I see this, but what is the performance with no feedback and only a 100 ohm grid stopper? I thought your claim was that adding local feedback to a stage extends the frequency response. For the no-feedback case the grid resistor would be ~100 ohms.
Please read my entire question from the previous post "Compared to a circuit using only a grid stopper, how can plate feedback with an added input resistor extend the frequency response?"
Cheers,
Michael
Last edited:
So the load on the driver increases because it has to swing more voltage into the distributed winding capacitance of the coupling transformer... That's creative!
Cheers,
Michael
This seems to be quite depleted topic now.
However I still want add that nobody should be concerned about worse frequency response when adding shunt feedback to tube stage, even when the grid series resistor is some 33...100 kohms, provided that the anode-grid resistor is also added and is some reasonable level, say below 1M.
It needs more voltage to drive tube's input impedance, since feedback is applied in an opposite polarity to the opposite side of the winding.
An externally hosted image should be here but it was not working when we last tested it.
one pot for screen voltage adjust and second for adjusting of "transformer % tap"
will it fly?
In my experience guitar amplifiers with feedback to reduce Zout simply don't work well with large cones. Triode-like or low Zout can work in specific cases where I don't want an "open" sound but not as a rule. High Zout is desireable with 10" and 12" cones. If you can find another way to reduce the hum will surely be beneficial to the sound. Otherwise you will have to use small cones to get some true trebles.
An externally hosted image should be here but it was not working when we last tested it.
one pot for screen voltage adjust and second for adjusting of "transformer % tap"
will it fly?
This works, but the supply voltage for the NFET must be higher than Ua if the DC-potential at the screen is near to Ua, like is the case with EL34 and 6L6GC..
Think about the voltage swing at the anode; there is the DC-component (about Ua) + AC component. Then the screen must follow above Ua level.
With sweep tubes when the Ug2 is some 150 V your circuit works as it is now.
there is not such thing as shunt or series or voltage or current feedback. All of them are incomplete descriptions. To fully describe the form pf negative feedback you need to specify both how it is derived and how it is applied. Possibly the most consistent way to describe negative feebdack is to say how the feedback voltage is derived and how it is applied. So you can have:
Series derived, series applied
Series derived, shunt applied
Shunt derived, shunt applied
Shunt derived, series applied.
These definitions are useful because the way NFB is derived affects the output impedance and how it is applied affects the input impedance. As a rule:
Series derived or applied NFB increases output or input impedance
Shunt derived or applied reduces output or input impedance
Cheers
Ian
Series derived, series applied
Series derived, shunt applied
Shunt derived, shunt applied
Shunt derived, series applied.
These definitions are useful because the way NFB is derived affects the output impedance and how it is applied affects the input impedance. As a rule:
Series derived or applied NFB increases output or input impedance
Shunt derived or applied reduces output or input impedance
Cheers
Ian
That's fine if it works for you. The problem I see is that many people are confused about negative feedback. This is only made worse by others using ambiguous terms when talking about it.
Cheers
Ian
Yes. That ought to appear as a banner headline when people first arrive on this site. The big problem is that some people are so confused about NFB that they either don't realise they are confused or (even worse?) they don't think it matters that they are confused. This leads them to think that they have 'deep knowledge' which is hidden from those who are not confused about NFB. All very gnostic!ruffrecords said:
It would be foolish to make such a mistake, but there is considerable support for the idea that local feedback is easier to implement well and has less potential to negatively impact the overall quality of the amp.
Then again there are different forms of Local Feedback to muddy the water.
Shoog
But whether you are talking about local, global, nested or any other type of negative feedback, in all cases you need to specify how the feedback is derived and how it is applied in order to both describe it and analyse it unambiguously.
There are only four cases so it is not exactly onerous.
Cheers
Ian
If anyone is interested, here is a link to an introduction to NFB I wrote some years back for another tube interest group.
http://www.ianbell.ukfsn.org/data/nfb101.pdf
Cheers
http://www.ianbell.ukfsn.org/data/nfb101.pdf
Cheers
Ian, I have just started a quick scan through your note. Good on feedback, but I think you have made a classic mistake on biasing. Bottom of page 11 you seem to equate DC resistance of a valve with AC resistance. Va/Ia is not equal to dVa/dIa=ra. The AC resistance is usually lower than the DC resistance, perhaps by a factor of 2.
Adding Rk then adds Rk to the DC resistance but (mu+1)Rk to the AC resistance. To some extent this compensates for the original error, as with sufficiently large Rk the AC resistance can then exceed the DC resistance - but they are still two different concepts.
Adding Rk then adds Rk to the DC resistance but (mu+1)Rk to the AC resistance. To some extent this compensates for the original error, as with sufficiently large Rk the AC resistance can then exceed the DC resistance - but they are still two different concepts.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.