LED indicator for A DIY PSU - UK Vintage Radio Repair and Restoration Discussion Forum

Jolly 7 · 11th Oct 2019, 12:25 am

I have an old 5mm red LED with a forward voltage drop of 2V. I need to use the LED to indicate the 37V DC feeding my LM317T regulator circuit. No more than 20 mA should be allowed to flow through the LED and I am using a one watt 2K resistor to achieve this (nearest to the calculated value of 1750 ohm I.e. 35/0.02).
I am using the LED with its series resistor connected in parallel to the input of the regulator. Can anyone please advise if there a better way ? The resistor's getting a bit warm in the current configuration, unless of course I bump up its wattage.

Bazz4CQJ · 11th Oct 2019, 12:56 am

Does the LED need to have 20mA flowing through it? Does it not light up quite well with 10mA?

B

Jolly 7 · 11th Oct 2019, 12:58 am

I guess 10 mA might be OK although this is one of the older LEDs, not the more recent 'ultra bright' ones.

broadgage · 11th Oct 2019, 3:56 am

Unless there is some particular reason to use an old existing LED, it would probably be worth use of a new one.
A modern ultra high brightness type, driven at just 2ma will probably be brighter than an old one driven at 20 ma.

Reducing energy consumption and heat production is generally worthwhile.

David G4EBT · 11th Oct 2019, 9:00 am

I'd agree - an ultrabright LED would be far better and they're cheap. So cheap that the minimum order quantity will usually be at least ten.

EG - 10 5mm ones in your choice of ten colours for £1.30 post paid from a UK seller:

https://www.ebay.co.uk/itm/Ultra-Bri...eEVqw8GIfTzkIg

They don't mention the current drawn so a series resistor will need to be selected on test.

The ones that I have, which are warm white, need only 1mA and even then are a tad too bright. They're especially good on battery powered equipment as they only draw 1mA, which is negligible. My variable voltage bench PSU only goes up to 25V, on which I've tested an ultrabright LED. to drop the 23 V leaving 2V for the LED I tried a 22K series resistor, but it was a little too bright to my eyes, so I tried a 27K, which brought the brightness down to a more acceptable level, and the current down to just below 1mA.

To drop 35V with such an LED, the series resistor would need to be in the region of 33K, possibly even 39K, but as I say, depending on the actual ultrabright LED used, it would be a matter of select on test. (With only 1mA flowing the series resistor can of course be a low wattage one. Even a 1/8th Watt one would be fine.)

Hope that's of interest.

snowman_al · 11th Oct 2019, 9:09 am

You are 'loosing' 35 volts at 20mA through the resistor, that is 0.7 Watts. Better to use a 2 Watt 1.8k resistor if you Must have 20mA. It will still get just as warm though...

But, I think 10mA would be plenty bright enough, as others are saying. That consumes just 0.35 Watts and well within the limits of a 3.3K or 3.9k 1 watt resistor.

Jolly 7 · 11th Oct 2019, 2:01 pm

Thanks for all the replies. It appears I do have a superbright red led from an unused Velleman sound to light kit. As opposed to the old LED, which drops 1.8V, the Velleman one only drops 1.6V and lights up a lot brighter just using the test leads of my multimeter. I will proceed to limit the current to 2 mA or thereabouts and fix it to my PSU cabinet.

kalee20 · 11th Oct 2019, 11:02 pm

What is the LM317T regulator feeding? If it's something which is guaranteed to take rather more than 20mA, you can just connect the LED (with a lower series resistor) across the LM317.

You save 20mA of current, and the LM317 has an easier life too!

Jolly 7 · 11th Oct 2019, 11:44 pm

The regulator outputs 1.2 to 25V and will be used as a bench PSU. It doesn't light up at all until the output is 4V, using a 20K resistor. I chose this value as it allows the LED to glow using 1.7mA at optimum brightness from the 37V feed.
The other issue I'm worried about is injecting LED noise into AM radio circuits, although I could be wrong.

kalee20 · 12th Oct 2019, 12:03 am

OK if the LM317 output is variable, forget my suggestion.

However, LED noise is likely to be small: I have never had any problems. More likely to get trouble with LM317 noise (which I HAVE seen).

broadgage · 12th Oct 2019, 12:19 am

Simple LEDs seldom produce significant noise, they are used in all sorts of sensitive equipment without ill effect.

LED lamps as increasingly used for lighting, CAN produce problematic noise, but this originates from the switched mode driver circuit, not the actual LEDs.

Jolly 7 · 12th Oct 2019, 12:46 am

OK, thanks for explaining. In relation to RF noise, does the stepdown transformer itself determine how much RF is being generated ? For example, I use the older heavier transformers as opposed to the really light modern ones found in almost all power supplies nowadays. If I were to use the latter type with a bridge rectifier instead of an SMPS, would I get a comparable low RF noise output ?

