Fast but small optocouplers?

J

Joerg

Jan 1, 1970
0
Hello Folks,

So far I've never had size constraints and my staples were logic devices
such as the HCNW137 series. Now I need one that ideally doesn't require
a logic supply on the output, rise/fall times well under 1usec even when
not fed more than 1mA into the LED, as small as possible. No more than
about 0.150" wide and 0.250" long. Isolation around 2kV RMS is fine.

This would be at the size pain threshold but as usual is way too slow:

http://www.avagotech.com/docs/AV02-0774EN

I went through scores of them and all in the several usec. Any
sub-microsecond ones?

Unfortunately the bases aren't pipe out on these so not speed tricks :-(

What I want to do is ferry a fast PWM across a barrier, 100kHz range,
needs to be somewhat accurate. I could do the transformer plus NRZ thing
but that gets old and not enough space.

J

Joerg

Jan 1, 1970
0
Phil said:
Phototransistors are horrible, as I find I've said before. A photodiode
optocoupler can be a good deal faster, but you need an amplifier on the
isolated side. Even a homemade phototransistor might work--you could
use a quick transistor with a Baker clamp, for instance.

The Baker clamp is nice but all the tiny ones I found had only four

J

Joerg

Jan 1, 1970
0
John said:
You could consider my almost-famous totem-pole optocoupler thing

https://dl.dropbox.com/u/53724080/Circuits/Optos/Faster_Opto_Totem.JPG

(this one modified for speed and low power consumption)

Well, I don't really want to build my own optocoupler
Transformer plus DC restore wouldn't be bad, if you can keep the max
duty cycle a bit below 100%.

I can, and thought about that because I've used this on medical gear.
But even really little transformers are big when they come with
isolation ratings. Plus I'd need more stuff to restore a clean DC, just
a rectifier wouldn't be accurate enough. We are really space-constrained
on this one.

It's going to be tough in the area you're budgeting. What's the load?

It's essentially hi-Z inputs (<100nA bias). I have to feed a voltage
into those, between zero and about a volt, and it has to cross the
barrier. I could also use a DAC but that's another part. Cost is not so
important but real estate is.

The signal itself is steady for a millisecond at a time but the change
to another value must happen in tens of usec, that's why the PWM has to
run fast. I can goose it a bit though, by having the controller step to
full scale or zero for a while. But then the firmware guys will throw
things at me because I'll keep piling on the work for them. This whole
PWM is already a "By the way, we'd also need ..." job for them.

H

hamilton

Jan 1, 1970
0
On 9/1/2012 11:15 AM, Joerg wrote:

http://www.vishay.com/docs/83674/sfh619a.pdf

SWITCHING CHARACTERISTICS
PARAMETER TEST CONDITION SYMBOL MIN. TYP. MAX. UNIT
Rise time
VCC = 10 V, IC = 10 mA, RL = 100 Î© tr 3.5 Î¼s
VCC = 10 V, IF = 16 mA, RL = 180 Î© tr 1 Î¼s
Fall time
VCC = 10 V, IC = 10 mA, RL = 100 Î© tf 14.5 Î¼s
VCC = 10 V, IF = 16 mA, RL = 180 Î© tf 20.5 Î¼s
Turn-on time
VCC = 10 V, IC = 10 mA, RL = 100 Î© ton 4.5 Î¼s
VCC = 10 V, IF = 16 mA, RL = 180 Î© ton 1.5 Î¼s
Turn-off time
VCC = 10 V, IC = 10 mA, RL = 100 Î© toff 29 Î¼s
VCC = 10 V, IF = 16 mA, RL = 180 Î© toff 53.5 Î¼s

H

hamilton

Jan 1, 1970
0
On 9/1/2012 11:15 AM, Joerg wrote:

http://www.vishay.com/docs/83674/sfh619a.pdf

SWITCHING CHARACTERISTICS
PARAMETER TEST CONDITION SYMBOL MIN. TYP. MAX. UNIT
Rise time
VCC = 10 V, IC = 10 mA, RL = 100 Î© tr 3.5 Î¼s
VCC = 10 V, IF = 16 mA, RL = 180 Î© tr 1 Î¼s
Fall time
VCC = 10 V, IC = 10 mA, RL = 100 Î© tf 14.5 Î¼s
VCC = 10 V, IF = 16 mA, RL = 180 Î© tf 20.5 Î¼s
Turn-on time
VCC = 10 V, IC = 10 mA, RL = 100 Î© ton 4.5 Î¼s
VCC = 10 V, IF = 16 mA, RL = 180 Î© ton 1.5 Î¼s
Turn-off time
VCC = 10 V, IC = 10 mA, RL = 100 Î© toff 29 Î¼s
VCC = 10 V, IF = 16 mA, RL = 180 Î© toff 53.5 Î¼s

Never mind !!!

J

Joerg

Jan 1, 1970
0
Phil said:
I'm talking about using a photodiode optocoupler, not a phototransistor
one. With phototransistors, your're doomed.

With photodiode ones I am doomed as well, then I need too many parts on
the receiving end to turn the PWM into clean DC. No available space.

J

Joerg

Jan 1, 1970
0
hamilton said:
Never mind !!!

RL of 180ohms and IF 16mA is pretty unrealistic in real life, that
sounds like written by marketing EEs. Plus it's still too slow.

J

Jamie

Jan 1, 1970
0
Joerg said:
The Baker clamp is nice but all the tiny ones I found had only four
Even a baker clamp as simple as it maybe can also be a challenge to
implement if you're trying for a clean switch.

