 |
Search Sci.Electronics.Basics |
|
|
|
|
|
|
 |
|
|
Sci.Electronics.Basics -> Voltage followers with outputs connected
There are 9 messages in this thread.
You are currently looking at messages 1 to 9.
|
Author: MRWDate: 10:16 04-01-07
|
|
Happy New Year!
What happens if I connect the outputs of four opamps configured as
voltage followers?
I am not sure if I'm applying the right analysis, but if I use
superposition and short out the output of the other voltage followers,
I get 0V, but if I run it on SPICE, it gives me the value of the lowest
output voltage (I am modeling the input voltage for each follower as
different voltages that are up to 100mV different from 3V).
Also, will output currents from the other opamps be a concern with this
setup?
Thanks!
|
|
|
|
Author: Costas VlachosDate: 10:43 04-01-07
|
|
MRW wrote:
> Happy New Year!
>
> What happens if I connect the outputs of four opamps configured as
> voltage followers?
>
> I am not sure if I'm applying the right analysis, but if I use
> superposition and short out the output of the other voltage followers,
> I get 0V, but if I run it on SPICE, it gives me the value of the lowest
> output voltage (I am modeling the input voltage for each follower as
> different voltages that are up to 100mV different from 3V).
>
> Also, will output currents from the other opamps be a concern with this
> setup?
>
> Thanks!
You can't connect them together like this! Op-amp outputs have very low
impedance and directly connecting two or more together is equivalent to
a short circuit. Each voltage follower will "compete" with the others,
trying to bring its output to the same level at its (+) input, resulting
in high currents flowing from one output to the other. Don't do it.
In your SPICE simulation, try to measure the output currents of the
op-amps, and note their magnitude and direction.
--
Regards,
Costas
_________________________________________________
Costas Vlachos Email: c-X-vlachos@hot-X-mail.com
SPAM-TRAPPED: Please remove "-X-" before replying
|
|
|
|
Author: MRWDate: 13:37 04-01-07
|
|
Thanks Costas!
Costas Vlachos wrote:
> MRW wrote:
> > Happy New Year!
> >
> > What happens if I connect the outputs of four opamps configured as
> > voltage followers?
> >
> > I am not sure if I'm applying the right analysis, but if I use
> > superposition and short out the output of the other voltage followers,
> > I get 0V, but if I run it on SPICE, it gives me the value of the lowest
> > output voltage (I am modeling the input voltage for each follower as
> > different voltages that are up to 100mV different from 3V).
> >
> > Also, will output currents from the other opamps be a concern with this
> > setup?
> >
> > Thanks!
>
> You can't connect them together like this! Op-amp outputs have very low
> impedance and directly connecting two or more together is equivalent to
> a short circuit. Each voltage follower will "compete" with the others,
> trying to bring its output to the same level at its (+) input, resulting
> in high currents flowing from one output to the other. Don't do it.
>
> In your SPICE simulation, try to measure the output currents of the
> op-amps, and note their magnitude and direction.
>
> --
> Regards,
> Costas
> _________________________________________________
> Costas Vlachos Email: c-X-vlachos@hot-X-mail.com
> SPAM-TRAPPED: Please remove "-X-" before replying
|
|
|
|
Author: John LarkinDate: 15:41 04-01-07
|
|
On 4 Jan 2007 10:37:41 -0800, "MRW" <mr.whatever@gmail.com> wrote:
>Thanks Costas!
>
>Costas Vlachos wrote:
>> MRW wrote:
>> > Happy New Year!
>> >
>> > What happens if I connect the outputs of four opamps configured as
>> > voltage followers?
>> >
>> > I am not sure if I'm applying the right analysis, but if I use
>> > superposition and short out the output of the other voltage followers,
>> > I get 0V, but if I run it on SPICE, it gives me the value of the lowest
>> > output voltage (I am modeling the input voltage for each follower as
>> > different voltages that are up to 100mV different from 3V).
>> >
>> > Also, will output currents from the other opamps be a concern with this
>> > setup?
>> >
>> > Thanks!
