 |
Search Sci.Electronics.Basics |
|
|
|
|
|
|
 |
|
|
Sci.Electronics.Basics -> Choosing voltage opamp
There are 29 messages in this thread.
You are currently looking at messages 1 to 20.
|
Author: VitaliyDate: 13:05 23-09-06
|
|
Hello,
I'm looking to add the 2nd stage to photodiode transimpedance
amplifier. All I want to achieve by 2nd stage is voltage gain and
capability to drive 50 Ohm loads. The first stage is OPA657.
1) a) Given power supply for opamp +-5V, should I select the opamp that
has Io,max=>200mA. I do expect Vpeak-to-peak,output to be 10V.
b) Is that output current achievable only at maximum supply voltage,
or as long as
2) I need to get an opamp with maximum input impedance (well, anything
>=1MOhm will do it). Should I get the one with JFET-input? Or any voltage amplifier
will do the job?
3) What is Iq per channel(Max)(mA)?
4) Would SOIC packaging be better in general? Given that pcb layout is
optimized.
Thanks in advance,
Vitaliy
|
|
|
|
Author: PeteSDate: 14:47 23-09-06
|
|
Vitaliy wrote:
> Hello,
>
> I'm looking to add the 2nd stage to photodiode transimpedance
> amplifier. All I want to achieve by 2nd stage is voltage gain and
> capability to drive 50 Ohm loads. The first stage is OPA657.
>
> 1) a) Given power supply for opamp +-5V, should I select the opamp that
> has Io,max=>200mA. I do expect Vpeak-to-peak,output to be 10V.
> b) Is that output current achievable only at maximum supply voltage,
> or as long as
>
> 2) I need to get an opamp with maximum input impedance (well, anything
> >=1MOhm will do it). Should I get the one with JFET-input? Or any voltage
amplifier will do the job?
>
> 3) What is Iq per channel(Max)(mA)?
>
> 4) Would SOIC packaging be better in general? Given that pcb layout is
> optimized.
>
> Thanks in advance,
> Vitaliy
All of the usual suspects (National, TI [including Burr-Brown], Linear
tech, AD and others) make devices specifically designed to drive 50 ohm
loads.
Check out the websites and use the parametric search function and
specify 50 ohm load and then choose a suitable device. We don't know
what type of load that is beyond 50 ohm, so the above is the only
advice I can give.
Cheers
PeteS
|
|
|
|
Author: JamieDate: 16:57 23-09-06
|
|
Vitaliy wrote:
> Hello,
>
> I'm looking to add the 2nd stage to photodiode transimpedance
> amplifier. All I want to achieve by 2nd stage is voltage gain and
> capability to drive 50 Ohm loads. The first stage is OPA657.
>
> 1) a) Given power supply for opamp +-5V, should I select the opamp that
> has Io,max=>200mA. I do expect Vpeak-to-peak,output to be 10V.
> b) Is that output current achievable only at maximum supply voltage,
> or as long as
>
> 2) I need to get an opamp with maximum input impedance (well, anything
>
>>=1MOhm will do it). Should I get the one with JFET-input? Or any voltage
amplifier will do the job?
>
>
> 3) What is Iq per channel(Max)(mA)?
>
> 4) Would SOIC packaging be better in general? Given that pcb layout is
> optimized.
>
> Thanks in advance,
> Vitaliy
>
google for "High current Op-amp"
there are lots of them out there..
--
Real Programmers Do things like this.
http://webpages.charter.net/jamie_5
|
|
|
|
Author: David L. JonesDate: 18:18 23-09-06
|
|
Vitaliy wrote:
> Hello,
>
> I'm looking to add the 2nd stage to photodiode transimpedance
> amplifier. All I want to achieve by 2nd stage is voltage gain and
> capability to drive 50 Ohm loads. The first stage is OPA657.
>
> 1) a) Given power supply for opamp +-5V, should I select the opamp that
> has Io,max=>200mA. I do expect Vpeak-to-peak,output to be 10V.
> b) Is that output current achievable only at maximum supply voltage,
> or as long as
You won't get 10Vp-p output from an opamp with +/-5V rails, you need a
larger supply voltage.
