Opto Mosfet (SSR) equivalent circuit

O

OBones

Guest
Hi all

I'm having a hard time finding Opto mosfets in electronics shops and the
manufacturer (or their distributor) doesn't want to sell them in small
quantities (10 or less). The model I'm interested in is the LH1512B from
Vishay which datasheet is here:

http://www.vishay.com/docs/83820/83820.pdf

As a result, I'm considering building a replacement circuit with a
regular optocoupler and mosfets.
The application is to use two of these Form C SSR to swith telephone
lines. As I require one line to be connected without power, I need the
Normally Closed behaviour of the Mosfet.
In the end, I just need two LH1512B so that I can have the
functionnality of a DPDT relay.

Can anyone help with the design of such an equivalent circuit ?

Thanks a lot in advance.

Cheers
Olivier
 
On Fri, 01 Apr 2005 11:11:20 +0200, OBones
<obones_gfd_@_gfd_altern.org> wrote:

Hi all

I'm having a hard time finding Opto mosfets in electronics shops and the
manufacturer (or their distributor) doesn't want to sell them in small
quantities (10 or less). The model I'm interested in is the LH1512B from
Vishay which datasheet is here:

http://www.vishay.com/docs/83820/83820.pdf

As a result, I'm considering building a replacement circuit with a
regular optocoupler and mosfets.
The application is to use two of these Form C SSR to swith telephone
lines. As I require one line to be connected without power, I need the
Normally Closed behaviour of the Mosfet.
In the end, I just need two LH1512B so that I can have the
functionnality of a DPDT relay.

Can anyone help with the design of such an equivalent circuit ?

Thanks a lot in advance.

Cheers
Olivier
Click to expand...
Have you thought about an off-the-shelf device such as an A/B 2 line
switch box http://www.sandman.com/lineshar.html for $27? No power
required....
 
OBones (Olivier) wrote...
I'm having a hard time finding Opto mosfets in electronics shops and the
manufacturer (or their distributor) doesn't want to sell them in small
quantities (10 or less). The model I'm interested in is the LH1512B from
Vishay which datasheet is here:

http://www.vishay.com/docs/83820/83820.pdf

As a result, I'm considering building a replacement circuit with a
regular optocoupler and mosfets.
The application is to use two of these Form C SSR to swith telephone
lines. As I require one line to be connected without power, I need the
Normally Closed behaviour of the Mosfet. In the end, I just need two
LH1512B so that I can have the functionnality of a DPDT relay.

Can anyone help with the design of such an equivalent circuit ?
Click to expand...
To get a normally-closed functionality, with no power applied, you'll
need to use depletion-mode FETs. These are rare beasts, but Supertex
makes a small selection. When I need a form B (normally-closed) SSR,
I turn to Aromat NAiS and their rather extensive line of form B types.
For example, the AQV414 is in stock at DigiKey, in all its variants.
http://dkc3.digikey.com/PDF/T051/1243.pdf You may want to examine the
rest of the Matsushita Aromat NAiS SSR line before choosing a part.


--
Thanks,
- Win
 
Ross Herbert wrote:

Have you thought about an off-the-shelf device such as an A/B 2 line
switch box http://www.sandman.com/lineshar.html for $27? No power
required....
Click to expand...
Thanks for that, but this is not available in France.
Further, what I want to do is a device that detects the ring tone on two
lines and switches the one that is ringing to the phone.
But if I take the line with the phone to make a call, it must also allow
to select one line to use to go out.
There exists ready made boxes for that, but they are sold for about 50
euros, which is way too expensive (considering what's inside).

On top of that, I like doing small electronics projects, the "done it
myself" syndrom ;-)

Thanks for the help though.
 
Winfield Hill wrote:

To get a normally-closed functionality, with no power applied, you'll
need to use depletion-mode FETs. These are rare beasts, but Supertex
makes a small selection. When I need a form B (normally-closed) SSR,
I turn to Aromat NAiS and their rather extensive line of form B types.
For example, the AQV414 is in stock at DigiKey, in all its variants.
http://dkc3.digikey.com/PDF/T051/1243.pdf You may want to examine the
rest of the Matsushita Aromat NAiS SSR line before choosing a part.
Click to expand...
Thanks for that, but NAIS products are difficult to find in France. I
found the AQW614 but it costs 13 euros per unit, which is way too much,
considering that Vishay or Nec products sell for about 2 euros (rec
retail price).
 