The last one I did use high speed S diodes with low cap and a
-0.5 supply for the emitter to get it down close to 0 volts and the
use of RC lead and lags to remove parasitic on the corners.

Jamie

J

Joerg

Jan 1, 1970
0
Jan said:
Use small transformer, send impulse only,
use oneshot with fixed pulse and lowpass to make DC again at rx end.

No need for a one-shot, a comparator will do (NRZ scheme). But
unfortunately transformers with isolation ratings are large.

J

Joerg

Jan 1, 1970
0
John said:
I had in mind using ordinary couplers here. Given no base resistors
allowed, this circuit at least switches one opto on hard against the
other one running at low current. Might make 1 us edges.

Good idea. But unfortunately the usual small ones are still well above
1usec.
Nasty problem. You'll need a higher-order filter to get that sort of
step response, and 100KHz PWM likely isn't fast enough. I glad it's
not my problem; I've got enough impossible things going on as-is.

Well, you can goose a PWM if needed. Like the kick-down switch in some
automatic transmission cars.

One way to do fast PWM is to have the two optos (or transformers)
drive a cmos flipflop, so the data is edge-sensitive, not level
sensitive.

Don't give in to the coders!

But they also need to have a life

Put parts on both sides of the board!

We already do, and it's all full. Like our drill sergeant said "The day
has 24 hours, and if that ain't enough you've still got the night!"

Demand more board area!

They have to be retrofitted into a given space, no luck there.

J

Joerg

Jan 1, 1970
0
Joerg said:
Hello Folks,

So far I've never had size constraints and my staples were logic devices
such as the HCNW137 series. Now I need one that ideally doesn't require
a logic supply on the output, rise/fall times well under 1usec even when
not fed more than 1mA into the LED, as small as possible. No more than
about 0.150" wide and 0.250" long. Isolation around 2kV RMS is fine.

This would be at the size pain threshold but as usual is way too slow:

http://www.avagotech.com/docs/AV02-0774EN

I went through scores of them and all in the several usec. Any
sub-microsecond ones?

Unfortunately the bases aren't pipe out on these so not speed tricks :-(

What I want to do is ferry a fast PWM across a barrier, 100kHz range,
needs to be somewhat accurate. I could do the transformer plus NRZ thing
but that gets old and not enough space.

Well, there is hope:

http://www.sharpsma.com/webfm_send/1118

J

Joerg

Jan 1, 1970
0
Jan said:
The sort of transformer I would use would be just 1 turn insulated wire
to ???spec in a ferrite bead.
Whats you insulation voltage spcec?
I mean in that case it is specced for the wire insulation.

The usual 2kV RMS. It's not that I can't make a small one, done it many
times. The issue is that on most such projects it has to come with "the
papers".

Meantime I've found this from Sharp, requires a voltage and no
uncommitted emitter but might do the trick here:

http://www.sharpsma.com/webfm_send/1118

J

Joerg

Jan 1, 1970
0
Jim said:
I vaguely recall someone had an LED and an optical receiver chip in a
single package... Avago maybe? I'll look thru my TAOS/Avago project
files and see if I can find a part number.

I'd appreciate a hint if they still have them. But this Sharp part
should do the job. Best of all, it's only 80c a pop and half a bazillion
in stock which is always a good sign.

The PWM filter is done as well but the smallest 1mH inductor I can get
is still 0.100" by 0.100". Big. Strange, one problems solved and then
there's always the next one. Like that li'l yellow light on the dash of
my car. But luckily it was just an evap alert caused by a slightly loose
gas cap and I was able to reset it via OBD-II. The shop would surely
have charged $50 plus. The OBD-II interface that just arrived was$12
including shipping

J

Jan 1, 1970
0
J

Joerg

Jan 1, 1970
0
John said:
I've done this sort of thing before. On a hi-mu toroid or bead, you
can use fat-insulated wire, far apart. Maybe 1:1, or maybe 2t primary,
1t secondary?

https://dl.dropbox.com/u/53724080/Circuits/Optos/XFMR_Isolator.JPG

I've done that as well when there was no other way. NRZ on the other
side, or DC clamper. But ... then you are on the hook to get it
certified and that's a ton of paperwork. BTDT, a lot.

S

Syd Rumpo

Jan 1, 1970
0
Hello Folks,

So far I've never had size constraints and my staples were logic devices
such as the HCNW137 series. Now I need one that ideally doesn't require
a logic supply on the output, rise/fall times well under 1usec even when
not fed more than 1mA into the LED, as small as possible. No more than
about 0.150" wide and 0.250" long. Isolation around 2kV RMS is fine.

This would be at the size pain threshold but as usual is way too slow:

http://www.avagotech.com/docs/AV02-0774EN

I went through scores of them and all in the several usec. Any
sub-microsecond ones?

Unfortunately the bases aren't pipe out on these so not speed tricks :-(

What I want to do is ferry a fast PWM across a barrier, 100kHz range,
needs to be somewhat accurate. I could do the transformer plus NRZ thing
but that gets old and not enough space.

http://www.nve.com/webstore/catalog/default.php?cPath=30_25

But not opto...

J

Joerg

Jan 1, 1970
0
Jan said:
Something wron g there,
rise and fall times worst case .8uS in 1.9 OHM????? typo?

Yup, typo. The test circuit in figure 1 would not be able to generate a
signal at all with 1.9ohms.

Do not confuse wuth common mode rejection in 30kV /uS

??

Replies
8
Views
979
W
Replies
3
Views
2K
W
J
Replies
2
Views
1K
J
Replies
2
Views
604
Replies
12
Views
919