Dave :)
|
|
|
|
Author: VitaliyDate: 19:31 23-09-06
|
|
I've been trying to do that.
In a lot of examples on the datasheets of amplifiers I'm looking for, I
see they are using 50Ohm input. However, I know for sure (test results)
that output of the first stage (photodiode amp) does not work with
50Ohm loads, but works with 1MOhm loads. Should I be concerned about
that?
Vitaliy
>
> All of the usual suspects (National, TI [including Burr-Brown], Linear
> tech, AD and others) make devices specifically designed to drive 50 ohm
> loads.
>
> Check out the websites and use the parametric search function and
> specify 50 ohm load and then choose a suitable device. We don't know
> what type of load that is beyond 50 ohm, so the above is the only
> advice I can give.
>
> Cheers
>
> PeteS
|
|
|
|
Author: VitaliyDate: 19:44 23-09-06
|
|
Hi David,
Is the maximum Vp-p I can get from an opamp with +/-5V rails 5V then?
I know I measured the output of transimpedance amplifier and I was
getting maximum ~8Vp-p :confused:
Vitaliy
David L. Jones wrote:
> Vitaliy wrote:
> > Hello,
> >
> > I'm looking to add the 2nd stage to photodiode transimpedance
> > amplifier. All I want to achieve by 2nd stage is voltage gain and
> > capability to drive 50 Ohm loads. The first stage is OPA657.
> >
> > 1) a) Given power supply for opamp +-5V, should I select the opamp that
> > has Io,max=>200mA. I do expect Vpeak-to-peak,output to be 10V.
> > b) Is that output current achievable only at maximum supply voltage,
> > or as long as
>
> You won't get 10Vp-p output from an opamp with +/-5V rails, you need a
> larger supply voltage.
>
> Dave :)
|
|
|
|
Author: VitaliyDate: 20:22 23-09-06
|
|
I'm not sure if I'm answering my own question, but I can put 1MOhm at
the input of opamp to the ground.
Vitaliy
Vitaliy wrote:
> I've been trying to do that.
>
> In a lot of examples on the datasheets of amplifiers I'm looking for, I
> see they are using 50Ohm input. However, I know for sure (test results)
> that output of the first stage (photodiode amp) does not work with
> 50Ohm loads, but works with 1MOhm loads. Should I be concerned about
> that?
>
> Vitaliy
> >
> > All of the usual suspects (National, TI [including Burr-Brown], Linear
> > tech, AD and others) make devices specifically designed to drive 50 ohm
> > loads.
> >
> > Check out the websites and use the parametric search function and
> > specify 50 ohm load and then choose a suitable device. We don't know
> > what type of load that is beyond 50 ohm, so the above is the only
> > advice I can give.
> >
> > Cheers
> >
> > PeteS
|
|
|
|
Author: DanaDate: 21:47 23-09-06
|
|
"Vitaliy" <vmykhayl@ee.ryerson.ca> wrote in message
news:1159055088.803133.110670@m7g2000cwm.googlegroups.com...
> Hi David,
>
> Is the maximum Vp-p I can get from an opamp with +/-5V rails 5V then?
> I know I measured the output of transimpedance amplifier and I was
> getting maximum ~8Vp-p :confused:
>
> Vitaliy
>
> David L. Jones wrote:
> > Vitaliy wrote:
> > > Hello,
> > >
> > > I'm looking to add the 2nd stage to photodiode transimpedance
> > > amplifier. All I want to achieve by 2nd stage is voltage gain and
> > > capability to drive 50 Ohm loads. The first stage is OPA657.
> > >
> > > 1) a) Given power supply for opamp +-5V, should I select the opamp
that
> > > has Io,max=>200mA. I do expect Vpeak-to-peak,output to be 10V.
> > > b) Is that output current achievable only at maximum supply
voltage,
> > > or as long as
> >
> > You won't get 10Vp-p output from an opamp with +/-5V rails, you need a
> > larger supply voltage.
> >
> > Dave :)
Are you sure about that.
