I Need to Adjust the Visible Light Range

I am working on a circuit to alarm when the internal Vcc in a device is gon e. A sounder draws 3 mA but I can mitigate that by pulsing it. I was plan ning on running an LED at 1 mA. The whole thing runs off a supercap that i s charged while Vcc is up. So far, so good.

Somehow I mentally screwed up thinking red LEDs were around 1.2 volts. May be that was an IR LED I was thinking about. I've been considering my optio ns for driving it based on that approximate voltage. I am planning on runn ing the supercap from 5 volts down to about 2 volts. using a resistor to s et the current has issues even with a 1.2 volt device. But now that I'm lo oking for a part number and how to drive it I find my mistake that red LEDs are typically 1.8V or even higher and that's at 2 mA, not the 20 mA I thou ght they based the forward drop on.

I had been thinking of using an LDO to control the current and brightness. LDO output to LED anode - LED cathode to junction with LDO ADJ and sense r esistor - sense resistor to ground. I've found devices with voltage refere nces as low as 0.5V. Should have worked well, but with the LED needing 1.8

5 volts there's not enough voltage left for the sense resistor.

So I am asking if I am right about the Vf of the LEDS (the bright ones in p articular) and if there are any better circuits for regulating the current. Even if I can't get regulated current down to 2 volts, it won't be any wo rse than a resistor once it goes out of regulation.

--
  Rick C. 

  - Get 1,000 miles of free Supercharging 
 Click to see the full signature
Reply to
Ricketty C
Loading thread data ...

Is a TLVH431 too botique for your project? Internal reference of 1.25V, operates down to 100 uA cathode current.

Reply to
bitrex

gone. A sounder draws 3 mA but I can mitigate that by pulsing it. I was planning on running an LED at 1 mA. The whole thing runs off a supercap th at is charged while Vcc is up. So far, so good.

Maybe that was an IR LED I was thinking about. I've been considering my o ptions for driving it based on that approximate voltage. I am planning on running the supercap from 5 volts down to about 2 volts. using a resistor to set the current has issues even with a 1.2 volt device. But now that I' m looking for a part number and how to drive it I find my mistake that red LEDs are typically 1.8V or even higher and that's at 2 mA, not the 20 mA I thought they based the forward drop on.

ss. LDO output to LED anode - LED cathode to junction with LDO ADJ and sen se resistor - sense resistor to ground. I've found devices with voltage re ferences as low as 0.5V. Should have worked well, but with the LED needing 1.85 volts there's not enough voltage left for the sense resistor.

in particular) and if there are any better circuits for regulating the curr ent. Even if I can't get regulated current down to 2 volts, it won't be an y worse than a resistor once it goes out of regulation.

That would give me a loss of 1.25 volts right off the top, 2.5 times more t han I get with a LDO with 500mV set point. The LDO is a pretty good soluti on. I just hit a wall when I realized the LED would use so much of the 2.0 volts min Vcc. Also with the LDO I get a switch built in from the EN so n o loading of a CMOS output. Unless you know a way to make this work with s ome minimal loss in voltage to the LED.

Thanks for the suggestion though.

--
  Rick C. 

  + Get 1,000 miles of free Supercharging 
 Click to see the full signature
Reply to
Ricketty C

No I mean you can use it as a comparator w/ hysteresis to detect when the Vcc drops below 2 volts and turn on the LED a la:

Change R6 to provide somewhat more than 100uA to the TLVH down to say

1.5 volts and remove the green LED from there. Hang a P-channel FET with a Vt of like -1 to -1.5 volts off the TL cathode too, red LED and current-limit resistor in the drain to V- to turn on when the supply hits 2 volts.
Reply to
bitrex

Off the top of my head, maybe you can make use of the Holtek LDO HT7521-1. I have to go out and can't come up with the details of how you can apply it.

2.1V output with 25mV drop-out. 100mA max. 2.5uA ground current.
Reply to
Pimpom

It should be noted that the energy stored into a (super)capacitor is proportional to the voltage squared. If it stored 100 % at 5 V, only

25 % remains at 2.5 V and 6 % at 1.2 V. Thus the question is, how much effort should be used to squeeze out the last percents of energy.

, Does the LED have to be red ?

Also check out blue LEDs with red/green/blue/white phosphors how many lumens they produce at a specific current. While the threshold voltage is nearly 3 V, check out if a low power buck/boost switcher would be available.

