Searching \ for '[EE] I need a circuit for a bidirectional optoisol' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: www.piclist.com/techref/index.htm?key=need+circuit+bidirectional
Search entire site for: 'I need a circuit for a bidirectional optoisol'.

Exact match. Not showing close matches.
PICList Thread
'[EE] I need a circuit for a bidirectional optoisol'
2006\09\25@103456 by Bob Axtell

face picon face
You MIGHT consider using RS485/RS422 interface devices and transfer
differentially.
These chips can handle a GND offset of as much as 7V. Its about as fast
as possible
for a noisy interface. About $0.50 USD per channel.

--Bob
>  
>> {Original Message removed}

2006\09\26@162915 by Roy

flavicon
face
part 1 883 bytes content-type:text/plain; (decoded 7bit)

To through a bit more light on my goal :-)

I'm am thinking of making the circuit referred to in this application
note

http://www.freescale.com/files/microcontrollers/doc/app_note/AN2342.pdf


The goal is to opto isolate the JTAG programmer section of my AVRdragon
programmer, will use to program ATMEGA ships with it.

One aspect I was thinking of changing is to do away with the transistors
in the bidirectional section of the app note and use inverters as shown
in the unidirectional Fig 4.

The reason for considering this is that the other JTAG lines will end up
with spare inverter gates.

So I have played around with the Linear Technology spice program - see
attached jpeg.


_______________________________________

Feel the power of the dark side!  
Atmel AVR

Roy Hopkins
Tauranga
New Zealand
_______________________________________



part 2 26183 bytes content-type:image/jpeg; (decode)


part 3 35 bytes content-type:text/plain; charset="us-ascii"
(decoded 7bit)

2006\09\26@235625 by Vasile Surducan

face picon face
On 9/26/06, Roy <spam_OUTroy.hTakeThisOuTspamihug.co.nz> wrote:
> To through a bit more light on my goal :-)

I think the biggest light is to forget it and program the stuff from a
laptop running only from battery as long the jtag is connected.

Vasile

2006\09\27@011233 by Mark Rages

face picon face
On 9/26/06, Roy <.....roy.hKILLspamspam@spam@ihug.co.nz> wrote:
{Quote hidden}

One problem is the circuit you show is not isolated!  :)  I'm assuming
this is to make spice happy.

Two notes:

1)  Check the opto speeds.  You can buy fast ones now, but the normal
ones are slow.

2)  The circuit places a 330 ohm load on the target or the programmer
each time the level changes. That is a pretty heavy load.
Double-check the datasheets to see if its OK.  If it's not, increase
R4 and R5.  Don't increase them too much, because they're effectively
in series with the signal and could slow it down.

Regards,
Mark
markrages@gmail
--
You think that it is a secret, but it never has been one.
 - fortune cookie

2006\09\27@021246 by Roy

flavicon
face
part 1 272 bytes content-type:text/plain; (decoded 7bit)

Ok have changed my circuit to the attached version.

Let me know why this would not work :-))

_______________________________________

Roy Hopkins
22 Grenada Street
Tauranga
New Zealand
3116
_______________________________________




part 2 22858 bytes content-type:image/jpeg; (decode)


part 3 35 bytes content-type:text/plain; charset="us-ascii"
(decoded 7bit)

2006\09\27@022832 by Mark Rages

face picon face
On 9/27/06, Roy <roy.hspamKILLspamihug.co.nz> wrote:
> Ok have changed my circuit to the attached version.
>
> Let me know why this would not work :-))
>

OK.  Say Target is pulling low.  U2 is "on".  The phototransistor in
U2 turns on the diode in U1, which turns on its phototransistor,
pulling Target low.  This is fine, right? Target was low anyway.

But when Target tries to change and go high, it will be pulling
directly against the transistor in U1.  Ouch!

Regards,
Mark
markrages@gmail
--
You think that it is a secret, but it never has been one.
 - fortune cookie

2006\09\27@043130 by Alan B. Pearce

face picon face
>So I have played around with the Linear Technology
>spice program - see attached jpeg.

You will find that the 8N137 opto is a lot faster than the 4N25, but an even
better one is the Agilent (now Avago Technologies) HCPL-2300 which has an
amplifier for the PIN diode included, and has very low LED current
requirement, and is claimed to be 8M baud. Also has UL, CSA and VDE
approvals for working with mains voltages.

2006\09\27@044718 by Alan B. Pearce

face picon face
>You will find that the 8N137 opto

whoops, that is (off course) supposed to be 6N137 ...

2006\09\27@050948 by Roy

flavicon
face
The intention was to use the 6N137 - should be fast enough for this
application.

