I want to make an isolator so that I can use a logic analyzer with 24VDC and 120VAC signals. For the 120VAC I don't need a very fast optocoupler, and in fact it will be a pretty slow signal anyway since I will use a cap on the opto output to maintain a constant "digital high" output as the AC wave crosses zero. But when monitoring 24VDC I want to be able to switch the cap out of the circuit and allow the isolator be as fast as the analyzer. I would like to do this with a minimum of components, so I would prefer to use the same opto for AC & DC. All the AC input optocouplers I can find are pretty slow. The fastest one I can find (Avago ACPL-244) a combined tON+tOFF of 6μS to 42μS, which if my math is right, means a probable max frequency of 42-167kHz. What's the point of a several-hundred-MHz test device if it's limited by a 42kHz isolator? I'd like to know if there's such a thing as an AC version of the high speed DC-only optos, whose tON and tOFF are measured in nS.

 

What kind of logic analyzer? Could you isolate the digital lines instead? Logic optos are a bit faster then linear ones if I remember correclty.

 

What kind of logic analyzer?

I'm not sure yet. I'm trying to choose. But let's say it's a bitscope325

Could you isolate the digital lines instead? Logic optos are a bit faster then linear ones if I remember correclty.

I'm not sure what you mean. The way read that, it sounds like you're suggesting I do what I'm already talking about doing, and I need a high speed AC opto to do it.

 

Look at 6n137, that is what I wanted to use with my SMPS, it has propagation delay of about 50ns. You would have to rectify the input to get it to turn on on both polarities of the AC.
Either that or I don´t have a clue what are you actually asking about. Maybe a schematic?

 

I'm thinking, "What do they drive fiberoptic cables with?" Must be something pretty fast that converts electricity into light.

I'll go poking around...

How about 60 ns with a Schmidt trigger output?

or some of these?

http://www.avagotech.com/pages/parametricsearch/?src=&siteCriteria=HCPL723

available at mouser.

The schmidt trigger kinda wrecks the AC function. Well, others on that page might fit the bill...
or have you already been there?

 

Look at 6n137, that is what I wanted to use with my SMPS, it has propagation delay of about 50ns. You would have to rectify the input to get it to turn on on both polarities of the AC.

Now there's the obvious solution that I didn't think of. This 200V Schottky bridge ought to do the trick, don't you think?

Either that or I don´t have a clue what are you actually asking about. Maybe a schematic?

For clarity, here's my schematic. If I can't find a super fast AC input opto, I'll just abandon this and go with the rectifier.

no capacitors drawn; use your imagination