Hello everyone,
I'm building an low-cost analog isolator for a class-d amp, by running 6N137s in PWM mode. This approach is already used by some commercial products like Si8663 etc. My circuit (attached) for the same consists of LM324, LM311, 2 x 6N137 etc.
Although the simulation went really well, I soon ran into problems while actually trying out the circuit:
1) I'm not able to reduce the hysteresis band any lesser than 200mV. How do we obtain 10mV or thereabouts ? Am I missing something?
2) As of now, there are no logic gates used for shaping the PWM pulses before filtering (on either sides). However, wouldn't the passive pullup arrangement of the 6N137s cause pulse width error? Or do I let the feedback mechanism take care of it? The switching frequency I obtained was few hundred kHz (variable).
Eagerly waiting for your valuable suggestions and feedback.
Regards.
I'm building an low-cost analog isolator for a class-d amp, by running 6N137s in PWM mode. This approach is already used by some commercial products like Si8663 etc. My circuit (attached) for the same consists of LM324, LM311, 2 x 6N137 etc.
Although the simulation went really well, I soon ran into problems while actually trying out the circuit:
1) I'm not able to reduce the hysteresis band any lesser than 200mV. How do we obtain 10mV or thereabouts ? Am I missing something?
2) As of now, there are no logic gates used for shaping the PWM pulses before filtering (on either sides). However, wouldn't the passive pullup arrangement of the 6N137s cause pulse width error? Or do I let the feedback mechanism take care of it? The switching frequency I obtained was few hundred kHz (variable).
Eagerly waiting for your valuable suggestions and feedback.
Regards.
Attachments
What limits the current through the LED sections of the optoisolators? Your diagram shows them directly connected to the 12V supply and pulled to ground through the comparator. MACCO's comment regarding lack of isolation is also valid.
MAACO and jhofland,
Thank you both for pointing out these errors. I have corrected them (attached).
For the time being, the actual circuit is being operated from the same set of 5V/12V supplies. I plan to use two supplies once it begins working well. I have also added the missing resistor for the LEDs. The drive current is 10mA.
I am trying this type of circuit because I have heard that logic optocouplers have good matching unlike the analog types.
Thank you both for pointing out these errors. I have corrected them (attached).
For the time being, the actual circuit is being operated from the same set of 5V/12V supplies. I plan to use two supplies once it begins working well. I have also added the missing resistor for the LEDs. The drive current is 10mA.
I am trying this type of circuit because I have heard that logic optocouplers have good matching unlike the analog types.
Attachments
Last edited:
- Status
- Not open for further replies.