Home made speedo

[Gareth Dickens]

Do you use km/h or mph speedos in South Africa?

We use km/h.

[FridgeFreezer]

Fairly sure you can get a BusPirate sent anywhere in the world, they're not expensive and you just need a USB port and a few bits of wire to hook it up. However, if you're not that way inclined you're probably much better off just buying a VDO Speedo & configuring it to suit.

Acewell do some nice multi-gauges for good prices.

[discomikey]

maybe it would be as simple to superimpose a series clock "face" onto a TD5 gauge? after all for me that should about do the trick as standard defenders have 3,54:1 diffs and can have 7:50 tyres too that should clock me about right. although i assume its not a simple fit for the transponder onto a series transfer box.

[FridgeFreezer]

I doubt it would be terribly hard to clamp the TD5 sender to the side of a Series box. The mounting is very similar.

[Gareth Dickens]

It will all come down to cost and hassle from here. If all else fails there are nice GPS speedos from Yankville, I mean America, for about 280 $. No hassle. Never need recalibration for tire changes, gearbox changes, diff ratio changes, get my drift? If you drown it you have not walked the water crossing before attempting it.

So that price would set the bar for me w.r.t. value for money.

[CwazyWabbit]

Well if it were an mph speedo I'd have programmed a TD5 one up for however many pulses per mile you needed and posted it to you, unfortunately I haven't seen any kph ones over here for a while now.

[missingsid]

Don't know iif S1 is different to S2/3 but my cable is held on with three screws not one like defenders. Space is also a consideration.

look on ebay/internet there are millions of different senders available Rover/Jaguar senders are small. Smiths even use an optical sender.

Putting a series face on another dial is OK but you need to match the start position and the swing of the needle.

Another option for a sender is a short cable with a sender on the end, these are available from the US for Hardly Dangerous motorcycles or even the transducer from a RRC.

[wood-gee]

Really interested to here about this td5 sender. I've been ticking over on how to do this since first posting this and my plan this week is to have a go at trying to achieve it via a magnet attached to a jubilee clip on the brake drum as a trigger like some of the guys suggested, (well first mounted on the end of a drill as I can test that more easily on the bench). In honestly this is largely because I have a magnet and a bicycle speedo sensor, which as far as I understand is a switching hall effect sensor (or reed switch?), either way, a switch that's actuated by a magnetic field, and also have arduino etc. so if it doesn't work I haven't lost any money. If it does work I was then going to get it set up in the car with an LCD to actually display it, and then at a later stage try and copy Simonr and drive the speedo using a servo motor (CwazWabbits link looks like that job may have got a lot easier).

Will post the code and details on how I did it if I have success so others can have a go.

[Gareth Dickens]

I would still really appreciate it a lot if anyone could compare drum/propshaft rotations to speedo drive (cable) rotations in 4th gear high range on a series gearbox (preferably series 3) or transfer box, I'm not quite sure where the speedo cable is driven from.

The same would be very useful for td5 box as well. Would just have to factor in the difference between the 3.54 and 4.7 diff ratios.

Will post the code and details on how I did it if I have success so others can have a go.

For me you have to dumb it down a lot.eg. Solder red wire from processor p/n: abc123 to pin 1 of sender p/n: def456 and add pics.

[wood-gee]

Ok, so I've had a play today and got the programming side of it mostly covered. Using a bicycle speedo sensor (reed switch), and a magnet which I attached to a drill to simulate the rotating magnet on the brake drum I managed to get it reading off speed to my laptop.

I connected my arduino so that there was 5V across the reed switch, with a resistor in series with it, meaning that when the switch is open, no current flows through the resistor, and thus no voltage. When it closes a current flows, so a voltage is present across the resistor, and I then measured this using the Arduino.

The code compares the current status of the switch with the previous, ie has it gone from closed to open, allowing it to detect the magnet passing it.

Next step is to mount magnets to the drum and mount the sensor there, and get hold of an LCD screen to display it on.

Costs so far are:

Arduino mega clone - £20.99 (Amazon)

Wires & a resistor - not much ('borrowed' from university)

Bicycle speedo sensor - nothing (a spare in my parts box) but could be bought for around £5

Self adhesive magnets to mount on drum - £5.10

Screen - £13.30 (yet to buy)

Will post details of how it's wired together once I've finalised how it's going in the car, and the code will be updated when I get a screen, but here is a copy of my current code for anyone that's interested or is wishing to have a play:

//define variables and set initial conditions

int highpin = 7;

int sensepin = 0;

int val = 0;

int valold = 0;

int sbit = 0;

int m=0;

int ttotal=0;

int i = 0;

int dt=0;

float circumference = 0;

float Disp =0;

float diffratio = 4.7;

float vel = 0;

float velmph = 0;

void setup () {

Serial.begin(9600); //initialise serial port

pinMode(7, OUTPUT); //set sensor powering pin to output

digitalWrite(highpin, HIGH); //set sensor powering pin to 5V

circumference=3.14*787; //calculate circumference of wheel (787 is approx 31 inches in mm)

Disp=(1*circumference)/diffratio; //displacement at prop of 1 revolution of wheel

Serial.print(Disp); //print effective circumference of wheel at prop

Serial.println();

}

void loop () {

val=analogRead(sensepin); //read sensor

m=millis(); //calculate time

dt=m-i; //calculate tie difference between this and last reading

i=m; //set current time to old to allow next dt to be calculated

if (val > 3 && valold < 3) {

sbit=1; //set to 1 if a pulse is detected

vel=Disp/ttotal;//calculate velocity

velmph=(vel*3600)/1600; //calculate velocity in mph

ttotal=0; //reset counter to 0 to restart counting

}

else {

sbit=0; //speedo pulse bit

ttotal=ttotal+dt; //calculate total time since last pulse

}

//print values to serial port for problem shooting

Serial.print(val);

Serial.print(",");

Serial.print(sbit);

Serial.print(",");

Serial.print(ttotal);

Serial.print(",");

Serial.print(vel);

Serial.print(",");

Serial.print(velmph);

Serial.println();

valold=val; //set value to old to allow for comparisson

}

[discomikey]

I would still really appreciate it a lot if anyone could compare drum/propshaft rotations to speedo drive (cable) rotations in 4th gear high range on a series gearbox (preferably series 3) or transfer box, I'm not quite sure where the speedo cable is driven from.

