any electronics whizzes out there?

UKworkshop.co.uk

Help Support UKworkshop.co.uk:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

nev

Established Member
Joined
21 Jan 2011
Messages
4,860
Reaction score
16
Location
The green and wetter end of the M4.
I would imagine it is simple enough :roll:
I would like to replace (inexpensively) the (water) temperature gauge (analogue pointy needle type) on my motorcycle with a bi or tri coloured mini led that would be, say, amber when warming up, green when at temp and red when over temp. or just green and red for normal and panic.
it would need to be 12v
preferably utilising the existing sender, which is just a varying resistor as it gets warm.
I know the resistances at given temps eg 60c = 104 ohms, 85c = 43.9 ohms, 110c = 20.3 ohms but still have to ascertain what the 'normal' operating temps are.

so a simple 12v circuit that when reading x,y or z ohms will illuminate a, b or c colour led (on a flylead). simples!

Can anyone point me at a circuit diagram or project page on t'interweb anywhere? I have been googling but most are using their own thermistor or temp sensor that needs to be attached somewhere.

mtia
nev
 
Grayorm":i2n17cuy said:
Can't help you Nev...but what's the bike?

its a bit of a project, trying for a street tracker type of thing on a strict budget i.e nothing :roll: and want as little on the bike as i can get away with without sacrificing usability.
it started life as vt500c but now looks like this...

2013-08-26 15.15.31.jpg


trying to get it functional and mot' d , then see if its any good and if it is then strip it and pretty it over the winter.
 

Attachments

  • 2013-08-26 15.15.31.jpg
    2013-08-26 15.15.31.jpg
    100.8 KB · Views: 1,124
nev":3jqsnfo4 said:
Grayorm":3jqsnfo4 said:
Can't help you Nev...but what's the bike?

its a bit of a project, trying for a street tracker type of thing on a strict budget i.e nothing :roll: and want as little on the bike as i can get away with without sacrificing usability.
it started life as vt500c but now looks like this...

View attachment 2013

trying to get it functional and mot' d , then see if its any good and if it is then strip it and pretty it over the winter.

Nice project Nev. Keep us informed with progress. I had a Honda Varadero 1000cc V twin for a short while, didn't enjoy it much, had it between big tourers. great for one up round town stuff.

Honda Varedero June06h.jpg


Why do you want a digital gizmo for temp? I feel the need to point out that it's not water cooled ???
 

Attachments

  • Honda Varedero June06h.jpg
    Honda Varedero June06h.jpg
    252.3 KB · Views: 1,123
Grayorm":2pe0jg17 said:
...

Why do you want a digital gizmo for temp? I feel the need to point out that it's not water cooled ???

I dont want a digital gizmo, i want no gizmo! want to loose the clocks altogether (apart from the reqd speedo) and i reckon i can get into the speedo casing with an led to let me know if I'm cooking.

Those cunning people at honda have you fooled Gray - it is water cooled - the fins are for show only. look closely and you may spot a radiator :)
 
nev":16dfo845 said:
Grayorm":16dfo845 said:
...

Why do you want a digital gizmo for temp? I feel the need to point out that it's not water cooled ???

I dont want a digital gizmo, i want no gizmo! want to loose the clocks altogether (apart from the reqd speedo) and i reckon i can get into the speedo casing with an led to let me know if I'm cooking.

Those cunning people at honda have you fooled Gray - it is water cooled - the fins are for show only. look closely and you may spot a radiator :)

Ha! so there is!
 
Essentially, it sounds like you need a microcontroller to read the input signal, although generally speaking they run at 5v or 3.3v, reading the resistance on those voltages should in theory be relatively simple with any chip that's got an ADC, however, you're running at 12v so you would need a voltage divider to drop it down to a safe level or you'll fry the chip, then do some data conversion afterwards.

You would require something like an Atmel atmega 328 chip, you can buy those singly for about £3-4 or you can buy one as part of a development board, £10-15 called an arduino. These are relatively simple to program (if you have any coding experience) and there is quite mature with free development software.

How would it all work?

http://forum.arduino.cc/index.php/topic,75130.0.html

Check there for details. As for making the lights work, one of the first things you learn to do with a microcontroller is to make a 'blinky' or in laymans terms 'Turn an LED on and off'.

An arduino board is about 8cm x 5cm x 1cm but if that's too big for you, you could build it much smaller using stripboard, it really would only require a handful of components, 2-3 capacitors, 5-6 resistors, a 16mhz crystal and a regulated 5v power supply.
 
