Convert Pulse to DC Voltage - Ideas Needed

[Liad]

Hey,


I need to find a way to convert a pulse that can be in any amplitude from 0 to 5v
and output DC voltage at the same amplitude.

any ideas? :)



Liad.

 

[chroot]

To... convert?

Do you just want a circuit that measures the amplitude of the pulse and adjusts a regulator to produce the same voltage, or what?

Please try to describe your intended circuit in more detail.

- Warren

 

[Liad]

yes...:)


any circuit that his input will be 1 pulse and his output will be a dc voltage in the same amplitude and will stay that way until the circuit will get another pulse in his input.

 

[chroot]

So, does this signal always return to 0V between pulses?

- Warren

 

[Liad]

you can say that...:S


ahm...itll only be one pulse and then...nothing, till the next pulse.:)

 

[berkeman]

You are describing a peak hold circuit. You can make one easily with a diode (and be sure to account for the forward voltage drop of the diode), or you can add an opamp to eliminate the diode drop error, assuming the opamp is fast enough for your pulse. You can add a reset input to the opamp circuit to dump off the voltage that is held on the output...

http://www.st-andrews.ac.uk/~www_pa/Scots_Guide/audio/part8/Page3.html

http://www.analog.com/UploadedFiles/Data_Sheets/PKD01.pdf

http://www.amptek.com/ph300.html

http://www.priorartdatabase.com/IPCOM/000006418/

 

[berkeman]

You could also use an analog switch to sample and hold the peak value, if you knew when it was going to happen. A second switch would be used to dump off the peak held voltage from the small storage cap.

 

[chroot]

"Nothing" is not a meaningful term. I think you mean "zero volts."

I'd suggest you use something like a peak detector. ADI's PKD01 seems appropriate:

http://www.analog.com/UploadedFiles/Data_Sheets/PKD01.pdf

You can wire up its integrated comparator to automatically reset the peak detector each time a pulse comes along over some small threshold voltage. You might need to couple it with a one-shot to control how long the reset is asserted, or you might be able to use something like an RC delay. Either way, it's maximally a two-chip solution. You should look around at other peak detectors and see if you can find an even easier solution.

- Warren

 

[batman29]

Hi,

I am new to this forum and I hope someone will give direct answers to my query.

1- I have 2-wires that carry pulse DC to 2-sensors (parking sensors). The car’s ECU converts 12V car battery current to 0V-8V PCD (40 KHz pulse signal, emitted approx every 125 ms).
2- I would like to replace the 2-sensors (parking sensors) with 2-bulbs (12V, 25 Watts) and use the same 2-wires that carry PDC to light the bulbs

Question:
a- Can I do this?
b- Will the 25 Watt bulbs light?
c- Will the 25 Watt bulbs light at the same intensity (lumens) that a 12V DC source?
d- Will the 25 Watt bulbs burn with use & time?
e- How can I do the 25 Watt bulbs to light (i.e. same lumens as 12V DC) when the bulbs are connected to 0V-8V PDC?

Please help

 

[vk6kro]

You can use the same wires, but you would need to rewire the other end of the circuit to supply 12 volts.

At present you are getting a brief burst of 40 KHz pulses at 8 volts peak and you need to convert this to 12 volts at 50 watts, but this just isn't possible. Not easily anyway.

 

[sophiecentaur]

A peak detector, a relay driver and a relay?

 

[batman29]

Many thanks

I do not want to cut the 2-wires (under the car console) that carries PDC to the rear sensors and solder 2-wires 12V to them and thereafter using the 2-cutted PDC wires to carry 12V current under the car floor harness.

Questions:

Can I just solder the 2-wires 12V onto the present 2-wires PDC. This means that the 2-wires will carry 12V current and sometimes 0V-8V PDC to the rear or both. The car’s system activates the PDC when the car engages reverse gear. However I wish to light 2-bulbs (25 Watts each) at the rear when the car is or it is not engaged "Reverse".