Assume you have a 1k resistor on the inverting input in series with the
incoming signal. On the feedback line from the output back to the input you
have a 10k res, and on the non inverting input you have a 1k (909ohm
resister if you do the math).
With an input sinewave of 1 volt at 1ma you will have a 10v p-p output
>
|
|
|
|
Author: redbellyDate: 00:30 24-09-06
|
|
Vitaliy wrote:
> I've been trying to do that.
>
> In a lot of examples on the datasheets of amplifiers I'm looking for, I
> see they are using 50Ohm input. However, I know for sure (test results)
> that output of the first stage (photodiode amp) does not work with
> 50Ohm loads, but works with 1MOhm loads. Should I be concerned about
> that?
It depends on how fast a response time you need. Photodiodes often use
50 ohm loads for impendance matching, so that a pulse or step signal
does not reflect back and forth in the cabling and cause "echo"
signals.
If you're okay with, say 1 usec or slower response time, I wouldn't
worry about using a 50 ohm load. You might even be okay with the
higher impendance at somewhat faster response times, you might need to
experiment a little. If you could get by with a 1k or 10k load then
you'd have an easier time finding an opamp.
You didn't say much yet about the nature of your signal, other than
it's optical. Are you looking at light pulses (and at what pulse
duration), modulated light (and at what frequency), or a more-or-less
cw signal that changes slowly in intensity (and on what sort of time
scale does the intensity change)?
Mark
|
|
|
|
Author: David L. JonesDate: 03:28 24-09-06
|
|
Vitaliy wrote:
> Hi David,
>
> Is the maximum Vp-p I can get from an opamp with +/-5V rails 5V then?
> I know I measured the output of transimpedance amplifier and I was
> getting maximum ~8Vp-p :confused:
It depends on the particular device.
As a general rule of thumb an opamp can get to within at least 2V of
either rail, so a +/-5V supply will give you at a minimum +/-3V. Your
opamp could aboviously get within 1V of the rail to give you 8Vp-p.
There are "rail-to-rail" opamps that claim to get very close to the
full output voltage rail, but usually not at any significant output
current.
You need a higher rail voltage regardless of the device you choose.
Dave :)
|
|
|
|
Author: David L. JonesDate: 03:30 24-09-06
|
|
Dana wrote:
> "Vitaliy" <vmykhayl@ee.ryerson.ca> wrote in message
> news:1159055088.803133.110670@m7g2000cwm.googlegroups.com...
> > Hi David,
> >
> > Is the maximum Vp-p I can get from an opamp with +/-5V rails 5V then?
> > I know I measured the output of transimpedance amplifier and I was
> > getting maximum ~8Vp-p :confused:
> >
> > Vitaliy
> >
> > David L. Jones wrote:
> > > Vitaliy wrote:
> > > > Hello,
> > > >
> > > > I'm looking to add the 2nd stage to photodiode transimpedance
> > > > amplifier. All I want to achieve by 2nd stage is voltage gain and
> > > > capability to drive 50 Ohm loads. The first stage is OPA657.
> > > >
> > > > 1) a) Given power supply for opamp +-5V, should I select the opamp
> that
> > > > has Io,max=>200mA. I do expect Vpeak-to-peak,output to be 10V.
> > > > b) Is that output current achievable only at maximum supply
> voltage,
> > > > or as long as
> > >
> > > You won't get 10Vp-p output from an opamp with +/-5V rails, you need a
> > > larger supply voltage.
> > >
> > > Dave :)
>
> Are you sure about that.
> Assume you have a 1k resistor on the inverting input in series with the
> incoming signal. On the feedback line from the output back to the input you
> have a 10k res, and on the non inverting input you have a 1k (909ohm
> resister if you do the math).
> With an input sinewave of 1 volt at 1ma you will have a 10v p-p output
Not with +/-5V rails he won't.
Dave :)
|
|
|
|
Author: DanaDate: 12:19 24-09-06
|
|
"David L. Jones" <altzone@gmail.com> wrote in message
news:1159083015.452886.300660@b28g2000cwb.googlegroups.com...
>
> Dana wrote:
> > "Vitaliy" <vmykhayl@ee.ryerson.ca> wrote in message
> > news:1159055088.803133.110670@m7g2000cwm.googlegroups.com...
