preamp topology


2005-07-05 10:53 pm
I'm wondering about preamp topology. Do they typically follow in the footsteps of their big brother - i.e. input stage, vas, output? There must be a lot of variation here, but I was just wondering if there is a typical method and if so, what is it.

Another question is about coupling. What makes for a better design: ac or dc coupling?



Disabled Account
2006-11-02 6:15 pm
It's all down to how much gain you need. That comes down to how large the signal is from the source, and how much gain does the power amplifier have.

If the source is a typical CD player output of 2 Volts and the power amp requires only 1 Volt to reach max output then the pramp needs no gain at all. In which case the preamp is most likely to be a volume control feeding into a unity gain buffer.

With the availability of unity-gain stable IC Op-Amps, Cathode followers and Source followers I would be surprised if many people made a complex three stage (input, voltage amp, output) preamp.

Regarding coupling: theoretically dc should be superior, however think about current loops for a while and even dc coupled outputs are subject to some form of capacitor interaction via the power supply. Also dc coupling the output often requires servos which have their own issues.

There may be merit in investigating a single voltage supply (as opposed to bi-polar) and use a correctly polarised coupling capacitor. Then again, I'm personally leaning towards transformers for output isolation and easy derivation of a balanced signal.
Just for a quick overview of preamp topologies, you might take a look at some of the schematics Nelson Pass has made available from his older commercial products. It's not an exhaustive list, by any means, but it may give you some ideas.
I must say that I am not a big fan of Douglas Self, but if you are of the opinion that heaps of negative feedback are what makes the sun rise in the morning, then your views may differ.
Preamps vary widely in complexity. It can be as simple as a pot--this being the so-called "passive" preamp. (Be aware that this assumes that you don't need any gain. You may also have impedance matching problems.) Or it can be hideously complex. For an example, see the pictures on the Mark Levinson website showing preamps filled with twenty or thirty opamps. Then consider that even a hot-rod opamp will have at least ten or twenty transistors (and it's nothing for an opamp to have fifty or a hundred active devices), then multiply that by the number of opamps you see. Now that's a lot of complexity, even if you can't see all the individual pieces.
Okay, so what to do?
First, start by deciding on a realistic amount of gain. What do you really need to drive your system, given your source(s)? Some preamps are as low as 6dB. Some are in the mid-twenties. If you only listen to CD and your CD player has enough output to drive your amp as loud as you'll ever want to go, then you can get by with a simple buffer or a passive preamp.
But start by determining how much gain you'll need. That will simplify your life greatly by ruling out certain topologies.



2005-07-05 10:53 pm
I figure gain will be -100db < A < +10dB. That'll cover my volume needs plus being able to drive the amp to full output when input is low. After checking various manufacturers integrated amps, they spec sensitivity as low around 150mv (rms, I think)! My amp design has around 32 dB gain, then another 10db at the preamp brings this up to around 42 dB. Somehow, it sounds too high though.

I will check out the Nelson Pass schematics, thanks.


2006-07-14 1:29 am
gearheadgene said:
I'm wondering about preamp topology. Do they typically follow in the footsteps of their big brother - i.e. input stage, vas, output? There must be a lot of variation here, but I was just wondering if there is a typical method and if so, what is it.

Another question is about coupling. What makes for a better design: ac or dc coupling?


Going back to your original question.........yes and no. Most hifi can be done unbalanced so a differential balanced input is only required if you want all your sources to be balanced. if your sources are unbalanced don't bother with differential input.

Sources come in a wide variety of signal levels. Phono (MM) is a few mV and has to be treated specially with RIAA equalization to sound anything near right. If you don't need phono forget it. Next are MP3 type players that I think are a few tens of mV. Then cassette machines, MD and CDPs are 2V, but that is the maximum, the typical level may be nearer 200mV.

You want to be able to drive your amp into clipping. You want to have the volume control at 50% most of the time. This probably adds up to between 10dB and 20dB gain.

It is usual to put the gain before the pot. There is no need for a power output stage, you are driving a few feet of shielded cable to the power amp, the output device, opamp?, needs to handle capacitative loads OK, the OPA2134 is a nice dual IC for this application.

Good Luck



2006-07-14 1:29 am
I forgot........

AC or DC coupling:

AC coupling requires small value input capacitors and larger value output capacitors. If the values are chosen with attention to the rolloff formula you pick a bandwidth much wider than audio needs, for example 5Hz and 50kHz. See this link for high and low pass filters which is what blocking capacitors are

AC coupling is normal for simple designs, it is easy and failsafe.

DC coupling is nice but not so easy to do. You get wider bandwidth, even though there isn't much audio below 40Hz. Unfortunately it only takes a few mV of DC offset on the volume pot for it to be scratchy when turned. You don't want there to be bangs when you plug the preamp into the poweramp, strains your drivers. You can chose single opamp ICs that have offset null pins and fit a trimmer pot to adjust the offset to zero, in theory. The single OPA134 will do that.



2006-07-14 1:29 am
on second thoughts.......

It is more usual to put the active gain stage after the pot, it then serves as a driver for the line to the power amp. Note that any hum or noise present at the stage output will be constant at all volume settings.
I am a no preamp convert... yes I do have an amp that actualy sounds better with a preamp, but every single other amp I have sounds better just passive.