 

[vk6kro]

No, there is no 12 volt supply to power the relay or the lamps.

Just a (presumably) low powered 8 volt pulsing circuit. It drives an ultrasonic transmitter which would be low powered. How do you get 12 volts at 50 watts from that?

At least one of the wires needs to bring 12 volts at a decent power level to the back of the car. If the ground is an option for the return path this would leave one wire spare.

 

[vk6kro]

We posted at the same time.

Can I just solder the 2-wires 12V onto the present 2-wires PDC. This means that the 2-wires will carry 12V current and sometimes 0V-8V PDC to the rear or both. The car’s system activates the PDC when the car engages reverse gear. However I wish to light 2-bulbs (25 Watts each) at the rear when the car is not engaged "Reverse".

No, you probably can't do this without a thorough knowledge of how the present system works.

However, you could borrow 12 volts from the trailer power outlet if the car has one. This is powered all the time.

You probably shouldn't be doing this. If you want lights then you should get an Auto Electrician to fit them for you. In most countries, lights on a car are controlled by regulations and you might see flashing blue lights if you get it wrong.

 

[batman29]

I am able to locate a 12V wire under the car’s console. I will branch it into 2-wires to match the 2-wires PDC and prevent any over heat that may happen in the carrying wires until they reach the rear of the car. So each wire will light a 25W bulb.

 

[batman29]

The final manipulation of the details will be done by an auto electrician. My main concern rests into finding out if the approach is physically/theoretical feasible. If the approach is feasible, then I will discuss the details and the proper use of tools to do the job with the expert.

I have study the electrical circuits (in details) of my vehicle. The ECU corner/back sensors produces the PDC (a phenomena that is new for me) and send 8-wires to the rear; 4-wires carry the 0V-8V PDC and 4-wires bring the information from the sensors into the ECU. Once the 4-wires with PDC leave the ECU these go through a number of connectors (female & male) until the wires reach the sensors itself. The wires travel as part of the floor harness.

The lights at the rear that I am talking about are called “rear fog lights” which are legal in Europe, Australia, and Canada. In the USA car may have it or not.

The old method of implementing “rear fog light” is to run an orphan wire under the car carpet floor, which implies detaches (damage) fasteners, drilling holes on the car chassis, etc. At the end of the job the car will rattle because the end product will not be perfect.

What happen if a bulb (25W) is connected to this 0V-8V PDC? Will light or not?

 

[jim hardy]

the RMS of an 8 volt square wave is 8 volts.

OOPS EDIT make that 4 volts.. for 8Vp-p

sooo ..If the driver has enough current capability, you'll deliver (4/12)^2 power to 12 volt bulbs. they'll be mighty dim.

presumably the ECU only activates the backup sensors when you're backing up

do it right and run a switched & fused wire , affixing it to frame with suitable adhesive and cable ties.

and be sure that your "12 v wire under the dash" has nothing to do with airbags...

and it's rude to burst into a thread

 

[batman29]

and it's rude to burst into a thread!

I am very sorry. It was not my intention. Just I thought the topics are related and why to open a new thread to talk about similar subjects. Sorry again

Using your formula the 25 W bulb will light with a power of 11.1 watts which I think is not "so mighty dim". Side marker bulbs are just 5 Watts.

 

[vk6kro]

It was an old thread, but this was on-topic, so don't worry about it.

If you have a pulse for 25 μS then nothing for 125 mS, then this has a duty cycle of 1:5000, meaning that the 8 volt pulse is effectively 1.6 mV . (8 volts / 5000 = 1.6 mV). If there are several 25 μS pulses, then the voltage will rise slightly (by 1.6 mV per pulse) but still not enough to be useful.

But it gets worse. The 12 volt lamps will have a resistance of only 3 ohms total (12 volts / 4.166 amps) and the Parking Distance Control will not be designed to cope with such a low resistance load.
At best, it will just drop the voltage internally and give a few millivolts out.
At worst it could be damaged or destroyed.

