LED's in the cathode..

[Shoog]

If you put a LED in the cathode of a simple triode or pentode CCS you are multiplying up the dynamic resistance of the LED by the mu. Not very impressive results will follow.

I made that mistake once. As in all things, its horses for courses and a LED is not a universally appropriate solution.

Shoog

 

[M Gregg]

If you put a LED in the cathode of a simple triode or pentode CCS you are multiplying up the dynamic resistance of the LED by the mu. Not very impressive results will follow.

I made that mistake once. As in all things, its horses for courses and a LED is not a universally appropriate solution.

Shoog

What did you use to replace the LED? Or was it just back to resistors?
IE did you find any world beating replacement..

Regards
M. Gregg

 

[Shoog]

Just a resistor. You can make quite a stiff little CCSink by using a resistor chain from +B to ground with the grid reference taken at any point between (typically 10-20V), which allows you to place a substantially larger cathode resistor in the triode tail with a substantial increase in plate resistance resulting.

This was an Allen Wright trick.

Shoog

 

[M Gregg]

OK,

I'll keep it in mind..I must admit that they seem work well with a FET MU stage..

I'm trying to think how to compare the change in sound to other mods..

I think those that say they don't want Sand...Forget LED's were not avilable in the valve era..and I am sure they would have used them if they had..
Then again Telequipment scopes used sand under tubes to alter the gain of the channels..

Thank's for the heads up!

Regards
M. Gregg

 

[mogliaa]

Be aware that replacing the cap / resistor combo with LED's (or zeners, or other constant voltage circuits) essentially converts a cathode biased circuit to fixed bias. The grid resistors must be sized accordingly to avoid grid current induced runaway.

Hi George,
Would a Gyrator not help here with the induced runaway as the anode voltage will be fixed by it?

Thanks
Ale

 

[M Gregg]

If you put a LED in the cathode of a simple triode or pentode CCS you are multiplying up the dynamic resistance of the LED by the mu. Not very impressive results will follow.

I made that mistake once. As in all things, its horses for courses and a LED is not a universally appropriate solution.

Shoog

Spot on,

Shoog, I have now experienced exactly what you have found...LOL

Yes its not so obvious...and very subtle...horses for courses..

Regards
M. Gregg

 

[DF96]

Surely the (LED x mu) component of the anode resistance will always be much smaller than the intrinsic Ra? I can imagine a problem occuring with
- high gm
- high mu
- far too low anode load
but a correctly designed circuit ought to be OK. Problems are most likely to be seen when running with a far too low supply rail voltage?

 

[Merlinb]

Surely the (LED x mu) component of the anode resistance will always be much smaller than the intrinsic Ra?

I think Shoog meant that for a CCS you want an unbypassed cathode resistor, since you want to maximise output resistance, whereas an LED doesn't have much resistance to add to the mix.

 

[DF96]

Yes you are right, I misunderstood him.

 

[dgta]

I'm not following this. You want to MAXIMIZE the output AC impedance of the stage? For what purpose? What kind of stage are we talking about?

 

[SY]

He wants to maximize the output impedance of a tube-based CCS (where high output Z is desired), not a voltage amplifier stage. It was a sudden conversational lurch that left a few people behind.

 

[dgta]

Ah, yes. Lurch indeed. But that's not what confused me. It was the "output" part. Most people think of a CCS as a 2-terminal device, so it doesn't have an "input" or "output". Even when its physical realization happens to be multi-terminal.

 

[Shoog]

The pertinent point is that every circuit needs considering on its own terms and that this is easy to miss when you are using other peoples ideas - I know I have used things inappropriately when I failed to understand the underlying concept.

Shoog

 

[DF96]

A CCS has an output, it just doesn't have an input. It outputs current, and you want the output impedance to be as high and linear as possible. People sometimes seem to forget about the 'linear' bit, but provided the impedance is high enough you can get away with this.

 

[TonyTecson]

with CCS the input is the source of voltage or B+, the output is current flow with very high impedance, think Norton.....

 

[dgta]

You guys are funny. But sure, make up your own definitions, why not.

Later on today I'm going to order some resistors with very linear output

 

[Shoog]

You guys are funny. But sure, make up your own definitions, why not.

Later on today I'm going to order some resistors with very linear output

Do you imagine that a CCS doesn't have a response curve ?

Now that really is funny

To my certain knowledge just about every component imaginable has a response curve which needs to be considered.

Shoog

 

[TheGimp]

Since current sources are composed of non-linear devices, how could their output be linear (over a large enough current range)?

Sim some current sources and look at the differences in distortion. It is quite revealing. You might also try varying the source voltage driving them to see how it effects distortion.

Or for those of you who don't believe in sims because they are only approximations, build several current sources and measure the distortion.

 

[TonyTecson]

as i understood it, this thread was about replacing the traditional resistor capacitor combo with LED's on common cathode circuits......and in this aspect LED's rocks......

why look a tube CCS's when sand is freely available?

 

[Shoog]

Because the use of LED's in other situations is inappropriate and it would be easy to go away from this thread thinking that they are the solution to everything. If nothing else this CCS aside will have given beginners pause for thought when they "imagine" that LED's can work everywhere.

Shoog