I'm trying to figure out how to configure the LEDs in the tach redline to go out with increase in oil temperature like in the e39 M5 tach.
I won't know how many LEDs for the redline I have to play with until I open up the tach itself.

I figured out the basics as what I need to do in order to achieve this. Basically the oil temp sensor will give out increasing voltage as oil temperature approaches operating temp. And low voltages when oil is cold. So I need to light up all LEDs at low voltages, and make them disappear as voltage gets higher.

I'm really stuck at this point since I'm forgot almost all of my electronics training. I'd like a pretty delicate solution using IC, and avoid wiring 20 resistors in parallel or something.

AND as a bonus, I want to incorporate a shift light feature into it also like the e46 M3 tach. I know JCW has something pretty much the same coming out, and I just want to move it inside the tach. I guess the trick is to get the frequency based signal (crank signal) to light up the LEDs while overriding the temp based redline at the same time.

ANY input and help would be appreciated, thanks!

 

Step one: Buy spare tach.

I am an electrician, and most of my work is on the wiring of nuclear control systems...I have some ideas, but it's late and I'm getting sleepy...I'll run them past you tomorrow sometime.

 

lol well I SHOULD buy a spare one in case I royally kill the one currently in the vehicle, BUT I'm a poor student AND I'm fairly confident of my soldering skills. Hopefully my confidence doens't come back and bite me in the butt!

and it'd be greatly appreciated if you could shoot me some ideas! Man I'm excited. hehehe

 

Do a search here and over at MINI2 (I think) someone had already modded the tach to act as a shift light but I forget who. The thread had a pretty nice video too.

 

The guy that did it here that we knew of was named Irelend S or something like that. He has since left us and nowhere to be found. THis topic was discussed heavily in the JCW shift light forum. If you get this project completed I think there will be a lot of people that will want a full explanation of how to do this yourself. so be prepared to write everything down with parts that were included. The guy that lived in Ireland demanded that people send him their tachs and that he would mod them then send them back however noone can drive without a tach, so that was shot. But if you complete it, you are the man!

 

Yeah I saw that vid you're talking about, but the guy who made it is MIA and nowhere to be found. I can do the install myself, I just need to figure out what parts to use and how to wire it up. And I'm not looking for anything fancy like he did, I don't need the programmable functions. Looks like I'll just have to figure it out myself if nobody's able to help me figure it out.

 

Found a schematic for building a LED tach, so I'm pretty much set now. Just need to change out a few things to get it working for the MINI. Hopefully I'll get this done by end of March. Woot!

 

A little update here...well I took my tach apart and got to see what is involved. There is no way to make a neat little package that can be sold to people so they can install it themselves. This is only for the serious DIYers mainly because it's simply not "plug and play".

The difficult issue I had with this is where to get the tach signal from. The rpm signal going to the tach is "hidden" in the CAN high and low signals. Basically these are encoded signals that give information regarding DME, DSC, and other powertrain management functions. So the rpm signal is there but I don't feel like trying to figure out WHICH part of the CAN signal is the rpm one, so on to my next idea...I believe this is what the other guy figured out if he was able to contain his mod JUST in the tach itself, with the user only having to connect the connector back in for it to work.

Looked at the JCW shiftlight install, and found out they wire the tach signal from the OBD II pins. This makes sense as RPM signal is available on an OBD II compliant diagnosis tester, not just BMWs. So hopefully this will give me a nice clean digital signal which would make my life alot easier....

And finally if that doesn't work then it looks like I have to tap into the negative side of the coil and build a circuit protection for it since the voltage can peak pretty high on the coils due to the resonance.

Ok so summing up. TONS of work involved. Needs to be built in 2 parts. One part to turn the frequency (rpm signal) into voltage, and then an another part to turn that voltage signal to light up the LEDs.

As for the LEDs, the tach has 5 LEDs for the lines, which 2 of them are not used in the S. So only 3 red LEDs are used for the 6.7k-8k range. There are total of 11 lines between 5.5k-8k. The tach has plastic separaters to keep the lights apart, but the 3 red LEDs are grouped as one, so if one LED goes on it will light up a large portion of the lines. Seems like the best option is to somehow stick the LEDs onto the lines themselves. The LEDs are small enough to fit between the lines without shining into the other lines.

Hopefully that made some sense. I'll post more results as I build the circuit and let you guys know the progress. The schematic I found needs to be tweaked depending on what kind of rpm signal I'll be using, so once I figure that out I'll post the schematic and let the electronic savvy guys take a shot at making their own before I post a full blown how to.