Reply to
upsidedown

is gone. A sounder draws 3 mA but I can mitigate that by pulsing it. I wa s planning on running an LED at 1 mA. The whole thing runs off a supercap that is charged while Vcc is up. So far, so good.

. Maybe that was an IR LED I was thinking about. I've been considering my options for driving it based on that approximate voltage. I am planning o n running the supercap from 5 volts down to about 2 volts. using a resisto r to set the current has issues even with a 1.2 volt device. But now that I'm looking for a part number and how to drive it I find my mistake that re d LEDs are typically 1.8V or even higher and that's at 2 mA, not the 20 mA I thought they based the forward drop on.

ness. LDO output to LED anode - LED cathode to junction with LDO ADJ and s ense resistor - sense resistor to ground. I've found devices with voltage references as low as 0.5V. Should have worked well, but with the LED needi ng 1.85 volts there's not enough voltage left for the sense resistor.

s in particular) and if there are any better circuits for regulating the cu rrent. Even if I can't get regulated current down to 2 volts, it won't be any worse than a resistor once it goes out of regulation.

,

re than I get with a LDO with 500mV set point. The LDO is a pretty good so lution. I just hit a wall when I realized the LED would use so much of the 2.0 volts min Vcc. Also with the LDO I get a switch built in from the EN so no loading of a CMOS output. Unless you know a way to make this work wi th some minimal loss in voltage to the LED.

The alarm is the opposite of this, coming on when the power source is absen t. The supercap is there solely to power the alarm circuit. The supercap is charged up to 5 volts while power is present and the alarm sounds down t o 2 volts, so I'd like the LED to be on at least that long.

It's a bit funny that the sounder is rated for 3 mA and I can run the LED a t 1 mA and still have it visible. I'm pulsing the sounder so the average c urrent will be less than the LED and the LED will be hard to see at the vol tage the sounder works!

I found a TLDR5800 which seems to be pretty bright and has an IV curve show ing 1.6 volts at 1 mA. But I don't know how to adapt the LTspice models to match that IV curve.

I'll ask in the LTspice group and see if anyone there can help. I'm sure t his question has been asked before, but it is hard to find anything in that group. There's a LTspice wiki, but it's not really a wiki since they don' t let users update the pages. So documentation on LTspice is not really ve ry good. Docs on the digital stuff are just wrong in some cases.

I can get an LDO that has a 0.5 V reference so using it to power the LED wi ll get me full current to nearly 2.1 volts. That's a lot better than just using a resistor over a range of Vbb 2.0 to 5.25 which works out to 1:10 ra nge of current. But again, finding a spice model is a different matter.

--
  Rick C. 

  -- Get 1,000 miles of free Supercharging 
 Click to see the full signature
Reply to
Ricketty C

gone. A sounder draws 3 mA but I can mitigate that by pulsing it. I was planning on running an LED at 1 mA. The whole thing runs off a supercap th at is charged while Vcc is up. So far, so good.

Maybe that was an IR LED I was thinking about. I've been considering my o ptions for driving it based on that approximate voltage. I am planning on running the supercap from 5 volts down to about 2 volts. using a resistor to set the current has issues even with a 1.2 volt device. But now that I' m looking for a part number and how to drive it I find my mistake that red LEDs are typically 1.8V or even higher and that's at 2 mA, not the 20 mA I thought they based the forward drop on.

ss. LDO output to LED anode - LED cathode to junction with LDO ADJ and sen se resistor - sense resistor to ground. I've found devices with voltage re ferences as low as 0.5V. Should have worked well, but with the LED needing 1.85 volts there's not enough voltage left for the sense resistor.

in particular) and if there are any better circuits for regulating the curr ent. Even if I can't get regulated current down to 2 volts, it won't be an y worse than a resistor once it goes out of regulation.

Turns out I don't think it is about finding the lowest drop out voltage. T o set the current the output voltage drives the diode and then the sense re sistor. The voltage on the sense resistor sets the loss. So the real limi tation is how low the output voltage can be set. The Digikey search doesn' t find what I need exactly. Setting the minimum output voltage to 0.5 or l ess and a few other items to narrow the search like quiescent current, it r eports 17 devices, all of which are LT and all cost far too much $5 or more . But they are about perfect for the job. They use a current source on th e ground pin to set the output voltage. Adding the right size resistor the re to keep the output voltage low and the right size resistor on the output to set the 1 mA current on the diode (which is wired between Vbb and the i nput) would work at champ!

formatting link
pdf

They even show this usage on page 19!

