I need to amplify an unbalanced signal by 18db. Can anyone check me on the following design, please
Configuring a 99V as a non-inverting amp would mean -
(a) SigIn goes to 99V + through a DC compensating resistor, Rin.
(b) 99V Out goes to SigOut, and also to the feedback resistor Rf.
(c) The other side of the feedback resistor goes to 99V -, as well as to ground through the voltage divider resistor, Rg.
Now the gain of the opamp is Vout = Vin (1 + Rf / Rg), or
(1) Vout / Vin = 1 + Rf / Rg
We want gain to be 18db, so 18db = 10 * log (Vout / Vin), or
(2) Vout / Vin = 10^(18/10)
Putting (1) and (2) together means 10^(18/10)= 1 + Rf / Rg, so
(3) Rf = (10^(18/10) - 1) * Rg, or Rf = 62 * Rg
so if Rg is 1K, Rf is 62K.
Now Rin is (Rg*Rf) / (Rg + Rf) = 62000K/63K = 984R (=1K closest).
So Rin is 1K, Rf is 62K, Rg is 1K. Is this correct? Is there anything else I should be considering?
99V as a fixed gain amplifier
Moderator: Joe Malone
-
- Posts: 184
- Joined: Sun Mar 04, 2007 11:46 am
- Location: Sydney, Australia
99V as a fixed gain amplifier
Chris P
---------------------------------------------------------
I do lots of things. I believe eclectic skills are best.
---------------------------------------------------------
I do lots of things. I believe eclectic skills are best.
-
- Posts: 184
- Joined: Sun Mar 04, 2007 11:46 am
- Location: Sydney, Australia
Hang on
Shouldn't (2) be Vout / Vin = 10^(18/20) (becuase 18db = 20 * log(Vout/Vin) )
That means if Rg=1k, Rf would be more like 7k (6K943) and Rin would be 870R (875R)
Plugging these real life numbers back in the equations,
db = 20 * log (1 + Rf/Rg) = 20 * log (1 + 7/1) = 18.06db. Close enough for rock and roll.
Shouldn't (2) be Vout / Vin = 10^(18/20) (becuase 18db = 20 * log(Vout/Vin) )
That means if Rg=1k, Rf would be more like 7k (6K943) and Rin would be 870R (875R)
Plugging these real life numbers back in the equations,
db = 20 * log (1 + Rf/Rg) = 20 * log (1 + 7/1) = 18.06db. Close enough for rock and roll.
Chris P
---------------------------------------------------------
I do lots of things. I believe eclectic skills are best.
---------------------------------------------------------
I do lots of things. I believe eclectic skills are best.
Online
- Joe Malone
- Site Admin
- Posts: 2082
- Joined: Wed Jan 24, 2007 11:35 pm
- Location: Brisbane, Australia
- Contact:
A DC compensating resistor on the way to the +in of the 99v isn't going to remove the 99v 1volt DC offset due to the -in being a emiter input. So is of little use. Even with 1% resistors you can get more than +/-0.5dB gain difference so I would make the -in to ground resistor a 25 turn trim pot in series with a cap so the DC offset is only x 1 and not x the AC gain.
Joe
JLM Audio
Capturing Audio without Injury
JLM Audio
Capturing Audio without Injury
-
- Posts: 184
- Joined: Sun Mar 04, 2007 11:46 am
- Location: Sydney, Australia
Thanks for that, Joe - I was treating the 99V like a black box theoretical opamp when I should have simply been looking at how the 99V gain is done in the BA / BAD etc schematics. I see what you mean and stand educated.
Chris P
---------------------------------------------------------
I do lots of things. I believe eclectic skills are best.
---------------------------------------------------------
I do lots of things. I believe eclectic skills are best.
-
- Posts: 184
- Joined: Sun Mar 04, 2007 11:46 am
- Location: Sydney, Australia
Re: 99V as a fixed gain amplifier
OR .....
SIGIN to 99V +
99VOut to SIGOUT and Rfeedback=10K and Cf=47p
99V- to Rf, Cf and 100uf electro anode
100uf cathode to Rser = 1450 ohms = 1k3 and 150R in series, and then to ground.
My maths says that Gain = 1 + Rf/Rser = 1 + 10000/1450 = 7.8966
db = 20*log(gain) = 20 * 0.8974 = 17.95db (18db)
Can anyone out there please confirm?
SIGIN to 99V +
99VOut to SIGOUT and Rfeedback=10K and Cf=47p
99V- to Rf, Cf and 100uf electro anode
100uf cathode to Rser = 1450 ohms = 1k3 and 150R in series, and then to ground.
My maths says that Gain = 1 + Rf/Rser = 1 + 10000/1450 = 7.8966
db = 20*log(gain) = 20 * 0.8974 = 17.95db (18db)
Can anyone out there please confirm?
Chris P
---------------------------------------------------------
I do lots of things. I believe eclectic skills are best.
---------------------------------------------------------
I do lots of things. I believe eclectic skills are best.