Help with DC Servo for Current Feedback Opamp

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Jump to Latest
klewis

klewis

Member
Joined 2007
I've built a preamp around the LME49713 current feedback op amp. The results so far look very good. I added a DC servo to manage the DC offset - which varies from -2mv to 19mv depending upon the volume pot setting. I used the LME49720 op amp for the servo. I simulated the following circuit and it show the servo correcting the offset. The actual circuit changes the offset to 32mv regardless of the volume pot setting. So, is some respects, it's working in that it's making the DC a constant value, but, it's not the 0 volts expected. Advice and critique would be greatly appreciated.

Regards,

Ken

I used the LMH6723 current feedback op amp model for the 49713 and the op275 for the 49720.
 

Attachments

  • pre-amp-with-servo.gif
    pre-amp-with-servo.gif
    55.7 KB · Views: 993
A

Artie

Member
Joined 2007
I've been reading up on CFB opamps too, and there's two things I've picked up on:

1. (This is taken directly from Linear Tech's "CFB Do's and Don'ts") "DO NOT PUT A SMALL CAPACITOR FROM THE INVERTING INPUT OF A CURRENT FEEDBACK AMPLIFIER TO ANYWHERE, ESPECIALLY NOT TO THE OUTPUT." The "all caps" is their form. Note your C4.

2. I don't believe a CFB opamp should ever be used in an inverting mode, due to the difference in input impedances. (This part hasn't been stated implicitly, but no app notes show an inverting circuit.)

Hope this helps.
 
klewis

klewis

Member
Joined 2007
Artie said:
I've been reading up on CFB opamps too, and there's two things I've picked up on:

1. (This is taken directly from Linear Tech's "CFB Do's and Don'ts") "DO NOT PUT A SMALL CAPACITOR FROM THE INVERTING INPUT OF A CURRENT FEEDBACK AMPLIFIER TO ANYWHERE, ESPECIALLY NOT TO THE OUTPUT." The "all caps" is their form. Note your C4.

2. I don't believe a CFB opamp should ever be used in an inverting mode, due to the difference in input impedances. (This part hasn't been stated implicitly, but no app notes show an inverting circuit.)

Hope this helps.
Click to expand...

Artie,

I read the app notes as well, so, I didn't explain the cap shown parallel with the gain resistor if for the simulation only, it represents the capacitance of the resistor and traces.

Audioman54 who helped develop the 49713 recommended the inverting configuration - like you I had the same impression. Have a look at the thread in the solid state forum "National Op amp inflation" you'll see his comments and my measurements from the preamp.

I tried simulating the it with the servo output going to the non-inverting pin of the preamp opamp and it offset the entire sinwave by about 4v positive.

Other thoughts - thanks.

Ken
 
klewis

klewis

Member
Joined 2007
pacificblue said:
Should the DC servo for an inverting amp not also be inverting and act on the non-inverting input?
It is also advisable to use anti-parallel diodes between the servo's input pins to avoid overvoltage.
Click to expand...
I think you have a good point here, though the non-inverting input behaves quite differently on a CF opamp compared to VF opamp. I'll post another schematic.
 
klewis

klewis

Member
Joined 2007
Here's the output plot for the inverting servo configuration. It moves the offset but, just barely, certainly not back to zero. I used resistor values as recommended by national in AN-1192 figure 15. These are for voltage feedback amps.
 

Attachments

  • servo-v2-plot.gif
    servo-v2-plot.gif
    65.3 KB · Views: 692
klewis

klewis

Member
Joined 2007
Since the non-invert resistor has a much smaller value on a CF amp vs VF amp, I thought I might reduce the resistor values in the servo by an order of magnitude in the inverting servo. I've made a run with this configuration and it seems to correct the offset. I will try to build this configuration tonight and see what happens.
 

Attachments

  • servo-v2-2.gif
    servo-v2-2.gif
    16.9 KB · Views: 503
P

pacificblue

Member
Joined 2008
The more I think about it the more I come to the conclusion that a DC servo won't work with a current feedback amp. The DC servo gives a voltage feedback, not a current feedback. You will have to inject a current into the non-inverting input to achieve an offset compensation.

I. e. connect a potentiometer (trimmer) between + and -. To the wiper connect a small capacitor to ground and a resistor to the non-inverting input. With the potentiometer you set the voltage to which the capacitor will be charged. The capacitor discharges through the resistor and amplifier's input stage to ground, which leads to the compensation current. Adjust with the potentiometer.
 
klewis

klewis

Member
Joined 2007
pacificblue said:
The more I think about it the more I come to the conclusion that a DC servo won't work with a current feedback amp. The DC servo gives a voltage feedback, not a current feedback. You will have to inject a current into the non-inverting input to achieve an offset compensation.

I. e. connect a potentiometer (trimmer) between + and -. To the wiper connect a small capacitor to ground and a resistor to the non-inverting input. With the potentiometer you set the voltage to which the capacitor will be charged. The capacitor discharges through the resistor and amplifier's input stage to ground, which leads to the compensation current. Adjust with the potentiometer.
Click to expand...

Pacificblue,

Thanks for thinking about this. I got a servo to work, the values for the resistors are quite different from servos for vf amps. But the resistor value changes move the Fc way up to around 1Khz. Which then boost the lower frequencies - not good. So, I'm currently simulating a third stage in the servo to bring the voltages back up... it's sort of working. I'll build one next weekend.

Ken
 
J

jcx

Member
Joined 2003
by wrapping the CFA in the fedback loop of a good "audio" op amp you can get your "DC Servo" action for free and gain other advantages from the multiloop composite amplifier

compensation is a little more challenging if you want to use more of the CFA gain in the loop but doing so "improves" the performance of the input op amp as well

Walt Jung describes several "multiloop" composit amplifier circuits in his "Op Amp Applications" book
http://www.analog.com/static/import... Amp Applications Book (PDF)/P2 Ch6_final.pdf
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Top Bottom