I don't get why this is a $4 chip though. 10uA current source, an amp and a pass transistor. Seems not much different from many LDOs in terms of wh at's inside.

--
  Rick C. 

  -+ Get 1,000 miles of free Supercharging 
 Click to see the full signature
Reply to
Ricketty C

Just for completeness...

A six pin PIC10F will run from 5.5V to 2V for 36p one off UK price. I don't think they have ADC, but they do have a comparator, so timing an RC gets you the Vcc and then a bit-banged PWM gets you the LED average current.

--
Cheers 
Clive
Reply to
Clive Arthur

gone. A sounder draws 3 mA but I can mitigate that by pulsing it. I was planning on running an LED at 1 mA. The whole thing runs off a supercap th at is charged while Vcc is up. So far, so good.

Maybe that was an IR LED I was thinking about. I've been considering my o ptions for driving it based on that approximate voltage. I am planning on running the supercap from 5 volts down to about 2 volts. using a resistor to set the current has issues even with a 1.2 volt device. But now that I' m looking for a part number and how to drive it I find my mistake that red LEDs are typically 1.8V or even higher and that's at 2 mA, not the 20 mA I thought they based the forward drop on.

ss. LDO output to LED anode - LED cathode to junction with LDO ADJ and sen se resistor - sense resistor to ground. I've found devices with voltage re ferences as low as 0.5V. Should have worked well, but with the LED needing 1.85 volts there's not enough voltage left for the sense resistor.

in particular) and if there are any better circuits for regulating the curr ent. Even if I can't get regulated current down to 2 volts, it won't be an y worse than a resistor once it goes out of regulation.

Thanks, but an MCU is not in the future of this design.

I did find NSM4002MR6 at ON. It's really just two crappy transistors in an arrangement that will use the BE voltage as a reference. A bit higher dro p out than I would like but it gets the job done with a lower current below about 2.5V, but not nearly as bad as just using a resistor.

At least they have a spice model.

I could fix the voltage on the LED if we just all adapted to being able to see into the IR range a bit lowering the voltage on LEDs.

--
  Rick C. 

  +- Get 1,000 miles of free Supercharging 
 Click to see the full signature
Reply to
Ricketty C

Simple charge pump voltage doubler with a couple of inverters and just limit the LED current by the size of the decoupling capacitors.

--
Regards, 
Martin Brown
Reply to
Martin Brown

You need to connect the supercap to the sounder and LED only when Vcc fails, so you need some kind of switching. The simplest would be using a NC relay powered from Vcc, which doesn't cause any extra voltage drops.

However, if the relay is not acceptable, you need to include the switch voltage drop in the calculations.

If you intend to pulse the buzzer, why not also pulse the LED.

A P-channel FET might be an option handling both the switching and perhaps also the LED current limiting in a single "power" component by adjusting the gate voltage.

Reply to
upsidedown

I completed the simulation using the NSM4002MR6 and it worked better than e xpected. Some of the graphs in the data sheet talk about Ic/Ib being only

  1. I guess that is not expressing the gain, it's pointing out the conditi ons of the test. I already have a PNP on the board, so rather than add ano ther line item to the BOM I configured two 2N2907 in the same arrangement a nd it works the charm. With 1 mA of current at 5 volts the current only dr ops off toward the bottom, still 940 uA at 2.25 volts and 700 uA at 2 volts . The simulated diode drop for the LED is 1.56 volts at 1 mA current. We' ll see how realistic this is.

Unless someone has a better circuit, this is what I'm going with. It's a h uge improvement over a resistor and only costs two transistors, two resisto rs and the transistors are already in the BOM.

--
  Rick C. 

  ++ Get 1,000 miles of free Supercharging 
 Click to see the full signature
Reply to
Ricketty C

The NSM4002MR6 looks pretty silly to me, why tie into a single source boutique part which probably costs much more than two jelly beans? You are right to stick to jelly-beans for undemanding apps like this.

Anyhow here is my take on the low headroom problem:

piglet

Reply to
piglet

one. A sounder draws 3 mA but I can mitigate that by pulsing it. I was pl anning on running an LED at 1 mA. The whole thing runs off a supercap that is charged while Vcc is up. So far, so good.

