Seeking direction to make an led tachometer for an obd1 car

Thread Starter


Joined May 23, 2024
Hello to those far more knowledgeable on the subject of electronics than I.

I guess I'll tell you what I'm aiming for and go from there. Ideally, I've been wanting to create a led tachometer that at least resembles the one from a honda s2000, I'll include an image for reference, but to make it a "heads up display". But knowing that it would best function at night and possibly very little during the day, more realistically or perhaps the first thing to do is to learn about how they're made, then trying to make one before I bite off more than I can chew amd become discouraged. Even though this idea or desire, does seem to have some staying power since it keeps popping up in my mind.

I just like the sweeping tachometer that lights up as the rpms increase and maybe some additional lights that light up at a certain rpm, a shift light. I've tried researching online to see if anyone else has made one and if they said how, but have found very little.

The only one I really found was someone who used ten leds in a circle. I thought about just expanding on thier idea, but am unsure of how I would get the resolution that I want. I want one led for every 100 rpm all the way up to 8500 rpm.

I could not illicit a response from them to ask any questions. Like what exactly the computer chip did in that circuit and if it was capable of expanding the number of leds to what I am looking to do. It would go from the ten they used to the 85 that I want to do. Or if that chip was limited to ten leds and I would have to look towards multiplying the same chip or look towards a different chip and if so, which one.

I'm also unsure of how to convert an analog rpm signal to make it work with the leds. Searching online can be a fruitless endeavor when you don't know the correct terms or names of items, that if known would make the search much shorter. But I'm also pretty sure that google just likes to make things difficult for me.

Constantly gives me results that make me feel like I'm getting close to the answers I'm looking for, but never actually giving me the answers I'm looking for. It can drive a person INSANE!!! I'm certain that is thier plan. Create one of the biggest tech companies ever, just to filter my search results in an attempt to slowly drive me insane! Think about it, it's brilliant! Nobody would ever believe me if I said that was thier goal the entire time.... I don't even believe me. Anywa.... What was that?!... the wind.

Anyways, all kidding aside, I guess if I broke it down to the most simple thing that I would like and appreciate some guidance with is to create a tachometer where the analog signal comes in and will sequentially light up leds in 100 rpm increments.

Don't hold it against me if this whole post could have been just the last paragraph. I just wouldn't want anyone to think that I just had an idea, haven't put much thought into how to get it done and want someone who has put a lot of effort into being knowledgeable on the subject to tell me exactly how to do it with as little effort into it as possible on my end. I enjoy learning how to do things and how things work and am willing to put in the work to create something. I think more and more, people seem to take for granted, those who are knowledgeable in the trades, making things work, building things, etc. I am one who appreciates the knowledge and wisdom gained by those who came and did with thier hands before I came around. In fact, I guess you could say I hope to join that fraternity one day. I'm 38 and am knowledgeable in a few different but similar areas, but electronics? Not so much. But I am starting to try and learn.

Thank you for reading my post, if you made it this far, and I appreciate any assistance, advice, guidance and knowledge that anyone is willing to give.


Use Charlieplexing or expanders or shift-registers for driving multiple Leds.
Convert analog to digital with ADC converter (contain almost every microcontroller like AVR/Arduino or PIC).

Note: There is also possible to use a more microcontrollers for Led expansion since they are pretty cheap these days.
Last edited:


Joined Jul 10, 2017
The best advice I can give you is to put this project on the back burner until you have learned an awful lot more about microcontrollers, A/D converters and electronics in general. That is quite a complex project for a beginner. I would hesitate to tackle this one unless there was a desperate need for it.
If I really had to, I would not use individual LEDs. I would use an Arduino microcontroller with built-in A/D, with the RPM meter showing on a TFT display as in your picture.
Look up on the internet similar Arduino projects that use a circular meter on a display.
Last edited:


Joined Oct 6, 2013
First question is what are the specifics of the car (make, year, engine, etc)?

Tachometers usually use some form of pulsed signal from the ignition system or ECU to generate a visual representation. I haven't heard of an analog signal to drive a tach so far (not that it isn't possible... it's just not normal).

