can I hire somebody here?

I put the F-V circuit together, it works and I have made a pretty neat sequential 10 LED shift light with it.... onwards and upwards, just need to build the differentiator now!

you are almost done wish you the best

I would like to take the tach signal wire, measure the engines acceleration rate using it.
IF the acceleration rate is too high trip an output wire to ignite my stand alone shift light (which is NOT hooked to anything, and this is not a RPM based shift light)

Click to expand...

Believe it or not I DO understand what you're trying to accomplish. A "pulse width modulation variation indicator". A light that will come on when your motor revs too fast in too short of a time.
Sooo, you need to tell the forum...

1.) What lower RPM do you want to ignore (if it revs too fast from 800 to 1200 and you don't want that
to apply, advise to ignore below 1200).
2.) How frequently you want the sampling (I would think ten times a second would do, see next line...)
3.) How much of a change trips the light (if the engine is at 2500, and in a tenth of a second increase to
2750, that would be a 10% differential, which is truly unlikely, because IF the engine COULD do that
it would flat line from idle in about two seconds). The differential should be less than 10%...
4.) Finally, indicate if you desire to have potentiometers to adjust the low RPM bypass level, and the
differential level, and indicate what would be the ideal ranges for such (ex.-bypass level range from
800rpm to 2000rpm, and the differential range from 4% to 8%), so you can fine tune the circuit to
your liking.

Then, the PWM (pulse width modulation) pros here might be able to extract a sampler/comparator circuit to fit your needs. I myself do not know how to make this circuit, and am just suggesting some
theory for you, so don't call me names or tell me I'm not trying to contribute.
Even though the number of pages and "beratement" you believe to have received seems excessive to you, you have also been "feeding" it as well. To be honest with you, with some of your language, I'm actually suprised they (the site auditors) haven't cut you off. I ask simple questions here as well, and many times the responses have nothing to do with my question, or it gets too complex for the simple inquiry I started, but you should do what I do. Try to keep your cool, try to capacitate EXACTLY what others are writing, take a deep breath before bouncing back, and above/beyond all..., remember..., without the "pros" here, you may as well not try to build your project because you won't get any insight towards it. Even a punk like myself without a high school diploma may be able to contribute, but if someone ticks me off, I won't try anymore to help them, and most of all, I'm not perfect (nor is anybody else who visits this forum). I don't have all the answers, and if I can help, I want to make sure I spend my own personal time figuring something out for someone, once, by knowing exactly what it is they want.

Hey Hippie, I just want to say thanks. You have posted a very useful and informative post IMO. I was rather rude and I guess I need to apologize to everyone here.
I did give thanks to Roff and Mbarazeen b/c they were pretty helpful and nice, without them I would not have been able to build that sequential shift light. I think I was primarily upset with TCMtech b/c he came off as a very knowledgeable person when it came to circuits and it didn't feel like he was really trying to help me. At any rate, again, thanks, I do understand what you are saying.

I bought the off the shelf unit which was just released (odd, as a side note, does anybody believe in a "collective sub-conscious" hehe)?

BUT even with buying a unit that is probably better than what can be made at home I still want to try and make one, I love learning new stuff and I've always liked electronics.

let me see if i can answer some questions you posed b/c I didn't know i would need something like that. I'm glad you posted



1.) What lower RPM do you want to ignore (if it revs too fast from 800 to 1200 and you don't want that
to apply, advise to ignore below 1200).

The minimum apply should be around 3000 RPM, if this is simple to build into it?



2.) How frequently you want the sampling (I would think ten times a second would do, see next line...)

I would think 10 times a second would be great!


