smoothing out voltage spikes and dips?

[scampo77]

Re: smoothing out voltage spikes and dips?

before this gets to far along it needs to be said that i will need to build a second one of these things that operates identically to the first. i was hoping to steal ref voltages and power supply voltages,

[ben7]

Re: smoothing out voltage spikes and dips?

I added a transistor into the circuit to amplify the Vref current capability, and added a diode so C3 won't discharge through the potentiometer.

R3 is what discharges the capacitor, but the capacitor still could charge up almost right away, so it only solves half of the problem.

I don't know if there should be a resistor to limit the base current on Q1, but I am thinking you could use a 2N3904 as Q1.

The voltage across the potentiometer equals Vref-Q1vbe (I think)

Could someone point out if I am going in the right direction or not?

[scampo77]

Re: smoothing out voltage spikes and dips?

i have worked on this and to no avail. i am slowly adding caps and changing things but im still not 100% clear what i should be changing and doing.

the diode?
the 9.35v?
do i change all my resistor values to match up with Ben7's diagram even though i cant change the 100 ohm float?

i have re done my schematic (please be kind its my first time) it is easier to see what i have done and what i was thinking, with the colors.

[scampo77]

Re: smoothing out voltage spikes and dips?

is this new diagram based on me keeping the Rsensor as a 0-100ohm float? and it is empty (one bulb lit up) at 5 ohm? and full (all 10 bulbs) at 85 ohm?

[ben7]

Re: smoothing out voltage spikes and dips?

is this new diagram based on me keeping the Rsensor as a 0-100ohm float? and it is empty (one bulb lit up) at 5 ohm? and full (all 10 bulbs) at 85 ohm?

Yes it is based on the 100ohm potentiometer. it should be like this: 1 bar-10ohm, 2 bar-20ohm, etc... the display is a 10 bar display so the readout shows approx. 10ohm per bar

D1-D3 can be 1N4148, 1N4001, etc... just any regular diode

D1 drops about 0.65v from the 10v source, making roughly 9.35v, that connects to the Vdd of the LM3914. The 9.35v is smoothed out by filter cap C2.

For C3, use a 100uF capacitor, and for R3, use a 10k resistor, and see how it works.

[budm]

Re: smoothing out voltage spikes and dips?

Hi Ben, what is the calculated Vref? 6V?

[ben7]

Re: smoothing out voltage spikes and dips?

Hi Ben, what is the calculated Vref? 6V?

Yeah, its still ~6v. I haven't changed the values of R1 and R2 yet

[budm]

Re: smoothing out voltage spikes and dips?

The transistor is hooked up as an Emitter follower, that means the Ib will be about 60mA if it does not have Base current limiter, just think of the Vref of the IC is a 6V regulator, if you have the Base connected directly to the out put of the regulator, the E-B junction and the 100 Ohms pot will be the load.
The pot that he is using based on the way his diagram is drawn, it looks like 2-pin pot, if that is the case, it means that the load will be varying from 0 ohm to 100 Ohms if it is not the 3-pin pot that allow you to tap the output at the wiper.
I am still not convinced on the accuracy of his wiring diagram to the actual unit.

[ReverendJones]

Re: smoothing out voltage spikes and dips?

I would like to point out that the 100 ohm sender used is also the value used in a lot of automobile fuel tanks.

Gauges of this style are available from $10 - http://www.ebay.com/itm/200773320938

There are also plenty of circuits on the net showing how to use a LM3914 with a 100 ohm sender for a fuel gauge. The one below includes trimmers for setting the high & low readings.

[budm]

Re: smoothing out voltage spikes and dips?

It does not make sense to have PIN2 (Ground) IC tied to B+ with 180 Ohms 5W resistor (wasted power, that 180 Ohms will have 10V on it when pot is at 0 ohm) and return to ground through ground through the variable 100 Ohm pots (the pot will dissipate power also), all the LED return current will also be flowing through pin2 of the IC. The SIG input pin is just looking for the Voltage input to compare it the the resistor string of the comparator ICs. Using the 100 Ohms or any value is not the problem.
I wonder if that circuit was actually built and tested.
When ALL LEDs are on, there will be about 100mA flowing out pin 2, and there will be voltage drop on that pot which will vary and depends on how many LEDs are on so the the Rhi value will keep on changing so as the the trip points of each comparator, it looks like a mess with pin2 ground connection like that.
The SIGNAL pin is also being fed by voltage that is set by the TRIM POT instead of sensing the variable voltage from the sensor signal.

[budm]

Re: smoothing out voltage spikes and dips?

Hi Ben,
What program do you use to draw the diagram? I hate launching my ALTIUM just to draw the simple diagram.

[scampo77]

Re: smoothing out voltage spikes and dips?

reverendjones. have had the worst luck with these gauges i have been through 3 so far and they just keep failing. i have bigger plans once i get both of these gauges working which is to design a auto leveling system so i dont have one tank run dry while the other is full.

plus it is exciting to have an excuse to ask these questions and understand the results.

[budm]

Re: smoothing out voltage spikes and dips?

So is the sensing pot a 2-leg or 3-leg device?
"these gauges i have been through 3 so far and they just keep failing" are they the circuits you had posted, or they are other design?

[ben7]

Re: smoothing out voltage spikes and dips?

Hi Ben,
What program do you use to draw the diagram? I hate launching my ALTIUM just to draw the simple diagram.

Probably wont be of much help, but I use ExpressSCH (its free). I take a screenshot, then open up a paint program and clip away the extra bits...

Would a 1k5 resistor work well as a Ib current limiter for Q1?

[budm]

Re: smoothing out voltage spikes and dips?

Not sure if that setup will work yet since we still need to find out how that 100 sensor resistor work, if it has two pins and its value is varied from 0 to 100 Ohms, or it has three pins with the wiper as an output so it will be like voltage divider, the load on the Emitter will always be 100 Ohms, if it is two pins variable resistor then the Emitter resistor can be 0 ohm when fully close which may damage the transistor.

I will see if I can build and verify what he needs base on his sensor.

[budm]

Re: smoothing out voltage spikes and dips?

Another question is how much current the sender/sensor resistor can handle without damaging it.

[scampo77]

Re: smoothing out voltage spikes and dips?

it is 2 pins and never goes to 0 ohm i think empty is around 5 ohm. i am hoping to build just a simple voltage divider to get a more accurate range of the gauge. before i install my new contraption i am building here. but when i tried i had the same problem was that the voltage shot all over the place not sitting still for even a second and it is why i started this thread.

then i got side tracked with new diagrams and now i am just wiring up the new way.

[scampo77]

Re: smoothing out voltage spikes and dips?

again i am very very new to this and would like just a bit of reassurance that what im not putting the diodes in the wrong direction or the transistor in upside down

i drew on this picture the way i understood it

[scampo77]

Re: smoothing out voltage spikes and dips?

where does pin 4 RLO go to?

[budm]

Re: smoothing out voltage spikes and dips?

Well that circuit will not work with your Rsensor since it is just a variable resistor not a 3-pin voltage divider type. Pin 4 Rlo is connected to ground.
That is why I ask if the Rsensor is 2-pin or 3-pin type.
May I ask what kind of setup you have before, before you want to use this LM3914 circuit.