Fader/Dimmer

[John Fields]

On Wed, 31 Aug 2005 04:17:33 GMT, Robert <hedrock@insightbb.com>
wrote:

On Tue, 30 Aug 2005 18:34:40 -0500, "DBLEXPOSURE"
celstuff@hotmail.com> wrote:

He said the 470uf cap wasn't needed.

---
Originally, the plan was for the cap to act like a differentiator
when the supply was turned on and send a positive going pulse into
the base of Q1 to turn it off and then to let it charge gradually,
pulling the base of Q1 down to ground gardually, thus turning Q1 on
gradually. Unfortunately, to get a time constant of, say, one
second, the cap would have to have a capacitance of:

T 1s
C = --- = ------ = 1.96E-6F ~ 2000ľF
R 510R

Not bad, but the hitch is that the base only has to go about 0.7V
more negative than the emitter in order to turn the transistor on,
and as soon as it even _starts_ turning on, current will be flowing
into the cap from the collector-emitter diode, further shortening
the turn-on time. In short, unless you really use a huge cap, it
won't make much difference what you put in there. Here's a solution
that works:



+18>--> | Q1
|S1 PNP
O--+------E C------+----> >---+
|R1 B | |
[100k] | | |
| C Q2 |+C2 |Rl
+------B 2N4401 [4700ľF] [50R]
| E | |
|+C1 |R2 | |
[47ľF] [510R] | |
| | | |
GND>-------+--------+--------+----> >---+

You can vary the length of the turn-on by changing either R1 or C1,
and the length of the turn-off with C2.
As shown, your turn-on ramp will be about half a second long.

--
John Fields
Professional Circuit Designer

 

[Robert]

Beautiful! It works great. Many thanks.

---
Originally, the plan was for the cap to act like a differentiator
when the supply was turned on and send a positive going pulse into
the base of Q1 to turn it off and then to let it charge gradually,
pulling the base of Q1 down to ground gardually, thus turning Q1 on
gradually. Unfortunately, to get a time constant of, say, one
second, the cap would have to have a capacitance of:

T 1s
C = --- = ------ = 1.96E-6F ~ 2000ľF
R 510R

Not bad, but the hitch is that the base only has to go about 0.7V
more negative than the emitter in order to turn the transistor on,
and as soon as it even _starts_ turning on, current will be flowing
into the cap from the collector-emitter diode, further shortening
the turn-on time. In short, unless you really use a huge cap, it
won't make much difference what you put in there. Here's a solution
that works:



+18>--> | Q1
|S1 PNP
O--+------E C------+----> >---+
|R1 B | |
[100k] | | |
| C Q2 |+C2 |Rl
+------B 2N4401 [4700ľF] [50R]
| E | |
|+C1 |R2 | |
[47ľF] [510R] | |
| | | |
GND>-------+--------+--------+----> >---+

You can vary the length of the turn-on by changing either R1 or C1,
and the length of the turn-off with C2.
As shown, your turn-on ramp will be about half a second long.

 

[Jasen Betts]

In article <ouo9h1lkfclg97nken75k601orplporho8@4ax.com>, Robert wrote:

Thanks again for the help. Finally got the time and parts to try it.
The fade out is great, but it doesn't fade on. If anything, there
might be a very slight delay before it comes on at full brightness. I
can't tell for sure, but not the fade I get when I power down. Any
more thoughts?

On Thu, 25 Aug 2005 05:53:09 -0500, John Fields

+18>--> | Q1
|S1 PNP
O--+------E C---+----> >---+
|+ B | |
[470ľF] | | |
| | |+ |
+--------+ [4700ľF] [50R]
| | |
[510R] | |
| | |
GND>-------+--------------+----> >---+

Q1
PNP
18V (+)------E C---+---(+) LOAD
B |
| |
| |+
| [4700ľF]
| |
[510R] |
| |
GND>-----------+-----+---(-) LOAD

putting that 470uF capacitor (or a larger one) back in will help
with the startup delay. as will increasing the resistor, but picking the
right resistor varies with the transistor you have and the operating
conditions...


longer delays can be most conveniently had by using a second
transistor to increase the sensitivity of the first and using a
larger timing resistor

a second resistor to discharge the capacitor will also help,


+18>--> | Q1
|S1 PNP
O--+----+-----E C--------+----> >---+
|+ | B | |
[470ľF] | | Q2 PNP | |
| [33K] E C------+ |
| | B | |
+----+---------+ | [50R]
| |+ |
[33K] [4700uF] |
| | |
GND>-------+-----------------------+----> >---+

something lika a 2N3638A looks like a good choice for Q2

Bye.
Jasen