LED Project... Is the diagram correct?

[Maxer]

Hi Max,
Clarification:
1) Of course the ICs have resistors inside. The LM3914 has a bunch of resistors between pins 4 and 6 that determine what input voltage lights its LEDs, and the 555 has resistors that determine the lowest and highest voltages of its timing cap.
2) Of course the circuit has changed from my original one, now the 1uF cap is the 555's timing cap, which never discharges to ground like it did in my circuit.
3) Inside the LM3914 between pins 4 and 6 are 10 resistors in series, that are used as a voltage divider. This voltage divider determines what input voltage turns on the LEDs. Now that the 1uF cap is used as the 555's timing cap, the cap's voltage will range from 1/3 of the supply voltage for its discharged voltage, to 2/3 of the supply voltage for its highest voltage. The 555's 1/3 and 2/3 of the supply voltage reference points are determined by a voltage divider of 3 resistors inside the 555 that are connected between the supply and ground.
4) We must duplicate the 555's voltage divider with another voltage divider for the LM3914 so that the 1st LED lights when the cap's voltage (input to the LM3914) is 1/3 of the supply voltage (the cap is discharged by the 555) and the 6th LED lights when the cap's voltage is a little less than 2/3 of the supply voltage (cap is at its highest voltage). It would be a lot easier to calculate voltages if the 555's cap voltage simply ranged from ground to a fixed voltage. It would also be a lot easier if you had 10 LEDs instead of only 6.
5) The 1st scan is longer because the cap must charge from zero volts (fully discharged from the circuit being unpowered) to 1/3 of the supply voltage before the timing actually begins. The LEDs will be off during that time.
6) You must measure the LM3914's voltage divider total resistance between pins 4 and 6 because the datasheet says that it could be anywhere from 8K to 17K ohms. We must know exactly what it is.
IC's are lousy at making accurate value resistors, but can make very accurate duplicates and ratios that are used in "normal" LM3914 and 555 circuits.

BTW, the emitter of the 2N3906 current source and its resistors must now connect to the supply voltage, instead of to the LM3914's pin 7. Pin 6 must also be disconnected from pin 7. Pin 4 must be disconnected from ground.

what if we change the 1uf cap to .1 uf? it would discharge quicker

 

[audioguru]

Hi Max,
Sure 0.1uF would discharge quicker. 0.01uF or 0.001uF would be even quicker! 100 puffs (slang for picoFarads)? or just 10 puffs?

In order for the timer's period to be about 1 second per scan, the capacitor's charging current must be low. The time calculation is roughly one divided by the current, times the capacitor's value. So if you reduce the value of the capacitor then the charging current must also be reduced by the same amount.
The problem with having a very low charging current is that pins 2 and 6 of the 555 need a small current to operate, and may steal all of the very low charging current, leaving the capacitor with nothing. Actually, the currents for pins 2 and 6 are opposing, so that the current required by pin 6 can be supplied by the current from pin 2. The datasheet spec's a current for pin 2 but not for pin 6, so maybe they are not equal.
Since the current for pin 2 is spec'd with a wide range, I wouldn't risk operating the 555 with a very low charging current and would stay with using a 1uF (low-leakage metalized poly) cap.

 

[Maxer]

Hi Max,
Sure 0.1uF would discharge quicker. 0.01uF or 0.001uF would be even quicker! 100 puffs (slang for picoFarads)? or just 10 puffs?

In order for the timer's period to be about 1 second per scan, the capacitor's charging current must be low. The time calculation is roughly one divided by the current, times the capacitor's value. So if you reduce the value of the capacitor then the charging current must also be reduced by the same amount.
The problem with having a very low charging current is that pins 2 and 6 of the 555 need a small current to operate, and may steal all of the very low charging current, leaving the capacitor with nothing. Actually, the currents for pins 2 and 6 are opposing, so that the current required by pin 6 can be supplied by the current from pin 2. The datasheet spec's a current for pin 2 but not for pin 6, so maybe they are not equal.
Since the current for pin 2 is spec'd with a wide range, I wouldn't risk operating the 555 with a very low charging current and would stay with using a 1uF (low-leakage metalized poly) cap.

internal delay of IC555 that will discharge cap much less then 1/3 Since the cap is now shorten threw pin 5 of ic555... If you think this schematic looks allright and no other modifications are needed please identify tha values of all resistors, my specs for LED's are posted above.... at least 20mA per Led would be good
thanks again
max

 

[audioguru]

Hi Max,
Are you saying that the 1uF cap will be discharged so quickly that its voltage will overshoot the 1/3 of the supply voltage that triggers pin 2, because the internal delay of the 555 will release the discharge transistor at pin 7 too late?
It could be, but I don't think the capacitor's discharge voltage will overshoot much. If it does, the voltage level for the 1st LED to light can be adjusted to be the same voltage.

