06-01-2025, 06:57 PM
(06-01-2025, 06:10 PM)Mark Hennessy Wrote: To get acceptable linearity for the audio side, you need a second opto-isolator incorporated into the negative feedback loop of the op-amp driving the LED.
That's what I imagined. Though the mind boggles at decent video, but I'm sure possible. Some opto-couplers are only just fast enough for RS422/485 and MIDI.
I have a live chassis HMV radiogram that uses a big transformer between the pickup cartridge and amp. The turnable and arm is amazingly actually earthed. Polarised 3 core mains cord.
I used 2 kV isolation quad pulse transformers (16 pin DIL footprint) for isolated industrial control PCBs on a mother chassis. CPU clock was also AC power. The other three were data in, data out and data clock. Each board then had a single 16bit DAC or ADC with instrumentation amp. About 1983. Thus galvanic isolation for each precision analogue I/O.
