Hi Patrick!
Hmm, how did you come to this conclusion? I'm somehow lost, because the tranconductance ration out of the datasheet ist 5/0.7=7.14.
If I go and multiply the Crss, Ciss and Coss values of the 2SK2013/2SJ313 pair with this value, I get higher capacitance data than a 2SK1530/2SJ201 combo.
What did I wrong in my calculation?
Thanks, Tino
Hi Patrick!
Hmm, how did you come to this conclusion? I'm somehow lost, because the tranconductance ration out of the datasheet ist 5/0.7=7.14.
If I go and multiply the Crss, Ciss and Coss values of the 2SK2013/2SJ313 pair with this value, I get higher capacitance data than a 2SK1530/2SJ201 combo.
What did I wrong in my calculation?
Thanks, Tino
Isn't the discussion about replacing 6 transistors with 1? Try and rerun your math with this in mind and see where it takes you.
No need for a PhD to demonstrate that, just get 5 people together that can sing. Get 4 of them to the sing in tune and the fifth person you choose with a very high pitched voice (ie 2 or 3 Octaves higher) and ask that person to sing out of key 2 to 3 octaves higher. I hope you can visualise that.
There use to be a show called "The Nanny" http://en.wikipedia.org/wiki/The_Nanny_(TV_series). I hope some of you know it. She had a high pitched voice that sounded horrible, espescially if she tried to sing.
If someone were to sing 3 Octaves higher and out of key with the rest of the group, it would make The Nanny sound like Barbra Steisand in comparison.
It wouldn't matter how well the the other 4 singers sounded it would sound bloody horrible.
Even if you increased the number of singers to 20 it would still sound horrible.
Once you got to a 100 singers in key to 1 out of key (2 to 3 Octaves higher) you would still perceive something annoying about the group but it wouldn't be so obvious what it was.
I hope that paints a clear picture.
Edit: For comparison. If you now asked that person to sing 1 octave higher in key (ie 2nd harmonic) it would sound bloody beautiful. If you you then asked that person to sing one octave higher but to sing a perfect fifth of the key note (ie third harmonic) it would still sound beautiful. Some might even prefer it.
Last edited:
Tino,
This is how I calculate :
1) I compare the capacitances of 2SK2013 to 2SK1530, and they are roughly 1:6. So I can parallel 6x 2SK2013 to get to the same capacitance as a single 2SK1530.
2) I look at the Yfs vs Id curve. As I used the 2SK1530 in Xed F5 at 2A, I use 2A as value. 2SK1530 has a Yfs of about 2S at 2A, but a single 2SK2013 has about 0.6S at 0.3A. So 6 of them would be 3.6S at 2A.
Hope I did not do my sums wrong this time.
Good to hear from you in any case,
Patrick
This is how I calculate :
1) I compare the capacitances of 2SK2013 to 2SK1530, and they are roughly 1:6. So I can parallel 6x 2SK2013 to get to the same capacitance as a single 2SK1530.
2) I look at the Yfs vs Id curve. As I used the 2SK1530 in Xed F5 at 2A, I use 2A as value. 2SK1530 has a Yfs of about 2S at 2A, but a single 2SK2013 has about 0.6S at 0.3A. So 6 of them would be 3.6S at 2A.
Hope I did not do my sums wrong this time.
Good to hear from you in any case,
Patrick
According to wikipedia Common source - Wikipedia, the free encyclopedia
For Common Source
Cm=Cgd(1-Vd/Vg) or Cm=Cgd(1+gm(ro//RL)
Does that mean what I said in post #65 was right (ie Cgd=Crss is the dominating specification)
If there is more to it than this, someone please explain.
For Common Source
Cm=Cgd(1-Vd/Vg) or Cm=Cgd(1+gm(ro//RL)
Does that mean what I said in post #65 was right (ie Cgd=Crss is the dominating specification)
If there is more to it than this, someone please explain.
For the MOSFETs in F5, Cgs see the full input voltage (Voltage across the drain resistors of the JFETs). Cgd sees the input voltage multiplied by (1 - open loop gain of the second stage).
Cgs is usually much larger than Cgd, so which one dominates will depend on gain, amongst others.
In a follower, Vgs remains essentially constant, and Vgd will more or less equal -Vin.
Capacitances only have an effect if there is voltage change.
Hope it is clear.
Patrick
Cgs is usually much larger than Cgd, so which one dominates will depend on gain, amongst others.
In a follower, Vgs remains essentially constant, and Vgd will more or less equal -Vin.
Capacitances only have an effect if there is voltage change.
Hope it is clear.
Patrick
According to Erno (JFETs Part 1) the input capacitance of a common source is Cin = Ciss - Av * Crss, with Av = stage gain (< 0 for source follower). With the gain for 6 SK2013 about twice the gain of one SK1530, you have to increase local FB to reduce gain or you end up with a higher Cin. (without having analysed the numbers in detail)
Peter
Peter
> With the gain for 6 SK2013 about twice the gain of one SK1530, you have to increase local FB to reduce gain or you end up with a higher Cin.
You are almost correct, except that the stage gain is not only determined by Yfs, but also Rsource and Rdrain. To be precise,
Stage Gain = Rdrain / (Rsource + 1/Yfs)
And the Yfs of 6 2SK2013 = 3.6S approx.
The Yfs of 1 2SK1530 = 2S approx.
Rdrain = 0.5*Rload in F5, so say 3 ohm.
Rsource = 0.22R in the circuit I published.
I'll leave you to work out whether the effective capacitance is a factor of 2 or a bit less.....
Patrick
You are almost correct, except that the stage gain is not only determined by Yfs, but also Rsource and Rdrain. To be precise,
Stage Gain = Rdrain / (Rsource + 1/Yfs)
And the Yfs of 6 2SK2013 = 3.6S approx.
The Yfs of 1 2SK1530 = 2S approx.
Rdrain = 0.5*Rload in F5, so say 3 ohm.
Rsource = 0.22R in the circuit I published.
I'll leave you to work out whether the effective capacitance is a factor of 2 or a bit less.....
Patrick
Last edited:
Hi Juma,
I have found your project very interesting.
Speaking about PS, what's the reason why you used capacitance multipliers to lower the V to 19+19v instead of the standard (for FirstWatt) 24+24 from a 18+18 trafo and a CRC?
I don't think that the Toshibas would suffer from some more power to dissipate. Is it a matter of "sweet spot" or what else?
In case, what would be the correct (i.e. non distructive...) value for the Rsource with a 24+24 Vsupply?
Thanks
Guido
I have found your project very interesting.
Speaking about PS, what's the reason why you used capacitance multipliers to lower the V to 19+19v instead of the standard (for FirstWatt) 24+24 from a 18+18 trafo and a CRC?
I don't think that the Toshibas would suffer from some more power to dissipate. Is it a matter of "sweet spot" or what else?
In case, what would be the correct (i.e. non distructive...) value for the Rsource with a 24+24 Vsupply?
Thanks
Guido
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.