>>
>> You can't connect them together like this! Op-amp outputs have very low
>> impedance and directly connecting two or more together is equivalent to
>> a short circuit. Each voltage follower will "compete" with the
others,
>> trying to bring its output to the same level at its (+) input, resulting
>> in high currents flowing from one output to the other. Don't do it.
>>
>> In your SPICE simulation, try to measure the output currents of the
>> op-amps, and note their magnitude and direction.
That's provided the opamps are different! Spice likes to make them all
the same. Absolutely identical opamps will share the load politely, in
some simulated universe.
You can connect multiple followers to a common output node if you put
a small resistor in series with each one. If the max opamp offsets
were +-1 mV, and you used 2 ohm resistors, the maximum possible
opamp-fighting-opamp current would be 1 ma, which might be tolerable.
John
|
|
|
|
Author: Costas VlachosDate: 19:25 04-01-07
|
|
John Larkin wrote:
> On 4 Jan 2007 10:37:41 -0800, "MRW" <mr.whatever@gmail.com> wrote:
>
>> Thanks Costas!
>>
>> Costas Vlachos wrote:
>>> MRW wrote:
>>>> Happy New Year!
>>>>
>>>> What happens if I connect the outputs of four opamps configured as
>>>> voltage followers?
>>>>
>>>> I am not sure if I'm applying the right analysis, but if I use
>>>> superposition and short out the output of the other voltage followers,
>>>> I get 0V, but if I run it on SPICE, it gives me the value of the lowest
>>>> output voltage (I am modeling the input voltage for each follower as
>>>> different voltages that are up to 100mV different from 3V).
>>>>
>>>> Also, will output currents from the other opamps be a concern with this
>>>> setup?
>>>>
>>>> Thanks!
>>> You can't connect them together like this! Op-amp outputs have very low
>>> impedance and directly connecting two or more together is equivalent to
>>> a short circuit. Each voltage follower will "compete" with the
others,
>>> trying to bring its output to the same level at its (+) input, resulting
>>> in high currents flowing from one output to the other. Don't do it.
>>>
>>> In your SPICE simulation, try to measure the output currents of the
>>> op-amps, and note their magnitude and direction.
>
> That's provided the opamps are different! Spice likes to make them all
> the same. Absolutely identical opamps will share the load politely, in
> some simulated universe.
>
>
> You can connect multiple followers to a common output node if you put
> a small resistor in series with each one. If the max opamp offsets
> were +-1 mV, and you used 2 ohm resistors, the maximum possible
> opamp-fighting-opamp current would be 1 ma, which might be tolerable.
>
> John
If the followers share the same input then yes you can do that. But the
OP mentions about 100mV max. input difference between followers... What
use would that circuit have in practice? If he wants to increase output
current capability, why the 100mV input difference? Just use a push-pull
or AoE's op-amp-only circuit idea (AoE, 2nd Ed., p253).
--
Regards,
Costas
_________________________________________________
Costas Vlachos Email: c-X-vlachos@hot-X-mail.com
SPAM-TRAPPED: Please remove "-X-" before replying
|
|
|
|
Author: EeyoreDate: 20:44 04-01-07
|
|
MRW wrote:
> Happy New Year!
>
> What happens if I connect the outputs of four opamps configured as
> voltage followers?
They'll have an argument with each other.
See 'offset voltage'. You'll need to use combining resistors. Is your intention
to obtain a greater output current ?
Graham
|
|
|
|
Author: John LarkinDate: 20:49 04-01-07
|
|
On Fri, 05 Jan 2007 02:25:35 +0200, Costas Vlachos
<c-X-vlachos@hot-X-mail.com> wrote:
>John Larkin wrote:
>> On 4 Jan 2007 10:37:41 -0800, "MRW" <mr.whatever@gmail.com>
wrote:
>>
>>> Thanks Costas!
>>>
>>> Costas Vlachos wrote:
>>>> MRW wrote:
>>>>> Happy New Year!
>>>>>
>>>>> What happens if I connect the outputs of four opamps configured as
>>>>> voltage followers?