_______________________________________

Feel the power of the dark side!  
Atmel AVR

Roy Hopkins
Tauranga
New Zealand
_______________________________________

> -----Original Message-----
> From: .....piclist-bouncesKILLspamspam.....mit.edu [EraseMEpiclist-bouncesspam_OUTspamTakeThisOuTmit.edu] On
Behalf
> Of Alan B. Pearce
> Sent: Wednesday, 27 September 2006 8:47 p.m.
> To: Microcontroller discussion list - Public.
> Subject: Re: [EE] I need a circuit for a bidirectional
> optoisolatedTTLinterface
>
> >You will find that the 8N137 opto
>
> whoops, that is (off course) supposed to be 6N137 ...
> -

2006\09\27@075438 by Gerhard Fiedler

picon face
Mark Rages wrote:

> On 9/27/06, Roy <roy.hspamspam_OUTihug.co.nz> wrote:
>> Let me know why this would not work :-))

> But when Target tries to change and go high, it will be pulling
> directly against the transistor in U1.  Ouch!

You'll probably find that making a bidirectional isolated interface is not
trivial :)  

You need a mechanism to disable the active drive of something like U1 when
its input (in this case the diode in U1) is not actively driven by the
associated signal (AVRDragon in this case).

Gerhard

2006\09\27@081020 by Alan B. Pearce

face picon face
>You'll probably find that making a bidirectional isolated
>interface is not trivial :)

Amen to that. Look at the Philips application note AN460 about isolated I2C.
They make a special chip (P82B96) to do it, and this app note tells you why.

http://www.standardics.nxp.com/support/documents/i2c/pdf/an460.pdf


2006\09\27@090856 by Vasile Surducan

face picon face
On 9/27/06, Alan B. Pearce <@spam@A.B.PearceKILLspamspamrl.ac.uk> wrote:
> >You'll probably find that making a bidirectional isolated
> >interface is not trivial :)
>
> Amen to that.

Explain me at least one good reason why you need an optoisolated
programmer interface.

Remember you'll never need to program a microcontroller directly
connected to power supply phase -or just to the earth- (for this job
you could disconnect the main connection while programming and pay a
little attention to the target design and supply )

The AN2432 from Fresscale it's a little silly one. What is more
expensive: building a bidirectional optoisolated debugger or to use an
1:1 isolator transformator for the target board  while
debugging/programming ?

Vasile

2006\09\27@125334 by Xiaofan Chen

face picon face
On 9/27/06, Vasile Surducan <KILLspampiclist9KILLspamspamgmail.com> wrote:

> Explain me at least one good reason why you need an optoisolated
> programmer interface.
>
> Remember you'll never need to program a microcontroller directly
> connected to power supply phase -or just to the earth- (for this job
> you could disconnect the main connection while programming and pay a
> little attention to the target design and supply )

You might want to do ICSP with Wisp628. You might want to debug it.
Sometimes the ciruits have various isolated sections, may be
high voltage or maybe not, but need to be isolated. That is why people
are doing isolated Wisp628A and ICD2.

Xiaofan

2006\09\27@130619 by peter green

flavicon
face


> -----Original Message-----
> From: RemoveMEpiclist-bouncesTakeThisOuTspammit.edu [spamBeGonepiclist-bouncesspamBeGonespammit.edu]On Behalf
> Of Vasile Surducan
> Sent: 27 September 2006 04:56
> To: Microcontroller discussion list - Public.
> Subject: Re: [EE] I need a circuit for a bidirectional
> optoisolatedTTLinterface
>
>
> On 9/26/06, Roy <TakeThisOuTroy.hEraseMEspamspam_OUTihug.co.nz> wrote:
> > To through a bit more light on my goal :-)
>
> I think the biggest light is to forget it and program the stuff from a
> laptop running only from battery as long the jtag is connected.
personally i wouldn't wan't to use a laptop whoose ground bus was tied to a
nasty voltage (which is i guess the reason for wanting the isolation).

you don't seem to have given any details on why your doing this but one
option might be to use a floating output isolation transformer to power your
device from during development work (thus allowing you to tie the logic
ground of your target to mains ground without breaking anything).

2006\09\27@133622 by Philip Pemberton

face
flavicon
face
peter green wrote:
> personally i wouldn't wan't to use a laptop whoose ground bus was tied to a
> nasty voltage (which is i guess the reason for wanting the isolation).

Me neither, but a laptop with a floating ground makes it a hell of a lot
easier to do ICD2+oscilloscope debugging without creating ground loops!

