# How does one determine the capacitance of a coupling cap?

Status
Not open for further replies.

#### potoole64

##### New Member
If I want to transfer a pulsing (square wave centered on a positve dc voltage) output from one ckt. to another, and return the centering voltage to 0 or ground by using a coupling cap and resistor, how would I determine the size of the cap and or the resistor? If the frequence is quite low, less than 100 Htz.

#### Torben

##### Well-Known Member
If I want to transfer a pulsing (square wave centered on a positve dc voltage) output from one ckt. to another, and return the centering voltage to 0 or ground by using a coupling cap and resistor, how would I determine the size of the cap and or the resistor? If the frequence is quite low, less than 100 Htz.

Hi potoole,

Can you provide an example schematic to show what you are trying to do and a diagram of the waveforms?

It sounds to me like you are intending to capacitively couple two digital circuits. Is this correct?

Torben

#### Willbe

##### New Member
square wave centered on a positve dc voltage
return the centering voltage to 0 or ground
a coupling cap and resistor
If the frequence is quite low, less than 100 Htz.

A differentiator will not preserve the square wave shape.

A level shifter will, but the downstream waveform then becomes AC.

Last edited:

#### potoole64

##### New Member
I'm trying to couple pulses from my motorcycle ignition to ckts., such as LM2917 chip tachometer ckt, or a monostable MV. I successfully coupled from the ignition to the LM2917 (using coupling cap, but just guessed at cap size), but am not able to couple the the monostable MV, with or without coupling cap.

Torben replied
Thank you, Torben. I looked it up, and found helpful information.

The ignition pulses (173517.jpg ) are displayed on a downloaded "sound card oscilloscope" and are from 16 to 18 Mseconds in time, they are step shaped. I don't know what the volts are; sound card scope can't show that. I believe the pulses are all above ground; sound card won't show that either. Pulses were "captured" with Radio Shack sound recorder then transferred to sound card scope on my PC.

IMG.jpg shows circuitry.

Thank you
P. O'Toole

#### Attachments

• 173517.jpg
50.1 KB · Views: 137
• IMG.jpg
895.9 KB · Views: 178

#### potoole64

##### New Member
More on coupling from mc ignition.
I want to apply these ignition pulses to my Parallax Basic Stamp (BS2), and use the COUNT command to count the pulses per second. I've been partially successful at this, but pulses are not counted accurately (nor with consistancy) going directly from ignition to BS2. I used a Schmitt trigger between ignition and BS2, but the count was very innaccurate.

#### Roff

##### Well-Known Member
If the basic Stamp has a comparator or A/D input, you can probably apply the signal directly to one of those inputs to trigger your counter, if you know the input signal's amplitude. Otherwise, you should be able to use the circuit below. The circuit is a differentiator, followed by a clipper, then by a comparator which has a little hysteresis to ensure clean switching.
Be sure to ground one of the inputs of the unused comparator in the LM393 package.

#### Attachments

• tach pulse sch.PNG
17.3 KB · Views: 278
• tach pulse waves.PNG
15.4 KB · Views: 177

#### MrAl

##### Well-Known Member
Hi there,

You may even be able to get away with an NPN transistor with a large base
resistor and nominal collector resistor. The ignition signal feeds the large base
resistor, and the collector goes to your circuit. This will help catch the large
spike and not the rest of the waveform, which will be oscillatory and give very
false readings if detected. A small amount of low pass filtering might be needed
to get rid of some noise as well as possibly some ringing, but i would say first
experiment with the base resistor (in series with the base, connected to the
ignition coil primary). Raising the base resistor value helps to keep the switch
level above most of the ringing, but going too high and it will miss the entire
wave. Starting with a design goal of about 50v should get you started.
Note you dont need any 'ac coupling' with this kind of circuit.

Last edited:

#### Willbe

##### New Member
I don't know what the volts are
minimum 10 kV, probably
I believe the pulses are all above ground

The spark plug center electrode goes negative; it's hotter than the shell so the electron flow is better this way, just like a vacuum tube with cathode being negative.

#### crutschow

##### Well-Known Member
The spark plug center electrode goes negative; it's hotter than the shell so the electron flow is better this way, just like a vacuum tube with cathode being negative.
My car and motorcycle have no distributor but use coils with dual outputs (each side of the coil secondary) which fire two plugs in series simultaneously (one is firing at the top of the exhaust stroke which has no effect on the engine). Thus one plug is positive while the other is negative.

#### Willbe

##### New Member
My car and motorcycle have no distributor but use coils with dual outputs (each side of the coil secondary) which fire two plugs in series simultaneously (one is firing at the top of the exhaust stroke which has no effect on the engine). Thus one plug is positive while the other is negative.

Do they have a way to make the spark that's occurring on the power stroke always the right polarity (they'd have to reverse coil primary current to do this, e.g. with two switching transistors)?

#### potoole64

##### New Member
Mr. Al said
You may even be able to get away with an NPN transistor with a large base
resistor and nominal collector resistor
You say a large base resistor, can you give me an idea how large? More than 100k?
I'm taking the pulses from the low side of the coil, which is in the twelve volt range.
I have several transistors, old and new, that I can try out. Any idea what type I should start out with?
Thanks
PO'T

#### potoole64

##### New Member
If the basic Stamp has a comparator or A/D input, you can probably apply the signal directly to one of those inputs to trigger your counter, if you know the input signal's amplitude. Otherwise, you should be able to use the circuit below. The circuit is a differentiator, followed by a clipper, then by a comparator which has a little hysteresis to ensure clean switching.
Be sure to ground one of the inputs of the unused comparator in the LM393 package.