May not help your existing project, but here's how I handled a similar situ ation in a recent design (that needed to know when the internal Vcc vanishe d): Basically, a solid-state-relay in SPST-NC 1 Form-B configuration.

Not the cheapest approach, but dirt simple.

Part = LCB716 Link:

formatting link

Reply to
mpm

gone. A sounder draws 3 mA but I can mitigate that by pulsing it. I was planning on running an LED at 1 mA. The whole thing runs off a supercap th at is charged while Vcc is up. So far, so good.

tuation in a recent design (that needed to know when the internal Vcc vanis hed): Basically, a solid-state-relay in SPST-NC 1 Form-B configuration.

And how did that help what you were doing?

--
  Rick C. 

  --- Get 1,000 miles of free Supercharging 
 Click to see the full signature
Reply to
Ricketty C

an expected. Some of the graphs in the data sheet talk about Ic/Ib being o nly 10. I guess that is not expressing the gain, it's pointing out the con ditions of the test. I already have a PNP on the board, so rather than add another line item to the BOM I configured two 2N2907 in the same arrangeme nt and it works the charm. With 1 mA of current at 5 volts the current onl y drops off toward the bottom, still 940 uA at 2.25 volts and 700 uA at 2 v olts. The simulated diode drop for the LED is 1.56 volts at 1 mA current. We'll see how realistic this is.

a huge improvement over a resistor and only costs two transistors, two res istors and the transistors are already in the BOM.

Yeah, On has a bunch of LED drivers but mostly they are digitally controlle d. At the bottom of the list I found that part which gave me the idea to r oll my own. Using the same 2N2907 I'm using in the current limiter for the supercap (which you gave me the idea for with your soft start in Win's thr ead on the USB power limit) it works as good if not better than the On part .

I have to say I'm not following your circuit, but then I've had 2 hours sle ep last night. Need to catch a few winks shortly.

Anyway, thanks for the suggestion.

--
  Rick C. 

  --+ Get 1,000 miles of free Supercharging 
 Click to see the full signature
Reply to
Ricketty C

You are welcome. I am sure my low headroom circuit has been done before, it is fairly straighforward to work out:

Q3 is the heart of the circuit, imagine for simplicity that it's Vbe is exactly 600mV and there is a reference voltage of 700mV above ground on the base. Then it will reduce drive to Q4 once it's emitter gets lifted above 100mV by LED current through the 56R.

Making the 700mV reference is the constant current source Q1 Q2 driving a current (determined by Q1 Vbe and 12k) into the 15k resistor. The tempco of Q1 makes that constant current decrease with increasing temperature but that roughly matches the tempco of Q3.

It is imprecise but does keep the LED current roughly constant over a wide range of supply voltage.

Perhaps this is the kind of circuit the late Jim Thompson might have come up with to solve your challenge!

piglet

Reply to
piglet

Charge pump.

--
John Larkin         Highland Technology, Inc 

Science teaches us to doubt. 
 Click to see the full signature
Reply to
jlarkin

A sounder draws 3 mA but I can mitigate that by pulsing it. I was planning on running an LED at 1 mA. The whole thing runs off a supercap that is charged while Vcc is up. So far, so good.

that was an IR LED I was thinking about. I've been considering my options for driving it based on that approximate voltage. I am planning on running the supercap from 5 volts down to about 2 volts. using a resistor to set the current has issues even with a 1.2 volt device. But now that I'm looking for a part number and how to drive it I find my mistake that red LEDs are typically

1.8V or even higher and that's at 2 mA, not the 20 mA I thought they based the forward drop on.

LDO output to LED anode - LED cathode to junction with LDO ADJ and sense resistor - sense resistor to ground. I've found devices with voltage references as low as 0.5V. Should have worked well, but with the LED needing 1.85 volts there's not enough voltage left for the sense resistor.

particular) and if there are any better circuits for regulating the current. Even if I can't get regulated current down to 2 volts, it won't be any worse than a resistor once it goes out of regulation.

You can make a low-dropout current source using an LM334 with its 60mV reference volateg, and an PNP pass transistor on top. Run at 1.8V delivering current to a 1.6-volt LED.

--
 Thanks, 
    - Win
Reply to
Winfield Hill

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.