The same would be very useful for td5 box as well. Would just have to factor in the difference between the 3.54 and 4.7 diff ratios.

For me you have to dumb it down a lot.eg. Solder red wire from processor p/n: abc123 to pin 1 of sender p/n: def456 and add pics.

considering the speedo drive is on the output shaft of the transfer box, the rotations it sees will be more than the wheel rotations by 4.7:1 (diff ratio) and then the amount of rotations of the speedo drive per rotation of the output shaft will need to be taken into account of

[Gareth Dickens]

considering the speedo drive is on the output shaft of the transfer box, the rotations it sees will be more than the wheel rotations by 4.7:1 (diff ratio) and then the amount of rotations of the speedo drive per rotation of the output shaft will need to be taken into account of

I agree. My Toyota box speedo drive rotates once for every 3.125 rotations of the prop (in 4th high range). Therefore, for every 1 wheel rotation I will have 4.7 / 3.125 (1.504) speedo drive rotations. I am hoping that by some miracle that the series box speedo drive ratio ends up the same. That would mean fitting the series speedo gauge I have to my Toyota box with no correction required. A really long shot, I know.

All I'm trying to find out at the moment is how large the difference is between the series box and toyota box.

[DaveSIIA]

Recollection is 2.2 turns of the propshaft flange / handbrake drum gives one turn of the speedo drive.

[Gareth Dickens]

Recollection is 2.2 turns of the propshaft flange / handbrake drum gives one turn of the speedo drive.

That would mean that my speedo will under read by about 30%, and that exludes the difference in wheel size.

[DaveSIIA]

Approach the problem from the other direction - which speedo gives the correct reading with the gearing you have?

Turns per Mile = 63360 / PI / Wheel diameter inches * diff ratio / speedo drive ratio

TPM is the number of turns the speedo cable makes in a mile.

e.g. for a 88" LR on 6.00x16 tyres TPM = 63360/3.1415927/28*4.7/2.2 = 1538.8

The TPM value is usually marked on the face of the speedo, around bottom dead centre. Standard values for Series LR speedometers are 1408 (7.50x16), 1504 (6.50x16) & 1536 (6.00x16). Early Range Rover Classic used 1152 and Defenders are around 1000. RRC and Defender mechanisms can be fitted in to a Series speedo housing with a bit of thought.

[wood-gee]

Update to this for anyone interested, got hold of an OLED display to make for a cleaner approach as it's of a size that I can put it in place of the charge warning light (had previously been using a 16x2 LCD display, but it meant attaching it on the size of the dash and it looked a bit whack). Pictures attached below:

Shows the land rover logo on start up, then the main screen with information (also currently have it to show a side profile of a SWB on start up just cos it seemed a fun idea but forgot to photo this.

Further plans are to add leds either side, one to replace the charge warning light, and the other as a further warning (as yet undecided), use a voltage divider to take a reading off the temperature sensor (this is what the 95'C placeholders for), after calibrating the sensor using a multimeter, thermometer and pan of hot water, voltage divider to measure battery voltage (the space that '1700 RPM' text is shown in will change,so you can cycle through modes showing battery voltage, max speed etc).

Please excuse the slightly tacky ali plate, I'm goign to redo it and spray it black when I take it out to put in the led's.

[Ex Member]

Looks great.

For whatever it is worth, my Lightweight's speedo is dead on so whatever is causing yours to be off can be fixed. You can fit a defender VSS to the series transfer case and a VDO speedo to the dash (which are programmable). I've see it done. For the crazier people, you can purchase devices that electronically drive the mechanical speedos, allowing you to keep the old speedo and look with modern control.

[FridgeFreezer]

That's very cool!

[Retroanaconda]

Fantastic mix of old and new, looks pretty cool

[wood-gee]

I've actually fixed the original speedo, which is spot on too, it turns out it was down to a britpart cable! (The inner is a little too short on them, so slides down the outer into the speedo drive, meaning it doesn't engage properly in the speedo, hence why it would work for a bit if i removed and replaced it then stop again, hope this is useful to anyone with a similar problem).

[Bowie69]

wood-gee, same problem here.... gotta wiggle it till it works, then fine for ages

[simonr]

Wood-Gee - that looks fantastic!

If you'd like a tidy, 3D printed bezel for the display - you draw it on the CAD and I'll print it for you for free!

Si

[wood-gee]

Simonr - Thanks for the offer, that's a great idea! I'll pm you when I've finalised the design if I decide to go down that route that's OK? (Could be a wee while yet).

[simonr]

No worries,- PM me when you have something! I've printed a few boxes & bezels for projects and they turn something built in a shed in to something that looks almost bought!

Si

[xychix]

if someone can create a simple programmable speedo system which uses a series gauge i would definately be interested in borrowing the idea, maybe paying for someone to bewitch the microcontroller system for me! haha

I'd have no problems coding the thing.

Just need to find a sensor.

I tend to buy stuff at banggood.com or dx.com (the mini / nano arduino clones).

Just need to find a sensor that measures rotations in a way the duino can read. the rest is peanuts.