LED as bias resistor for LTP

[anson]

I tried using the idea that I've seen circulating on the forum to use LED's instead of resistors for biasing preamp 12ax7 stages. I tried using an LED on the first stage of my amps and it sounded pretty good, a little clearer than before, but a little too much bass so I returned it back to it's original 1.8k resistor and 22u cap.

I used an LED in place of the 470 ohm resistor in the LTP of my amp, and the sound improvement was amazing. Very clear and ringing, and less congested sounding, to my ears and brain. I jammed on my amp all night because it sounded so sweet. But It seems that the global NFB is a little less effective than it was before. I'm not entirely sure how a LTP works, or if the LED I inserted made the NFB ineffective for one of the halves of the phase splitter. Any Ideas?

Otherwise, I'm led to more experimentation with using LED's on 12ax7 cathodes.


Anson

[daz]

I can't tell you why, but i would say try using a higher output tap fo NFB. I tried all 3 and found the highest one, the 16 ohm tap gave me more NFB gain. Using the 4 ohm i got the kind of lack of NFB that you cited. So if it's not connected to the highest OT tap try it and you may get back what you lost. Also you could try a smaller resistor.

I need to try this too, and in fact i will tomorrow.

[mooreamps]

but a little too much bass so I returned it back to it's original 1.8k resistor and 22u cap.


Anson

You can abate the extra bass problem by dropping the value of the plate coupling cap to 4.7 nF ; as a start. There is also the values of the components on the input circuit, along with components that compromise the noise gate.

-g

[mooreamps]

I used an LED in place of the 470 ohm resistor in the LTP of my amp, and the sound improvement was amazing. Very clear and ringing, and less congested sounding, to my ears and brain.

Anson

Yes, it does, doesn't it. I'm assuming you are using triodes for your LTP. In my next prototype, we will be looking at pulling out the same 470 ohm bias resistor ; replacing with fixed bias using a pair of pentodes in the LTP.

-g

[Merlinb]

I'm not entirely sure how a LTP works, or if the LED I inserted made the NFB ineffective for one of the halves of the phase splitter. Any Ideas?

By putting in the LED, you effectively reduce the tail resistance a bit, which would alter the balance of the PI. However, it's such a small change that I wouldn't expect it to make a great deal of difference. Also, the LED will increase the input impedance to the PI, but again, only a bit!
What I suspect you are hearing is the change in bias on the LTP, since the drop across ther LED is unlikely to be preceisely the same as the resistor gave you! That will have a much greater impact on the tonality of the stage.

[daz]

I just tried it but there was only a very subtle difference that i didn't like as much. The feel got a bit harder and the top was a bit harsher. But again, i'm talking very subtle, but thats what i noticed.

[anson]

True, the LED did change the bias voltage of my LTP, from 1.2 v(with a 470 ohm resistorr) to 1.6 v with the LED. But I've tried higher voltages on the LTP in the past and the effect wasn't the same.

Merlin. are you sure about there being in INCREASE input impedance? I Just had a look at the long tail par article at Aikenamps.com, and from what I gather there, the increase in input impedance that's normally seen at the input of a LTP is due to: 1. the signal at the junction of the 1 meg input resistor and the 470 ohm cathode resistor which is a result of the unbypassed cathode resistance, and 2. the signal from the global NFB loop. So, my first thought is that the LED would not allow a signal to pass through it the way that the 470 ohm resiistor does, and so that would actually make for an increase in current accross the grid resistor and therefore a decrease in input impedance. What do you think?

I had a .0022 coupling cap going into the LTP, which was doing a good job of passing the bass frequencies before hand, but after I put the LED in the LTP, I noticed that my resonance control was virtually useless, but the presence was pretty much ok. My line of thought now is that perhaps the LED caused a reduction in input impedance which resulted in the removal of alot of bass frequencies, consequently giving the global NFB circuit very little bass to boost.

[Merlinb]

Merlin. are you sure about there being in INCREASE input impedance?

Yep. An LED doesn't attenuate the signal reaching the bottom of the grid leak like a resistor does, allowing the input impedance to reach its ideal maximum value of 2xRg. Ok, it's a pretty tiny increase, which is why I don't think it makes any audible difference.

My line of thought now is that perhaps the LED caused a reduction in input impedance which resulted in the removal of alot of bass frequencies, consequently giving the global NFB circuit very little bass to boost.

I would think that highly unlikely, given the algebra. I suspect there is something else at work here, or your ears are playing tricks on you. Sure you didn't alter anything else that might affect the resonance circuit? (I don't know what schem you're using).