Why do you have a resistor to ground at pin 5 of the 555?
It would just lower the capacitor's maximum charge voltage directly, and lower the cap's minimum voltage half as much.

 

[Maxer]

Hi Max,
Are you saying that the 1uF cap will be discharged so quickly that its voltage will overshoot the 1/3 of the supply voltage that triggers pin 2, because the internal delay of the 555 will release the discharge transistor at pin 7 too late?
It could be, but I don't think the capacitor's discharge voltage will overshoot much. If it does, the voltage level for the 1st LED to light can be adjusted to be the same voltage.

Why do you have a resistor to ground at pin 5 of the 555?
It would just lower the capacitor's maximum charge voltage directly, and lower the cap's minimum voltage half as much.

well if using ref voltage of LM3914 v have to lower voltage of 555 too.
added modification of res to pin 5 of 555. so now i have to know which pins are on when cap is charging... and add led's to those pins since there's only 6 leds.. what are the values of the resistors and pot you added??
I could be wrong but i want to order these parts from all one place....
so here is the list for 1 circuit board.
6 Leds
1 LM3914 IC
1 LM555 IC
1 18 pin socket
1 8 pin socket
1 capacitor 100uf/25V
6 270ohm resistors
2.7K ohm resistor
620 Ohm resistor
transistor 2N3906
2 capacitors .01 uf/25v
Experimental soldering board
-------
parts not sure about
1UF/25v Capacitor
3 resistors ( don't know the value????)
1 Pot (don't know the value???)
----------------------------
audio if you think i should add anything to this order that i might need to change or adjust please let me know i would like to order this stuff so i can start to make it and experiment to see how things work... along the way i can be corected, i will post what happens. I figure if i see it i will have a better understanding of this ramp generator work since this is my first eletronic project.
also the values of the resistors that i don't know if your not sure either just tell me what different values i should order....
thanks
max

 

[Maxer]

another IC55 schematic that migh work my friend made...
this might require you to change up the LM3914 diagram

 

[audioguru]

Hi Max,
Sorry for my delay.
I don't like the way your new diode pulls down the 555's pin 5 so low that the flip-flop inside the 555 may "latch" and not be able to be triggered by pin 2.

To avoid the problem of a delay on the 1st cycle of the 555, I have been thinking to leave the 555 and the current source transistor powered whenever the bike is running, with pins 2 + 6 and the cap held at a voltage just a little lower than pin2's threshold voltage. When you activate a turn signal switch, the LM3914 will be powered and immediately indicate the 1st LED. At the same time, the voltage limiter on the cap will be released and it ramps normally along with the other LEDs.

I will sketch a "voltage limiter" and the resistors for the current source soon.

 

[magno_grail]

Sorry, if I am jumping in late, but it appears you are going about this the hard way. I am presuming you want the LEDs to sequentially turn on, extinguish, then repeat with a constant interval inbetween LED switching?
You need a linear sawtooth generator which can be done on a 555 or OP amp. An OP amp might be more linear over charging with a constant current transistor.
The 3914 only has to measure the voltage of the sawtooth. By adjusting the HI/LO ends of the divider relative to the sawtooth you can have as many or few LED's in the bank up to 100, I believe. The NS data sheet shows examples of stacking the 3914's.
You can keep the discharge time down by using a small capacitance and large charge resistor.
The 3914 is somewhat tricky to set up. I used a 3915 for a programmable ignition. Instead of LEDs I had a bank of resistors discharging the timing capacitor with the 3915 signal input being the capacitor voltage. Very sensitive to the voltage divider HI/LO inputs and the internal reference supply. The LED current is determined by the current drawn from the reference supply (1.2V). You do not have to use that to power the voltage divider. You could actually hook it up across the 555 divider but there will be interaction between the two. The 3914 is around 1k/step and the 555 about 5k between sense and trigger (2&6).
I do not get the reason for the switch on the mode input. For dot mode I believe it is left open and bar mode it connects to the second highest LED cathode through a 10k resistor. Check the National data sheet.

