Login

Recent searches

No recent searches

Search options

Only available when logged in.
fosstodon.org is one of the many independent Mastodon servers you can use to participate in the fediverse.
Fosstodon is an invite only Mastodon instance that is open to those who are interested in technology; particularly free & open source software. If you wish to join, contact us for an invite.

Administered by:

Server stats:

10K
active users

fosstodon.org: AboutStatusProfiles directoryPrivacy policy

Mastodon: AboutGet the appKeyboard shortcutsView source codev4.3.7

Public
Bornach

Testing the switching characteristics of cheapo PC817 optocouplers from Ebay.
555 timer astable is generating a 37kHz square wave test input.
Rise/fall time is way out of spec for use in . But the recommended component, 6N138, is much more expensive.

Instead have the drive the base of an NPN and by adding appropriate resistors, I get much improved switching speeds. Is this good enough for a MIDI input circuit?

On a prototyping breadboard are a 555 timer wired up as an astable oscillator and two PC817 optocouplers, one of which is connected to additional discrete components: the emitter is driving the base of an NPN BJT. Resistors have been specifically chosen to minimise the rise/fall times of the signal measured at the transistor's collector
Oscilloscope shows the 555 timer output as a yellow square wave, and the optocoupler collector-emitter voltage as a blue waveform that is anything but square. At 37kHz the output signal is clearly not usable
The oscilloscope is now hooked up the the collector of the NPN transistor whose base is driven by the 2nd optocoupler. The blue waveform is a near perfect square wave, with rise/fall times well within 2us spec for a MIDI application.
Public *
mike805

@bornach somewhere I saw a "fast optocoupler" circuit where the opto was the bottom transistor in a cascode, driving the emitter of a transistor. Since there was very little voltage change across the opto there was no Miller effect.

I know there are some optos that have a fast output amplifier built in.

As I understand it that's the problem with optos, the phototransistor is huge to collect light so it has a lot of Miller capacitance.

Public *
Bornach

@mike805
[IMSAI Guy] explains the cascode amplifier trick
youtu.be/APUhm2tsCiE

See Figure 13 High bandwidth optocoupler configuration
onsemi.com/pub/Collateral/AND8

I might need to try that for comparison

YouTube#798 Optocoupler Optoisolator Cascode ConfigurationBy IMSAI Guy
#cascode#amplifier#optocoupler
Public *
Bornach

@mike805
I tried the trick with the PC817 . IMSAI Guy only showed how the voltage of the transistor base affects the speed of the falling edge. But notice how the rising edge remains largely unaffected.

By comparison, the common mode NPN amplifier with 150 ohm resistor across base-emitter, yielded noticeably sharper rising/falling edges

GIF
Quiet public *
Bornach

Possible explanation for why you tend not to see optocoupler chip packages that combine a phototransistor/photodiode in a cascode configuration with a regular transistor in one integrated circuit, see this answer in Electronics (27 April 1978, page 154)
worldradiohistory.com/Archive-

By contrast, optoisolator ICs such as the 6N138, arrange for the phototransistor emitter or photodiode anode to drive the base of an NPN BJT

Electronics (27 April 1978, page 154) Why not a cascode optocoupler? Here's why not On this page on March 2, 1978, S. Ashok suggested that the cascode connection of phototransistors has been overlooked. Not so, replies John Carroll of Dynamics Measurement Corp. in Winchester, Mass. — it has  been tried and found wanting. "I couldn't make it work well enough to be worth bothering with, and I suspect others have run into the same roadblock," he says. The problem, according to Carroll, is that a phototransistor acts like a conventional transistor with a photodiode across the collector junction. Even though the cascode circuit holds the collector-to-emitter voltage constant, the base-to-emitter voltage must change to switch the transistor on and off. The photodiode charges the junction capacitances until the transistor turns on, and then the base current must discharge the same capacitances to turn the transistor off again. With rather poor photon collection and a very high beta transistor, response times in the circuit tend to be in the range of tens of microseconds to milliseconds. There just isn't very much current available to charge and discharge these capacitances. Carroll suspects that a cascode scheme might work better with a photodiode instead of a phototransistor — but then there would be a problem building up the extremely small output current into a logic swing with good speed.
Bornach

Finally got around to editing video of my experiments in improving response times of slow low-cost ICs such as the

youtu.be/s1l86nTIWEE
The improved speed was sufficient to pass data (31250 bits/s) through it

YouTubeCan slow low-cost optocouplers be used in MIDI applications?By bornach
#cascode#electronics#transistor
Aug 27, 2023, 11:38 PM·Public
0boosts·1favorite
ExploreLive feeds
About