Every stage you add to an amp adds is own distortion and signature to the sound... build a proper 3 stage amp and it is all the gain you need.

I would only use a preamp on realy crappy amps.. does add some detail.
Or some amps that needs realively large voltage swing on the input like some mosfet jobbies...
Food for thought, Nordic. I had thought of that, actually! And, I have to agree that less is more - probably, but I don't have the practical experience to back that statement up.

Although I'm designing a class-D amp, it doesn't really matter the type as long as the gain is large enough to be able to drive to full output.

Toe problems remain:
1. volume control
2. variations in input types.

So let's see - CD inputs are around 2Vrms, pretty hot. But, other line-input stuff may be as low as 150mV or so. That's a 10dB spread - give or take a couple of dB.

How objectionable would it be to have a 10dB input gain select switch on the machine? Just a thought - since it brings those 2 extreme ranges in line with one another.

Then, volume control - what a can of worms. I'm reluctant to use the likes of the CS3310 or TI or National equivalent programmable gain amp. However, I am puzzling over some programmable gain stages that may be fairly linear - jury is out on this one, though, I need to put more work into it.


2006-07-14 1:29 am
variable input sensitivity

if you are willing to go to the trouble why not include an active input stage for each input channel with a gain trim adjuster, that way everything lines up exactly but can be changed as you change sources.

Most preamps though the volume control range, pots typically have 40dB usable range, takes care of the differences between source levels.



2011-05-31 5:32 pm
Pre Amplifiers

I am a currently trying to design and build a preamplifier through out the summer. I saw that there are lots of information that Douglas Self provides but they are really complicated and would take me longer than the summer which i have available.

I need to know that best pre amplifier that i could build in the alotted time period that i have ( this summer i.e. now till september 2). This pre amp should not be too simple yet not at complicated. I am fairly new to preamplifier topologies since i know that there are know integrated amplifiers. I really hope that someone could be of help to me. If there are any additional information that i should make available let me know so i could reply immediately.

You can start by making a power supply suited for preamp.
+/-15 VDC can easily be made with
one transformer 2x15 VAC, say 10 VA (a rather small trafo)
one positive 15v regulator: LM7815
one negative 15v regulator: LM7915
They both have 3 pins: in, out,ground

If you have a well regulated +/-15 volt power supply
then it is easier to build a good and fairly basic preamplifier.
For example using resistors instead of current sources.

I get it, you do not want the IC way. You think of transistors.
Because for example opamp OPA2134 is a dead simple way to get a good preamp.
Here you have 2 different ways.

1. Discrete opamp with 4 small transistors.
2. One opamp OPA134

Both have equal gain around x4.
22k/6.8k + 1 gives x4.23

OPA134 has a little bit lower THD distortion.
But both have performance very low distortion.
The transistor opamp should be more fun to build.


  • preamp_1b.png
    9.3 KB · Views: 247


2011-05-31 5:32 pm
WOW thanks soo much for the reply lineup and rayfurell. I have a few questions though about the system in terms of its performance, specifications and usage:

What type of preamplifiers are they?

What are their limitations and specifications interms of input and output impedance.

Could they drive any power amplifier or it should be built for a particular application?
They are line preamps.
They should take sources, like CD-players, mp3 players etc,
and give a moderate gain x4.2
and drive any power amplifiers from the output.
I have added one 100 Ohm resistor, so the output impedance is like 100 Ohm

The input impedance is the 22 kOhm resistor.
Can be changed to 47 kOhm or whatever.

You could have one potentiometer before the input.


2011-05-31 5:32 pm
Thank you sooo much and may the Gods bless u. I have three questions ( hope im not over doing it : )

How well is the Signal to Noise ratio?
Whats the diiference (if any) between a line pre amp and the usual pre amp

Can the circuit be adjusted accordinglly ( i was thinking of having a selector switch so diffent signals to be proccessed by first :

-Using the selector switch to select the type of source/input signal that is going to be pre amplified .
-When the particular signal is selected, approriate proccessing of the signal could be done e.g. amplification, filtering etc then finally go through the pre amplifing stage befor fedding it to the power amplifier.

This came to mind because in know that different input sources( e.g a mic vs a guitar) would require greater or less (as the case may be) gain levels for pre amplification. Would the gain of 4.2 work on any source signal ?
If not could the selector switch technic could actually be implemented in the design topology?
Mic or guitar might need a special amplifier, and so phono from turntable
BEFORE the switching into the Line amp.
They are called mic.amp and guitar.amp and RIAA.phono.amp
Line amp is more to gather all different sources
and make a little gain if needed
and buffering the signal suitable for power amp.

I have made a more detailed schema here.
Uses a dual 8-pin opamp OPA2134. Two amps in one IC.

A. S1 is a 4 positions rotary switch.
B. Goes to AMP1 to be amplified x4 (+12dB is the usual for Line preamps)
C. Potentiometer is between the AMPS
D. Finally from pot goes to AMP2 which is a buffer with gain x1

This topology is the usual and best.
You can see it here, too: High Quality Audio Preamp

Treble, Bass, Balance you can find in his other preamp project.
But real hifi sources does not often need any such filtering of signal.

(For a headphone amplifier you may want a balance
because we can hear better with one ear.)


  • preamp_1d_topology.png
    5.5 KB · Views: 203