I have built a isolation amp that I often use for measurements when working with high-voltage audio circuits (ESL's and related amps).
It is based upon the hcnr200, basically a modified version of the high-speed lowcost circuit from the Agilent datasheet. Powered by two 9v blocks it makes a very handy standalone tool for this purpose.
But it is rather noisy, I doubt if it will reach 70dB dynamic range. THD at 1kHz is around 0.25%, not spectacular either. When playing music through it you don't need double blind testing to notice that it severely degrades sound. Oh well, the high speed circuit may not be the best in linearity but it will do 750 KHz (-3dB) which is more important for me in this application. Don't know what can be achieved with more refined circuitry.
I did do some experimentation with these devices in high-voltage amps in the past but abandoned the idea after problems with instability and poor linearity. Especially the capacitive coupling is troublesome. 0.5pf may not seem much but the photo diode currents are very tiny and easy to disturb if there is any large AC signal present across the coupler. You will also need separate power supplies for both sides which introduces additional complexity and capacitive coupling.
I don't know what use you have in mind but believe that for most audio applications it's easier to use transformers.
It is based upon the hcnr200, basically a modified version of the high-speed lowcost circuit from the Agilent datasheet. Powered by two 9v blocks it makes a very handy standalone tool for this purpose.
But it is rather noisy, I doubt if it will reach 70dB dynamic range. THD at 1kHz is around 0.25%, not spectacular either. When playing music through it you don't need double blind testing to notice that it severely degrades sound. Oh well, the high speed circuit may not be the best in linearity but it will do 750 KHz (-3dB) which is more important for me in this application. Don't know what can be achieved with more refined circuitry.
I did do some experimentation with these devices in high-voltage amps in the past but abandoned the idea after problems with instability and poor linearity. Especially the capacitive coupling is troublesome. 0.5pf may not seem much but the photo diode currents are very tiny and easy to disturb if there is any large AC signal present across the coupler. You will also need separate power supplies for both sides which introduces additional complexity and capacitive coupling.
I don't know what use you have in mind but believe that for most audio applications it's easier to use transformers.
we also have this
Texas Instruments ( Burr-Brown ) application note:
For the Principal Operation idea of these isolation amplifier chips
see my Attachment.
What I have seen are chips with bandwidth like 50 kHz
.. but there may be considrably faster chips now, with higher resolutions, too
Analog Signal > ADConverter > capacitive barrier > DAConverter > Analog Signal
Texas Instruments ( Burr-Brown ) application note:
For the Principal Operation idea of these isolation amplifier chips
see my Attachment.
What I have seen are chips with bandwidth like 50 kHz
.. but there may be considrably faster chips now, with higher resolutions, too
Analog Signal > ADConverter > capacitive barrier > DAConverter > Analog Signal Attachments
Here is a public article (a bit old maybe)
by Burr-Brown workers:
April 2000 - Applications of Signal Isolation
http://archives.sensorsmag.com/articles/0400/70/main.shtml
regars from Lineup, Sweden
by Burr-Brown workers:
April 2000 - Applications of Signal Isolation
http://archives.sensorsmag.com/articles/0400/70/main.shtml
regars from Lineup, Sweden
I'm currently trying HCNR200 in an application that requires passing an isolated audio signal to a mains live circuit. The noise floor is barely acceptable. Linearity seems good. Frequency response seems also good.
The big problem is CMRR, the 230V common-mode component does not suffer enough attenuation, it's clearly audible. The problem seems to be input-to-output parasitic capacitance. Could somebody with experience on linear optocouplers provide some hints for CMRR improvement?
I'm considering TIL300 since a CMRR of 130dB at 60Hz is claimed, and a balanced approach involving two optocouplers. Interestingly, HCNR200 datasheet and application notes doesn't provide any data on CMRR.
The big problem is CMRR, the 230V common-mode component does not suffer enough attenuation, it's clearly audible. The problem seems to be input-to-output parasitic capacitance. Could somebody with experience on linear optocouplers provide some hints for CMRR improvement?
I'm considering TIL300 since a CMRR of 130dB at 60Hz is claimed, and a balanced approach involving two optocouplers. Interestingly, HCNR200 datasheet and application notes doesn't provide any data on CMRR.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.