 

[audioguru]

Hi Magno,
That's right, we're using a 555 as a linear sawtooth generator with a transistor current source charging its cap. We could have used your opamp idea but the 555 does the same and has the "extra to your circuit" comparator built-in.
An advantage that your opamp idea would have is that its low-point comparator could have a very low voltage threshold (if the opamp is a type that can operate wth its inputs and outputs near its negative supply voltage). This low threshold will minimize the extra long 1st cycle of the sawtooth generator. Unfortunately, a 555 is not designed to have its pin 2 threshold voltage directly changed nor very low.
It is not possible to connect the LM3914's resistor divider between the 555's resistors because pins 2 and 6 of a 555 are the control pins and the resistors are on the opposite side of a differential circuit as reference voltages, so the resistors for pin 2's threshold voltage are not available.
We are operating the LM3914's mode pin correctly. Your connection applied to your cascaded ICs.

 

[magno_grail]

I had to take another look at the 555 internal connections. The 2&6 pins are comparitors to the 1/3 and 2/3 Vcc voltage divider. Pin 5 is the direct connection to 2/3 Vcc. There is no external connection to 1/3 Vcc.
The sawtooth will run between 1/3 and 2/3 Vcc. You could connect the HI end of the 3914 to 555 pin 5 then adjust the LO end to get the range you need for the number of LEDs and correct for the discharge cycle time.

 

[audioguru]

Hi Magno,
I am sorry for mis-spelling your name before, I have corrected it.
I would rather not connect the top of the LM3914's divider to the 555's pin 5 because it would change the 555's timing period. Besides, Max has only 6 LEDs so the top of the LM3914's divider will have to be at a much higher voltage anyway. Actually, since the top of the LM3914's divider will have to be above the supply rail, it is probably best to reduce the voltage at the divider's low end, and just use the last 6 outputs of the LM3914.

 

[Maxer]

Hi Max,
Sorry for my delay.
I don't like the way your new diode pulls down the 555's pin 5 so low that the flip-flop inside the 555 may "latch" and not be able to be triggered by pin 2.

To avoid the problem of a delay on the 1st cycle of the 555, I have been thinking to leave the 555 and the current source transistor powered whenever the bike is running, with pins 2 + 6 and the cap held at a voltage just a little lower than pin2's threshold voltage. When you activate a turn signal switch, the LM3914 will be powered and immediately indicate the 1st LED. At the same time, the voltage limiter on the cap will be released and it ramps normally along with the other LEDs.

I will sketch a "voltage limiter" and the resistors for the current source soon.

Allright I came up with this finally and ordered the stuff i need. But i still want to try out with a voltage limiter. Also I looked at my bikes wiring when the switch is on for right or left turn signal it runs threw a turn signal relay and then to the bulbs and the indicator lights. will that relay effect this?

 

[audioguru]

Hi Max,
This project is becoming a can of worms.
The 2.2K resistor that you put across the cap steals current from the current source so the cap will no longer charge linearly, and may not even charge high enough to allow the 555 to work.

You mentioned a relay. I didn't know bikes use an expensive and failure-prone relay flasher when an inexpensive, simple electronic circuit can flash reliably. In a car, its clicking reminds you to turn it off, but on a motorcycle can you hear it?
Why use this complicated 555 circuit when the flasher can do it very well?
The flasher can apply power to the current source transistor which linearly charges the cap. The LM3914 is also powered and indicates the rising cap voltage on the LEDs. When the flasher clicks off, the LEDs also turn off and a P-channel junction FET quickly discharges the cap. When the flasher applies power, the FET is turned off.
Really simple, do you like it?

 

[Maxer]

Hi Max,
This project is becoming a can of worms.
The 2.2K resistor that you put across the cap steals current from the current source so the cap will no longer charge linearly, and may not even charge high enough to allow the 555 to work.