The results of feeding 12 volts into the output of the PDC are certain to be catastrophic. A new one will cost you hundreds of dollars.

 

[pantaz]

... I have study the electrical circuits (in details) of my vehicle. The ECU corner/back sensors produces the PDC (a phenomena that is new for me) and send 8-wires to the rear; 4-wires carry the 0V-8V PDC and 4-wires bring the information from the sensors into the ECU. Once the 4-wires with PDC leave the ECU these go through a number of connectors (female & male) until the wires reach the sensors itself. The wires travel as part of the floor harness.

Modifying or interrupting this wiring may have unintended and possibly negative effects on the vehicle's operation. Automotive computers have become intricately linked to just about every part of the car. You will have a much easier time simply adding a new, independent circuit.

... What happen if a bulb (25W) is connected to this 0V-8V PDC? Will light or not?

Not with any appreciable brightness, if at all.

 

[batman29]

Many thanks for your comments. You have scared me!

OK what about this approach:

1- The corner/back sensor ECU activates 4-wires with pulsed direct current, i.e. 1-sensor corner LH, 1-sensor corner RH, 1-sensor back LH, and 1-sensor back RH

2- I will cut the 2-sensor corner wires LH & RH and therefore they will not be connected to the sensor ECU at all.

3- I will solder these 2-wires with a 12V source wire. The 12V wire will be activated after the dipped beam is switched ON and the wires will carry 12V current to the rear of the car ( a relay and a fog switch will be needed)

4- On the rear bumper the sensor wires (corner LH & RH) are detached from the physical sensors and the 12V bulb is attached instead.

The final product will be like this: The 2-rear fog light will light only when the dipped beam is turned ON. The 2-back sensors (LH & RH) will be activated when the sonar/buzzer switch is turned ON and the car is engaged in Reverse. In summary, only 2-back sensors will be in operation when doing a parking manoeuvre.

 

[vk6kro]

It sounds like it would work, but it is probably a bad idea.

You would be disabling a useful feature of the car for the sake of a few bits of wire.

Why not just get an Auto Electrician to drill a couple of holes and run some wires for you?

 

[batman29]

There are some subtle drawbacks with running a wire under the carpet floor, refastening and re-clipping it, hole drilling, and putting the protector detached panel (quarter trim) back in place. By experience I may said that the auto electrician is most concerned in bringing light to the rear side of the car without caring to much about the workmanship of the final job. This will be my third vehicle since 2004 and I used indeed the service of experienced auto electricians. Vehicles were returned to me with mysterious rattles and floor carpets badly attached. In the process of running a wire under the carpet floor it is very likely that fasteners and clips are damaged and even broken and the auto electrician would not have time to order spare parts because this will take another day. Wires in contact with the car chassis were not isolated or camouflaged.

It is a matter of fact that rattles noises in any of my vehicles have almost driven me crazy (paranoid) and I even bought a hearing instrument from a doctor/physician to locate the rattles. In other words, I cannot cope with rattles so I usually ask my wife (not that she cares about the rattles) to drive while I am in the rear seat or cargo area using my doctor instrument to locate the rattles! This behaviour is also driven my wife crazy. Maybe the simplest solution would be to get some psychological help with my paranoia!

 

[vk6kro]

I can understand the reaction to rattles.
In previous cars, I have had rattles under the dash that I couldn't locate.
But then, I can't stand bouncing basketballs either.

Maybe you could find an auto electrician who will let you watch while he does the work. At least then you will know where the wire runs, so if it rattles, you can fix it.

 

[sophiecentaur]

Maybe the simplest solution would be to get some psychological help with my paranoia!

Just turn the radio up loud.

