LED Cathode Biasing in 5e3

[NegotiateNothing]

Thinking over the possibilities of using an led in place of standard resistor / cap in a 5e3 build I’m working on.

Seems like a low cost, simple and drop-in method to get superior functionality (Blencowe says essentially a resistance with a perfect cap) and lacks the need for replacement / servicing of the cathode cap.

If you split the first stage you have 2x 1.5k and 22uf. I’d like to switch one to Robinette’s “lead channel” mod of 2.7k and .0068uf.

I don’t think the output stage would fare well with LEDs, at least not without a string of them.

Can someone help me along with some rough calcs?

From what I understand:

Rough estimates of color vs. V drop:

1N4148 diode = 0.7V (approx)
Infrared = 1.2V
Red = 1.8V to 2.0V
Green = 1.9V to 2.2V
Yellow = c. 2V
White = 3.2V (roughly)
Blue = 3.6V to maybe 4V

Most standard LEDs have max current capacity of 20mA.

But you can put LEDs in series to achieve higher voltage (additive).
 
What should I be shooting for with the 5e3?

[shooter]

What should I be shooting for with the 5e3?

the last one I build had 16vdc at cathode drawing ~35mA per tube

[sluckey]

No need for calcs. Just build the circuit and experiment. I would start with red. Play with yellow and green also. Blue and white are too much voltage drop for the preamp. Let your ears tell you which you prefer.

[pullshocks]

The AX84 Blues Preamp had LED biasing on V2.  The V2 triodes were run in parallel.


https://ax84.com/archive/ax84.com/static/corepreamps/Blues/AX84_Blues_Preamp_Schematic.pdf


I had one green and one red, as per the schematic.  Supposedly there was some mojo in the slightly different bias from the 2 colors.


Worked fine.

[NegotiateNothing]

No need for calcs. Just build the circuit and experiment. I would start with red. Play with yellow and green also. Blue and white are too much voltage drop for the preamp. Let your ears tell you which you prefer.

That goes for V1 and V2?

If cathode is split as in the Hoffman boards, instead of the fender lazy 820R- one led each?

Seems so cheap, easy and ideal you’d wonder why it’s not in far more common use?

Am I missing something? Anyone want to provide their experiences with LED biasing / impact on “tone”?

[NegotiateNothing]

What should I be shooting for with the 5e3?

the last one I build had 16vdc at cathode drawing ~35mA per tube

V1? V2 or V3?

3 LEDs in parallel should handle that with some wiggle room.

[NegotiateNothing]

The AX84 Blues Preamp had LED biasing on V2.  The V2 triodes were run in parallel.


https://ax84.com/archive/ax84.com/static/corepreamps/Blues/AX84_Blues_Preamp_Schematic.pdf


I had one green and one red, as per the schematic.  Supposedly there was some mojo in the slightly different bias from the 2 colors.


Worked fine.

Thanks I’ll take a close look at this and report back.

Edit: looks like they only used LED biasing in the 12au7 cathode follower. Is there a reason for that and not LED biased throughout?

Plopping in a red LED feels too easy....

I feel as though there are some additional considerations I may be missing.

For example, in Robinette’s lead channel mod he changes to .0068uf and 2.7k on the normal channel for more versatility.

How could I manifest that change in an LED swap?

How do the different LEDs effect the cut-off frequencies and gains as in Blencowes example here:

http://www.valvewizard.co.uk/OtherStuff.html

[pullshocks]

LED biasing was discussed on the AX84 forum many times over the years.


The old AX84 forum has been moved to this site.  Very little activity currently, but you can search for old discussions.  Merlin himself was a participant.


https://perpetualinnovation.us/MyBBDir/index.php

[PRR]

Plopping in a red LED feels too easy.... I feel as though there are some additional considerations I may be missing.
How do the different LEDs affect the cut-off frequencies and gains as in Blencowes example here: http://www.valvewizard.co.uk/OtherStuff.html 

Read the page you linked.

ways in which LEDs can be incorporated, usually with no impact on the tone of the amp.

> 0.0068uf and 2.7k on the normal channel for more versatility.

If you need frequency-slicing, an LED will not do that.

[PRR]

The old AX84 forum has been moved to this site.https://perpetualinnovation.us/MyBBDir/index.php

VERY spam-ridden.

Top section, a post about "Chandigarh Call Girls" has been up since May.

[shooter]

V1? V2 or V3?

sorry, a mis-read on my part.  those "spec's" are for the PA tubes

[NegotiateNothing]

ways in which LEDs can be incorporated, usually with no impact on the tone of the amp.

> 0.0068uf and 2.7k on the normal channel for more versatility.

If you need frequency-slicing, an LED will not do that.

Yes I read that, and it said “usually”- wasn’t clear to me if I could “frequency-slice”


But I guess doesn’t totally answer why if its such an ideal replacement it’s not more commonly used. I suppose “custom” and the aforementioned desire to frequency slice / adjust tone.

[shooter]

adjust tone.

seems most guitarist's hunt good tone way more than perfect load lines     

[NegotiateNothing]

adjust tone.

seems most guitarist's hunt good tone way more than perfect load lines     

There’s many ways to skin a cat.

Goal is more to eliminate electrolytics / limited lifetime parts.

[NegotiateNothing]

Could someone kindly explain where the ideal locations would be in this layout to bypass with an LED and what might not be an ideal location, and why?

(Electrically speaking but also in terms of tone, mod considerations, etc.)

http://el34world.com/Hoffman/images/Hoffman_5E3_small2.gif

[brewdude]

If it were me, I would replace V1a’s (the first/input gain stage) cathode resistor and cap with a red LED and make my first judgement based on playing it for a while.  Then I would try different colors in that same position and see if there is a clear winner... if this satisfies my curiosity I would likely stop here. 

If not I would move on to the next stages cathode and repeat. 


It’s kind of a tedious iterative process, but this is all analogous to seasoning applied to taste. 


Years ago I was playing with LED biasing, but at the time I was more interested in the VVR on the amp—which LED biasing didn’t track right as the voltage was reduced—and never really got back to give it another try.

[sluckey]

There are only 3 places in this amp where you can do this. See pic. I doubt you will find anyone on this forum who has actually done this to a 5E3. Most people want the 5E3 to be true to the original. A few will use separate R/C for cathode bias on V1a and V1B. And a few others will do a "Tweedle Dee" mod.

This is such a simple mod and very easy to do on a Hoffman turret board. If you are serious about knowing how it will sound then you are the only one that can answer that. Just build it stock and then replace the R/Cs one at a time and compare what you hear. It'll only take a couple minutes to actually do the mod. Heck, even use a SPDT switch so you can quickly switch between R/C and LED. Or if you want to be more elaborate use a SP4T rotary switch with R/C, red, yellow, and green LEDs. Should only take a few minutes to determine your preference. Please report your findings in this thread.

[NegotiateNothing]

Right. And since I’m doing the lead channel in V1B, I have two spots...

I think the switch will be the only sensible way to really tell the difference.

Swapping manually becomes too confusing.