> > > Hi David,
> > >
> > > Is the maximum Vp-p I can get from an opamp with +/-5V rails 5V then?
> > > I know I measured the output of transimpedance amplifier and I was
> > > getting maximum ~8Vp-p :confused:
> > >
> > > Vitaliy
> > >
> > > David L. Jones wrote:
> > > > Vitaliy wrote:
> > > > > Hello,
> > > > >
> > > > > I'm looking to add the 2nd stage to photodiode transimpedance
> > > > > amplifier. All I want to achieve by 2nd stage is voltage gain
and
> > > > > capability to drive 50 Ohm loads. The first stage is OPA657.
> > > > >
> > > > > 1) a) Given power supply for opamp +-5V, should I select the
opamp
> > that
> > > > > has Io,max=>200mA. I do expect Vpeak-to-peak,output to be
10V.
> > > > > b) Is that output current achievable only at maximum supply
> > voltage,
> > > > > or as long as
> > > >
> > > > You won't get 10Vp-p output from an opamp with +/-5V rails, you need
a
> > > > larger supply voltage.
> > > >
> > > > Dave :)
> >
> > Are you sure about that.
> > Assume you have a 1k resistor on the inverting input in series with the
> > incoming signal. On the feedback line from the output back to the input
you
> > have a 10k res, and on the non inverting input you have a 1k (909ohm
> > resister if you do the math).
> > With an input sinewave of 1 volt at 1ma you will have a 10v p-p output
>
> Not with +/-5V rails he won't.
So help me out here. If he has 5 and -5 rails, , and an input sinewave comes
in, so lonag as the op amp is properly biased, will not the output swing
between 5 and -5 (not exactly 5 do to loss), in essence giving a 10vpp
signal depending on how the output is referenced to ground.
>
> Dave :)
>
|
|
|
|
Author: Rene TschaggelarDate: 12:46 24-09-06
|
|
Dana wrote:
> "David L. Jones" <altzone@gmail.com> wrote in message
> news:1159083015.452886.300660@b28g2000cwb.googlegroups.com...
>>Not with +/-5V rails he won't.
>
>
> So help me out here. If he has 5 and -5 rails, , and an input sinewave comes
> in, so lonag as the op amp is properly biased, will not the output swing
> between 5 and -5 (not exactly 5 do to loss), in essence giving a 10vpp
> signal depending on how the output is referenced to ground.
Only rail-to-rail-output opamps come close to the rail
at the expense of some other parameter. Meaning RR-Output
is not exactly free.
Rene
--
Ing.Buero R.Tschaggelar - http://www.ibrtses.com
& commercial newsgroups - http://www.talkto.net
|
|
|
|
Author: DanaDate: 13:56 24-09-06
|
|
"Rene Tschaggelar" <none@none.net> wrote in message
news:4516b746$0$13443$5402220f@news.sunrise.ch...
> Dana wrote:
>
> > "David L. Jones" <altzone@gmail.com> wrote in message
> > news:1159083015.452886.300660@b28g2000cwb.googlegroups.com...
>
> >>Not with +/-5V rails he won't.
> >
> >
> > So help me out here. If he has 5 and -5 rails, , and an input sinewave
comes
> > in, so lonag as the op amp is properly biased, will not the output swing
> > between 5 and -5 (not exactly 5 do to loss), in essence giving a 10vpp
> > signal depending on how the output is referenced to ground.
>
> Only rail-to-rail-output opamps come close to the rail
> at the expense of some other parameter. Meaning RR-Output
> is not exactly free.
>
> Rene
True. But I think Dave was saying that with a +/- 5v supply, the output
would only be able to get a 5v p-p signal. I would think it would get close
to both the 5 and -5 supply giving close to a 10v p-p output.
Like the original poster mentioned when he observed an approx 8v p-p output.
And back to the original poster, yes if he required an absolute minimum of
10v p-p out, yes his rail supplies would have to be in line with that
requirement.
