I built a circuit to deliver a 40mA pulse on top of a 10mA DC offset.
The 40 mA pulse is for 100uS.R4 & C1 set the approximate 100uS time
constant. I'm thinking there must be a way to rapidly discharge C1
when the switch opens using fewer components then R11, X1, and Q6 plus
an RC filter for the SCR gate. I tried just a PNP still takes
milliseconds same with a JFET. This current scheme is the quickest
about 20uS to bring VC1 down enough to turn Q2 off.

The schematic is here.

http://i8.tinypic.com/6lawo5d.png

Any suggestions?

"Hammy" <spamme@hotmail.com> wrote in message
news:c1f5o395vhcv0rcg0ilc9ftkgl5cmkpva7@4ax.com...
>
> I built a circuit to deliver a 40mA pulse on top of a 10mA DC offset.
> The 40 mA pulse is for 100uS.R4 & C1 set the approximate 100uS time
> constant. I'm thinking there must be a way to rapidly discharge C1
> when the switch opens using fewer components then R11, X1, and Q6 plus
> an RC filter for the SCR gate. I tried just a PNP still takes
> milliseconds same with a JFET. This current scheme is the quickest
> about 20uS to bring VC1 down enough to turn Q2 off.
>
> The schematic is here.
>
> http://i8.tinypic.com/6lawo5d.png
>
> Any suggestions?

Transistor Q6 should have a turn-on resistor (Q6 base to ground), instead of
having the load turn it on, when the switch is released.

Brian Ellis

Hammy wrote:
> I built a circuit to deliver a 40mA pulse on top of a 10mA DC offset.
> The 40 mA pulse is for 100uS.R4 & C1 set the approximate 100uS time
> constant. I'm thinking there must be a way to rapidly discharge C1
> when the switch opens using fewer components then R11, X1, and Q6 plus
> an RC filter for the SCR gate. I tried just a PNP still takes
> milliseconds same with a JFET. This current scheme is the quickest
> about 20uS to bring VC1 down enough to turn Q2 off.
>
> The schematic is here.
>
> http://i8.tinypic.com/6lawo5d.png
>
> Any suggestions?

Use a 555 one-shot to contol your output transistor.
Replace everything to the left of Q5 & its resistors
with the one-shot.

Ed

On Tue, 08 Jan 2008 19:38:26 GMT, ehsjr <ehsjr@bellatlantic.net>
wrote:


>Use a 555 one-shot to contol your output transistor.
>Replace everything to the left of Q5 & its resistors
>with the one-shot.
>
>Ed

Yes I guess that's what I'll do.

I was trying to do it on the cheap with penny transistors and thick
film resistors. The SCR at half the price of a TLC555 kind of defeats
that purpose though.

Thanks all for the suggestions

I thought I'd give it one more try.

I used a jfet P-channel to discharge the cap at switch off. It drops
the cap voltage to zero in 230uS not spectacular but good enough. Load
current drops almost instantaneously. I tried a JFET before and it
wasn't fully discharging the cap. I must have had a wire crossed
somewhere because it works well now. With a lower RDSON JFET I could
improve that but for 23 cents for the FET I can't complain.

New schematic.

http://i18.tinypic.com/6z8fygx.png

The load is really a photo mosfet driver not two 1n4001's, R10 is for
current sampling. So the total parts count is 11 and price about 50
cents.

Thanks again

"Hammy" <spamme@hotmail.com> wrote in message
news:bob8o31dp206isnsehfbpr6r8ht7kvdlub@4ax.com...
>
> I thought I'd give it one more try.
>
> I used a jfet P-channel to discharge the cap at switch off. It drops
> the cap voltage to zero in 230uS not spectacular but good enough. Load
> current drops almost instantaneously. I tried a JFET before and it
> wasn't fully discharging the cap. I must have had a wire crossed
> somewhere because it works well now. With a lower RDSON JFET I could
> improve that but for 23 cents for the FET I can't complain.
>
> New schematic.
>
> http://i18.tinypic.com/6z8fygx.png
>
> The load is really a photo mosfet driver not two 1n4001's, R10 is for
> current sampling. So the total parts count is 11 and price about 50
> cents.
>
> Thanks again

It looks like J1 is a dead short to ground. You probably want an N-MOSFET
instead, and you'll need a resistor to pull the gate to ground when the
switch closes.

Regards,
Bob Monsen

On Wed, 09 Jan 2008 06:43:55 GMT, "Bob Monsen" <rcmonsen@gmail.com>
wrote:

>It looks like J1 is a dead short to ground. You probably want an N-MOSFET
>instead, and you'll need a resistor to pull the gate to ground when the
>switch closes.
>
>Regards,
> Bob Monsen
>
>
No that's a P-channel JFET. It will only conduct when the switch is
open Vgsoff (max) =4V.When the switch is closed the JFET is off. When
the switch opens it provides a discharge path for the capacitor
RDSON*C. Once the cap discharges the JFET drain is at zero volts, the
switch is open so there is nothing for it to conduct. RDSON for the
J176 is 250 ohms.

I didn't take direct current measurement but there is no need. I just
measured the rise time of the cap which is 750uS through the 820 Ohm
R4. The fall time is roughly 1/3 that so RDSON must be roughly 1/3 R4.
It just gets warm (not hot) after repeated switching for testing.

"Hammy" <spamme@hotmail.com> wrote in message
news:g5b9o3d1o3r75p5jcobsmnajcjkof251df@4ax.com...
> On Wed, 09 Jan 2008 06:43:55 GMT, "Bob Monsen" <rcmonsen@gmail.com>
> wrote:
>
>>It looks like J1 is a dead short to ground. You probably want an N-MOSFET
>>instead, and you'll need a resistor to pull the gate to ground when the
>>switch closes.
>>
>>Regards,
>> Bob Monsen
>>
>>
> No that's a P-channel JFET. It will only conduct when the switch is
> open Vgsoff (max) =4V.When the switch is closed the JFET is off. When
> the switch opens it provides a discharge path for the capacitor
> RDSON*C. Once the cap discharges the JFET drain is at zero volts, the
> switch is open so there is nothing for it to conduct. RDSON for the
> J176 is 250 ohms.
>
> I didn't take direct current measurement but there is no need. I just
> measured the rise time of the cap which is 750uS through the 820 Ohm
> R4. The fall time is roughly 1/3 that so RDSON must be roughly 1/3 R4.
> It just gets warm (not hot) after repeated switching for testing.

Odd, I must have looked at it wrong before. Sorry about that.

Regards,
Bob Monsen