Design Help 2 - Please NO PIC

The relay driver transistor should be an npn too.

John

as a few

delay

stor

ure

the end

lay is

pect

a when

of

ll the

Isn't that going to reverse the on off?

The pnp, as drawn, will never turn on, because it needs negative base drive and there's none available. If you want the relay on when pin 3 is high, the driver should be an npn.

And you could eliminate two more parts, the 10K and its diode. Just let the first transistor discharge the cap.

Or replace the 555 with a simple schmitt trigger inverter.

John

e 555

has a few

tty

f delay

sistor

The

ssure

at the end

relay is

uspect

me of

all the

No. If you really had a conventional PNP output transistor in the circuit as your drawing shows, the collector base diode would be forward biased and pull the base high enough to reverse bias the base- emitter junction into avalanche breakdown, which would probably have destroyed the transistor.

For the long delay time that you want, a CMOS version of the 555 might be attractive.

I hope your 100uF electrolytic capacitor is a tantalum device - back in the 1970's, when I would with these sorts of time constants, tantalum electrolytic capacitors offered much lowr and more stable leakage currents than aluminium-based devices.

-- Bill Sloman, Nijmegen

s a few

elay

tor

re

the end

ay is

ect

when

of

l the

These might help:

Have fun,

Michael

he 555

t has a few

etty

of delay

nsistor

=A0The

essure

at the end

relay is

suspect

ome of

r all the

Guess I'd better check my parts again because I have something on the breadboard that works. I thought I had a pnp there but maybe I forgot what I put there. I'd better look at the part number on the transistor in question :-)

I'll try it without the 10k and the diode. Easy enough to pull them off. I'm assuming you mean on the output of the first transistor.

Working with what I have in the parts bin at the moment, 555s,

few

end

suspect

when

the

Not with the 1K in the way. There might be a very slow beta degradation, which is irrelevant in a circuit that doesn't work anyhow.

Common aluminums have self-discharge time constants measured in hours or days, sometimes weeks. Try it.

John

few

end

suspect

when

the

This guy integrates 10nA on a 4,700uF electrolytic to drive a motor:

Cheers, James Arthur

For the long delay time that you want, a CMOS version of the 555 might be attractive.

But it does not have the drive capacity of the bipolar device, so you need a sensitive relay or a transistor driver.

I hope your 100uF electrolytic capacitor is a tantalum device - back in the 1970's, when I would with these sorts of time constants, tantalum electrolytic capacitors offered much lowr and more stable leakage currents than aluminium-based devices.

You can get aluminum electrolytics with very low leakage and high stability for much lower cost than tantalum.

Paul

few

delay

pressure

end

suspect

when

the

Yup, note his capacitor quality check: no significant self-discharge after "a few days."

John

few

delay

pressure

end

suspect

when

the

No, now I see what you mean. The c-b-e path could conduct through the relay coil in addition to the base drive through the 1K.

That might zap the transistor, but pnp's seem to often have higher b-e zener voltages than npn's, so the current may be low. Again, it doesn't matter much if you damage a transistor that doesn't work anyhow.

John

the 555

=A0It has a few

pretty

s of delay

ransistor

=A0The

pressure

or at the end

he relay is

I suspect

t 14ma when

some of

for all the

"Common aluminums have self-discharge time constants measured in hours or days, sometimes weeks. Try it."

I will, I always thought that the aluminums had a high leakage current. I've got a Nichicon HE series 1000uF and 50V. The leakage current is listed as 500uA. This would imply a 'resistance' of 100 k ohm and a time constant of 100 seconds.

OK in 20 minutes the voltage has dropped by less than 10%! (And typically there is a 'fast droop in voltage after the initial charge.) Will this get worse with age? Why such a conservative spec if they are much better than this?

George Herold

s a few

elay

tor

re

the end

ay is

ect

when

of

l the

I suggest you don't need a transistor to invert the trigger pulse. Just connect the switch to V+ and 2 resistors to create a voltage divider to ground. Thy can be quite large as the trigger current of the 555 is very low. The midpoint of the divider should be at the specified trigger level (about 5 volts I think). When he switch opens the resistor to ground pulls the trigger line low. The NE555 should be able to run a 150 mA relay directly (200 mA drive capability).

Cheers

Cheers

Cheers

You evidently missed the part where the switch has one terminal accessible, and "ground" is bolted to an engine block.

Hope This Helps!

I've seen his latest 555 rendition, and I think I like it. ;-)

Hey, OP! Does it work?

Cheers! Rich

Because they're NOT much better than this at their maximum rated temperature, which is sometimes what they're run at. If it's rated to

Here it is:

Leakage current is .01CV = .01 * .0001F * 50V = 500uA. Leakage current is guaranteed to be no more than that over the temperature range of

-40C to +105C.

I heard Bob Pease speak a few years ago, he said there are engineers who call him (or National) to solve problems - they were getting bit because they designed based on what's in the "typical values" column of data sheets. He said you should always design based on the worst case figures, as they're the only ones that are guaranteed. I was shocked, shocked I tell you! to learn that engineers would design based on anything other than worst-case specifications.

Remember this leakage current thing when this thing is in the engine compartment of your idling car in a traffic jam on a hot summer day.

the 555

=A0It has a few

pretty

s of delay

ransistor

=A0The

pressure

or at the end

he relay is

I suspect

t 14ma when

some of

for all the

We did, back in the early 1970's at Kent Instruments in Luton, England. Tantalum was quite a bit better. At that time, I knew the price of every transistor I was designing in, and we wouldn't have paid the extra for tantalum capacitors if they hadn't been worth the money.

-- Bill Sloman, Nijmegen

Thank you. It does work :-)

ec

Ahh high temperatures, thanks Ben. I'll try a bit of heat on the cap and see if that makes it discharge faster. BTW I left the cap discharging through a 1G ohm resistor over night and came in to find the voltage had dropped from 18.3V at 5:30 pm to 15.6V at 8:30 AM. It's nice to know you can make a cheap RC time constant of a day or more if you ever need such a thing. A few years ago I 'designed' (stole from H and H) a variable ramp generator with periods from 1 ms to 1000 seconds. I used a 100uF tantalum for the long times. I looked at the leakage specs on the aluminum caps and never even thought to try them in the circuit.

George Herold