optocoupling power stage - hiss with 6N137

[ionutgaga]

I'm wondering why 6n137 is rated at 100v/us in some datasheets, or 10000v/us in other places. Certainly, after i remove other variables (like pcb, power fet, inductor, etc), a small hiss is presented at output, even if i drive only power stage with clean square wave 250khz (offcourse, with deadtime).
It is a posible cause for hiss this 6n137 optocoupler?

Description on final stage: 6n137+tc429+power fet+floating 15v supply

 

[Eva]

The hiss may be just due to the sum of time errors of all components (jitter). Power supply noise is a direct contributor too. When doing open loop tests, note that output filter resonance may boost some frequencies unles it happens well above 20Khz.

6N137 performance depends on manufacturer. I think that Avago (formerly Agilent (formerly Hewlett packard)) supplies the one rated at 10KV/us. You may consider trying other fast shielded optocouplers.

Anyway, if erratic switching due to false optocoupler triggering is actually taking place, the oscilloscope should show it. False triggering only happens during transients and may create a self oscillating system (this has happened to me).

 

[ionutgaga]

Thanks Eva for your quick answer. This is my sch:

 

[ionutgaga]

And this is my layout. The yellow traces is from top ground plane (unfigured for clarity reason). May i ask to you to analyse this and give your valuable advice?

 

[fredos]

Synchronize your floating power supply to amplifier clock....This should fix your probleme!

 

[ionutgaga]

Thanks Fredos!
But i'm already checked that... even if i synch, even if i adjust the Fosc of smps, same white noise... (ok, small, but perceptible)

 

[ionutgaga]

This is top layer:

 

[ionutgaga]

And bottom:

 

[lumanauw]

Could you try change C20 to 4n7 and R11 to 1k8?

 

[ionutgaga]

In fact, i tryiet with semireglable in place of R11, modifying Fosc, but give more noise (other type of noise, more like radio interference or something like this). Indeed, i tryiet before to sync SG with 62.5kHZ, but result if the same with RC with values showed in the sch.

 

[lumanauw]

This may be troublesome, but you can try to replace 6N137 with transformer (PulseEng gate driver or something else), to know for sure if it is the optocoupler that is making the noise.

 

[ionutgaga]

I'm already ordered HCPL2611 and HCPL0611 (smd version of 2611) used in PowerSoft and Crown comercial PA. But i have to wait a couple of days before i will try to change my 6N137 optocouplers. In the meantime i will try to figure out what other else is the source of the white noise...

 

[lumanauw]

If you got a soundcard and FFT program, you can iterate the noise problem. If it is audible, then it should show up in fft below 20khz. Try to change something, then look at the fft whether it becomes louder or fainter, or if it is shifting frequency. Then you can try to approximate what is causing the problem.

 

[ionutgaga]

Thanks lumanauw for your valuable advice!
I will try this maybe tommorow, when i repaired my final stage (i'm abusing a little: +/-130v at 2 ohm, only a few minute, then...)

 

[fredos]

What happen if you replace your 2 floating supply with 9V battery? Be sur to completly diconect floating PSU to avoid coupling inside transformer...If it solve your probleme, try using 2 small separate transformer...Even if you adjust flaoting power supply to same frequency of your amplifier, this do not syncronize them!

 

[ionutgaga]

Some photos of my prototype (incomplete, only power stage without floating smps...)

 

[ionutgaga]

and top:

 

[ionutgaga]

Some measurement. +/-90v, 4 ohm, 10mhz oscilloscope

TC429 out, 2v/div, 0.5uS

 

[ionutgaga]

-vcc to gnd
1v/div, 1us

 

[ionutgaga]

+vcc to gnd
1v/div, 1us