If I am using a voltage regulator to bring a voltage down from 9 to 5
volts should I still put a capacitor before a TTL IC to smooth out the
current?

D...@gmail.com wrote:
> If I am using a voltage regulator to bring a voltage down from 9 to 5
> volts should I still put a capacitor before a TTL IC to smooth out the
> current?
> 

The short answer is yes.

You don't tell us if you have only this one IC.

If you have a regulator and a single IC, then you should have the bulk 
output capacitor on the regulator and a 0.01 - 0.1uF cap *at the power 
pins* of the TTL device.

Cheers

PeteS

D...@gmail.com wrote:
> If I am using a voltage regulator to bring a voltage down from 9 to 5
> volts should I still put a capacitor before a TTL IC to smooth out the
> current?

It depends.  Your regulator should have a cap at the output.  If your
IC is a short distance (a couple of inches) from the regulator, then
no.  But if you've got a long run, the trace or wire has inductance,
which will produce a transient on the Vcc line.  Remember that Vcc for
the IC is the voltage at the pin.  In that event, you should use a cap.

Good luck
Chris

D...@gmail.com wrote:

> If I am using a voltage regulator to bring a voltage down from 9 to 5
> volts should I still put a capacitor before a TTL IC to smooth out the
> current?

It's not really to 'smooth out the current' so much as to provide a low
impedance path for switching currents.

You don't really need one on *every* IC though. Just make sure each chip is near
to a decoupling cap.

Graham

On Sat, 28 Oct 2006 20:53:35 +0100, Eeyore
<r...@hotmail.com> wrote:

>
>
>D...@gmail.com wrote:
>
>> If I am using a voltage regulator to bring a voltage down from 9 to 5
>> volts should I still put a capacitor before a TTL IC to smooth out the
>> current?
>
>It's not really to 'smooth out the current' so much as to provide a low
>impedance path for switching currents.
>
>You don't really need one on *every* IC though. Just make sure each chip is near
>to a decoupling cap.

---
Put one on _every_ IC, and as close to the IC as you can unless you
have some overriding reason not to and you know what you're doing.
Decoupling caps are cheap.  Trying to find out why your circuit
doesn't work because you followed Graham's advice isn't.


-- 
John Fields
Professional Circuit Designer 

John Fields wrote:
> On Sat, 28 Oct 2006 20:53:35 +0100, Eeyore
> <r...@hotmail.com> wrote:
> 
> 
>>
>>D...@gmail.com wrote:
>>
>>
>>>If I am using a voltage regulator to bring a voltage down from 9 to 5
>>>volts should I still put a capacitor before a TTL IC to smooth out the
>>>current?
>>
>>It's not really to 'smooth out the current' so much as to provide a low
>>impedance path for switching currents.
>>
>>You don't really need one on *every* IC though. Just make sure each chip is near
>>to a decoupling cap.
> 
> 
> ---
> Put one on _every_ IC, and as close to the IC as you can unless you
> have some overriding reason not to and you know what you're doing.
> Decoupling caps are cheap.  Trying to find out why your circuit
> doesn't work because you followed Graham's advice isn't.
> 
> 

Although I generally agree with the rule of putting a cap per device 
unless there's some reason not to, this crap between you two is getting 
old. Get a fucking room and duke it out or something.

I personally don't give a rats ass what your (or whose) problem is, but 
it gets in the way of serious discussions. On those threads where there 
is no serious discussion, no big deal. On those threads where there is, 
both of you have something to contribute but poison it with this crap. 
I'd like to think both of you are adults. If one of you raises above the 
fray, then at least I can simply plonk the other.

Cheers

PeteS

PeteS wrote:

> John Fields wrote:
> > Eeyore <r...@hotmail.com> wrote:
> >>D...@gmail.com wrote:
> >>
> >>
> >>>If I am using a voltage regulator to bring a voltage down from 9 to 5
> >>>volts should I still put a capacitor before a TTL IC to smooth out the
> >>>current?
> >>
> >>It's not really to 'smooth out the current' so much as to provide a low
> >>impedance path for switching currents.
> >>
> >>You don't really need one on *every* IC though. Just make sure each chip is near
> >>to a decoupling cap.
> >
> > ---
> > Put one on _every_ IC, and as close to the IC as you can unless you
> > have some overriding reason not to and you know what you're doing.
> > Decoupling caps are cheap.  Trying to find out why your circuit
> > doesn't work because you followed Graham's advice isn't.
> >
> >
>
> Although I generally agree with the rule of putting a cap per device
> unless there's some reason not to, this crap between you two is getting
> old. Get a fucking room and duke it out or something.
>
> I personally don't give a rats ass what your (or whose) problem is, but
> it gets in the way of serious discussions. On those threads where there
> is no serious discussion, no big deal. On those threads where there is,
> both of you have something to contribute but poison it with this crap.
> I'd like to think both of you are adults. If one of you raises above the
> fray, then at least I can simply plonk the other.

