Attn: Electrical Engineering Types - PeachParts Mercedes-Benz Forum

KarTek · #1 05-04-2015, 09:03 PM

There was a posting a while back where someone made an electronic pop tester with a high pressure transducer. I would like to sort of alter that concept somewhat and make an electronic timing light.

I have a 2000 PSI pressure transducer and I would like to make it trigger a bright LED to use on my injection pump machine. (Which I will expand on at another time).

Basically, I'm looking to do this (picture). What I'm not sure of is the circuitry needed to change what is, I'm sure, a slow rising pulse, into a square wave that will drive the LED.

Any ideas?

HappyHappy · #2 05-04-2015, 11:05 PM

Quote:

Originally Posted by KarTek

There was a posting a while back where someone made an electronic pop tester with a high pressure transducer. I would like to sort of alter that concept somewhat and make an electronic timing light.

I have a 2000 PSI pressure transducer and I would like to make it trigger a bright LED to use on my injection pump machine. (Which I will expand on at another time).

Basically, I'm looking to do this (picture). What I'm not sure of is the circuitry needed to change what is, I'm sure, a slow rising pulse, into a square wave that will drive the LED.

Any ideas?

One of the easier ways to do that might be to make use of the basic 555 op amp to overdrive a basic 555 op amp based circuit to clip the spike into what would be in effect a square pulse. You could probably do a Google search and come up with a thousand circuits that might fit the parameters you're working within.

Google Forrest Mims, he used to publish little simple handbooks of simple circuit blocks, Radio Shack used to sell them for a couple bucks each.

There are possibly even monolithic chips that exist that are designed to do just that, signal conditioners, A?D converters, etc. The real impactful parameter will be the frequency of the signal and the time frame needed to capture the event and translate it, faster costs more usually.

HappyHappy · #3 05-04-2015, 11:43 PM

Here's a super simple solution that might work for the signal conditioning aspect

SINE-WAVE-TO-SQUARE-WAVE CONVERTER - Circuit Diagram - TradeOfic.com

driving the LED directly is probably out of the question but should be pretty simple as well.

KarTek · #4 05-05-2015, 06:04 AM

Quote:

Originally Posted by HappyHappy

One of the easier ways to do that might be to make use of the basic 555 op amp to overdrive a basic 555 op amp based circuit to clip the spike into what would be in effect a square pulse. You could probably do a Google search and come up with a thousand circuits that might fit the parameters you're working within.

Google Forrest Mims, he used to publish little simple handbooks of simple circuit blocks, Radio Shack used to sell them for a couple bucks each.

There are possibly even monolithic chips that exist that are designed to do just that, signal conditioners, A?D converters, etc. The real impactful parameter will be the frequency of the signal and the time frame needed to capture the event and translate it, faster costs more usually.

Funny you say that, SOMEWHERE I have an original Forrest Mims electronic handbook from circa 1977. I'm pretty sure it has a circuit called a wave shaper or some such thing. But of course, I can't find it anywhere...

That circuit in your second post looks promising. The pressure transducer should arrive in the next day or so and I'll be able to hook it up and see what kind of wave the injection event produces on my oscilloscope.

Stretch · #5 05-05-2015, 06:18 AM

I'm not the best person to advise about electrickery but I don't think a LED is going to be the ideal light source for a strobe light.

Funola has made something recently that does more or less (what I think) you are doing.

Stretch · #6 05-05-2015, 06:34 AM

My digital RIV timing light - Engine testing

KarTek · #7 05-05-2015, 10:50 AM

Quote:

Originally Posted by Stretch

I'm not the best person to advise about electrickery but I don't think a LED is going to be the ideal light source for a strobe light.

Funola has made something recently that does more or less (what I think) you are doing.

That's good info. In my case, I'm working in a very controlled environment. The furthest the light has to shine is about 2" and I can darken the area to observe the degree wheel.

TnBob · #8 05-05-2015, 12:03 PM

A narrow beam LED will suffice.
555's are under a dime on ebay

97 SL320 · #9 05-05-2015, 06:14 PM

Something else to consider. There will be some delay between the pressure sensor reaction / circuit operation and LED flash.

Depending on how fast the pressure spike is and sensor reaction time, you might be flashing later then you think.

Sine wave converters generally fire as the wave starts downward, trying to fire on the peak before falling would be messy as you would need to know where you were on the wave before it happened.

A pulse stretcher / contact debouncer might be better in conjunction with a forming circuit that has 0 volt out until your sensor hits near but less than peak might be better. But, this is still triggering on the up swing and varying pressures could cause a timing variation.

With all of this delay occurring, you could delay things farther so they are synced back up on the next firing. A similar situation exists with electronic spark control on a distributor. Base timing is set at say 5* BTDC but electronic advance can pull it to 35 - 38* BTDC. Since the distributor pickup is set at 5 BTDC, to get the advance you end up delaying spark so much it is early ( the desired 35 - 38* ) for the next cylinder.

mytimeyet · #10 05-06-2015, 11:26 AM

Here is a guy that built an injector tester, for gas but might give you some ideas.
https://www.youtube.com/watch?v=x0baCPvD8QE

BillGrissom · #11 05-06-2015, 03:06 PM

I built the electronic pop tester. The xducer I used gives an amplified voltage signal output (~5 V max, I recall), which is the best type to use. Simplest might be a "hex inverter" chip, but it trips at a fixed TTL level (~1-3 V). It can drive ~50 mA, or you might parallel the 6 outputs for more current. An op amp "comparator" can trip off an adjustable level (you set w/ a potentiometer). Most output ~20 mA. Some op amps are sold as a comparator, but you could use a generic 747 one.

Your LED might require more thought. Will the square wave signal be just a trigger, or must it also provide power? I doubt 50 mA would make much light. Also, insure negligible delays from components or "RC filter" effects of cables. It might be easier to leverage a "timing light" for spark-ignition engines. Used ones are $2 at garage sales. New ones (ebay) are ~$30 and allow adjusting a delay/advance which might prove useful. With that, one can determine say 38 deg spark advance even though the timing tab on the engine only goes to +/-10 deg.

97 SL320 · #12 05-06-2015, 07:32 PM

Yep, a comparator is what I was thinking of. The trick would be to find the proper reference voltage as that would affect apparent timing.

KarTek · #13 05-06-2015, 09:25 PM

Thank you all very much for the input! In this application, I'm not concerned at all with delay in the circuit because I'm comparing the cylinder injection events to each other, not to a revolution of the crank shaft.

When I put the pump on the machine, I have the locking tool in place and I set the degree wheel to "0". From there I will check the timing on all 6 elements, comparing them to each other to make sure they are all as close to 60 deg. apart as possible.

Then, I can make adjustments to the pump to correct the timing. I have the Druck transducer now and I need to get a DC power supply, a bread board and components. I also need to pick up the "plumbing" to connect the transducer to the injection line.

I'll update this thread with info as it comes along.

Stretch · #14 05-07-2015, 06:30 AM

I don't know much about the speed of electronics but I would imagine that the delay in most electronic circuits would be tiny compared with the rate at which events happen within a mechanical injector pump - the IP runs at half the speed of the crank - so we're probably looking at events happening within a range of 5 cycles to say about 50 cycles per second...

...50 cycles per second (50 Hz) has a meaning on this side of the water as it is the frequency of (home) mains electricity but I understand it is 60Hz where you are so I guess that'll have a minimal effect on your circuitry where you are (assuming you are using mains power for your measurement systems)...

funola · #15 05-07-2015, 08:53 AM

If yourr sensor output is a varying voltage level proportional to pressure, you don't want a level detector circuit. Google peak detector or rising to falling edge detector.