I am in no way an electronics whizz, but...

If you want the simplest device possible, you could use a potential divider circuit that switches a red led on when the temp gets above normal.

http://www.mrcorfe.com/DAOS/Year11/ElecControl/PotentialDividers.php

You would only need a couple of extra components (LED and resistor) ...maybe a couple more to protect over voltage.

I know the theory (being a physics teacher), but I'm not so hot on the nuts and bolts of it. Have a google around for potential divider circuits and thermistors.
 
Thanks Reggie and Mike.
Reggie - with your suggestion i would guess i would require some kind of interface between pc and chip to program? and a tad more knowledge than i currently possess but do-able if necessary.
I do like the simplicity of Mikes suggestion and in my head it makes sense. also looks like only a few pennies to expend to try it out.
just need to figure out what the resistances are at normal and hot, make sure i have a regulated ps and find out what voltage is required for chosen led, add a variable pot for fine tuning and away we go?
Thinking cap on :)
 
nev":2qxx9quj said:
Thanks Reggie and Mike.
Reggie - with your suggestion i would guess i would require some kind of interface between pc and chip to program? and a tad more knowledge than i currently possess but do-able if necessary.
I do like the simplicity of Mikes suggestion and in my head it makes sense. also looks like only a few pennies to expend to try it out.
just need to figure out what the resistances are at normal and hot, make sure i have a regulated ps and find out what voltage is required for chosen led, add a variable pot for fine tuning and away we go?
Thinking cap on :)

You can experiment with a variable resistor to tune the system, but I would suggest that for your operational system you use a fixed resistor. So experiment with the variable one until it's working OK, then measure the value that works and then find a fixed resistor of that value to replace it with. It will make your system more robust and less likely to go wrong due to dust and vibration.
 
Keeping it simple something like an LM3914 bargraph display driver will do the job
http://www.maplin.co.uk/lm3914-bargraph-display-2909

It sequentially lights LED's as the input voltage rises. Use a variable resistor in line with the input to get the signal voltage from your temperature sender down to the required level then use whichever outputs suits the temperature range for your needs i.e. first three joined together for low, next four for warm and the last three for high.

Gerry
 
If you bought the atmega328 chip on an arduino board, it would already have a programming interface to the PC built into the board.
 
Hi,
£20 isn't bad considering...
I've just seen your post and being an electronics engineer it got me thinking, the simplest circuit would involve three comparators, a chain of three resistors, a bit of logic to decode for the leds and a power supply for everything. You would spend a good few hours assembling the circuit and (if you have little or no experience of these things) there's every chance it will not work first time. It'll also take me at least an hour (optimistic!!) to design it for you and then some time (deffo more than an hour)to draw the circuit it out, work out a layout and explain everything. There's also the cost of your components on top of all this.
Given all that £20 for a ready built unit is bl**dy cheap :wink: Buy it.

Cheers,
Steve.
 
I agree that £20 is cheap, considering the amount of work you'd have to do for a one-off. I could do this with a 50p microcontroller, but then you have to factor in the power supply components(always a problem in vehicle electrics) and other stuff. Now if you wanted 1,00 of them....
 
20 quid is more than reasonable (for any hand made item), its just that like most people I dont like spending money if i can do it myself (hammer) . I did build electronical things many many years ago when I was school age - I think it was 'Simple Electronics' magazine projects like fuzzboxes, Dalek voice changers etc. great fun for teenager 8) so I am not completely new to soldering etc.

But you are correct even if its a fiver or a tenner for the components, by the time of shopped, gathered sundries like heatshrink etc and a few quid for fuel I will probably have surpassed the 20 quid price.

But ... i wont have the satisfaction of having created myself. :(
 
When do you want it though? Now or in 6-12months? If you don't own a soldering iron, that'll set you back 20quid, then solder on top of that, you'll also need some flux, a pair of flush cutters, some tweezers etc. If you learnt how to use eagle, you could get seeedstudio to knock you up some boards relatively cheaply but you'd still have to source the parts and populate the board yourself, then there's troubleshooting any design issues you didn't spot, getting boards re-made, or doing the boards yourself, in which case you'll need etchant, fr4 pcb and access to a laser printer. Alternatively, you could use stripboard but it would be bigger and chunkier and would like like it was from 1974 or you could use protoboard which will look like a box of dead spiders.

I would say spend the £20 a buy a unit, save your creativity for an issue that needs solving that doesn't already have a cheap off the shelf alternative.
 
Back
Top