> you don't seem to have given any details on why your doing this but one
> option might be to use a floating output isolation transformer to power your
> device from during development work (thus allowing you to tie the logic
> ground of your target to mains ground without breaking anything).

Personally, I'd design the device to work of low-voltage AC or DC during
testing/debugging, then ramp up to full AC without the debugging kit. But
that's just me.

--
Phil.                         |  (\_/)  This is Bunny. Copy and paste Bunny
RemoveMEpiclistspamTakeThisOuTphilpem.me.uk         | (='.'=) into your signature to help him gain
http://www.philpem.me.uk/     | (")_(") world domination.

2006\09\27@143924 by Roy

flavicon
face
part 1 657 bytes content-type:text/plain; (decoded 7bit)

I can see what you are saying and have now added a couple of resistors
as per attached jpeg. :-)

May also increase R3 & R6 to make the transistors a weaker signal.

Also my understanding is that the AVR will control it output so that it
is either driving the circuit or in a tri state receive mode thus
allowing the target to then drive the circuit.

I presume the target would then switch back to a receive mode to allow
the AVR to take control again.


_______________________________________

Feel the power of the dark side!  
Atmel AVR

Roy Hopkins
Tauranga
New Zealand
_______________________________________



part 2 20992 bytes content-type:image/jpeg; (decode)


part 3 35 bytes content-type:text/plain; charset="us-ascii"
(decoded 7bit)

2006\09\27@145715 by Roy

flavicon
face
I like the idea of using low voltage high impedance AC supply during
debugging, may adopt that idea.

Would also still look at using opto isolation for extra safety.

PS how do I make outlook us bottom posting?

_______________________________________

Feel the power of the dark side!  
Atmel AVR

Roy Hopkins
Tauranga
New Zealand
_______________________________________


> Personally, I'd design the device to work of low-voltage AC or DC
during
> testing/debugging, then ramp up to full AC without the debugging kit.
But
> that's just me.
>


2006\09\27@150856 by Roy

flavicon
face
I have already said I will probably adopt the idea of using low voltage
AC source for debugging - this alone may be enough.

On reason is from a past experience were I accidentally put a TRAIC in
the wrong way around and the 5v supply that was direct of the phase
suddenly presented 230v AC to my PIC when the TRIAC burnt out (I was
using a PIC at the time), this failed then my programmer died and
finally my RS232 interface on my computer died also.

I was using a 1:1 isolating transformer in this case but phase to
neutral still present 230v when things Mr Murphy calls in to take a
look. :-(

I was worth it to see my twenty year old son jump when there was the
flash and bang lol.

But there are lots of reasons mainly to do with debugging for designs
with different power supply systems than the programmer.  

_______________________________________

Feel the power of the dark side!  
Atmel AVR

Roy Hopkins
Tauranga
New Zealand
_______________________________________


{Quote hidden}

2006\09\27@155523 by Vasile Surducan

face picon face
On 9/27/06, Roy <roy.hEraseMEspam.....ihug.co.nz> wrote:
> I have already said I will probably adopt the idea of using low voltage
> AC source for debugging - this alone may be enough.

You don't need anything else than  isolate the main phase from your
target design, in any other easiest way than this stupid (and
unuseless) bidirectional optocoupler. (think before answer right
here).


>
> On reason is from a past experience were I accidentally put a TRAIC in
> the wrong way around and the 5v supply that was direct of the phase
> suddenly presented 230v AC to my PIC when the TRIAC burnt out (I was
> using a PIC at the time), this failed then my programmer died and
> finally my RS232 interface on my computer died also.


Well if you did that, the optoisolated ICD will not help you now...
:)
A triac must be driven using a photo triac optocoupler (like some MOC
series) if you still need RS232 to be connected to your system. Just
because the RS232 ground is connected to the earth ground via PC
supply cable.

Just for info, I've made some PIC flame supervisors where the PIC was
floated at 200DC from the earth ground. I didn't use any optoisolated
WISP, or ICD or other programmers and never ever killed any PIC.
And BTW, WISP is the last device which need optoisolator when is
connected to a laptop *which run powered only from own batteries* .


>
> I was using a 1:1 isolating transformer in this case but phase to
> neutral still present 230v when things Mr Murphy calls in to take a
> look. :-(

There is no phase and no neutral after a 1:1 transformer Roy... just
one floated voltage isolated from the earth ground. Think a little
deeper and analyze all options where the parasytical currents may flow
in such design. Then you'll understand (hopefully) hat did I mean.

Vasile

2006\09\27@180857 by Gerhard Fiedler

picon face
Roy wrote:

> PS how do I make outlook us bottom posting?

