.4 V DC to 5 V DC possible?

[zach48191]

Here is the part of the circuit that we are stuck at. Hope this makes it easier.

As for TedA could you take that image I have and draw what your trying to say with the resistor so I can understand it visually since im not really great with conceptual instructions. The LED are 4 red and 4 yellow the issue is not the LED I think as they are only 1 way. So the way they are in the circuit is the same that we put them in the other way will not turn them on. I have drawn the correct format that we used since we are using the dot mode to make it easier to see which LED we have selected and in the image I have shown pin 9 not connected. I am sorry for the misunderstanding and guessing game. Also Ted I was wondering with your previous statement of linking the LEDs together to create up to 25v DC, could you draw something or link me an example of what your trying to suggest so I can understand it much clearer? Basically I just want the LED once on to boost it to atleast 5V DC when no lights are on it is hovering at roughly .2V on pin 14. as it moves up the chain of LEDs the voltage increases at pin 14.

As for Cocacola I do not understand the question and I have researched it and it still is confusing. I have looked at the internal component of the LM 3914 and still It is basically taking the voltage which is .7V coming in from pin 5 from the LM34 and then it comes out that certain pin at a lower voltage of .3-.4 depending if on which LED is lit. The higher LED that is on the higher the voltage gets which ranges from .2 - .5V which /5 is when pin 18 LED is lit on. Also that chip was recommended earlier would not work since when pin 14 is not on the voltage at that pin is .2V, so it would turn on the light regardless of which LED is on.

 

[CocaCola]

and then it comes out that certain pin at a lower voltage of .3-.4 depending if on which LED is lit.

Nope, not what is happening...

The outputs of the LM3914 do not output (source) any voltage, they sink (ground) the LED...

Look at the circuit again, all the LEDs are directly connected to +9 volts though resistor R8 to the anode (+) leg, where is the cathode (-) aka negative leg connected?

BTW, not a fan of the R8 limiting resistor implementation, especially in bargraph mode, even in dot mode there is a latency delay between transitions... It works, but, yeah not my cup of tea...

 

[(*steve*)]

OK, Zack, measure the voltage when the LED is OFF, and again when it is ON.

You will find that there is *way* more than a 0.4V difference. It is that difference you need to use to turn on a transistor. Once you see it, you'll slap yourself around the head a few times

 

[CocaCola]

You will find that there is *way* more than a 0.4V difference. It is that difference you need to use to turn on a transistor. Once you see it, you'll slap yourself around the head a few times

Me thinks he might need to adjust his measurement points as well, put the negative voltage probe on the LM3914 pin 14 and the positive prob on the 9 volt source and give that circuit a few on/off cycles like Steve suggested...

Then look over this, PNP transistor configuration...

From here... http://www.kpsec.freeuk.com/trancirc.htm

And consider a pullup on the base of the transistor configuration above, and you need that protection (flyback) diode with a relay...

 

[Wabajig]

The above schematic is great, but so you don't get a false trigger tie a 100k ohm resistor from base to +Vs. I suppose you will use the relay where the R Load is. Also your relay should be rated for your AC lamp supply. Here in the USA its 125VAC. Good Luck, John

 

[TedA]

zach48191,

Thanks for providing some more hints about what you are doing. There are still outstanding questions, though.

And a new one: Do you really want the heater to go off only over a narrow temperature range? If the LM3914 is in dot display mode, the heater will go back on if the temperature goes higher and the LED on pin 14 goes off again.

And it would really help to know about the relay coil. Do you have ratings? You might measure the resistance, if there are no resistance or current ratings.

In your last post, the drawing showed something that might be a transistor, sketched to the right side. Do you know what sort of transistor? What sort of transistors do you have or can easily get?

Ted

 

[CocaCola]

The above schematic is great, but so you don't get a false trigger tie a 100k ohm resistor from base to +Vs. I suppose you will use the relay where the R Load is. Also your relay should be rated for your AC lamp supply. Here in the USA its 125VAC. Good Luck, John

Yep, that's why I added "And consider a pullup on the base of the transistor configuration above"

And a new one: Do you really want the heater to go off only over a narrow temperature range? If the LM3914 is in dot display mode, the heater will go back on if the temperature goes higher and the LED on pin 14 goes off again.

You should be able to just tie all the above outputs together to solve that issue until it goes out of range? I'm not 100% familiar with the chip but does the last output stay triggered when you exceed the range you are measuring?

On that note I have to agree that in the end it's not the most reliable method to do this using traditional thermostats or building a µ-chip controlled circuit, heck you can purchase over the counter reptile thermostats designed for just this purpose, for a reasonable price...

 

[(*steve*)]

If he uses the bargraph mode of the chip, and the relay is pulled in when the LED is off, then the heater will be on when the temperature is below the temperature which lights the LED.

However the led may flicker on and off around this temperature which may not be a good thing for the heater.

A more complex circuit might turn the heater off when led "n" goes on, but doesn't turn it on again until led "n-1" goes off. This would have the room (or whatever) temperature oscillate in some range.

There's quite a bit more logic required to do this though.

 

[zach48191]

We got the power by using a 741 op amp and inverting it w had just forgot to supply it with negative voltage on pin 4. Thanks for all the input though!!!

Our new problem its that anytime the led at pin 14 is on it is .4 v and if the led is off the voltage is 1.8v. WE need our transistor which was an npn to turn off at 8v and turn on at or lower than 7v because the light is always on regardless of with led it on. Because the voltage never falls below 5v... We tried a pnp transistor and the light always stays on...

