Hi! Just a question on the NFB. In my first build using feedback the efective bandwith had increased Now i am working on another prototype amp an SE with EL34 which seems to be working fine on the bench .Howewer testing with feedback seems that on this one the bandwith tend to decrease whe i aply feedback. Is that possible? Be pacient i am a beginner...
Tks
Tks
Hi,
No the bandwidth can not decrease with increasing NFB, as long as the feedback is negative the bandwidth will always increase. If you have decreasing bandwidth the feedback must be positive. I will give a simple explanation:
The bandwidth is the frequency where the gain starts to drop. What we do with NFB is to decrease the gain so the drop point move to a higher frequency. Se the attached picture.
The product of Gain x Bandwidth always stays the about same for an amplifier, we use NFB to lower the gain, the gain decreases from (in the picture) 200 to 50. Since Gain x Bandwidth is constant, the bandwidth has to increase when we lower the gain by using NFB.
In the picture the gain without feedback was 200 and the bandwidth 10 k. 200 x 10k = 2000k.
We lower the gain to 50 by using NFB, then the bandwidth must increase to 40k, since GxBw is constant at 2000k (50x40=2000)
(the picture is just to illustrate what happens, and not from any real amp)
No the bandwidth can not decrease with increasing NFB, as long as the feedback is negative the bandwidth will always increase. If you have decreasing bandwidth the feedback must be positive. I will give a simple explanation:
The bandwidth is the frequency where the gain starts to drop. What we do with NFB is to decrease the gain so the drop point move to a higher frequency. Se the attached picture.
The product of Gain x Bandwidth always stays the about same for an amplifier, we use NFB to lower the gain, the gain decreases from (in the picture) 200 to 50. Since Gain x Bandwidth is constant, the bandwidth has to increase when we lower the gain by using NFB.
In the picture the gain without feedback was 200 and the bandwidth 10 k. 200 x 10k = 2000k.
We lower the gain to 50 by using NFB, then the bandwidth must increase to 40k, since GxBw is constant at 2000k (50x40=2000)
(the picture is just to illustrate what happens, and not from any real amp)
Attachments
Last edited:
Thanks. I knew it could not be right.
Any way i reverse the polarity on the primary of the out trafo and things went back ok .I mean the bandwidth now increase increasing the NFB and the gain decrease (see attached file) so it should be right.
Now i have another problem. I wanted to try removing the by pass cap in the pre stage.Unfortunately with the cap removed and NFB connected i get like a strong oscillation. Not sure what it can be. With the cap all is ok.
I do not have a schematic to post but it is an SRPP pre stage connected to a traditional output stage. The NFB is global from opt positive to cathode on the pre stage only resistor used. As posted otherwise all seems ok .Nice bandwidth ,decent square wawe etc only thingh there is a phase that is rotated 180 degrease compared with the input signal.Not sure if is ok or may be a problem..
Any way i reverse the polarity on the primary of the out trafo and things went back ok .I mean the bandwidth now increase increasing the NFB and the gain decrease (see attached file) so it should be right.
Now i have another problem. I wanted to try removing the by pass cap in the pre stage.Unfortunately with the cap removed and NFB connected i get like a strong oscillation. Not sure what it can be. With the cap all is ok.
I do not have a schematic to post but it is an SRPP pre stage connected to a traditional output stage. The NFB is global from opt positive to cathode on the pre stage only resistor used. As posted otherwise all seems ok .Nice bandwidth ,decent square wawe etc only thingh there is a phase that is rotated 180 degrease compared with the input signal.Not sure if is ok or may be a problem..
Here is a schematic (is essentially the same as my first build but modified to allow use of different power tubes)
The Nfb is very traditional and works when the cathode by pass capacitor is there but when i tried to remove it i get like a strong hiss from the speakers
Dont think this is normal must be some Thing wrong…(or not)
I only change the connection polarity on the primary of the opt to get it working right
Now i am out two weeks for work and i cannot make any test but would be interessant to listen your advice
The Nfb is very traditional and works when the cathode by pass capacitor is there but when i tried to remove it i get like a strong hiss from the speakers
Dont think this is normal must be some Thing wrong…(or not)
I only change the connection polarity on the primary of the opt to get it working right
Now i am out two weeks for work and i cannot make any test but would be interessant to listen your advice
Attachments
Apologies it is grounded!
I just edit my other build schematic and i made this wrong also the value of the cathode resistor on the power tube is in Fact 390 Homs not 360 (other editing mistake)
Yes i do have a basic set up with a sound card and a computer program
I am not very good at this but i manage to make some basic measurement like the frequency Response (attached in previous post) and also square wawe. All looks fine the ampli also sound decent. the only problem are that Using the dual trace oscilloscope i note that the output signal is rotate of 180 degree compared with input and i dont know if this is normal or not
And the other problem is that i cannot remove the pre stage by pass capacitor and ..seems strange to me
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.