You mentioned a relay. I didn't know bikes use an expensive and failure-prone relay flasher when an inexpensive, simple electronic circuit can flash reliably. In a car, its clicking reminds you to turn it off, but on a motorcycle can you hear it?
Why use this complicated 555 circuit when the flasher can do it very well?
The flasher can apply power to the current source transistor which linearly charges the cap. The LM3914 is also powered and indicates the rising cap voltage on the LEDs. When the flasher clicks off, the LEDs also turn off and a P-channel junction FET quickly discharges the cap. When the flasher applies power, the FET is turned off.
Really simple, do you like it?

damn i wish i would of looked at my bikes wiring schematic earlier because that would of saved me some time and money. well we can try it that way.
i don't hear no noise on the bike even if there was i wouldn't be able to hear it because exasht noise, wind and helmet. But it does Flash. so a light comes up when indicator is turned on. Relay is used for all 4 signals (left front, left back, right front, right back) When i hit the switch for right or left it just grounds the relay and it activates it. If i use the flasher to apply power to current source transistor, will have to mess with the flasher's wiring? or can i still use the wiring from the previous indicators? (pos / neg) sorry about dumb questions but i am still learning you guys are far more advance then me. do i need to find out what ype or flasher relay is it or anything?
thanks i guess we will have to draw a new diagram for this method
max

 

[audioguru]

Hi Max,
One thing about the flasher relay on cars: the flash rate becomes very fast when a light bulb is burnt out. So it might be fast on your bike when using low-current LEDs instead of a high-current light bulb. If you use the LM3914 in its Bar mode, the rate might start fast then slow during each scan.

For cars, a "heavy-duty electronic" flasher was available to drive additional lights on trailers, etc. Their rate doesn't change with load changes. Maybe they are still available and will fit your bike.

If your bike had 12V light bulbs, this circuit (the current source transistor circuit, the LM3914 and the new FET control) connects to their wiring. If it had LEDs, you must connect before their current-limiting resistor.

 

[Maxer]

Hi Max,
One thing about the flasher relay on cars: the flash rate becomes very fast when a light bulb is burnt out. So it might be fast on your bike when using low-current LEDs instead of a high-current light bulb. If you use the LM3914 in its Bar mode, the rate might start fast then slow during each scan.

For cars, a "heavy-duty electronic" flasher was available to drive additional lights on trailers, etc. Their rate doesn't change with load changes. Maybe they are still available and will fit your bike.

If your bike had 12V light bulbs, this circuit (the current source transistor circuit, the LM3914 and the new FET control) connects to their wiring. If it had LEDs, you must connect before their current-limiting resistor.

I don't think it flashes "faster" when one is burnt out but it does flash faster when you change out the stock bulbs (they were never LED's) some people say that when they added LED's to just Flash they would flash faster then stock bulbs did. Also i think if one rea bulb is broken then other rear one doesn't work. i will reconfirm this tommaro i have to get to bed but i think either way fast or not as long as the Bar mode comes on where people can see its a faste bar mode not just a flash it should be good to go. so lets just try that with the new FET control. and i will give you specs of how bulbs work on the bike for sure tommaro
thanks
Max

 

[Maxer]

Hi Max,
One thing about the flasher relay on cars: the flash rate becomes very fast when a light bulb is burnt out. So it might be fast on your bike when using low-current LEDs instead of a high-current light bulb. If you use the LM3914 in its Bar mode, the rate might start fast then slow during each scan.

For cars, a "heavy-duty electronic" flasher was available to drive additional lights on trailers, etc. Their rate doesn't change with load changes. Maybe they are still available and will fit your bike.

If your bike had 12V light bulbs, this circuit (the current source transistor circuit, the LM3914 and the new FET control) connects to their wiring. If it had LEDs, you must connect before their current-limiting resistor.

I don't think it flashes "faster" when one is burnt out but it does flash faster when you change out the stock bulbs (they were never LED's) some people say that when they added LED's to just Flash they would flash faster then stock bulbs did. Also i think if one rea bulb is broken then other rear one doesn't work. i will reconfirm this tommaro i have to get to bed but i think either way fast or not as long as the Bar mode comes on where people can see its a faste bar mode not just a flash it should be good to go. so lets just try that with the new FET control. and i will give you specs of how bulbs work on the bike for sure tommaro
thanks
Max

Allright this is what my friend told me he owns the same bike, he has LED's in pegs but they only flash no sequencing (barmode)

stock bulbs are 12volt and if one bulb breaks they all will flash faster.
if you put led's in your footpegs they will flash faster if there is no relay inline.
basicly if the stock configuration is altered, it will always not flash correctly unless the output of the bulbs stay exactly the same

 

[magno_grail]