Depending on the car and what signals are available from the ECU this can range from a simple programming exercise to a whole world of new things.


Joined Nov 6, 2012
Use Your Ears, and your built-in-Butt-Dyno.

Shift right before the Acceleration starts to drop-off, then remember that sound and feeling,
this is the way it's been done for the last 100-years or so, it works great !!!!!

This project is an incredible amount of work and expense just for some unique "bling".


Joined Oct 2, 2009
Very doable, so don’t be discouraged.
The trick is to look beyond 85 discrete LEDs and think big, such as a big TFT graphics display.

How big do you want the screen to be, it terms of physical dimensions?


Joined Jan 23, 2018
OK, there are a number of different challenges here. The first is picking up an RPM pulse signal. That is simple and easy for vehicles with a single spark coil, while it is entirely different for those engines with multiple spark coils, as well as diesel engines.
After gaining an RPM frequency, the rest can be analog voltages or counts per unit of time. An analog mechanical meter with afixed scale is the very simplest, an analog bar graph, linear or circular is the other common option. There are analog bar graph driver ICs available but the common ones do not produce that many segments
Last edited:
How about cascading LM3914 dot/bar graph displays? That could turn an analogue input into a bar graph display that could have LEDs arranged in a circle.
Sniffing the RPM from the distributor HT lead or a toothed crank trigger wheel with a VR or Hall sensor. TI do a helpful 5 page note explaining how to convert frequency to voltage using a 555 timer. I’d put a link but don’t know how to do this on my phone


Joined Oct 2, 2009
LM3914/5 is old school suitable for VU meters. Go with a full color graphics display and let your imagination run wild. Heck, even an Amazon Fire table would make a fantastic display.


Joined Jan 23, 2018
Creating a static display that looks really good is just a matter of learning how. The complicated part is then the code that makes it change in response to the external signals . For rapidly changing values like speed and RPM it is a bigger challenge if there is a counter, because to show rate (events per unit of time) some amount of time must pass between reading values. So the fast way to do that is an A/D conversion of the analog representation of the variable. Doing all of this at the same time with software demands fast code. And then it is complicated.

The challenge is doing all of that and updating the display at the same time.

The other choice is a display driven by the variables being measured, with all of the inputs simply being displayed as they arrive. So then the 81 step tachometer is just a voltmeter display calibrated in RPM. An 8-bit A/D channel is adequate for that with range to spare, incrementing in 50 RPM steps to be a bit smoother.
For the pulses to voltage an LM2917 is difficult to beat for that.
Attempting to convert pulses to an analog voltage with a 555 is really a waste of time, and a rather poor choice.
On my laptop now. See which should take you to the link do download the pdf I mentioned.

To quote from it "a 555 timer configured to convert an input frequency to a respective output voltage. This
can be easily applied to any application that requires the sensing of an input waveform, especially for revolutions
per minute (rpm) measuring in automotive applications".

Note that it's from September 2023, by TI for whom I have a lot of respect - not a hobby site. If you are interested to learn about electronics rather than taking the easy route with a processor and TFT display then I maintain that this, combined with "old school" cascaded LM3914s is a possible way to build your project.

I don't always agree with other people's ideas, but I try not to trash them, preferring to offer a parallel alternative.


Joined Jan 23, 2018
Certainly a 555 CAN be made to provide a analog output from a pulse train, BUT it lacks the input protection portion of a dedicated IC created for that purpose. I have also seen an LM339 quad comparator used as an F/V converter, but it was not as good, in fact, it took more parts and was a pain to work with.


Joined Oct 7, 2019
100 rpm all the way up to 8500 rpm.
You need 85 LEDs if you are going that direction.
If you use the LM2917 it would take 9 ICs. I have never used more than 2 in a meter.
Most of the "analog" options I can think of will require better than 1% accuracy.

A LCD display and a small computer can easily do 85 lights and far better than 1%.

Here is an example of an "analog" looking clock on a LCD.