OBones wrote...
Winfield Hill wrote:

To get a normally-closed functionality, with no power applied, you'll
need to use depletion-mode FETs. These are rare beasts, but Supertex
makes a small selection. When I need a form B (normally-closed) SSR,
I turn to Aromat NAiS and their rather extensive line of form B types.
For example, the AQV414 is in stock at DigiKey, in all its variants.
http://dkc3.digikey.com/PDF/T051/1243.pdf You may want to examine the
rest of the Matsushita Aromat NAiS SSR line before choosing a part.

Thanks for that, but NAIS products are difficult to find in France.
I found the AQW614 but it costs 13 euros per unit, which is way too
much, considering that Vishay or Nec products sell for about 2 euros
(rec retail price).
Click to expand...
If you "like doing small electronics projects," as you say, I'd grab
the part you can find rather than agonize over the part you can't get,
even if you do see a (false) indication of a large price difference.

If you use the NAiS part and your project becomes successful so you
have a reason to care about every Euro spent, you'll easily find a
competitive-price source for these parts. Matsushita competes over
price and delivery head-to-head worldwide with the other SSR types.


--
Thanks,
- Win
 
Winfield Hill <hill_a@t_rowland-dotties-harvard-dot.s-edu> writes:

OBones wrote...

Winfield Hill wrote:

To get a normally-closed functionality, with no power applied, you'll
need to use depletion-mode FETs. These are rare beasts, but Supertex
makes a small selection. When I need a form B (normally-closed) SSR,
I turn to Aromat NAiS and their rather extensive line of form B types.
For example, the AQV414 is in stock at DigiKey, in all its variants.
http://dkc3.digikey.com/PDF/T051/1243.pdf You may want to examine the
rest of the Matsushita Aromat NAiS SSR line before choosing a part.

Thanks for that, but NAIS products are difficult to find in France.
I found the AQW614 but it costs 13 euros per unit, which is way too
much, considering that Vishay or Nec products sell for about 2 euros
(rec retail price).

If you "like doing small electronics projects," as you say, I'd grab
the part you can find rather than agonize over the part you can't get,
even if you do see a (false) indication of a large price difference.

If you use the NAiS part and your project becomes successful so you
have a reason to care about every Euro spent, you'll easily find a
competitive-price source for these parts. Matsushita competes over
price and delivery head-to-head worldwide with the other SSR types.
Click to expand...
Has anyone seen a "normally closed" photo-mos that can switch half an
amp or so? (At 110VAC)? I have an application that could use such a
beast if it existed. It needs to switch 24VDC too, so cannot use a
normal SSR!

Failing that, is it practical to build it out of discretes?

Thanks,

--

John Devereux
 
John Devereux wrote...
Has anyone seen a "normally closed" photo-mos that can switch half
an amp or so? (At 110VAC)? I have an application that could use such
a beast if it existed. It needs to switch 24VDC too, so cannot use
a normal SSR!
Click to expand...
You could use five 100mA units in parallel. :>) Actually, for
120Vac, or even 24Vdc, one is usually willing to accept a volt
or so of switch Ron or forward drop, for example, with the SSRs
we've been discussing. By contrast, if you use the SSR to turn
on a triac, the added voltage drop will not be too painfully high,
given that the current through the controlling SSR will be low.
I am a big fan of the Sharp zero-crossing opto-triac parts, which
come in insulated large-power-transistor-like packages.


--
Thanks,
- Win
 
Winfield Hill <hill_a@t_rowland-dotties-harvard-dot.s-edu> writes:

John Devereux wrote...

Has anyone seen a "normally closed" photo-mos that can switch half
an amp or so? (At 110VAC)? I have an application that could use such
a beast if it existed. It needs to switch 24VDC too, so cannot use
a normal SSR!

You could use five 100mA units in parallel. :>) Actually, for
Click to expand...
Well that's cheating!

120Vac, or even 24Vdc, one is usually willing to accept a volt
or so of switch Ron or forward drop, for example, with the SSRs
we've been discussing. By contrast, if you use the SSR to turn
on a triac, the added voltage drop will not be too painfully high,
given that the current through the controlling SSR will be low.
I am a big fan of the Sharp zero-crossing opto-triac parts, which
come in insulated large-power-transistor-like packages.
Click to expand...
A few volts drop in itself is no problem... Although the heat
generated at a few hundred mA might be.

The application is a replacement for a board with some normally closed
relay outputs. In some installations the relays switch 24V into a PLC
input. In other installations it directly switches a 110VAC solenoid
valve. The replacement board has to be able to cope with either
scenario.

The "relays" actually need to be changeover contacts, but I have seen
opto-fet parts that can do the "NO" side already. It is the "NC" parts
that are still limited to 100mA or so, it would seem.


