S.E.X. driver bias

[Dr. Toobz]

I did a very neat and tidy re-build of my S.E.X. amp, as all the experiments I had going rendered it like a bird's nest. I also have turned back the clock to the stock amp with MQ iron upgrade - no C4S, cathode feedback, 6EM7 tubes, etc. for now. I did forget to order one particular part from Mouser, however, and this leaves me wondering.

I currently have substituted a 1k resistor for each 1270 ohm bias set resistor on the Va halves of the 6DN7s. I don't have anything closer at the moment, nor can I get any closer by paralleling resistors, either. I do have a handful of HLMP-6000 LED's, however. I'm just wondering, would it be OK to set the bias with the LED's and if I were to substitute those for the 1.27k resistors, would I have to do anything else anywhere along the line? The reason why I ask is that I'm not sure of how much voltage the resistor is dropping, so I can't compare it with the LED. If they were the same, I obviously wouldn't have to do anything, right? The 1k resistors in there now seem to make the amp louder than it should be and more prone to distortion when the volume is turned all of the way up. I can order the 1.27k parts, but want to use this as an opportunity to play around with LED bias while I'm at it. Supposedly, LED bias is cleaner sounding in the bass than resistors?

RadioShack has 1.5k resistors (just as "off as 1k, so no good) or 1.2k 1/4W, but you have to buy $13 worth (pack of 500) to get them. So if I can get away with the LED's I already have on hand, that would be great.

[Paul Joppa]

The voltage check in the manual gives 2.5 volts as the bias (A6, B6) voltage. I do not recommend you go any lower. You could use two HLMP-6000 in series to get about 3.1 volts, which will raise the plate voltage to about 100 volts. That will only slightly reduce the plate current and should sound excellent. You can experiment with other LEDs, possibly a yellow or green one with a 2.5-3.0v drop at 2mA current would work. We prefer the low-efficiency red LEDs because they are quieter and have a lower dynamic impedance than most other available devices.

Another possibility is to use a TL-431 shunt regulator chip wired as a 2.500-volt "zener". That's how the stock Paramount 300B driver works.

[JC]

I wonder, would a ~500 Ohm resistor in series with the HLMP LED get you in the ballpark?

[Dr. Toobz]

I wonder, would a ~500 Ohm resistor in series with the HLMP LED get you in the ballpark?

I've wondered that, too. If we're talking about a current draw of 2mA and the thing is dropping 1.6V, that would be 800 ohms, no? If so, I guess the LED+500R resistor would put me really close (1.3k) and well within the 5% confidence interval/tolerance for a typical resistor.

[JC]

It seems to me that you could probably go as low as 470 Ohms in series with the LED, but you would probably not want to go any lower.

Hopefully, Paul will get a chance to check back in on this .

[Dr. Toobz]

Well, the LED's in series seem to work very well. I can't tell if there is any difference in sound quality between these and the resistors - even with the 1k in one channel and the LED's on the other, I really couldn't hear an obvious difference. Since the awesome-sounding Crack amp uses LED's to bias the 12AU7 cathodes, this is a mod I can be comfortable with leaving as-is. The S.E.X. amp has gotten back the warm, "floaty" midrange I remember from when it was new. So I indeed don't seem to be a fan of CCS's.

Has anybody ever biased a tube using an inductor (like a 150H choke) instead of a resistor or diode? I would think the reactance from a choke would act sort of like negative feedback, yes?

[JC]

Just so I'm clear, did you use two LEDs in series to replace the cathode resistor?  If so, have you measured the resulting Voltage at the cathode?  The plate?

[Dr. Toobz]

Yes, two HLMP-6000's in series, with the striped ends toward ground. I didn't get a chance to measure voltages before cleaning up; maybe I'll take some measurements next time.

[Doc B.]

The voltage at the cathode would be 2x1.56V = 3.12V

[JC]

Thanks, I'm just curious, so if you get around to it sometime, it would be interesting to know.

As would any impressions you care to share after you've had a chance to listen to it for a while, of course!

[Paul Joppa]

...
Has anybody ever biased a tube using an inductor (like a 150H choke) instead of a resistor or diode? I would think the reactance from a choke would act sort of like negative feedback, yes?

Yes, a choke load is a great way to implement a cathode follower, if you need to reduce the heat wasted in the cathode resistor or current source. In a voltage amplifier the large audio-frequency impedance will reduce the gain to nothing - it's the opposite of a cathode bypass capacitor.

A choke also makes a great plate load in a voltage amplifier. It behaves like a current source in the audio band, but with less voltage drop. Not good for a SEX amp; the plate voltage with a choke is too high for the driver - unless you get even more complex and add a shunt regulator. And of course, if you don't like the current source load you probably won't like the choke load either...  :^)

I agree with the 500 ohms in series with an LED; it should work pretty well.