555 Timer? Input: sawtooth 0-8 volts, Output: pulse at given voltage 1-7 volts

[gmcjetpilot]

555 Timer? Input sawtooth 0-8 volts, Output pulse variable duration

This is what I want to do.

Input: sawtooth from 0-8 volts with freq of 30 Hz to 60 Hz

Output: a pulse when input is at a user set voltage between 1-7 volts.
The pulse output is adjustable voltage (3-9v) and duration (5uS to 2mS)

I'm flexible, it is not super critical. I just need a stable trigger even if it has some delay.

If some one can draw out a circuit I would appreciate it.
The 555 timer seems to be the way to go?

Thanks

 

[(*steve*)]

the easy way would be to use the sawtooth to generate the variable pulse width by comparing the input voltage (the sawtooth) with a variable reference voltage (using an op-amp wired as a comparator). The output of this could be passed via a pot (to attenuate the output signal) to another op-amp wired as a buffer.

The pulse width would vary with input frequency (which may or may not be what you're after)

 

[gmcjetpilot]

the easy way would be to use the sawtooth to generate the variable pulse width by comparing the input voltage (the sawtooth) with a variable reference voltage (using an op-amp wired as a comparator). The output of this could be passed via a pot (to attenuate the output signal) to another op-amp wired as a buffer.

The pulse width would vary with input frequency (which may or may not be what you're after)

I would like control over the pulse duration.
Do you have some op-amps in mind. Thanks

 

[(*steve*)]

Changing the reference voltage via a pot on the comparator would allow you to alter the pulse length (up to 1/f) for a given frequency. If the frequency changed, the pulse length would also change.

For such low frequencies, use anything cheap. 741's are probably appropriate. A +/- 9V supply would be appropriate with the input signal referenced to ground.

Note that if your sawtooth isn't perfect, you may well have multiple transitions.

 

[55pilot]

Note that if your sawtooth isn't perfect, you may well have multiple transitions.

A simple hysteresis circuit would solve that. Almost every comparator datasheet has the circuit along with the calculations.

Any reason you are recommending using an op-amp rather than a comparator like LM311?

---55p

 

[(*steve*)]

Any reason you are recommending using an op-amp rather than a comparator like LM311

Because it sounded easier A dedicated comparator would be better -- how much better may be academic.

 

[gmcjetpilot]

Thanks Steve and 55p.

Lets switch gears a little, but still doing some comparator. Lets not compare Freq but compare DC VOLTAGE..... and make a trigger right at a given voltage. The signal is now a sawtooth from 0-8 volts, with a steady DC freq of 30-60 Hz... I want to have a trigger send a pulse when the ramp is at 4 volts. May be it would be possible to trigger at three different volts along this ramp...... say 4 volts +/- 0.015 volts.... if possible. This might be a bit of a timing circuit as well. Supreme accuracy is not needed because it will be "calibrated" manually using a freq counter and scope. First the plan than the background.


** NEW PLAN **

I need to make a PULSE of about 5 volts with a short duration in the range of, 10 to 20 uS, variable by user at the right point along the sweep (0-8 volt saw tooth). The pulse duration can not be too long and some adjustable control would be nice. For narrow sweeps shorter pulse is needed. For wide sweeps wider pulse is better for identification. However since the sweep rate and voltage is the same, only one pulse duration might be fine. May be have a width and narrow setting.

The sawtooth sweep ramp is 0-8 volts and sweep typically is 30-60 Hz. It can go slower, but it is too slow to see on the scope. The voltage corresponds to a Freq along the sweep.. I just need to set some voltage that will trigger the pulse I need. It would be cool to have two or three of these pulses.


** BACKGROUND **

This is for a Sweep freq generator to align AM and FM radios. It is an HP 8601A, circa 1970's. The sweep output is a sawtooth and corresponds to the sweep RF freq.

The use is alignment of older AM and FM radio IF transformers. AM is typically 455 Khz center freq and ~6 Khz band width. FM is 10.& Mhz and ~200 Khz band width. Markers are usually some kind of harmonic beat marker, which works well for FM, but for AM it is too wide and swamps the display. Beat markers are wavy lines about 20Khz wide (or more). This is not good for AM radio IF alignment (~6Khz wide). So for narrow band width sweeps, harmonic type or beat type markers are too wide.