> --
> Ing.Buero R.Tschaggelar - http://www.ibrtses.com
> & commercial newsgroups - http://www.talkto.net
|
|
|
|
Author: David L. JonesDate: 17:29 24-09-06
|
|
Dana wrote:
> "Rene Tschaggelar" <none@none.net> wrote in message
> news:4516b746$0$13443$5402220f@news.sunrise.ch...
> > Dana wrote:
> >
> > > "David L. Jones" <altzone@gmail.com> wrote in message
> > > news:1159083015.452886.300660@b28g2000cwb.googlegroups.com...
> >
> > >>Not with +/-5V rails he won't.
> > >
> > >
> > > So help me out here. If he has 5 and -5 rails, , and an input sinewave
> comes
> > > in, so lonag as the op amp is properly biased, will not the output swing
> > > between 5 and -5 (not exactly 5 do to loss), in essence giving a 10vpp
> > > signal depending on how the output is referenced to ground.
> >
> > Only rail-to-rail-output opamps come close to the rail
> > at the expense of some other parameter. Meaning RR-Output
> > is not exactly free.
> >
> > Rene
>
> True. But I think Dave was saying that with a +/- 5v supply, the output
> would only be able to get a 5v p-p signal. I would think it would get close
> to both the 5 and -5 supply giving close to a 10v p-p output.
> Like the original poster mentioned when he observed an approx 8v p-p output.
> And back to the original poster, yes if he required an absolute minimum of
> 10v p-p out, yes his rail supplies would have to be in line with that
> requirement.
I was not saying a +/-5V rail will give you 5Vp-p, go have another
read.
A regular op-amp will only give about 8Vp-p as the OP found out.
Only (output) rail-to-rail opamps will get close to the rail voltage.
Normal rule of thumb for ordinary op-amps is you design for a 2V margin
on the rails, unless you inspect the datasheet and can take it further.
Dave :)
|
|
|
|
Author: redbellyDate: 17:37 24-09-06
|
|
Dana wrote:
> True. But I think Dave was saying that with a +/- 5v supply, the output
> would only be able to get a 5v p-p signal. I would think it would get close
> to both the 5 and -5 supply giving close to a 10v p-p output.
> Like the original poster mentioned when he observed an approx 8v p-p output.
> And back to the original poster, yes if he required an absolute minimum of
> 10v p-p out, yes his rail supplies would have to be in line with that
> requirement.
>
Most op-amps clip the signal when the output is within 2 volts (or so)
of the + or - supply. Hence the need for special "rail to rail" op
amps, which can get much closer to the supply voltage before clipping
(tens or hundreds of mV?)
The problem is worse the higher the output current, and the OP's desire
for 200 mA is a high current for an opamp.
Mark
|
|
|
|
Author: Johnny BoyDate: 05:39 25-09-06
|
|
"David L. Jones" <altzone@gmail.com> wrote in message
news:1159082931.013808.167160@d34g2000cwd.googlegroups.com...
> Vitaliy wrote:
> > Hi David,
> >
> > Is the maximum Vp-p I can get from an opamp with +/-5V rails 5V then?
> > I know I measured the output of transimpedance amplifier and I was
> > getting maximum ~8Vp-p :confused:
>
> It depends on the particular device.
> As a general rule of thumb an opamp can get to within at least 2V of
> either rail, so a +/-5V supply will give you at a minimum +/-3V. Your
> opamp could aboviously get within 1V of the rail to give you 8Vp-p.
> There are "rail-to-rail" opamps that claim to get very close to the
> full output voltage rail, but usually not at any significant output
> current.
> You need a higher rail voltage regardless of the device you choose.
>
> Dave :)
>
You should listen to David, he knows what he's saying - even the very best
of rail-to-rail op-amps won't go close to the supply rails, except if
they're very lightly loaded, 1M etc. They still usually fall short of the
negative rail by 5-50mV and 0.5-1.5V short of the positive rail.
... Johnny
|
|
|
|
Author: JamieDate: 14:32 25-09-06
|
|
Vitaliy wrote:
> I've been trying to do that.
