23 replies to this topic

[Tomf]

OH thanks i have dont really know where i can get the regulators so i searched ebay again lol. i couldn't find the u7810 chip or the LM317 chip but i could find the LM317T chip, will this work?

LM317T ebay link

[Tomf]

BUMP

[syholl]

The LM317T is an excellent one to use, and is the one i normally fit, but you have to build a little circuit with capacitors and resistors to get it to work. For a simpler one, go for something like this:

http://www.maplin.co...e...age&doy=1m6

Usually the product code contains the voltage, so the 7th one down will be the 10v one. Maplin are very cheap, and postage is reasonable as well. Much better than that Ebay one (sorry, i didn't notice that ridiculous price!!!)

You can download the PDF, which shows a couple of capacitors connected. You don't need these for a simple thing like a fuel gauge, it will work fine without them.

[Tomf]

Great thank you very much

Just to confirm before i buy it, is it this one that i need, its the 7th one down but i cant see anywhere where it says 10V.

Description: TS7810CZ Code: N38CA

or do i have to build up my own circuit with capacitors and resistors to get the correct voltage input and outputs?

Edited by Tomf, 01 June 2007 - 07:53 AM.

[dklawson]

As Syholl said, the LM317T will be fine. I've used them before.

However, capacitors are not necessary for the LM317T in this application. They are typically added to filter out noise and transients for very clean DC output. Download the PDF I provided a link to earlier and it will show you that there are only two resistors required to form the feedback loop and set the output voltage.

Before leaving the subject of why the capacitors are not needed I'd like to explain a little bit more about "why not". Remember that you are building a simple DC power supply to replace the Smiths voltage stabilizer. The Smiths stabilizer is a bimetallic switch. Current flowing through the stabilizer causes a resistance wire to heat up, deflect a bimetallic strip, which then open a set of contact points interrupting current flow to the gauges. When these points open and the current flow stops, the bimetallic strip cools which in turn allows the contacts to close again. This happens fairly quickly causing voltage swings up and down... delivering an time averaged output of 10 volts. You don't see this on your gauges because they work through the heating of yet more bimetallic strips. The thermal mass of the system keeps the response time slow. Anyway, the filter capacitors are NOT needed with the LM317 or u7810 because they inherently deliver a smoother, more steady output than the Smiths stabilizer.

[Tomf]

Thank you .

So if i buy the LM317T, i will just have to use the correct resistors and no capacitors? If that is the case it should be very easy to wire up and get the gauge working.

[dklawson]

The formulas for determining the resistor values and pictures of what this will look like are at:
http://home.mindspri...eStabilizer.pdf

After calculating the resistors you think you'll need, buy several of each rating. Resistors (like everything else) are subject to "tolerances" so clip them in place and measure the circuit's output voltage before soldering anything. You may find swapping a couple of the components you buy will give you results closer to 10V. Aim for an output between 9.9 and 10.1 volts.

[Tomf]

Thanks ill order the stuf now then and when they come ill mess about with the resistors and a multi-meter and see when i get a 9.9 - 10.01 output

Thanks for the help, hopefully i will have that sorted very soon the

[Tomf]

EDIT: sorry dont worry i have found the answer now

Edited by Tomf, 01 June 2007 - 11:17 PM.