 

More info if anyone's intersted...got to scope out the tach signal that goes to the OBD II connector on a BMW (I didn't have time to do it on my own car since I'm at school). It is a 14v digital signal, which is PERFECT for what I need to do. The 530i gave a signal of 50Hz at 1000RPM, I don't know how different it would be for a 4 cylinder motor. Anyways so one of my worries has been solved, as this would make a very neat package as all wiring will be contained in the passenger compartment.

Monday I'll be scoping out the MINI and should have all the calculations done by next weekend and soldering shall commence as soon as I get my iron shipped from home.

 

If anyone's still interested more info. Scoped out the OBDII rpm signal and got 33.333Hz at 1000rpm. This is linear, so 33.333Hz x whatever rpm is the frequency you get. The signal is about 14v, and very clean so this should make it easier.

For those with electronic know how, here are some part #s so you can get started. You'd want the LM2917 chip for the frequency to voltage conversion. Calculation is Vout=Vs x R1 x C1 x Fin
Vout is the output voltage after the conversion, Vs is the source voltage (12v), C1 and the R1 are the main variables here that you can fiddle with to get the Vout that you want. Fin is the input frequency so you'd probably want to figure out what your Vout will be for max Fin and min Fin.

Now for outputing the Vout to LEDs you need the chip LM3914 and some SM LEDs. Fairly simple, as you just need to fiddle with a few resistors to set the amperage going to the LEDs.

The surface mount LEDs you can get in a few colors, yellow, yellow-orange,red, blue, and white are teh colors I saw were available.
Go to www.mouser.com for all these parts.

 

Ok finally finished w/ the schematic. It is a basic design as any other LED tachs out there with just some adjustments to the resistance and capacitance values. I don't understand the freq to voltage chip fully, so I won't fully know what will happen until I build a prototype and test it out.

To explain the schematic...*deep breath*

The chip on the left (LM2917) does the frequency to voltage conversion. Main things you have to worry about on the chip are R1, and C1. If you change them, the output voltages will change linearly. For example, if you increase R1 and C1, the output voltage will increase in small increments. You don't want to make these values very big because you'll get inaccuracies in the conversion (I'm still not too sure what the threshold is). R1 is a variable resistor to fine tune the output voltage to make sure the LEDs light up when they should.

I have set C1 and R1 to match the output voltage. The frequency for the lower end of the threshold is 213Hz @6.4kRPM (when the LEDs will start lighting), and for the upper end its 230Hz @ 6.9kRPM (when ALL LEDs should be lit). Using Vout=Vcc x C1 x R1 x Fin. Vcc is 12.4V(batt voltage) and Vout I set at 12.4V for 6.9kRPM and 11.49V for 6.4kRPM. The reason why I made the output voltages so high is because I needed the biggest voltage difference between the two freqencies as possible. 17Hz difference isn't alot to work with so I needed to expand the voltage as big as I can for such a narrow frequency range. I don't know how else to increase the difference so I just increased the Vout at 6.9krpm. This gives me 0.90V to work with which HOPEFULLY will be enough to fully distinguish each 10 LEDs. The problem that might arise from such a low voltage difference between the upper and lower RPMs is that the chip loses accuracy and might start "bleeding" current into other LEDs. For example, when LED2 should be lit, LED3 might glow faintly. Hopefully it'll work though :\

The output voltage comes from pin10, which ends up being the input to the chip on the right (LM3914). This side is fairly simple, as the only important part is resistor 7 which controls the current to the LEDs therefore controls their brightness. Again I put a variable resistor for easy adjustment. Pin 9 is an another adjustment someone might want to fiddle with. This controls whether the LEDs light up in bar mode (what I have it set to) or dot mode. I'm not exactly sure what the dot mode is, but it's very easy to just put in a single throw switch to have the pin be open for dot mode. For bar mode just put it directly to Vcc.

So there ya go. I'm going to fiddle w/ the capacitors and resistors a bit to get them to match the values that are available, as right now I have some weird #s. Or just make them all variable resistors and just fiddle w/ it to get what I need. I'm going to try the circuit on a breadboard by next week and hopefully it'll work. Any inputs and anyone who can clean up the circuit would be appreciated.

EDIT: Sorry the schematic is such poor quality. I tried to upload my original but it was too big (exceeded 800x800), PM me if you want the original, and there is a small error in it also. Pin 11 and pin 9 on LM2917 shouldn't be connected. That dot on pin 9 by the 470 Ohm resistor is for the resistor and not the line going from pin 11 to ground

 

NICE Barn!

