's probably fine that way, but it's not as sure a thing as...

... you can run everything to a single star point instead, as long as you have separate wires from each ground return point to the star.

... right?

See the attached article by R.G. Keen

Exactly - Keen is showing the single start point, right? He likes the stage local nodes at each stage's filter supply cap, which seems as good a choice as any if it's convenient, but in the case of the 5F11, all the stages before the 6V6's are being supplied by that one, last filter cap at #12. That would seem like a lot of wire into that one point, which is why I was thinking of making a node of each stage and tying them each directly to the star. Thoughts?


What I mean is #13 and #14 can be grounded together (by their own separate wires to the common point)


You're saying their own separate wires to the star point, right?


Yes because the NFB is also part of the pre-amp circuit, which is more sensitive

I'm thinking by "pre-amp" maybe you mean only the 1st (half) 12AX7 gain stage? Whereas I was thinking "pre-amp" meant both (half) 12AX7 gain stages before the signal hits the 6V6's.

If that's right, then you're saying that #6, #7 and #10 are part of the "pre-amp" whereas #18 is part of the 2nd (half) 12AX7 gain stage?

For #10, I'm not clear on how that's part of the first (half) 12AX7? For #6 and #7, they might be considered part of either (or both) the 1st or 2nd 12AX7, but I read this from Randall Aiken's "Star Grounding" doc (http://www.aikenamps.com/StarGround.html):

You should always solder a wire from the grounded pot connections back to the common local ground of the stage the pot is used in. For example, the grounded pin of the volume pot, if it is located at the grid of the second tube section, should go to the local common point for that second tube section's cathode resistor and bypass cap.

So he's saying #6 and #7 should form a node with #10, right?

For #18, I'm still not seeing how it's not directly part of only the 2nd (half) 12AX7. Consider this from Merlin (http://valvewizard2.webs.com/Grounding.pdf):

If global feedback is used then the speaker ground should be returned to the local star of whichever stage the feedback happens to be injected into, usually the phase inverter (e.g., fig.13.14).

Thgoughts?

Tonebrulee! Sorry I didn't see back on your heater picks, I see you are doing them the modern way (i.e. not grounding pin 9 on 12a_7 for example), sorry for not seeing that. I still wish someone would chime in on doing the input like a single HI style.

No problem andy.

Not familiar with "single HI style" - can you explain?

I've made some suggestions and whether you want to go with this is ultimately over to you. There are several methods for grounding. I've given you a suggestion for a split ground, which works fine with tweed amps (I've built several totally hum-free amps this way), because your drawing seemed to infer splitting the ground, and I was sharing my experience about how to split it because you asked anyone for suggestions. I've also used the star grounding ('R.G.'s') method and the galactic ground ('Merlin's') method in various amps.

However the best way to find out about this stuff is to build the amp and experiment for yourself. If you're in doubt about the end result, then build the amp in such a way that you can easily switch things around if that is a concern to you.

I've made some suggestions and whether you want to go with this is ultimately over to you. There are several methods for grounding. I've given you a suggestion for a split ground, which works fine with tweed amps (I've built several totally hum-free amps this way), because your drawing seemed to infer splitting the ground, and I was sharing my experience about how to split it because you asked anyone for suggestions. I've also used the star grounding ('R.G.'s') method and the galactic ground ('Merlin's') method in various amps.

However the best way to find out about this stuff is to build the amp and experiment for yourself. If you're in doubt about the end result, then build the amp in such a way that you can easily switch things around if that is a concern to you.

Oh tubeswell - truly, all very much appreciated! Hope I didn't give any impression otherwise - as I said I'm trying to understand it all, that's the only reason for all the questions.

I think I can understand why a full-blown star ground might not be necessary for tweed amps - simple circuit with relatively low gain. But I wanted to start off with the star b/c studying it has been a great education in that it touched on all the issues around grounding and ground-related hum. Plus it's the one method said to be sure to give a good result (if implemented correctly). I'd like this first build to work with relatively little hassle (e.g. trying to debug ground-related hum as a newbie) before graduating to less deliberate grounding schemes that can require more experimentation.

With grounding, it seems like there are a lot of ways potentially to get it right, but maybe even more ways to get it wrong! I'm just trying to narrow down the possibilities of getting it wrong this first time 'round.