Sorry about that.

It seems that JF wishes to stalk me.

I stand by my advice that you can 'distribute' decoupling caps on a typical pcb to the
extent of ~ 3 ICs per cap typically provided the ICs are close together ( trace
inductance is the issue here )..


Graham

On Sat, 28 Oct 2006 22:34:42 GMT, PeteS
<p...@ntlworld.com> wrote:

>John Fields wrote:
>> On Sat, 28 Oct 2006 20:53:35 +0100, Eeyore
>> <r...@hotmail.com> wrote:
>> 
>> 
>>>
>>>D...@gmail.com wrote:
>>>
>>>
>>>>If I am using a voltage regulator to bring a voltage down from 9 to 5
>>>>volts should I still put a capacitor before a TTL IC to smooth out the
>>>>current?
>>>
>>>It's not really to 'smooth out the current' so much as to provide a low
>>>impedance path for switching currents.
>>>
>>>You don't really need one on *every* IC though. Just make sure each chip is near
>>>to a decoupling cap.
>> 
>> 
>> ---
>> Put one on _every_ IC, and as close to the IC as you can unless you
>> have some overriding reason not to and you know what you're doing.
>> Decoupling caps are cheap.  Trying to find out why your circuit
>> doesn't work because you followed Graham's advice isn't.
>> 
>> 
>
>Although I generally agree with the rule of putting a cap per device 
>unless there's some reason not to, this crap between you two is getting 
>old. Get a fucking room and duke it out or something.
>
>I personally don't give a rats ass what your (or whose) problem is, but 
>it gets in the way of serious discussions. On those threads where there 
>is no serious discussion, no big deal. On those threads where there is, 
>both of you have something to contribute but poison it with this crap. 
>I'd like to think both of you are adults. If one of you raises above the 
>fray, then at least I can simply plonk the other.

---
Well, if I find that someone makes a technical error I'll choose how
to address the correction of the error as well as whether or how to
comment on the maker of the error.

If you find that offensive I apologize, and I'll apologize now for
any other unpleasantness you may experience from my posts in the
future.

However, since I can't guarantee that I'll "raise above the fray",
as you put it, I suggest you just plonk me and be done with it. :-)


-- 
John Fields
Professional Circuit Designer 

On Sun, 29 Oct 2006 01:48:13 +0000, Eeyore
<r...@hotmail.com> wrote:

>
>
>PeteS wrote:
>
>> John Fields wrote:
>> > Eeyore <r...@hotmail.com> wrote:
>> >>D...@gmail.com wrote:
>> >>
>> >>
>> >>>If I am using a voltage regulator to bring a voltage down from 9 to 5
>> >>>volts should I still put a capacitor before a TTL IC to smooth out the
>> >>>current?
>> >>
>> >>It's not really to 'smooth out the current' so much as to provide a low
>> >>impedance path for switching currents.
>> >>
>> >>You don't really need one on *every* IC though. Just make sure each chip is near
>> >>to a decoupling cap.
>> >
>> > ---
>> > Put one on _every_ IC, and as close to the IC as you can unless you
>> > have some overriding reason not to and you know what you're doing.
>> > Decoupling caps are cheap.  Trying to find out why your circuit
>> > doesn't work because you followed Graham's advice isn't.
>> >
>> >
>>
>> Although I generally agree with the rule of putting a cap per device
>> unless there's some reason not to, this crap between you two is getting
>> old. Get a fucking room and duke it out or something.
>>
>> I personally don't give a rats ass what your (or whose) problem is, but
>> it gets in the way of serious discussions. On those threads where there
>> is no serious discussion, no big deal. On those threads where there is,
>> both of you have something to contribute but poison it with this crap.
>> I'd like to think both of you are adults. If one of you raises above the
>> fray, then at least I can simply plonk the other.
>
>Sorry about that.
>
>It seems that JF wishes to stalk me.
>
>I stand by my advice that you can 'distribute' decoupling caps on a typical pcb to the
>extent of ~ 3 ICs per cap typically provided the ICs are close together ( trace
>inductance is the issue here )..

---
And I'll stand by mine, which is to stay on the conservative side to
avoid problems.  Once the design is done and the PCB layout
finalized, _then_ the necessity of cap-per-chip can be determined
and caps omitted where appropriate if the design is cost-sensitive.

 
-- 
John Fields
Professional Circuit Designer