I use Outlook for regular mail, and 40tude Dialog for newsgroups/mailing
lists. IMO Outlook is not good for this.

You can configure it to add the customary quoting character '>' when
replying to plain text mails, but you have to do most of the rest manually.
I find it sucks :)

Gerhard

2006\09\27@191818 by stef mientki

flavicon
face
Alan B. Pearce wrote:
>> So I have played around with the Linear Technology
>> spice program - see attached jpeg.
>>    
>
> You will find that the 8N137 opto is a lot faster than the 4N25, but an even
> better one is the Agilent (now Avago Technologies) HCPL-2300 which has an
> amplifier for the PIN diode included, and has very low LED current
> requirement, and is claimed to be 8M baud. Also has UL, CSA and VDE
> approvals for working with mains voltages.
>
>  
What I really don't understand is why these 10MB devices are so
difficult to find,
if you realize that optical communication through glass fiber goes upto
10GHz and maybe faster.
Stef

2006\09\28@050413 by Michael Rigby-Jones

picon face


{Quote hidden}

There is a world of difference between the technologies though. 10Gbit fibre optics use lasers that are often not directly modulated.  e.g. they use a mach-zehnder interferometer, or an electro-absorbtion modulator.  You can get directly modulated 10Gbit lasers, but they are for short distances, and still require a relatively sophisticated driver. A lot of work also goes into the receive chain as well. Photo-transistors are far too slow, they are either PIN diodes or for greater sensitivity APD's (avalanche photodiodes).  The optical components in a 10Gbit link tend to be physicaly large, uber expensive and power hungry.

Regards

Mike

=======================================================================
This e-mail is intended for the person it is addressed to only. The
information contained in it may be confidential and/or protected by
law. If you are not the intended recipient of this message, you must
not make any use of this information, or copy or show it to any
person. Please contact us immediately to tell us that you have
received this e-mail, and return the original to us. Any use,
forwarding, printing or copying of this message is strictly prohibited.
No part of this message can be considered a request for goods or
services.
=======================================================================

2006\09\28@080854 by Dave Lag

picon face
Michael Rigby-Jones wrote:
{Quote hidden}

There seem to be some "cheap" gbics around a $100.
What would be inside those?
D


2006\09\28@083020 by Michael Rigby-Jones

picon face


{Quote hidden}

The short reach Gigabit ethernet modules use an interesting device called a VCSEL, (Vertical Cavity Surface Emmiting Laser).  These are cheaper to make than traditional edge emmiting lasers, but are currently limited to the shorter wavelengths (gbic is around 850nm IIRC), and have lower output.  Ideal for cheap short range stuff using multimode fibre, though in relation to the original thread would still be expensive to package up to use as an opto-isolator by the time you add a PIN diode and drive electronics.

Regards

Mike

=======================================================================
This e-mail is intended for the person it is addressed to only. The
information contained in it may be confidential and/or protected by
law. If you are not the intended recipient of this message, you must
not make any use of this information, or copy or show it to any
person. Please contact us immediately to tell us that you have
received this e-mail, and return the original to us. Any use,
forwarding, printing or copying of this message is strictly prohibited.
No part of this message can be considered a request for goods or
services.
=======================================================================

2006\09\28@083100 by Alan B. Pearce

face picon face
rather than an optoisolated interface, you may get away with using a chip
such as the Maxim MAX3535E isolated RS485 transceiver. 2500Vrms isolation
claimed, all done inside the chip.

2006\09\28@101950 by Dave Lag

picon face
Michael Rigby-Jones wrote:
>>There seem to be some "cheap" gbics around a $100.
>>What would be inside those?
>
> The short reach Gigabit ethernet modules use an interesting device called a VCSEL, (Vertical Cavity Surface Emmiting Laser).  These are cheaper to make than traditional edge emmiting lasers, but are currently limited to the shorter wavelengths (gbic is around 850nm IIRC), and have lower output.  Ideal for cheap short range stuff using multimode fibre, though in relation to the original thread would still be expensive to package up to use as an opto-isolator by the time you add a PIN diode and drive electronics.
>
> Regards
>
> Mike

I was almost jumping to conclusions when I saw the DB-9 on the
electrical end of some gbics.
I had to remind myself that token-ring and T-1 also used DB9,15 for
their own nefarious purposes.
Must be some kind of standard (electrical) interface tho?...
gargoyle says... >  GBIC rev 5.5 sff-8053
Hmmm... runs on 5 volts Stef!
:)  D

More... (looser matching)
- Last day of these posts
- In 2006 , 2007 only
- Today
- New search...