Yep - I understand your concern to get things right - its part of the obsessive nature of this hobby. However its also normal to be troubleshooting a build, so its wise not to build up your expectations too much before the build is complete. If your'e getting keen on the star ground (no pun intended), then follow R.G.'s method to the letter (which means isolating all the jacks from the chassis etc)

@tubeswell - Cool. R.G. has another more detailed doc here (http://www.geofex.com/article_folders/stargnd/stargnd.htm), in case you haven't see it. Excellent example case using a Vox Pacemaker.

No problem andy.

Not familiar with "single HI style" - can you explain?

Yeah, I got thrown off when I saw one end of the heaters looking like it would go to your star ground stuff in your photo.
I don't fully understand it, buuut!
A typical fender has a hi and lo with 68k gridstoppers and the 1M resistor to ground. Through some wizardry :) aka maybe the parallel nature the hi goes to the grid through roughly 34k resistance and the lo is like 134 or something (?).
Soooo, no one likes the lo input, well some do but... So since your amp has one input, I thought it might be a nice deviation from that oldie scheme (which by doing the heaters the "right" way you're already diverting).
As an aside, there are lots of ways to get from input to grid, but it seems lots of builds gravitate to the 68k hi lo style.
I need to look at a pic of the original fender amp you're building to see how they designate the inputs, cause it looks like there is no hi lo, just three like inputs.
Good luck with the build!!
*edit, am I the only one who needs to go back and fix all my "your/'es"? :)*

Hey folks - here's a list of the documents I've found most helpful in coming to understand the multi-faceted topic of grounding. You've probably already noticed that grounding can spark some strong opinions, so you'll find some conflicting information across the document. But overall, studying them closely I found them really illuminating. Cheers - hope others find them helpful.

Star Grounding - Randall Aiken (http://www.aikenamps.com/StarGround.html)

Star Grounding in Tube Amplifiers - R.G. Keen (http://www.geofex.com/article_folders/stargnd/stargnd.htm)

Grounding (PDF) - Merlin Blencowe (The Valve Wizard) (http://valvewizard2.webs.com/Grounding.pdf)

Other sources of hum in star grounding (PDF) - R.G. Keen (http://www.geofex.com/Article_Folders/star%20grounding%20amps.pdf)

Basic Vacuum Tube Amplifier Design - oldstockaudio.com (http://www.oldstockaudio.com/DIY&Parts/basic_amplifier_design.htm)

Grounds - Hoffman Amps (EL34 World) (http://www.el34world.com/charts/grounds.htm)

A typical fender has a hi and lo with 68k gridstoppers and the 1M resistor to ground. Through some wizardry :) aka maybe the parallel nature the hi goes to the grid through roughly 34k resistance and the lo is like 134 or something (?)

Hey folks - Andy raises an issue here that I've always puzzled over and would love to understand. Looking at the schematic below for the input section of a 5E3, I don't understand what is determining the HI vs. LO input behavior? I'm thinking it has something to do with one jack in each pair seeing a 68K/68K voltage divider (resistance being lowered to 34K to get the LO input?) where the other jack only sees as single 68K resistor? Any help appreciated - thanks.

http://files.skaye.fastmail.fm/img/vb-5e3-inputs.png

See Andy's post two up.

See Andy's post two up.

Right, I was quoting Andy's post, but he wasn't sure about the explanation, so I'm asking for clarification. Thoughts?

The input jacks have a tip switch that opens when the plug is inserted.

With the hi-jack, when you insert the plug, the tip switch goes open/off and the signal wire sees 1M to ground and 34k (68k || 68k) to the grid (because the Lo jack's tip switch is still 'on'). In this mode, the 1M to ground offers a lower AC load (and forms a better impedance bridge to the geetar p'up's output impedance), so you get a stronger signal appearing at the V1 grid, with a fuller bandwidth, and the 34k in series with the grid interacts with the miller capacitance of V1, to cull higher RF-type interference, whilst still allowing a good amount of highs from the geetar.