 

[batman29]

After finding out that connecting 12V, 25 Watts bulbs into a PDC (8V) source will produce catastrophic results in my vehicle, it has come to my mind, just for curiosity, what would happen if LED (light emitted diodes) were used instead of filament /incandescent bulbs? Diodes may work with 3V-5V source and very small amperes. PDC seems to be suitable for LED lighting, isn’t it? Any comments will be highly appreciated. If this is possible I will not need to cut wires at the other end (i.e. front)

 

[vk6kro]

In that case you go from "certainly won't work" to "might work if you are lucky".

Once again, you have very short pulses of power and a long time when there is no power, so you need to store the power while it is there and use it when it is not there.

You need to find out how much power this circuit can deliver.
One way would be to put a 47 ohm resistor across it and see what the voltage drops to.
It is 8 volts now. If it drops to 7 volts, you could calculate the internal resistance of the circuit.
Like this:
The current in the 47 ohm resistor would be 7 volts / 47 ohms or 149 mA.
The internal resistance then has 149 mA flowing in it and 1 volt across it.
So, the internal resistance is 1 volt / 0.149 amps or 6.7 ohms.
Say so if you don't understand this. It is important.

This will tell you how much current you could use to light up some LEDs.
The duty cycle is still 1:5000 so you have to divide this 149 mA by 5000 to get an average current.
This is a uselessly small current of 29 microamps in this example. You wouldn't even see that it was on.

This leaves you with the chance of pulsing the LED. You can pulse a LED and your eye sees a bright light but holds the image for a short time. I don't know how well this works, but it might be your only hope.

 

[pantaz]

... PDC seems to be suitable for LED lighting, isn’t it? ...

No.

I would go into more detail, but you seem determined to mess with the car's computer system regardless of our advice.

Best of luck to you. I'm done.

 

[batman29]

Many thank for your comments

I am conversant with Ohm’s law and Kirchhoff’s law so I understand your DC calculations. I have also found some information on the web where PDC is used to drive LED (experimental stage). There is also a company that produces LED lamps driven by PDC.

I would like to make clear that by working only in the rear bumper (where the sensors are located) I am not touching at all the ETACS-ECU of the car neither cutting any wires at the front. The job will be just to detach the sensor head and replace it by a number of LED (7 or 8) which will be soldered (circular) in the manner that is done the little LED torches that are bought in places like Wall-Mart.

My concern now is to find out how to specify the LEDS (e.g. mA, V, and mW) so as a whole they can deliver 1700 lumens or 136 candelas that is equivalent to a 25 Watt 12V bulb luminescence lamp.

I must point out that I am very grateful for your comments that have distilled understanding in this problem. All the advice given has been taken very seriously by me and if I have persisted with the subject it is to find alternatives solutions to the same problem.

 

[vk6kro]

The main conclusion from the calculations is that there really isn't enough power available in the PDC to run any lamps or LEDs.

Why don't you run the lights from 12 volts taken from the tail light wires?

 

[batman29]

I would like to comply with international regulations about rear fog lights on vehicles and this requires that rear fog light can only be activated when the dipped beam headlight is ON, not without it.

I know I have some gaps in my thinking to fully understand the electronic behavior of LED when the LEDs are driven by PDC. Please be tolerant if my reasoning is faulty.

The back sensor ECU of the vehicle produces PDC (pulse signal) 0V-8V, 40 kHz pulse signals, emitted approx. every 125 ms). That is what it says in the electrical manual for this vehicle. According to the previous data it means 40,000 flashings per second. The vehicle ECU takes 0.125 s to release a batch of pulses and therefore a batch of pulses consist of 40,000 x 0.125 = 5000 flashing.

The standard red LED has these characteristics: 20mA (LED), 2V (40 mW). Calculations proceed like this: 2 = 0.02 (R), R = 100 Ω (each LED). A number of 4-LED will use the 8V at each flash.