--

John Devereux
 
John Devereux wrote...
It is the "NC" parts that are still limited to 100mA or so,
it would seem.
Click to expand...
That's because that's all the telecom market needs.


--
Thanks,
- Win
 
On Fri, 01 Apr 2005 14:08:01 +0200, OBones
<obones_gfd_@_gfd_altern.org> wrote:

Ross Herbert wrote:

Have you thought about an off-the-shelf device such as an A/B 2 line
switch box http://www.sandman.com/lineshar.html for $27? No power
required....

Thanks for that, but this is not available in France.
Further, what I want to do is a device that detects the ring tone on two
lines and switches the one that is ringing to the phone.
But if I take the line with the phone to make a call, it must also allow
to select one line to use to go out.
There exists ready made boxes for that, but they are sold for about 50
euros, which is way too expensive (considering what's inside).

On top of that, I like doing small electronics projects, the "done it
myself" syndrom ;-)

Thanks for the help though.
Click to expand...
I'd be more inclined to use an analog switch from Analog Devices
http://www.analog.com/en/subCat/0,2879,768%255F833%255F0%255F%255F0%255F,00.html

They make SPDT units up to 4 switches/package.
 
Ross Herbert wrote:

I'd be more inclined to use an analog switch from Analog Devices
http://www.analog.com/en/subCat/0,2879,768%255F833%255F0%255F%255F0%255F,00.html

They make SPDT units up to 4 switches/package.
Click to expand...
Thanks for the link, but how can I guarantee the electrical isolation in
this case ?
I really want the two circuits to be separated and that's what opto
isolators (or relays) provide.
Basically, I don't want transients from the telephone line to be fed
back into the control circuit.

Thanks for any info you can provide

Cheers
 
On Sun, 03 Apr 2005 12:30:25 +0200, OBones
<obones_gfd_@_gds_altern.org> wrote:

Ross Herbert wrote:

I'd be more inclined to use an analog switch from Analog Devices
http://www.analog.com/en/subCat/0,2879,768%255F833%255F0%255F%255F0%255F,00.html

They make SPDT units up to 4 switches/package.

Thanks for the link, but how can I guarantee the electrical isolation in
this case ?
I really want the two circuits to be separated and that's what opto
isolators (or relays) provide.
Basically, I don't want transients from the telephone line to be fed
back into the control circuit.

Thanks for any info you can provide

Cheers
Click to expand...
I'll just select this 2 switch device for example
http://www.analog.com/UploadedFiles/Data_Sheets/149061683ADG884_0.pdf

Look at the applications given on page 1. Now if there was going to be
any problem with channel x-talk or interference then I doubt that they
would be used for audio & video signal routing, modems or whatever.
The x-talk between switches is typically -60dB or -120dB.
 
Ross Herbert wrote:

I'll just select this 2 switch device for example
http://www.analog.com/UploadedFiles/Data_Sheets/149061683ADG884_0.pdf

Look at the applications given on page 1. Now if there was going to be
any problem with channel x-talk or interference then I doubt that they
would be used for audio & video signal routing, modems or whatever.
The x-talk between switches is typically -60dB or -120dB.
Click to expand...
That looks nice, but I have a question:

The phone line can have up to 200volts on it, and my control circuit
works in 12Volts.
Now, it also happens that there are high voltage transients on the line.
How can I guarantee that these 1000Volts transient won't "jump back"
into the control circuit ?
At least with opto couplers, I know that this won't happen. The mosfet
side will be burnt, I know, but that's less problematic than burning the
whole system.

Maybe I'm being dumb here, but I have a hard time "getting away" from
fully isolated designs, especially when one side (control) is DC and the
other side (line) is AC.

Thanks a lot for your help, it is very much appreciated

Cheers
Olivier
 
Winfield Hill wrote:

If you "like doing small electronics projects," as you say, I'd grab
the part you can find rather than agonize over the part you can't get,
even if you do see a (false) indication of a large price difference.
Click to expand...
Yes, but even those other parts are really hard to come-by. It seems
that none of the hobbyist use OptoMosfets here in France.


If you use the NAiS part and your project becomes successful so you
have a reason to care about every Euro spent, you'll easily find a
competitive-price source for these parts. Matsushita competes over
price and delivery head-to-head worldwide with the other SSR types.
Click to expand...
I understand that perfectly, but I'm mostly a bobbyist, and I don't
really intend to sell this project. It's something I want to do for
myself, but it is a bit too specific to ever reach the market. And my
work field is not electronics either.

