Pulse inverting etc

"Periproct" schreef in bericht news:dsssgh$ror$ snipped-for-privacy@nwrdmz02.dmz.ncs.ea.ibs-infra.bt.com...

Seems you have to convert the signal from the sensor to the format normally given by the speedo, but you forgot to mention the pulse comming from the speedo. Is that a 5V positive pulse? Below just a quicky to give you an idea what the circuit you need may look like. R1 may be 10k, R2 220k up to 1M depending on the load your sensor can handle. Remind I want to give you an idea. It's not a tested design. +5V +------+----------+-------- | | | | | .-. - | | | ^ | | |4k7 | | '-' | | | | | +------- OUT || R1__ | |-+2N5457 | IN---||---|___|-+-++-->| ___ |/ || | | |-+-|___|--| BC549 10nF | | | 22k |>

| | | | .-. | .-. | | | - | |1k2 | R2| | ^ | | | '-' | '-' | | | | | --------------+--+-----+----------+------ GND

created by Andy´s ASCII-Circuit v1.24.140803 Beta

petrus bitbyter

Thanks a lot for your response.

I think I was too quick off the mark with my question.

Having fitted the signal from the speed sensor (I guess some Hall sensor somewhere on the transmission) to the Sat Nav instead of the signal from the ECU to speedo I only drove a short distance to test it. I have now tested further and found the two signals behave exactly the same. The Sat Nav doesn't register the car is moving before 20mph and stops registering movement when the car drops below 10mph. I didn't check the speedo signal with the scope but will tomorrow.

The pulses I saw on the scope are very narrow and 'pointy' and I'm wondering if I need to not only invert them but square them off and maybe increase them to 5 volts. I'm also thinking that the Sat Nav may respond earlier to more 'pulses per mile'. IE, it won't register the signal before it gets a certain number of pulses/second. Can I have a device to double (or triple maybe) the pulses?

So you're saying a signal that's between the positive supply and 1V less than that ? or is it still 12V to 11V with the battery charged up to 13.5V ?

what supply is available to run the converter?

yes. but the problem needs to be clearly specified. often people say "12V" around cars when they mean the posotive supply which varies and is usually more than 12V that's why I'm asking again.

To change the pulse rate you may need to modify (or replace) the speed sender.

Bye. Jasen

It can be done but it's not easy you're basically looking at a microcontroller with a rather convoluted phase locking loop program.... if you have ABS and can tap into the signals from one of the senders, or can add another sender that may give better results.

"Periproct" schreef in bericht news:dsu15s$qhq$ snipped-for-privacy@nwrdmz01.dmz.ncs.ea.ibs-infra.bt.com...

First things first. The circuit above is supposed to make 5V positive pulses. But it's still uncertain whether or not the circuit will do as expected. Next question: "Will Sat Nav will accept them?" Once we know the shape and size of the pulses required, we can look for adding extra pulses. That will require a more complicated circuit or a (PIC or AVR) micro.

petrus bitbyter

Unfortunately, no ABS. I had thought about the posibilty of changing one hub to ABS or adding an ABS sensor if it is possible. Another idea would be to bolt a Hall effect sensor to the dif and attaching one or more magnets to the prop shaft.

A break through. I've put a signal generator into the Sat Nav and fed it a 5 volt square wave and it works a treat. The Sat Nav now reads right down to 1 mile per hour (at the appropriate frequency). I wound the voltage right down and the Sat Nav still responded to quite a low voltage but I omitted to measure it. (20Hz = 10mph, 200Hz = 100mph)

My pulses at low speed look something like this:-

_____________ _________________ ______________________ + 12 \ / \ / V V

• 4

Hope that came out OK. Do you reckon the Sat Nav is trying to read the long

I've checked voltages and I have the full battery voltage coming from the sensor. The negative pulses are 8.25 volts. The Sat Nav has a diode on its speed pulse input (marked E 2L, band towards the sensor) and the voltage of the pulse on the Sat Nav side of that is around 2.6 volts.

My thoughts are that I now need a PCB with a voltage regulator that I can run from the car battery supply. Something to get rid of the full battery voltage and leave me with my negative going pulse. Then something to invert to pulse and maybe clean it up so I have a nice 5 volt square wave to feed into the Sat Nav.

With the Sat Nav calibrated to my present speed signal I can make it read right up to 200 miles/hour with the frequency generator so I'm still quite keen on the idea of pulse doubling. My thinking is that more pulses will improve the response/accuracy of the Sat Nav. And does it need it!

I've had a google and found a couple of ideas for pulse doubling - it's just the other bits I'm going to have trouble with.

I got that slightly wrong somehow the first time. Full battery voltage from the sensor with negative going 8.25 pulses.

I'm thinking that would be the easy bit. Just a regulator providing 5 volts from the car battery.

Yup, the full battery voltage varying depending on if it's charging etc.

Hi all.

Many thanks for all your feedback.

As you may have guessed I'm a bit of an 'all the gear, no idea' person. I grab any test gear I see going cheap but I'm lacking the knowledge to build the circuits.

To sum up, my Sat Nav reads the car speed from the speed pulse which is a

I know know that I can feed a 5 volt square wave into the Sat Nav and it works a treat.

I just need a teeny bit more info on how I can take my 12 volts with negative pulses and convert to nice clean, square positive pulses. I can probably work it all out from manufactures data sheets if someone can point me at the right chip/chips to do the job.

Many thanks to Petrus for the diagram. It just needs me to do a bit of work to read it. And thanks to all the others for their patient advice. Nigel

"Periproct" schreef in bericht news:dsvnev$2n9$ snipped-for-privacy@nwrdmz03.dmz.ncs.ea.ibs-infra.bt.com...

Gues you'd rather have to get rid of the 4V or so DC as the voltage on the Sat Nat input never falls down low enough to be seen by it. You can do so by using a capacitor. The circuit below once again makes 5V positive pulses. When this works, you can go ahead to look for frequency doubling. AFAIK you will need either a PLL or a micro.

AFAIK the appnote 3327 is meant for fixed frequencies. Don't expect it to be usefull for you. The LM567 circuit are oscillators. Don't see an input to control them by the speed sensor.

petrus bitbyter

Snipped

Thanks a lot. I really appreciate your circuit diagram. (They display perfectly when pasted into notepad) Tomorrow I'll make a start.

I thought I'd found just the thing there but never mind. The pulse doubling may not be necessary. I am just beginning to dabble in PIC programming so I may eventually be able to sort that out for myself. When I was 'googling' I came across the suggestion doubling by triggering one pulse on the leading edge of the original pulse and another on the trailing edge. At slow speeds my pulses are a long way apart so there would be two pulses and a gap so I don't know how that would work. The Sat Nav seems to take a second or so to respond to changes of frequency on its input so it might work. Something for me to try. Nigel

you should really use a fixed pitch font.

. . "+12" . ----------+-----+----------------+--- . | | | . - | | . ^ | | . | | | BC556 . | |/ BC546 |< .---[100K]--++---| +-| | |\ . | | | | out . | +---|

only works with a 50% duty cycle square wave input

oscilator not doubler.

Bye. Jasen

That'll pick up every variation in the "12v" supply to whih the inp[ut is referenced...

Bye. Jasen

Just a regulator will be a problem it's called 12v but is sometimes much more and having a transient or fault destroy your work is no fun

posibly something like this: 100ma ----0-_-~-o--[100R]----+-------[7805]---- 5V@50ma "12v" 1W | | _|_| | |/^\ 15V | ~T~ 2W | GND | | -----------------------+----------+-------

Bye. Jasen