Maxer,
If I had a better idea of what you were trying to do I could have given more useful information. Various make bikes use different methods for flashing the signals. Usually the older bikes use a thermal relay like the antique cars. Current passing through heats a bi-metalic strip which opens the circuit, then cools off and closes again. Low wattage bulbs or one burned out will cause the other bulb to light but not flash (not enough current). That is why LED bulbs do not work in these bikes. Others, like my '97 Triumph use the ECU to flash the bulbs using a relay to carry the current. No change if a bulb burns out or replaced with LEDs.
Older bikes use the opposite signal bulbs to provide the ground path for the dash indicator.
Pre-80's British bikes are positive ground. Everything else, I believe, is negative.
Another point, most high ouput LEDs have a low visibility angle, some as little as 20 degrees. Someone in the next lane might not see your signal if they are too close (disregarding the fact that will not see it anyway as they are too stupid).

 

[Maxer]

Maxer,
If I had a better idea of what you were trying to do I could have given more useful information. Various make bikes use different methods for flashing the signals. Usually the older bikes use a thermal relay like the antique cars. Current passing through heats a bi-metalic strip which opens the circuit, then cools off and closes again. Low wattage bulbs or one burned out will cause the other bulb to light but not flash (not enough current). That is why LED bulbs do not work in these bikes. Others, like my '97 Triumph use the ECU to flash the bulbs using a relay to carry the current. No change if a bulb burns out or replaced with LEDs.
Older bikes use the opposite signal bulbs to provide the ground path for the dash indicator.
Pre-80's British bikes are positive ground. Everything else, I believe, is negative.
Another point, most high ouput LEDs have a low visibility angle, some as little as 20 degrees. Someone in the next lane might not see your signal if they are too close (disregarding the fact that will not see it anyway as they are too stupid).

Hey mango_Grail
Mine uses a Relay but i don't know if its thermal or not. but i do know when one is not working or broken other one still flashes.
what i am tring to do is have a sequnce LED flash rather then just on and off... here is what i want
Oooooo
OOoooo
OOOooo
OOOOoo
OOOOOo
OOOOOO
then back to
Oooooo
and when mode pin is connected to Pin 4 V+ on LM3914
it should have a different sequence like this
Oooooo
oOoooo
ooOooo
oooOoo
ooooOo
oooooO
and i will add a small switch on the project box i place the circuit in. and the box will be placed in my trunk i can change it periodicaly if i want to.. or have the switch in another place in the front but not sure yet.
so i would have 1 circuit per side.
Led's i will be using will have only 20 deg light angle, but the way the LEd's will be set up it should be okay because for front i will be using lower MCD's and higher disspation angles maybe like 30 - 40.
Audioguru brough this to my attention earlier in another posting.
i also have other lighting Mods that are going to be done but i after i figure this one out.

i am looking at the schematic right now and it says "turn signal relay" and it has pos and neg terminals...
neg is connected to the switches. switches activate the turn signals offcourse and also the flashing light on the speedometer display.
and postive wire is used as ground.. again i am not sure because i get confused between "convention flow" and "electron flow" since they are opposite. so i try to imagine the way current would active flow rather then how its named, and i just started some basic electronic courses.
here is my wiring schematic i circled where the flasher relay is. To make it easier i highlighted wires from RElay red is Negative and Blue is Postive also its a 98GSXR 750 its first Fuel injection model out with suzuki.
as seen in the diagram eventually pos side of relay is connected to neg, and bike doesn't shortcuit so i am guessing its a thermal relay? i gotta learn more about these types of relays...
crap i can;t post it because its about 582,347 bytes
can i upload it any where and post the link here?

 

[Maxer]

Hi Max,
This project is becoming a can of worms.
The 2.2K resistor that you put across the cap steals current from the current source so the cap will no longer charge linearly, and may not even charge high enough to allow the 555 to work.

You mentioned a relay. I didn't know bikes use an expensive and failure-prone relay flasher when an inexpensive, simple electronic circuit can flash reliably. In a car, its clicking reminds you to turn it off, but on a motorcycle can you hear it?
Why use this complicated 555 circuit when the flasher can do it very well?
The flasher can apply power to the current source transistor which linearly charges the cap. The LM3914 is also powered and indicates the rising cap voltage on the LEDs. When the flasher clicks off, the LEDs also turn off and a P-channel junction FET quickly discharges the cap. When the flasher applies power, the FET is turned off.
Really simple, do you like it?