But thanks for your help, it is very much appreciated

Cheers
Olivier
 
On Sun, 03 Apr 2005 14:04:39 +0200, OBones
<obones_gfd_@_gds_altern.org> wrote:

Ross Herbert wrote:

I'll just select this 2 switch device for example
http://www.analog.com/UploadedFiles/Data_Sheets/149061683ADG884_0.pdf

Look at the applications given on page 1. Now if there was going to be
any problem with channel x-talk or interference then I doubt that they
would be used for audio & video signal routing, modems or whatever.
The x-talk between switches is typically -60dB or -120dB.

That looks nice, but I have a question:

The phone line can have up to 200volts on it, and my control circuit
works in 12Volts.
Now, it also happens that there are high voltage transients on the line.
How can I guarantee that these 1000Volts transient won't "jump back"
into the control circuit ?
At least with opto couplers, I know that this won't happen. The mosfet
side will be burnt, I know, but that's less problematic than burning the
whole system.

Maybe I'm being dumb here, but I have a hard time "getting away" from
fully isolated designs, especially when one side (control) is DC and the
other side (line) is AC.

Thanks a lot for your help, it is very much appreciated

Cheers
Olivier
Click to expand...
How can you guarantee.....
By using standard techniques to minimise transient line peaks as all
other telco interfaces must do. The protection must be at the
telephone line entry point on your electronics device. The subject of
telephone line interfacing and surge protection requires specialised
knowledge if the relevant specifications are to be met. For example
this article
http://www.elecdesign.com/Articles/Index.cfm?AD=1&ArticleID=3564 while
relating to ADSL nevertheless covers the ground common to all telco
POTS lines.

In order to design your own 2 line sharing device with auto selection
you are taking on what appears be a simple project but it is more
complex than you think.

You may be better off buying an off the shelf line sharer which has
the functions you require
http://natcomm.com.au/australia/index.cfm?page=product_details&id=37&product_id=37

For DIYers'

Take a look at Tommi Engdahl's website
http://www.tkk.fi/Misc/Electronics/circuits/teleinterface.html
and more on the subject
http://www.rane.com/pdf/note150.pdf
http://www.epanorama.net/documents/telecom/telesurge.html
 
OBones wrote...
Well, I have the schematic ready, it's just that I can't find those
OptoMosfets. I may end up replacing them by a DPDT relay, but it uses
more amps, and that's not desirable (well, I could live with it)
Click to expand...
What was wrong with my suggestion of using a low-current form B
ssr (with a series resistor) to drive a triac? This will work
for both low-voltage dc and high-voltage ampere-range ac.


--
Thanks,
- Win
 
OBones wrote...
Winfield Hill wrote:

OBones wrote...

Well, I have the schematic ready, it's just that I can't find those
OptoMosfets. I may end up replacing them by a DPDT relay, but it uses
more amps, and that's not desirable (well, I could live with it)

What was wrong with my suggestion of using a low-current form B
ssr (with a series resistor) to drive a triac? This will work
for both low-voltage dc and high-voltage ampere-range ac.

The problem is not the current in the ssr (phone signals are around
60mA, way under the 100mA limit), the problem is to actually find those
ssr in electronics store around here in France.
I would prefer to use two Form C SSRs, but if I can find 1 dual Form A
and 1 dual Form B, i'd use them.
Problem is, noone seems to use this among the hobbyists...
Ah well.
Click to expand...
How many did you want?


--
Thanks,
- Win
 
Winfield Hill wrote:

How many did you want?
Click to expand...
Two for a start, and maybe 2 more.
I went online to digikey, found the part I wanted for about 4$ per piece
which is a reasonnable price. But then I went to the shipping and
handling costs: 11$ shipping, 6$ handling (or something similar).
Heck, 25$ for two chips, it's a bit too much.
I'll keep nagging the official distributors in France, maybe one will be
so annoyed that it will send me two chips ;-)
 
"OBones" <obones_gfd_@_gfd_altern.org> a écrit dans le message de
news:d2u0vk$940$1@reader1.imaginet.fr...
Winfield Hill wrote:

How many did you want?

Two for a start, and maybe 2 more.
I went online to digikey, found the part I wanted for about 4$ per piece
which is a reasonnable price. But then I went to the shipping and
handling costs: 11$ shipping, 6$ handling (or something similar).
Heck, 25$ for two chips, it's a bit too much.
I'll keep nagging the official distributors in France, maybe one will be
so annoyed that it will send me two chips ;-)
Click to expand...
I have to pass an order at digikey too.
We could share costs. Write me in private.


--
Thanks,
Fred.
 
You must log in or register to reply here.

Welcome to EDABoard.com

Sponsor

Back
Top