Please Correct my analysis of this circuit..

[iampoor1]

Hi!

I have recently been attempting to further my knowledge of circuits, and I have found a great way to do that is by looking at them, and attempting to figure them out based on my (limited) knowledge. I know I am VERY wrong on some parts here (Hence this thread).  I know some of this information might be redundant...but I would appreciate any recommendations for resources to read (Loving Dan Lancasters Active filter design, enough info, with enough math on the side..). Anyways Ill stop rambling and start posting.  8)

Here is the circuit. It is the transformer pre for Project 30 over at ESP

Okay so lets get started.

R7: Not sure what this does? Provides loading?

R8/C2: These form a zobel network, correct? Is there a simple formula for figuring this out? I dont totally understand zobel networks....

C1: I know this is the compensation cap. How is the value determined? It depends on the opamp  and the high frequency damping which differs for each family of op-amps correct? I know that most audio grade op-amps are now internally compensated, but it can become important at super high frequency/high bandwidth applications to have control over this parameter. Is there any advantages for audio?

Vr1/R9/C3 control the feedback and set the voltage gain. The article says the preamp has a gain of 40db which is a voltage gain of 100. How the heck do you get a voltage gain of 100 from this circuit? Wouldnt a 1k resistor with a 100k pot provide that? What is the capacitor for and why is the voltage divider 1k/50k? I know I am missing something obvious!

C4 blocks DC at the output. WHat happens with a larger value here? More bass? (Or is this a dumb question?)

Also, is there any advantage to using say a small resistor on the output before C4? Say 100ohms?

Thanks for taking the time to read through all of this! Again...I know I am missing a ton of obvious things! If my searches have missed very obvious articles, please link me. I have finally decided to suck up my pride and just ask....

 

[ruffrecords]

You are mostly right. The amplifier gain will only be in the region of 50 times but I expect the transformer provides 6dB or so of gain. Most mic pres are designed to present a load of around 1500 ohms to the mic. The 6K8 resistor on the secondary is reflected to the primary and if this is to look like around 1500 ohms then the turns ratio must be about 1:2 which gives 6dB of gain.

Cheers

Ian

 

[sahib]

R7 indeed loads the transformer but does not act alone. R8-C2 combination is also a load.

Simple formula for the corner frequency is the well known 1/2pi R C.

There are two corner frequencies there.

First one is at 1/2pi R7+R8 C. This is a Pole at  60.4kHz. At this frequency signal is subjected to attenuation. You'll get a decline in the plot.

The second corner frequency is at 1/2pi R8 C. This is a Zero at 24.7kHz. At this frequency the signal will straighten up. You'll get a straight line in the plot.

So the peaking that occurs on the output of the transformer at high frequency is damped.

The compensation cap C1 is a long story. Best to look up the data sheet and the manufacturers recommendation.

Ian has explained the additional voltage gain that comes from the transformer. However, the C3 blocks DC as in the case of C4. That means the opamp now has full DC feedback. It is said to increase the stability. However, it also limits the lower frequency by the same formula above.

Above info also tells you that there is no point in increasing the value of C4 unnecesarily.

You generally insert a resistor (like 100 ohms) after C4 but this is to isolate the capacitive loading such as driving a long signal cable.

 

[ruairioflaherty]

You are asking all the right questions.  If you haven't already got it treat yourself to a copy of "Small Signal Audio Design" by Douglas Self, it is the single best book I own on the subject.

 

[iampoor1]

You are mostly right. The amplifier gain will only be in the region of 50 times but I expect the transformer provides 6dB or so of gain. Most mic pres are designed to present a load of around 1500 ohms to the mic. The 6K8 resistor on the secondary is reflected to the primary and if this is to look like around 1500 ohms then the turns ratio must be about 1:2 which gives 6dB of gain.

Cheers

Ian

Hey Ian, thanks for that catch! Totally forgot it was a step up transformer! Is the reflected impedance the square of the turns ratio divided by the resistor? If the resistor was replaced with a pot with nominal values, the input impedance could be variable, correct? Or do most commercial designs implement this a different way?

Thanks!

 

[iampoor1]

R7 indeed loads the transformer but does not act alone. R8-C2 combination is also a load.

Simple formula for the corner frequency is the well known 1/2pi R C.

There are two corner frequencies there.

First one is at 1/2pi R7+R8 C. This is a Pole at  60.4kHz. At this frequency signal is subjected to attenuation. You'll get a decline in the plot.

The second corner frequency is at 1/2pi R8 C. This is a Zero at 24.7kHz. At this frequency the signal will straighten up. You'll get a straight line in the plot.

So the peaking that occurs on the output of the transformer at high frequency is damped.

The compensation cap C1 is a long story. Best to look up the data sheet and the manufacturers recommendation.

Ian has explained the additional voltage gain that comes from the transformer. However, the C3 blocks DC as in the case of C4. That means the opamp now has full DC feedback. It is said to increase the stability. However, it also limits the lower frequency by the same formula above.

Above info also tells you that there is no point in increasing the value of C4 unnecesarily.

You generally insert a resistor (like 100 ohms) after C4 but this is to isolate the capacitive loading such as driving a long signal cable.

Okay thanks! Are the cornor frequencies put into place to reduce bandwidth at inaudible frequencies to transformer headroom? If the signals are inaudible, whats the benefit? (I know this probably doeskin make much sense, just been trying to wrap my head around this the past few days!)

I actually havent seen anything about it on the manufacturers datasheet, just read that 22pf is "right".

Thanks for that!

Okay, so by providing a load resistor, it "outweighs" any capacative loading on the cable by always providing a greater load? (Or is this totally wrong too)!?

Just started reading Douglas Self's book, loving it!!

Thanks for all the help so far!

 

[sahib]

Frequency being inaudible does not mean that you should let it loose.  However, generally the application of Zobel network on a mic input transformer is with regards to the ringing. This will have an effect in the form of hf boost. It all comes down to how you respond to the subjective outcome. I have just searched to see if there has been a discussion on the forum so that we don't repeat the same things. Read through this http://www.groupdiy.com/index.php?topic=32441.40

For the compensation capacitor, yes there is a detailed mathematical work out but  if the manufacturer says 22pF is right then it is right. 

A capacitive load  can cause oscillation on the output stage when the frequency is right. By introducing a resistive element in series you are simply isolating the output stage from the rest of the load.