 

I clicked on the file thinking I'd see a picture or a short video, but all I got were some lines.

 

lol I'm not QUITE there yet. I will post a pic or a movie once I get the parts and build it on a breadboard and IF it works.

 

This sounds so cool. Any chance that it might be commercially available once you figure it out?

 

You have no idea how big a help this, I am working on a project of my own right now and was in the phase of figuring out the signal to voltage earlier today, I'd love to see your final product, hoping to have my mod ready for my track day (as a bit of an inveailing) then I'll post my results, thanks for all the info.

 

I have thought about this as I'm designing a PCB to go with it once the schematic is finalized. The chips themselves will be mounted outside the tach, and be hidden by the OBDII port or whereever. I'm planning on running a wire with a connector at the end for the LEDs which must be inside the tach. The LEDs will probably be mounted on a PCB also for a clean install.

If I do sell this, I can only sell the PCBs and you guys have to do the LED install yourself, which is the hard part actually. I WILL NOT take someone's tach and do the full install as that'd be too complicated and will cost way more than I believe its worth. Except maybe very special instances Even though I could use the $$ as I'm a jobless student lol, but I'll be posting a DETAILED how to once I'm finished complete with the PCB design so anyone with decent hands on ability can do the mod.

EDIT: It will not be more than $100 if I decide to sell this, but again dont' bank on me selling it as I'll try to make this as easy as possible. Also final FINAL version won't be ready for about a month as I want the PCB to fit really nicely into the tach without any drilling or gluing.

Motoron, glad to be of help Hope your project turns out well in time for the track.

 

Thats a lot of work building a unit when one already exists look here
http://www.datatool.co.uk/consumer/p...oduct=revlight
it is designed for motorbikes but works perfectly well on any 12v system
it has the right ammount of led's for the mod.

next get a spare tach and take it apart,
underneath each line cut and fit a piece of white card and glue it in as a divider,
carefully remove the led's from the datatool unit fit cables to make them remote mounting and glue them into the slots.
the circuit board will fit nicely behind the tach with some modification to the casing.

buy an autometer tach adapter http://autometer.carshopinc.com/prod...s_id/9144/9117
and fit it in to your coilpack positive and you will have a nice clean rpm signal
wire this to your datatool and hey presto M3 rev lights
oh and also you can fit an external pushbutton to change your revlights display and for programming.

I knew the guy who first done this and he built one for me.

 

I thought of doing that a long time ago, but what's the fun in that? Not to mention I didn't want to spend $100 and then tear it apart. It only took me about a week to figure all this out, so not really alot of work (which I don't mind, this stuff is pretty fun for me) And it's also $100 while my mod probably won't be more than $50, just labor intensive. But it is programmable, which is one thing it has over my cheapie mod.

 

yep I totally understand your feelings,
but the advantages are not only is it programmable you can change the display and it is very accurate and stable,
believe me I tried the homemade route but for the extra few dollars it is worth it.
also for those with little electronics experiencs it aint that hard to do if you take your time.

 

It definitly is an option for some, but I like the challenge and the satisfaction of knowing I made something from scratch And the programmable functions can be added to the circuit, up to a point. The sequential, dot display modes are easy to implement, as well as switches for setting different rpm ranges for it to operate. Again it's all up to the user, and I probably won't implement these in my schematic (maybe a switch for JCW and/or 8k redline settings), but to make it FULLY programmable that is beyond my scope for now.

Thanks for the input, and hope those who want a simplier soln find it useful.

 

I don't understand a bit of this but thumbs up for custom mods!

 

No problem and I hope I didn't come across as trying to undermine you,
I totally respect that you are doing this for yourself just wanted to point out that I took the lazy and easy option,
best of luck with it and most important of all enjoy.

 

haha no worries, not at all. I just like fiddling and overcomplicating things (I should be german )

I ordered the parts tonight and will be testing it out as soon as I get the parts. Total cost of parts is $26. This doesn't include a copper board for making the PCB, etching soln, and a drill to drill holes in. But if you don't want to make it all nice and neat you can just wire it up with a board with holes in it. Shouldn't be any interference problems with the wires at such a low frequency.

So there ya go, shift light for less than $30 given you have a soldering iron. End result might be a bit cheaper as I bought potentiometers for most of the resistors to fiddle with it to see what values work best.

 

I thought this is why you are making an M3/M5 style tach and not doing the american how many different colors can it be style tach