>>>>>
>>>>> I am not sure if I'm applying the right analysis, but if I use
>>>>> superposition and short out the output of the other voltage
followers,
>>>>> I get 0V, but if I run it on SPICE, it gives me the value of the
lowest
>>>>> output voltage (I am modeling the input voltage for each follower
as
>>>>> different voltages that are up to 100mV different from 3V).
>>>>>
>>>>> Also, will output currents from the other opamps be a concern with
this
>>>>> setup?
>>>>>
>>>>> Thanks!
>>>> You can't connect them together like this! Op-amp outputs have very low
>>>> impedance and directly connecting two or more together is equivalent to
>>>> a short circuit. Each voltage follower will "compete" with
the others,
>>>> trying to bring its output to the same level at its (+) input,
resulting
>>>> in high currents flowing from one output to the other. Don't do it.
>>>>
>>>> In your SPICE simulation, try to measure the output currents of the
>>>> op-amps, and note their magnitude and direction.
>>
>> That's provided the opamps are different! Spice likes to make them all
>> the same. Absolutely identical opamps will share the load politely, in
>> some simulated universe.
>>
>>
>> You can connect multiple followers to a common output node if you put
>> a small resistor in series with each one. If the max opamp offsets
>> were +-1 mV, and you used 2 ohm resistors, the maximum possible
>> opamp-fighting-opamp current would be 1 ma, which might be tolerable.
>>
>> John
>
>If the followers share the same input then yes you can do that. But the
>OP mentions about 100mV max. input difference between followers... What
>use would that circuit have in practice? If he wants to increase output
>current capability, why the 100mV input difference? Just use a push-pull
> or AoE's op-amp-only circuit idea (AoE, 2nd Ed., p253).
I've done that: four opamps, each with a gain of about +4, using 1%
resistors to set the gains so the outputs match pretty well. Each then
has a 200 ohm resistor to the output. The result is a nice, fast 50
ohm drive.
John
|
|
|
|
Author: EeyoreDate: 22:39 04-01-07
|
|
John Larkin wrote:
> Costas Vlachos <c-X-vlachos@hot-X-mail.com> wrote:
> >
> >If the followers share the same input then yes you can do that. But the
> >OP mentions about 100mV max. input difference between followers... What
> >use would that circuit have in practice? If he wants to increase output
> >current capability, why the 100mV input difference? Just use a push-pull
> > or AoE's op-amp-only circuit idea (AoE, 2nd Ed., p253).
>
> I've done that: four opamps, each with a gain of about +4, using 1%
> resistors to set the gains so the outputs match pretty well. Each then
> has a 200 ohm resistor to the output. The result is a nice, fast 50
> ohm drive.
Was there a specific reason you didn't use one amp as a x4 stage and use the
others as followers ? Concerned about the HF characteristics presumably ?
Graham
|
|
|
|
Author: John LarkinDate: 00:39 05-01-07
|
|
On Fri, 05 Jan 2007 03:39:15 +0000, Eeyore
<rabbitsfriendsandrelations@hotmail.com> wrote:
>
>
>John Larkin wrote:
>
>> Costas Vlachos <c-X-vlachos@hot-X-mail.com> wrote:
>> >
>> >If the followers share the same input then yes you can do that. But the
>> >OP mentions about 100mV max. input difference between followers... What
>> >use would that circuit have in practice? If he wants to increase output
>> >current capability, why the 100mV input difference? Just use a push-pull
>> > or AoE's op-amp-only circuit idea (AoE, 2nd Ed., p253).
>>
>> I've done that: four opamps, each with a gain of about +4, using 1%
>> resistors to set the gains so the outputs match pretty well. Each then
>> has a 200 ohm resistor to the output. The result is a nice, fast 50
>> ohm drive.
>
>Was there a specific reason you didn't use one amp as a x4 stage and use the
>others as followers ? Concerned about the HF characteristics presumably ?
>
>Graham
I don't recall exactly, but I think it was an input common-mode-range
issue. I was trying to get a clean, very fast +-10 volt drive with
just +-12 volt supplies, and couldn't find a single opamp that would
work. They were THS3062's, maybe.
Saves one opamp, too!
John
|
|
|
|
1 |
|
|
|
|
Contact | Electronic Portal
|
|
|