AM radio 455Khz IF
http://img185.imageshack.us/img185/4544/amalignment2.jpg

Also injecting the sweep signal and marker signal into the radio is not ideal. There are "post injection marker adders". This avoids the problem of putting the marker signal through the radio, but the post injection marker adder is still harmonic. So it still suffers from being too wide. FM radio with 10.7Mhz and 200Khz band width is more tolerant of wide beat markers, it just a smaller percentage of the freq response trace. However even on FM there are issues with side bandwidth markers. The center marker is fine since it is in the middle of a Horizontal flat line. The side bandwidth markers are on lines that are more vertical than horizontal. So the beat markers are long wavy lines.

FM radio 10.7Mhz IF
http://img85.imageshack.us/img85/7579/markers.jpg

Digital pulse marker would work better, using the Z (intensity) input of the scope. The issue is HOW TO TRIGGER THEM? You want them to trigger right at the freq you need them at or slightly before so they are centered on the Freq. These intensity markers don't care about vertical or horizontal lines.


WORK AROUND "CALIBRATION"

I came up with a way to "CALIBRATE" the sweep on the scope display so the center freq is right at the center of the scope display. With the scope in X-Y mode, the X horizontal is sweep out and Y vertical is the RF signal.

http://img193.imageshack.us/img193/7088/calabration2.jpg

Through a detector it makes this sweep.

http://img717.imageshack.us/img717/2792/calabration.jpg

Using a diode detector and adding a beat marker with no RADIO under test, I get a good idea where that Freq is along the sweep, even if the center Freq is 455Khz and the sweep is only 300Khz wide. In the "calibration set-up", the sweep is perfectly flat, so the marker is not hard to see. I can center the sweep on the scope with the horizontal adjustment. It may not be perfect but it is with in a dozen Hz or so. That is totally acceptable.

IT WOULD BE NICE TO MANUALLY PUT A MARKER THERE, AS WELL AS TWO MORE +4Khz and -4Khz. For FM it would be +/- 100 Khz. Again this could be done by using the sweep gen, second signal gen (a very accurate digital one) and moving the markers on the display manually.

NOTE. The sweep Gen is a HP 8601A. It's analog but accurate. The sweep Freq is stable and can monitor it with my Freq counter and it has two sweep outputs (for the ramp sweep signal). Stability and accuracy are good; however as good as it is, their is tolerance. The Freq can be monitored but sweep symmetry can be off say 5 Khz or more. This is not good enough for AM radio. BTW AM radio alignment typically does not use sweep alignment but a single modulated signal, peaking the audio. Sweep is more accurate but not critical for AM radio. FM radio of course being wide bandwidth is more critical to be sweeped to get a good visual symmetry. However I found sweep alignment of AM offers many advantages, not the least of which is you don't need to have the speaker on full volume listening to 400 Hz AM modulation. The second signal gen I am using for a marker is a HP 8640B which locks onto the Freq.

After calibration, the CENTER FREQ is at the scope's center graticule. Now I can just sweep and line up on the scope's center graticule ....

http://img15.imageshack.us/img15/6553/amifalign.jpg

With these markers (manually set) I can see the trace better with more accuracy.

 

[55pilot]

The simplest way for me to do it would be to do it in a PIC micro with a built-in comparator. The external voltage is resistor divided so the trigger point is at the PIC's trigger point. The code busy wait's for the comparator to trip. It then sets the output for a set number of instructions and then goes back to looking at the comparator.

If you know any other microprocessor, you can do the same code with an external comparator and the micro of your choice.

---55p

 

[gmcjetpilot]

The simplest way for me to do it would be to do it in a PIC micro with a built-in comparator. The external voltage is resistor divided so the trigger point is at the PIC's trigger point. The code busy wait's for the comparator to trip. It then sets the output for a set number of instructions and then goes back to looking at the comparator.

If you know any other microprocessor, you can do the same code with an external comparator and the micro of your choice. ---55p

Pic micro? OK You are the second person to suggest a Micro Controller. I was hoping to do it with discrete components, but may be I should jump into the micro-controller. I have little to no experience with them. Some one mentioned the PICAXE-08. They seem to be quite simple. I think it has to power to do what you say.