3.) How much of a change trips the light (if the engine is at 2500, and in a tenth of a second increase to
2750, that would be a 10% differential, which is truly unlikely, because IF the engine COULD do that
it would flat line from idle in about two seconds). The differential should be less than 10%...

most motors make very little power under 2500 RPM, even a mild race motor can rev from 3000 to 6000 RPM in 1-2 seconds without problem. it is really hard to say but this would need to be adjustable for fine tuning. I would actually think a differential adjustable from 5-10% would be about right



4.) Finally, indicate if you desire to have potentiometers to adjust the low RPM bypass level, and the
differential level, and indicate what would be the ideal ranges for such (ex.-bypass level range from
800rpm to 2000rpm, and the differential range from 4% to 8%), so you can fine tune the circuit to
your liking.

all that sounds good, but I think the RPM bypass level could be adjustable from 2000-4000 RPM, and the differential range from 4% to 8% would probably work.
Is what I think would also be very useful is if the "trip time" was also adjustable from maybe 1/10 second to 1/2 a second. It would actually be nicer IF the light was constantly ON and dimmed (ie, voltage reduced) as the duration of trip time increased (trigger 1/10 second, 50% power reduction by 1 second, 100% power reduction by 2 seconds).this "light" should be assumed to be a very high power (10 amps!)

this is actually sounding rather complicated at this point, but you asked and if I was dreaming it is how I would have it operate.

Even us people that you think have given you a hard time are really trying to help. If you go back to the start and reread the whole thread you'll see that you have changed your mind a couple of times about what you want and been extra secretive on this. Almost like you want someone to design this for you and then you'll market/sell them.

As I stated before you want to measure when the engine revs faster than a certain amount in a certain time. A tach output will only give you what the RPM's of the engine are at a specific point in time. Where are you going to get the value to measure your tach output against? You want basically a difference measurement but you only have one value to measure.

A traction control system measures engine RPM against wheel or transmission output RPM.(I know not what you want, just an example) Thats do able, two values. But you need to come up with some thing to measure your engine RPM against. Maybe throttle position?

generally traction control would compare the front and rear wheel speeds against each other.

that is not exactly what I'm after as I could go to a salvage and grab all that stuff easily.

I have a frequency to voltage converter that outputs from .2 to 1.2 V dependent on RPM, I also made what i thought was a working circuit that would measure the rate of change in voltage. It worked in yenka but as soon as i powered it up the chip got real hot and I think it fried. In this diagram I just used a POT to use as the "source voltage" b/c it is easy to move. I took some time working with the capacitors and resistors to get a nice "feel" for rate of change to trigger and length of time it was on. this is a far cry from the dream circuit in my last post (which IF I was really dreaming it would take the tach signal input and output a mirror of it to feed the coil through the circuit and retard the timing or"delay" the pulses exponentially in relation to the rate of change over X rate of change)

anybody that might know why this didn't work?

you have to reselect the values for resistors, also how did 22uf came to the feed back? for differenciator capacitor only comes at the input. use 1 uf instead of 220uf and calculate a ressitor for feed back (instead of 1 Ohm) using the formula already posted. use Vin/dt= ?? better amplify the out put of the Freq-volt convertor using the same op-amp stage so the your range would now be from 2-12 (when at a gain of 10) thus the sensitivity of the circuit is improved.

for the last stage select some 100 k instead of 4.7k. and use 100 k pot at the place of 5 Ohm.

How much did the unit cost that you ended up buying? Was it more than the $20 you originally offered for someone to design this for you?
If you had no secrets and offered to pay the right money there is enough talent here to have done your project ages ago.
I just found this thread and it got my attention because I just mede a very similar thing to what you want but my item is designed to protect my engine if something in my driveline blows while racing and was custom made for my car only.
I have a 9 second turbo charged injected car with a hall sensor ignition. But my project is based around a Pic which gives me more fine tuning.

I just found and read this thread right through. It seems to me that Joe rocket did explain what he was after from the first page.


This one from the last page :
" As I stated before you want to measure when the engine revs faster than a certain amount in a certain time. A tach output will only give you what the RPM's of the engine are at a specific point in time. Where are you going to get the value to measure your tach output against? You want basically a difference measurement but you only have one value to measure.

