LED phase splitter, in need of knowledge

[mcshaner2k]

Hello everyone, I recently finished building a Gibson BR6F circuit from an old piece of test equipment that just happened to have the same tube configuration.

6SJ7 in V1, and a 6SN7 PI going into a pair of 6k6's in PP. Just a volume pot, pretty simple and it sounds quite nice. I modified V1 by adding a 1k resistor and 1uf cathode bypass cap. I also played around with the bias voltage and was able to get some more gain from the amp. It's a relatively clean amp unless dimed out.

I stumbled across this video on YouTube of someone who modified the circuit with two LED's as a phase splitter. Apparantly, he was able to yield more gain from this mod.

http://youtu.be/fvk6_kaD2VQ


I was intrigued by the video and was hoping someone here could explain to me how this works. Maybe point me in the right direction on where to implement this mod. My knowledge on this subject is limited. . Thanks!

Ive attached the schematic I used as well.

[xtian]

Can you tell me where in the video he talks about the LEDs?

[Malcolm Irving]

At about 1:28 in the video, he mentions using two LEDs to bias the phase splitter. I expect that means he replaced the 3.3k cathode resistor on the 6SN7 phase splitter, with two LEDs in series. I don't think that would give any more gain necessarily, as that cathode resistor caries the current of both triodes of the phase splitter, which sum to a constant. So you already have a constant bias voltage with the resistor.

[mcshaner2k]

After looking at the video closer the LED's seem to be connected to the cathode of the PI like Malcolm suggested. Looks like a 1.6k resistor from cathode to 2LED's in series, then to ground I would suppose.

So I wonder what would the orientation of the LED's be +,-,etc?

I still don't understand what the point of this is.. I might just try it anyway and see what happens:)

[sluckey]

Cathode to ground. Provides fixed bias voltage drop across LEDs.

[rp]

Acts like a fully bypassed resistor, there's no more perfect a cap then no cap, more of a hifi thing, plus no e-cap to slowly degrade. Report back what you hear, I've always been curious if there's any sonic value in a guitar amp. You don't see it used much, maybe there's a reason?

[SilverFox]

If you search Meas Boogie schematics you'll find they use current regulating diodes on some of the preamp tubes. I've seen it in the past but couldn't find a good sch. for this post. Another one I've seen is the use of a FET on the cathode.

silverfox.

[mcshaner2k]

Interesting.. I've done quite a bit of research on this and like rp says it seems to be a hifi thing. It pops up a lot in the hifi forums when googling. I read somewhere that Supro used them as well; however, I was unsuccessful in finding schematics.

I'm thinking I'll wire it to a switch to see if there's an audible difference.

[rp]

Interesting.. I've done quite a bit of research on this and like rp says it seems to be a hifi thing.

could just be there's not much need for fully bypassed resistors in MI amps, more of a tone shaping need with caps, the the typical 25/25 (or marshall 320!) might as well be wide open - good place to try a led.

[gingertube]

It is a paraphase splitter. That means it will work best with the cathodes fully bypassed.
This is approximated in the original circuit as the 2 equal'ish triode signal currents are out of phase and so no net AC signal at the cathode.

From voltages on the schematic I calced the cathode DC voltage at +2V. This is just about spot on the forward voltage of a Green LED. So a simple Green LED in each cathode to 0V would work fine (anode of LED to the tube cathode.
As stated above LEDs have a very small dynamic resistance (AC impedance), typically in the order of 10 ohms, so they act as if they had a big cap across them.

I think this circuit was probably rev B where Rev A had a 6N7 phase splitter. The 6N7 is a dual triode but with a single common cathode connection.

Cheers,
Ian

[mcshaner2k]

Wow, great info! I appreciate the feedback.

I'm in the middle of another project right now, as soon as I finish I'll be giving this a try.

Will report back with my findings.

[stretch2011]

Where the led's lack capacitance wouldn't you loose the frequency filtering?

[Theashe]

Where the led's lack capacitance wouldn't you loose the frequency filtering?

Here's a post I found on AX84 about this. Not saying whether or not he's right, but it's what I found:

http://ax84.com/bbs/dm.php?id=238742

[mcshaner2k]

Sorry I took so long to respond.

I wired LED's and typical cap/resistor on a toggle to the cathodes of the PI. I was plesantly surprised at the tonal change from the LED's. The LED's definitely added gain and some mids, maybe a slightly fuller sound.

I took it a step further and and a toggle and LED to the cathode of v1 (6sj7). Same thing, more gain, fuller mids.

I wasn't expecting much from this, but I have to admit I'm pretty happy with it so far. I used 5mm green LED's, haven't tried other colors yet. It's a relatively dirty amp so I'm curious to see how this sounds on some other future builds.

Maybe try some different color LED's and Diodes on a rotary switch??

Anyways, just wanted to report back with my findings.

[rp]

Thanks for the report. Given you heard a positive difference I think I'll grab a few LEDs. There's still an old school E-parts shop in town, no HV stuff but they do have plenty of LEDs.