3 Phase Sine Wave Generator - OPAMPs too hot.

[ngen33r]

I am working on reproducing this hobby circuit from zapstudio.com. I wanted to make it surface mount and add some features that I need. I have one of these kits and it works, but the opamps are obsolete and I wanted to make some changes. I rolled a new PCB and it works but the opamps get screaming hot. I measured the outputs with a scope and I am getting some 200KHz-300KHz oscillations with very low amplitude in the output riding on the sine wave. The original design used some film caps, I used all ceramic and the original L272 opamps are not available so I used TI N5532A opamps. Any ideas how to make this run cooler like the original? I have attached both the original and my schematic for reference. If it would help I can also get some screen grabs of the scope traces.

Specifications: 3PS-02A60 Output voltage: 12V p-p (4.24V RMS), each phase output, P0, P1, P2 to common, G. Output voltage: 20.8V p-p (7.35V RMS), phase-to-phase outputs. Maximum current: 60mA RMS, each phase output, P1, P2, P3. Output frequency: 60Hz, ±0.2%. P0 = 0 deg. P2 = 120 deg. P3 = 240 deg. Harmonic distortion: less than 1%. Input power: +9 to +12VDC and -9 to -12VDC, 150mA maximum.

Here is a link to the data sheet:

http://www.zapstudio.com/Assets_Support/3PS-02A60%20Kit%20%20Manual.pdf

 

[KlausST]

Hi,

many OPAMPs don´t like to drive capacitve loads. Thus remove C11, C12, C13.

Better use these capacitors to stabilize the supply in parallel to C9, C10, since electolytics are some how slow and act more like an inductor above some frequency.
Datasheet says: "Place 0.1-μF bypass capacitors close to the power-supply pins"

Klaus

--- Updated Jul 25, 2022 ---

Maximum current: 60mA RMS, each phase output

No way.
The NE5532 have a minimum short circuit current of 10mA.
Typically it is 38mA, which more is like 25mA RMS.

with 25mA RMS you get about 0.4W of power dissipation per package.
Even 25mA RMS may rise the temperature to about 60 °C.
(didn´t use a calculator. So please verify)

Klaus

 

[ngen33r]

Hi,

many OPAMPs don´t like to drive capacitve loads. Thus remove C11, C12, C13.

Better use these capacitors to stabilize the supply in parallel to C9, C10, since electolytics are some how slow and act more like an inductor above some frequency.
Datasheet says: "Place 0.1-μF bypass capacitors close to the power-supply pins"

Klaus

--- Updated Jul 25, 2022 ---

No way.
The NE5532 have a minimum short circuit current of 10mA.
Typically it is 38mA, which more is like 25mA RMS.

with 25mA RMS you get about 0.4W of power dissipation per package.
Even 25mA RMS may rise the temperature to about 60 °C.
(didn´t use a calculator. So please verify)

Klaus

Thank you for the quick reply. I am not sure why I left out the decoupling caps. If I remove C11 C12 C13, how do I get the earth reference for a single phase? The 60mA RMS is for the L272 opamps in the original design. I will pull those caps and see how it effects the temperature and oscillations.

 

[crutschow]

If you need to keep the output caps, then you need to put a series resistor (try 100 ohms) at each op amp output, or use a higher current op amp that is designed to tolerate high capacitive loads.

The L272 is a high-power (1A output) opamp.
What made you think you could replace it with a low-power one?

 

[KlausST]

Hi,

If I remove C11 C12 C13, how do I get the earth reference for a single phase?

I don´t know what you mean.
The caps don´t bring any Earth reference. It´s the OPAMP that controls the voltage and refers it to GND.
(Because the IN+ pins are referenced to GND)

Klaus

 

[ngen33r]

If you need to keep the output caps, then you need to put a series resistor (try 100 ohms) at each op amp output, or use a higher current op amp that is designed to tolerate high capacitive loads.

The L272 is a high-power (1A output) opamp.
What made you think you could replace it with a low-power one?

High current opamps seem to be obsolete. Can you recommend an alternative that is SOIC-8. Component shortages also make sourcing alternates difficult. I do have 20-30 different opamps laying around that I can try if needed. Removing C11-C13 solved the heating and oscillation issue. This circuit is used to safely test 480V 3PH zero crossings. I am not driving any loads so I don't need a high current output. I also really do not need the earth ground as I am only looking for when Phase A crosses Phase C and using a comparator.

Also I have seen some posts requesting such a circuit so I can make the EasyEDA project public if needed.

 

[KlausST]

Hi,

if you don´t need high current, then stay with NE5532.
Again: the capacitors have nothing to do with EARTH reference.

Klaus