Continue to Site

# adding high voltage AC and DC signals

Status
Not open for further replies.

#### iman28aban

##### Newbie level 3
Hi and salute to everyone who just read this post, I am an electrical and electronic engineering student and have a question about adding high voltage AC and DC signals.
AC->150 volts Sin(2*pi*100*t)
DC->50 volt

I know that operational amplifiers(OP-AMP) can be used for purpose of adding(summing) two signals. However, the problem is the fact that most of OP-AMP have limitations on their rail to rail output(e.g. +/- 40 volts for LM143).
So is there any other way to increase OP-AMP rail to rail output which could satisfy the high voltage swing?
what do you friends suggest?

Signals usually have a certain load resistance or impedance. Yours has none.
When the load resistance or impedance is known then the output current can be calculated, so your circuit produces no output current.

If we knew the load resistance or impedance then we can select a suitable power audio amplifier IC to drive a suitable audio stepup transformer to produce the 150VAC. The output winding of the transformer can have its center-tap at 50VDC.

iman28aban

### iman28aban

Points: 2
Thanks for your answer Audioguru, the project as the name suggests, is in the field of power electronics. the load can be up to 1 ampere in the output side.
The goal of this project is approaching a asymmetrical of high voltage sin-wave with frequency of 100 Hz(each steps considered 1 volt per 16.67uS.)
voltages within one period of the signal is written below:
0 --> 1 --> 2 --> ... --> 199 --> 200 ->> 199 --> ... -> 1 --> 0 --> -1 --> -2 --> ... --> -98 --> -99 --> -100 --> -99 --> -98 --> ... --> -1 --> 0

First of all I came up with the solution of using series and parallel resonant converter[1] in order to attain a pure sin-wave from a predefined square-wave(e.g. -15 to +15 volt pick to pick), hence multiplying it(-150 to +150 volt pick to pick sin-waveAC) can be achieved via transformer with coefficient of "0.1" at final stage of the Inverter.

Everything seems perfectly great except the fact that it sounds as if impossible to add a DC voltage of 50 volts to this 150volt AC sinusoidal wave form in terms of using OP-AMP as high voltage to support this.
The reason of adding 50volt DC to the aforementioned 150vp-p is that it generates the target wave form indicated in blue.
I also decided to perform square-wave via Pulse Width Modulation and using 4 switch Inverter as well as AVR microcontroller as the core. I simulate it via ISIS and it works pretty fine.

[1] Daniel W. Hart. Introduction to Power Electronics. Resonant Converters chapter. (401-414. Pages). Prentice Hall PTR Upper Saddle River, NJ, USA ©1996

Just hook a 50V DC supply to one end of the transformer secondary (Possibly with a butch electrolytic across it if the power supply is single quadrant)?

Granted the iron would need to be designed to avoid saturation, but this sort of thing was once a very standard way of, for example, modulating a radio transmitter (They used to use many KV, but series injection transformers have a long history).

Its kind of ugly, but you could use an additional secondary would for 50V, feeding a diode and smoothing cap, phase it right and I think you can even cancel the DC in the winding.

Regards, Dan.

iman28aban

### iman28aban

Points: 2
Just hook a 50V DC supply to one end of the transformer secondary .
That is how I would do it.
The transformer will almost certainly need an air gap to prevent the 1 amp dc bias from saturating the core, but its a very old and well proven method.

iman28aban

### iman28aban

Points: 2
I accept as true that transformer would saturate while transforming DC voltages from primary side to secondary.
That's why I wanted to add this DC voltage in the secondary side rather than primary.

So what other choices do you recommend?
Isn't there really a possible way to create such an specific wave-form?

I appreciate both Dan Mills & Warpspeed

There probably is, but for the application as you have presented it, the transformer solution suggested by Dan is simple, well proven and very robust.

Status
Not open for further replies.