With the Lo-jack, when you insert the plug, the tip switch goes off and the signal wire sees 68k to the grid and 68k to ground (because the hi-jack's tip switch shorts across the 1M - leaving only the 68k from the grid to ground). In this mode, the AC load on the geetar p'up is bigger (i.e.: its a 1:1 voltage divider - and the guitar's signal is attenuated by 50%), and the 68k which in series with the grid interacts with the miller capacitance of V1 to roll off more highs (than it does with 34k), so the guitar signal is cleaner and tamer. As it happens, with a 12Ay7 or a 12AX7, 68k is about the highest resistance you can put in series with the grid (of your typical unbootstrapped V1 stage) without noticing a lot of hi-frequency roll-off. If you used something bigger than 68k (of your typical unbootstrapped V1 stage), you would quickly notice a lot more highs being rolled off, because of the effect of the interaction with the miller capacitance of the tube.

So if you just want the 'hi-jack' option, you use one switching jack and 33k (which is about the same as 34k) to the grid and 1M to the ground.

@tubeswell - thanks, that makes sense. I marked up the schematic below showing what I think you're saying is the signal path for High and Low. For High, the signal has two paths to grid (blue & purple), each seeing a 68k resistor, so the signal as a whole sees 68k || 68k, or 34k total, and also sees the 1M resistor to ground (green). For Low, the signal has just the one path to grid through one 68k resistor (blue) and one 68k resistor to ground (green), the two resistors acting as 1:1 voltage divider. Have I got that right?

http://files.skaye.fastmail.fm/img/vb-5e3-inputs-hi-lo-2.png

If that's correct, could you explain what's going on below with the 3 different inputs on the 5f11? Are they all the same, each seeing 68k to grid?

http://files.skaye.fastmail.fm/img/vb-5f11-inputs.png

Input #1 you have the three 68k resistors in parallel and the grid of the tube returned to ground through the guitar. This gives the maximum signal to the grid. In input #2 you have two 68k resistors in parallel with the grid to ground through the #1 68k resistor. This gives 2/3 the signal as #1. In input #3 you have the 68k with the other two resistors in parallel to ground. This gives 1/3rd the signal.

@tubeswell - thanks, that makes sense. I marked up the schematic below showing what I think you're saying is the signal path for High and Low. For High, the signal has two paths to grid (blue & purple), each seeing a 68k resistor, so the signal as a whole sees 68k || 68k, or 34k total, and also sees the 1M resistor to ground (green). For Low, the signal has just the one path to grid through one 68k resistor (blue) and one 68k resistor to ground (green), the two resistors acting as 1:1 voltage divider. Have I got that right?

http://files.skaye.fastmail.fm/img/vb-5e3-inputs-hi-lo-2.png

If that's correct, could you explain what's going on below with the 3 different inputs on the 5f11? Are they all the same, each seeing 68k to grid?

http://files.skaye.fastmail.fm/img/vb-5f11-inputs.png

in your 5e3 pic, those are two separate hi lo couples. One set is Normal channel with hi lo, and the other is Bright channel with hi lo.
In the other pic, that's what I was concluding too, just 3 like inputs (pretty sure).

in your 5e3 pic, those are two separate hi lo couples. One set is Normal channel with hi lo, and the other is Bright channel with hi lo.

OK, with the Normal channel at the top? Normal being less bright b/c of the additional .0005uF cap (purple line) the signal sees on it's way to the shared tone control?

If that's right, does the fact that Normal channel sees the Tone control before it's Volume control (at least it looks that way in the shematic), while the Bright channel sees it's Volume control before the Tone account for any additional difference in brightness?

http://files.skaye.fastmail.fm/img/vb-5e3-inputs-hi-lo-normal.png

In the other pic (of the 5F11 input stage), that's what I was concluding too, just 3 like inputs

@printer2 says otherwise - I'm guessing his read is correct?

Input #1 you have the three 68k resistors in parallel and the grid of the tube returned to ground through the guitar.

@printer2 - can you explain the signal-to-ground path when you say the Input #1 signal is "returned to ground through the guitar"?

Mislabeled the inputs as Channel # 1, 2, 3. Should be Input #1, Input #2, Input #3.

http://i406.photobucket.com/albums/pp142/printer2_photo/Harvardinput.jpg

In a normal two input circuit we have two 68k resistors and the 1M resistor. No mater what there is a connection between the grid and the bottom of the cathode resistor (the one on the bottom of the tube).