Because the flashings are positive and the current can be either 0 mA or 20 mA the LED will light only when the led is hit with the flash. I may be wrong and please correct me; human eyes will be not able to see the light flickering. I may be also wrong but there is no reason (apparently) to average the current because the LED will be hit with current either 0 mA or 20 mA not anything else.

 

[vk6kro]

The pulses emitted are at the frequency of 40 KHz, but they cannot be emitted continuously. This is because the receiver must be able to listen for echoes from other vehicles etc without interference from the transmitter.

More likely, a brief burst of 2 or 3 pulses is transmitted and then there is nothing while the receiver listens for echoes.

This is the problem if you want to use these pulses to light a LED or a lamp. Because there will be very few of them and long periods of listening, there is very little power available to do anything.

You can eaasily settle this for your own satisfaction. Just put an oscilloscope on the transmit output and observe the waveform.
Or, connect a LED with an appropriate resistor across the output and see what level of brightness you observe.

 

[batman29]

(Pulse signal, 0V-8V, 40 kHz pulse signals, emitted approx. every 125 ms)

Making sense of the statement above in addition to your description of how the signals are transmitted to the sensor heads (rear bumper) I may say the following:

The corner/back sensor ECU has 20 pins and only 11-pins are engaged all the time. 5-pins are connected to 5-wires that route to the rear of the car. The pulses travel through 4-wires and returned to the ECU by only 1-pin wire (with information about the obstacle distance when the car is doing a parking maneuver)

From reading again the information above it appears that the ECU releases a batch of pulses and then waits for 0.125 seconds to send another batch of pulses and so on. While 4-wires are waiting 0.125 seconds any response from the sensor heads are channeled together into a different 1-wire and sent to the ECU for evaluation. In summary the LEDs are energized (0 mA-25 mA) every 1/8 of a second. Is this reasoning correct? Will 1/8 of a second be enough to prevent noticeable flickering?

The reasoning above may have some weaknesses but I would like to be theoretically certain that it could work before dismantling the rear bumper and the rear side quarter panel of the car. Just experimenting (e.g. oscilloscope) will need the rear of the car dismantling in addition to the custom built set of LEDs (Red) to replace the 25W bulbs.

 

[vk6kro]

I suspect that what you have is a few pulses 25 uS wide emitted 8 times a second.

The problem is not limited to flickering. The problem is that there just isn't enough power to run any lights or LEDs.

You may have something entirely different to this and the only way you can find out is examine the output of the ultrasonic transmit transducer.

One way to do this would be to hold an ultrasonic receiver transducer close to the transmit one and examine the output of the receive transducer with an oscilloscope. This would not require you to remove the bumper of the car.
The oscilloscope would have to be near full gain, but you should get something.

 

[batman29]

OK I understood the inability of the sensor wires of carry enough power to light the LEDs.

Will the power of the pulses DC be enough to drive a miniature relay?

 

[jim hardy]

if you're determined to use the ultrasonic sensor

i'd place an ultrasonic receiver nearby and let it drive a relay to switch, say, a GE-4522 aircraft landing lamp. let the ultrasonic receiver trigger a timer circuit so the light won't follow your 8pps ultrasonic. that way you haven't intruded at all in the car's electronics.


http://www.aircraftspruce.com/menus/el/replacementlampsflashtubes.html

that'll require a wire to power the lamp because it draws about twenty amps.
run it yourself, under the car so you'll be happy with the installation.

that link carries LED high power lamps too but note they're pricey.

 

[vk6kro]

OK I understood the inability of the sensor wires of carry enough power to light the LEDs.

Will the power of the pulses DC be enough to drive a miniature relay?

No, it won't.

 

[batman29]

May I kindly ask you to provide your comments (e.g. Pros, Cons and Workability) of the approach suggested by our member (Jim Hardy)?