 

[CocaCola]

Our new problem its that anytime the led at pin 14 is on it is .4 v and if the led is off the voltage is 1.8v.

Have you bothered to read my post? Where are you taking these measurements?

I repeat with emphasis... For good measure take off the LED on that pin for this... And for double good measure forget everything you 'know' about the circuit for this test...

put the negative voltage probe on the LM3914 pin 14 and the positive prob on the +9 volt source and give that circuit a few on/off cycles like Steve suggested...

What you should see is when pin 14 is 'on' pin 14 goes low aka it grounds/sinks to negative and you should see +9 on the volt meter since you just completed a circuit and the positive probe is on the 9 volt supply, when 'off' pin 14 is floating (it's an open collector) thus for your application and reliability it should be pulled high though a pull up resistor so it's not just floating, in this state you should see 0 on the volt meter as there is no completed circuit... You now should realize that pin 14 swings from 0 when 'on' and and +9 when 'off' when there is a pull up resistor added to the circuit...

WE need our transistor which was an npn to turn off at 8v and turn on at or lower than 7v because the light is always on regardless of with led it on. Because the voltage never falls below 5v... We tried a pnp transistor and the light always stays on...

I can't make heads or tails of that...

The diagram I posted above for a PNP transistor will work as you want this design to work (not how you keep thinking it works) it just needs the addition of a pull-up resistor (use a 100K) between pin 14 (or the base of the transistor) and +9 volts to make sure it shifts to high logic when off and isn't left floating...

 

[CocaCola]

Zach, build this and give it whirl...

Using a NO (normally open) or NC (normally closed) relay, you can flip flop the on/off status...

Examples...

Using a NO relay the device being switched will be off when the pin 14 LED is off...

Using a NC relay the device being switched will be on when the pin 14 LED is off...

 

[zach48191]

ok I'll try it tomorrow and see how it turns out

 

[TedA]

zach48191,

I understand that you are still learning this stuff. The problem is really analog in nature, and quite beyond what you learn in basic digital electronics.

The linked circuit and LM3914 present a fairly complex nut to crack. The LED driver outputs are constant current sinks; the output voltage depends a great deal on what is connected to them. In the linked thermometer circuit, when the individual LED connected to a given output pin is on, the voltage to ground is partially dependant on how many other LEDs happen to be on. When the LED is off, the voltage is determined by leakage currents, so it is not well defined, unless you add some extra source of current.

The digital portion of the LM3914 is not well described in the block diagrams on the data sheet. There is information in the text, but you have to figure it out.

I don't see any utility in using a 741, unless you have a very sensitive relay it can drive directly. You don't need much voltage gain or, depending on the relay, that much current gain. You need output current. Something that can drive your relay is what you should use. An NE555 might actually work pretty well.

The cocacola circuit should work, but there are some differences from what you were doing.

The "Here is the part of the circuit that we are stuck at. Hope this makes it easier" schematic showed pin 9 of the LM3914 open, giving a dot mode display. So only one LED is on at once. I don't know if you will be happy with a bargraph style display. I wonder also how much current the 9v supply can provide; lighting more LEDs takes more juice.

If the relay control is taken from a single LED output, using the bargraph mode solves the logic problem of what happens when the temperature is pretty far from the setpoint.

The relay will be actuated as the temperature rises, so the normally closed contacts would be used to operate the heater.

The correct transistor bias resistor values must be determined. It would help to know something about the relay coil when doing this. The value chosen for R8 matters. The 1k value sketched-in on your version of the schematic Is too large for the LED to get the full programmed current, but will allow the LM3914 to pull down the LED cathode close to ground.

It is very likely that the relay will chatter, or at least switch on and off more than you want. A little positive feedback from the relay driver to the temperature sensor signal, or reference voltage should fix this. One or two resistors in the right places can provide the feedback.

If you really want the dot mode display, it is not too hard to interface the relay driver to a group of LEDs, to allow this. If you would prefer for the relay to be actuated as the temperature falls, a second transistor can make this happen.

Ted

 

[CocaCola]

The cocacola circuit should work, but there are some differences from what you were doing.

As has been discussed overall more logic needs to be added to make this a reliable and full featured device as I believe is the intended end design... My circuit above is by no means a final solution, more so a step in the learning ladder... The reason I suggest building the above circuit is that hopefully it will trigger that little light bulb in the mind, and thus set straight what the output pins of the LM3914 are actually doing and we can get passed the .4v stuff...

 

[zach48191]

Here is our circuit that we built. We are using a function generator as a test for the differed temperatures in a much more stable manner than the LM34DZ as it was defective and we had no time to get a new one since our school has no more of them. The light turns on at the certain light we want but we had to put a switch at the end of the emitter to make it reset the relay to open the switch back up on it. The relay was a JY5H-K5VDC Relay it has a minimum voltage to close the switch of 3.5VDC and .5VDC or less to open the switch. We solved the problem of having the light turn on all the time with the OP amp with an inverter and voltage divider. This made the voltage on the LED when lit higher than 3.5 VDC and lower than 3.5VDC when off. I Thank everyone for helping me out and the probably 80 hours I spent in the lab fixing the problems and trying some of your ideas. Tomorrow I am turning it in and the teacher also said that what I had to work with and my knowledge is very impressive that I even got this far. I am glad with what I have as it is not perfect, but overall it does everything I want but need a little more work in turning off when it goes above and below the pin 14.

 

[CocaCola]

You would have had a much easier time if you actually used pin 14 of the LM3914 for triggering, instead of trying to detect a fluctuation in the +9 rail through a 680Ω R9 resistor...