>
> In a lot of examples on the datasheets of amplifiers I'm looking for, I
> see they are using 50Ohm input. However, I know for sure (test results)
> that output of the first stage (photodiode amp) does not work with
> 50Ohm loads, but works with 1MOhm loads. Should I be concerned about
> that?
>
> Vitaliy
put an old stand by TL082 type op-amp infront of it.
|
|
|
|
Author: VitaliyDate: 11:32 28-09-06
|
|
Johnny Boy wrote:
>
> You should listen to David, he knows what he's saying - even the very best
> of rail-to-rail op-amps won't go close to the supply rails, except if
> they're very lightly loaded, 1M etc. They still usually fall short of the
> negative rail by 5-50mV and 0.5-1.5V short of the positive rail.
> ... Johnny
OK, I doublechecked the requirements, I don't have to get the exact
10Vp-p. Maybe I was getting such high output voltages (for OPA657)
because of 1M load.
Jamie wrote:
> Vitaliy wrote:
>
> > I've been trying to do that.
> >
> > In a lot of examples on the datasheets of amplifiers I'm looking for, I
> > see they are using 50Ohm input. However, I know for sure (test results)
> > that output of the first stage (photodiode amp) does not work with
> > 50Ohm loads, but works with 1MOhm loads. Should I be concerned about
> > that?
> >
> > Vitaliy
>
> put an old stand by TL082 type op-amp infront of it.
But that would kill my BW (the BW on that is only 4MHz, and I need
40MHz), right?
redbelly wrote:
> If you're okay with, say 1 usec or slower response time, I wouldn't
> worry about using a 50 ohm load. You might even be okay with the
> higher impendance at somewhat faster response times, you might need to
> experiment a little. If you could get by with a 1k or 10k load then
> you'd have an easier time finding an opamp.
>
> You didn't say much yet about the nature of your signal, other than
> it's optical. Are you looking at light pulses (and at what pulse
> duration), modulated light (and at what frequency), or a more-or-less
> cw signal that changes slowly in intensity (and on what sort of time
> scale does the intensity change)?
I have broadband light source w/ rapid scanning optical delay line
(RSOD) and phase modulator. Are you suggesting this would affect the
required slew rate?
Thanks for all advices,
Vitaliy
|
|
|
|
Author: VitaliyDate: 12:56 28-09-06
|
|
Johnny Boy wrote:
>
> You should listen to David, he knows what he's saying - even the very best
> of rail-to-rail op-amps won't go close to the supply rails, except if
> they're very lightly loaded, 1M etc. They still usually fall short of the
> negative rail by 5-50mV and 0.5-1.5V short of the positive rail.
> ... Johnny
OK, I doublechecked the requirements, I don't have to get the exact
10Vp-p. Maybe I was getting such high output voltages (for OPA657)
because of 1M load
Jamie wrote:
> Vitaliy wrote:
>
> > I've been trying to do that.
> >
> > In a lot of examples on the datasheets of amplifiers I'm looking for, I
> > see they are using 50Ohm input. However, I know for sure (test results)
> > that output of the first stage (photodiode amp) does not work with
> > 50Ohm loads, but works with 1MOhm loads. Should I be concerned about
> > that?
> >
> > Vitaliy
>
> put an old stand by TL082 type op-amp infront of it.
But that would kill my BW (the BW on that is only 4MHz, and I need
40MHz), right?
redbelly wrote:
> If you're okay with, say 1 usec or slower response time, I wouldn't
> worry about using a 50 ohm load. You might even be okay with the
> higher impendance at somewhat faster response times, you might need to
> experiment a little. If you could get by with a 1k or 10k load then
> you'd have an easier time finding an opamp.
>
> You didn't say much yet about the nature of your signal, other than
> it's optical. Are you looking at light pulses (and at what pulse
> duration), modulated light (and at what frequency), or a more-or-less
> cw signal that changes slowly in intensity (and on what sort of time
> scale does the intensity change)?
I have broadband light source w/ rapid scanning optical delay line
(RSOD) and phase modulator. Are you suggesting this would affect the
required slew rate?
Thanks for all advices,
Vitaliy
|
|
|
|
|
|
|
|
|
Contact | Electronic Portal
|
|
|