Thank you, ROFF, for your ckt.
Your ckt. diagram shows diodes 1n4148, which is a high speed switching diode. I have several 1n4001 type general purpose diodes. Do you think they will work ok in place of the 4148 type?
I also have an LM338 four comparator chip, which I have used for a simple A/D converter. One of the comparators in the LM338 should be the same as one in the LM393 in your ckt., do you think?

Thanks
PO'T

#### MrAl

##### Well-Known Member
Hi again,

potoole64:
You can start with 100k base resistor connected from coil to base,
and a 1k collector to +12v and an NPN transistor. You also need
a small signal diode across the base emitter with cathode to base,
anode to emitter (to clamp negative excursions). Probably a good
idea to use a small cap across the base emitter too, like 0.01uf.
If the output doesnt switch correctly lower the value of the 100k
resistor to 47k, then to 22k if need be. The emitter also must
connect to the car ground.

You can check with a 9vac transformer plugged into the 60Hz
power line at home. 60Hz is equivalent to the following rpm's
for these three types of car engines:
4 cylinder, 1800rpm
6 cylinder, 1200rpm
8 cylinder, 900rpm

If you like you can also experiment with using the input to try
to pick up the high voltage pulse energy at the ignition coil
secondary, without actually connecting it. See if the base
resistor picks up the signal when it is moved near (not touching)
the coil top terminal. This would be something to try if you
felt like experimenting a little. If it doesnt work, more
gain might do it.

Last edited:

#### Roff

##### Well-Known Member
Thank you, ROFF, for your ckt.
Your ckt. diagram shows diodes 1n4148, which is a high speed switching diode. I have several 1n4001 type general purpose diodes. Do you think they will work ok in place of the 4148 type?
I also have an LM338 four comparator chip, which I have used for a simple A/D converter. One of the comparators in the LM338 should be the same as one in the LM393 in your ckt., do you think?

Thanks
PO'T
1N4001 should work fine at the low frequency you specified.
LM338 is a voltage regulator. LM339 is a quad comparator. You should connect one input of each unused comparator to ground (circuit common).
You should probably first try MrAl's circuit, which is much simpler. My circuit was designed when I had no idea what your DC level or p-p amplitude was. I doubt that the pulses are all above ground, as you originally suggested. If they were, MrAl's circuit would need AC coupling on the input.

Last edited:

#### potoole64

##### New Member
1N4001 should work fine at the low frequency you specified.
LM338 is a voltage regulator. LM339 is a quad comparator. You should connect one input of each unused comparator to ground (circuit common).

Actually, I meant that I have a LM339

PO'T

#### Roff

##### Well-Known Member
Actually, I meant that I have a LM339

PO'T
I edited my previous post.

#### potoole64

##### New Member
These are all great answers and information. A few new things to experiment with.
Thanks everybody
PO'T

#### potoole64

##### New Member
If the basic Stamp has a comparator or A/D input, you can probably apply the signal directly to one of those inputs to trigger your counter, if you know the input signal's amplitude. Otherwise, you should be able to use the circuit below. The circuit is a differentiator, followed by a clipper, then by a comparator which has a little hysteresis to ensure clean switching.
Be sure to ground one of the inputs of the unused comparator in the LM393 package.

In the circuit that you suggested, what is the purpose of the second diode that is connected from the comparator's '+' input to ground, with the diode's anode connected to the '+' input ? [/FONT]

I was able to use the resultant positive pulses from the differenting ckt, right from the 'top' of R2, apply them directly to my BS2 and derive an appropriate pulse count. Since I'm applying the ignition pulses to a 3:1 voltage divider, the ignition pulses are dropped down to less than 4 volts. Then they are applied to the differntiator/clipper ckt with the resultant positve pulses applied to the BS2. For some reason the diffed. pulses were greatly attenuated at the input to the comparator, and were apparently too small for the comparator to sense. I'm wondering if the second diode has something to do with the sharp attenuation of those pulses. If the diffed. pulses continue to work ok, that might be all I need to use.

The original ignition pulses are 'negative' pulses that drop from the MC's 12v to ground for a few milliseconds then back to 12v, with an approximate frequency range of 17 to 75 pps

thank you
PO'T

#### Roff

##### Well-Known Member
In the circuit that you suggested, what is the purpose of the second diode that is connected from the comparator's '+' input to ground, with the diode's anode connected to the '+' input ? [/FONT]

I was able to use the resultant positive pulses from the differenting ckt, right from the 'top' of R2, apply them directly to my BS2 and derive an appropriate pulse count. Since I'm applying the ignition pulses to a 3:1 voltage divider, the ignition pulses are dropped down to less than 4 volts. Then they are applied to the differntiator/clipper ckt with the resultant positve pulses applied to the BS2. For some reason the diffed. pulses were greatly attenuated at the input to the comparator, and were apparently too small for the comparator to sense. I'm wondering if the second diode has something to do with the sharp attenuation of those pulses. If the diffed. pulses continue to work ok, that might be all I need to use.

The original ignition pulses are 'negative' pulses that drop from the MC's 12v to ground for a few milliseconds then back to 12v, with an approximate frequency range of 17 to 75 pps

thank you
PO'T
The diode in question was put there because, when I designed the circuit, you didn't know the amplitude of the input signal. The diode limits the amplitude of the pulse at the '+' input to about 0.7V, to protect the comparator from possible overvoltage. If you know that the signal at R2 is adequate to trigger your comparator, and won't subject it to overvoltage, then use that point.

Status
Not open for further replies.

Replies
1
Views
1K
Replies
20
Views
2K
Replies
8
Views
2K
Replies
18
Views
13K
Replies
7
Views
853