I'm wondering: has anyone here ever used the HCPL-7800 isolation amplifier for audio? Those 'optocouplers' use a delta-sigma ADC and DAC, with digital optoisolation inbetween, all in one package.
They offer low nonlinearity and 100kHz bandwidth. Seems adequate for 'normal-quality' audio.
I would need them for prodividing a floating audio input in a headset audio amplifier to break ground loops. And this seems cheaper and smaller than using isolation transformers. The other option is to use an isolated DC/DC to power the entire amplifier.
I have used the search, and that came up with someone who had them on the shelf, but did not use them yet.
Any comments?
They offer low nonlinearity and 100kHz bandwidth. Seems adequate for 'normal-quality' audio.
I would need them for prodividing a floating audio input in a headset audio amplifier to break ground loops. And this seems cheaper and smaller than using isolation transformers. The other option is to use an isolated DC/DC to power the entire amplifier.
I have used the search, and that came up with someone who had them on the shelf, but did not use them yet.
Any comments?
I got some of these from Agilent and found that they were a bit tricky to implement -- if you look at the charts inthe PDF you'll find that the gain drops off quite quickly around 10kHz. you have to be scrupulous about siting the ground plane as Agilent shows in the PDF.
It's a good device for operation in a very noisy environment due to its differential input. they aren't that easy to get, you have to go through one of the OEM suppliers like Avnet (unless you can get some samples from Agilent.)
in the draft of an article I wrote a couple years ago I originally tried to use HP (now Agilent) HCNR201 analog optocoupler (these are very much like the Siemens, Vishay IL300 on which there are excellent application notes.) I found that I could get excellent linearity with the devices, but they weren't particularly easy to get, and the fact that you were compensating a photo-diode could lead to some oscillation if care weren't taken in layout etc. ergo hoc I abandonded the use of the analog opto-couplers and opted to put the isolation after the microprocessor. they are definately capable of 10 to 12 bit resolution, but may be a bit slow. the Vishay application notes tell you how to implement for bipolar signals.
btw, the analog opto-couplers are much, much, much better than the voltage dependent resistances for linearity and noise.
It's a good device for operation in a very noisy environment due to its differential input. they aren't that easy to get, you have to go through one of the OEM suppliers like Avnet (unless you can get some samples from Agilent.)
in the draft of an article I wrote a couple years ago I originally tried to use HP (now Agilent) HCNR201 analog optocoupler (these are very much like the Siemens, Vishay IL300 on which there are excellent application notes.) I found that I could get excellent linearity with the devices, but they weren't particularly easy to get, and the fact that you were compensating a photo-diode could lead to some oscillation if care weren't taken in layout etc. ergo hoc I abandonded the use of the analog opto-couplers and opted to put the isolation after the microprocessor. they are definately capable of 10 to 12 bit resolution, but may be a bit slow. the Vishay application notes tell you how to implement for bipolar signals.
btw, the analog opto-couplers are much, much, much better than the voltage dependent resistances for linearity and noise.
About how much noise are we speaking? I will be using this on a motorbike, so there is a lot of wind noise anyway. If the noise floor is at -40dBFS, I would be happy, although that is totally unacceptable for regular HiFi.
I could use IL300's (which are also cheaper), but I am concerned about the CTR differences between the various parts. Too much CTR difference would destroy the balance of the stereo signal.
I could use IL300's (which are also cheaper), but I am concerned about the CTR differences between the various parts. Too much CTR difference would destroy the balance of the stereo signal.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.