A traction control system measures engine RPM against wheel or transmission output RPM.(I know not what you want, just an example) Thats do able, two values. But you need to come up with some thing to measure your engine RPM against. Maybe throttle position? "


Why not do the sampling at a high speed rate then the clock cycle of the device can also become the time comparison ?

zorbzz said:

How much did the unit cost that you ended up buying? Was it more than the $20 you originally offered for someone to design this for you?
If you had no secrets and offered to pay the right money there is enough talent here to have done your project ages ago.
I just found this thread and it got my attention because I just mede a very similar thing to what you want but my item is designed to protect my engine if something in my driveline blows while racing and was custom made for my car only.
I have a 9 second turbo charged injected car with a hall sensor ignition. But my project is based around a Pic which gives me more fine tuning.

Click to expand...

It was a LOT more than $20, but b/c it was new on the market with no feedback they said they would sell it at cost to me. I would have paid more than $20 if I knew, or if anybody would have said something (part of the $20 offer was 1: me being cheap and 2: I wanted something simple i could understand hehe ... but I have learned enough I think I would like to examine or now have a little better ability to try a harder circuit).

Care to explain or show your circuit and tell me about the PIC?

meatpie said:

I just found and read this thread right through. It seems to me that Joe rocket did explain what he was after from the first page.


This one from the last page :
" As I stated before you want to measure when the engine revs faster than a certain amount in a certain time. A tach output will only give you what the RPM's of the engine are at a specific point in time. Where are you going to get the value to measure your tach output against? You want basically a difference measurement but you only have one value to measure.

A traction control system measures engine RPM against wheel or transmission output RPM.(I know not what you want, just an example) Thats do able, two values. But you need to come up with some thing to measure your engine RPM against. Maybe throttle position? "


Why not do the sampling at a high speed rate then the clock cycle of the device can also become the time comparison ?

Click to expand...

Hi, in the start of the thread he said no PIC and no programing.

shortbus= said:

Hi, in the start of the thread he said no PIC and no programing.

Click to expand...

Hmm , That is true and you would be guessing that's what I meant.

when I went to sleep I had a simple idea, basically two F-V converters, one with hysteriasis (sp?) and feed both outputs into a 324 op-amp. easy, right?

seems that is a very easy and cheap way to do it.

OK, so then my op-amp is only going to be putting out a comparable voltage to my trigger voltage, I would then need to feed it into as mosfet right (so I get full voltage?) (main line should be 14v, but the trigger voltages are 5v) .. basically I need to use the op-amp to trigger a relay, in other words, how do I make a hi-amperage relay with electronics?)

I need the transistor to be BEFORE my circuit, not after, all the setups I have used in yenka the load needs to precede the transistor (works using an n-channel mosfet)

yea, here is how, i can use an lm324 for the opamp, what transistor PN would somebody suggest (8 amp)?

**broken link removed**

i think you are in a wrong direction, whats the idea of using two F-V conv? you are still on the same project? or you wana do some thing else different?

same deal, what's wrong with doing it this way? If one F-V converter has hysteresis timed slightly different than the other at a slow rev they should remain near equal in output voltage but at a fast rev one will surpass the other in output voltage tripping the opamp.

how can you achive rate of change of freq?? i remember your primary idea is not to trip for RPM but to trip when the change in RPM occurs above a certain rate.

? frequency changes with RPM, voltage is linear in the chip you suggested but it also says that hysteresis capabilities, if one chip has that and the other not on too fast of a rate of RPM change won't they offer different voltages?

Both voltages will alter exponentially with the pulse width. If you set one converter for a lower voltage than the other, they will never change in their difference (both will vary no matter what speed the motor turns). You need to store (a sampled) signal, sample the signal again and compare it to the previously stored signal, react as required, store the newly sampled signal and then repeat (sample / compare / react / store)...