 

[sophiecentaur]

I will repeat what I wrote earlier on. If you have a supply of 12V with sufficient current capability then use a detector (diode and amplifier), a relay driver and a relay. That will do exactly what's needed. It will be the least 'invasive' solution and easily removed if you don't want it any more.
As Jim Hardy says, you can even use the ultrasound signal to avoid any direct connection to the 'electronics'. If it were up to me, I'd far prefer to do the hard work in the comfort of an indoor work bench and minimal effort on the car itself.

 

[batman29]

Many thanks

My background in electronic components is minimal; however I think I am able to follow your instructions. Certainly I will not be able to construct (myself) any ultrasound transmitter/receiver but I may order an already made ultrasound receiver to be located in the rear bumper. I am able to branch a line from a 12V source.

Reading the literature some people have built a transmitter and its corresponding receiver that will switch on a relay that will close a circuit to use a 12V source of current. I may say that the head of the sensor in my vehicle not only sends ultrasound beams but also receives the bouncer beams. Therefore the sensor head acts as a transmitter/receiver.

The problem I see from this exercise is that the rear fog light mostly is switched on when the car goes forward and the sensor head although it releases ultrasound to the air the beams are not bounced back (no obstacle) and therefore the best ultrasound receiver will not receive anything to be activated. I may be wrong here but I have the suspicion that the receiver will not work. If the ultrasound beams from the transmitter spray in all directions then the receiver may work.

Assuming that my logic is faulty I will appreciate it if someone can recommend a good brand of ultrasound receiver and where to contact.

A few months ago I designed (on paper) the use of RF transmitter (remote) to activate the rear fog light by draining 12V line from the 12V socket available in the cargo area of the car. I bought, very cheap, a Chinese remote kit (control unit ready to be used + transmitters) with a set of coloured wires to be connected directly to the 12V load. I have to use 2-transmitters one in the front to operate the rear fog light lamp (switch) and the other in the rear to operate the rear fog light. The problem with this arrangement is that I have to press 2-buttons to activate the system and I need only 1-button to perform the job. There are also some logistics troubles as the results of having 2-control units and also the transmitter is very small (about 1-third of the smaller mobile telephone), etc.

 

[jim hardy]

let's get real simple here...

what does this rear fog light do?

every car I've owned, including my '49 Buick, had back-up lights. Nowadays they're usually incorporated into the taillight assembly, though in '49 they were separate.

so if the object is to provide more illumination behind you when backing up,,,,,
the wire to your existing backup light will have capacity to drive one of those 45 watt LED aircraft landing lights and it will be energized when you're in reverse.

Back in the days when cars had four headlights (1960's) I used to put 4522's in the inboard two for mega-high beams. But the factory wiring wasn't heavy enough for those 250 watt lamps so i added separate wire and a relay.

in lamp terminology PAR means the glass is round parabolic reflector, the two digits following are its diameter in eighths of an inch.
The GE 4522 landing light is a PAR45, ie 5&5/8 inch diameter.
An old quad car headlight is a PAR46 , 1/8 inch larger. The PAR45 with a piece of inner-tube wrapped around it fits right into an automobile quad-headlamp ring.

The 4522 lamp illuminated cowpastures a half mile away . If that LED landing light has anywhere near comparable lumens, it will really impress your friends.

one of those LED landing lights should give you no electrical problems, just you got to mount it. You shouldn't even need a relay.


but be sure it can't shine at other drivers. i LOATHE those Xenon headlamps in the hands of imbeciles who won't dim them.

over and out,,, old jim

 

[sophiecentaur]

Many thanks

My background in electronic components is minimal; however I think I am able to follow your instructions. Certainly I will not be able to construct (myself) any ultrasound transmitter/receiver but I may order an already made ultrasound receiver to be located in the rear bumper. I am able to branch a line from a 12V source.

Whatever you choose to do, make sure that you put a low value of FUSE in series with the wire you take off the positive supply, wherever you connect it. I should keep it to 5A or less. That will ensure that any mistakes you make in wiring or construction will only result ina blown fuse!
Have fun.

 

[jim hardy]

@ sophie :: amen.