The next diagram is the Harvard, notice no 1M resistor. Now in the third diagram we are plugged into the # 1 jack and we are no longer shorting the # 1 input grid resistors to ground. All three resistors are in parallel so they act as one 22k resistor (first diagram below).

Since we need a complete path from the grid to ground for the tube to be biased properly and we do not have a resistor to do it for us when we are in the # 1 input, the guitar has to act as our resistor. Because we have the volume control of the guitar directly on the grid (through the 22k resistor) there is 100% of the guitar signal to the grid that the volume control is putting out.

Now in the next diagram we are going into input # 2. The switch on that input is open now and the circuit has the # 2 and # 3 input resistors in parallel with each other and where they join at the grid, the # 1 input resistor goes to ground. This gives us a voltage divider and the grid only sees 2/3 the signal from the guitar.

When we go into input # 3 we have input #1 and # 2 resistors in parallel with each other and a resistor divider as shown. This gives us 1 /3 the signal from the guitar to the grid.

Thanks @printer2, that's a great explanation, but I should have been clearer in what I was asking. My question was about your original comment that Input #1 is "returned to ground through the guitar", or as your wrote it above, for Input #1 "the guitar has to act as our resistor".

i.e. What specifically is ground for Input #1, and where in the Input #1 circuit does the signal see that ground?

Thanks @printer2, that's a great explanation, but I should have been clearer in what I was asking. My question was about your original comment that Input #1 is "returned to ground through the guitar", or as your wrote it above, for Input #1 "the guitar has to act as our resistor".

i.e. What specifically is ground for Input #1, and where in the Input #1 circuit does the signal see that ground?


Ground for the input is the chassis that the input jacks are mounted to. The 1.5k resistor has a wire running from the non-cathode end to the jack. It is the same point for the other inputs but they do not rely on the guitar to complete the grid circuit.

Cool thanks Printer!
Sorry Tonebrulee for the super diversions, some good info!!

Ground for the input is the chassis that the input jacks are mounted to. The 1.5k resistor has a wire running from the non-cathode end to the jack. It is the same point for the other inputs but they do not rely on the guitar to complete the grid circuit.

Ya know, I was really puzzled by this for a while b/c I couldn't understand how inputs #2 and #3 could have a connection to ground directly from the signal path, but input #1 could work without one. Then I had an AH-HA moment when I realized the 68k resistors to ground for #2 and #3 aren't the RETURN ground for the signal, they are there to SHUNT some of the signal to ground only. Input #1 has no shunt to ground, so the full signal from the guitar is going straight to the V1 grid. The RETURN ground for all three inputs is the same: it's the cable ground which connects the minus-side of the pickup to the sleeve of the input jack. Have I got that right now?

The big lesson here for me is that while "ground" is portrayed as a kind of generic thing where it's all equal, there are actually different reasons for connecting a circuit to ground - in this case it was the difference between a signal ground RETURN and a signal SHUNT to ground.

You made my head spin with your explanation. The idea of shunting signal to ground is different than how we normally look at a resistor divider, mind you once I read through it a couple of times it seems to make sense.

Not sure if it changes how you look at things but the ring of the jacks on the diagram are all at ground potential.

The idea of shunting signal to ground is different than how we normally look at a resistor divider

Really? I'm surprised - it seems closely related to the idea of a tone control where you shunt to ground some portion of the signal. The divider just does it with a resistor instead of a cap, so the effect is to bleed off some amount of signal across it's full bandwidth (i.e. making it less "hot" overall) rather than just a selected portion of the full bandwidth (i.e. shelving the signal at some particular frequency).

Not sure if it changes how you look at things but the ring of the jacks on the diagram are all at ground potential.

Yup, got that.

Really? I'm surprised - it seems closely related to the idea of a tone control where you shunt to ground some portion of the signal. The divider just does it with a resistor instead of a cap, so the effect is to bleed off some amount of signal across it's full bandwidth (i.e. making it less "hot" overall) rather than just a selected portion of the full bandwidth (i.e. shelving the signal at some particular frequency).



Yup, got that.

Not wrong just harder for some to visualize the shunting, too much for some fragile minds to wrap around. You are right, it is just two sides of the same coin.