Jolly 7 · 12th Oct 2019, 1:06 am

Quote:

Originally Posted by kalee20

OK if the LM317 output is variable, forget my suggestion.

However, LED noise is likely to be small: I have never had any problems. More likely to get trouble with LM317 noise (which I HAVE seen).

Was the LM317 noise detrimental for AM radio circuits ?

ajgriff · 12th Oct 2019, 1:46 am

Be radical and use a neon with a series resistor across the mains after your on/off switch. So much nicer visually and no brightness variation.

Alan

11th Oct 2019, 12:25 am	#1
Jolly 7 Octode Join Date: Feb 2018 Location: Southampton, Hampshire, UK. Posts: 1,061	LED indicator for A DIY PSU I have an old 5mm red LED with a forward voltage drop of 2V. I need to use the LED to indicate the 37V DC feeding my LM317T regulator circuit. No more than 20 mA should be allowed to flow through the LED and I am using a one watt 2K resistor to achieve this (nearest to the calculated value of 1750 ohm I.e. 35/0.02). I am using the LED with its series resistor connected in parallel to the input of the regulator. Can anyone please advise if there a better way ? The resistor's getting a bit warm in the current configuration, unless of course I bump up its wattage. Last edited by Jolly 7; 11th Oct 2019 at 12:26 am. Reason: clarity

11th Oct 2019, 12:56 am	#2
Bazz4CQJ Dekatron Join Date: Apr 2009 Location: Oxfordshire, UK. Posts: 4,934	Re: LED indicator for A DIY PSU Does the LED need to have 20mA flowing through it? Does it not light up quite well with 10mA? B __________________ Saturn V had 6 million pounds of fuel. It would take thirty thousand strong men to lift it an inch.

11th Oct 2019, 12:58 am	#3
Jolly 7 Octode Join Date: Feb 2018 Location: Southampton, Hampshire, UK. Posts: 1,061	Re: LED indicator for A DIY PSU I guess 10 mA might be OK although this is one of the older LEDs, not the more recent 'ultra bright' ones.

11th Oct 2019, 3:56 am	#4
broadgage Nonode Join Date: Jan 2010 Location: North Somerset, UK. Posts: 2,129	Re: LED indicator for A DIY PSU Unless there is some particular reason to use an old existing LED, it would probably be worth use of a new one. A modern ultra high brightness type, driven at just 2ma will probably be brighter than an old one driven at 20 ma. Reducing energy consumption and heat production is generally worthwhile.

11th Oct 2019, 9:00 am	#5
David G4EBT Dekatron Join Date: Feb 2003 Location: Cottingham, East Yorkshire, UK. Posts: 5,760	Re: LED indicator for A DIY PSU I'd agree - an ultrabright LED would be far better and they're cheap. So cheap that the minimum order quantity will usually be at least ten. EG - 10 5mm ones in your choice of ten colours for £1.30 post paid from a UK seller: https://www.ebay.co.uk/itm/Ultra-Bri...eEVqw8GIfTzkIg They don't mention the current drawn so a series resistor will need to be selected on test. The ones that I have, which are warm white, need only 1mA and even then are a tad too bright. They're especially good on battery powered equipment as they only draw 1mA, which is negligible. My variable voltage bench PSU only goes up to 25V, on which I've tested an ultrabright LED. to drop the 23 V leaving 2V for the LED I tried a 22K series resistor, but it was a little too bright to my eyes, so I tried a 27K, which brought the brightness down to a more acceptable level, and the current down to just below 1mA. To drop 35V with such an LED, the series resistor would need to be in the region of 33K, possibly even 39K, but as I say, depending on the actual ultrabright LED used, it would be a matter of select on test. (With only 1mA flowing the series resistor can of course be a low wattage one. Even a 1/8th Watt one would be fine.) Hope that's of interest. __________________ David. BVWS Member. G-QRP Club member 1339.

11th Oct 2019, 9:09 am	#6
snowman_al Octode Join Date: Dec 2018 Location: Llandeilo, West Wales, UK. Posts: 1,092	Re: LED indicator for A DIY PSU You are 'loosing' 35 volts at 20mA through the resistor, that is 0.7 Watts. Better to use a 2 Watt 1.8k resistor if you Must have 20mA. It will still get just as warm though... But, I think 10mA would be plenty bright enough, as others are saying. That consumes just 0.35 Watts and well within the limits of a 3.3K or 3.9k 1 watt resistor. __________________ Never Leave Well Enough Alone...

11th Oct 2019, 2:01 pm	#7
Jolly 7 Octode Join Date: Feb 2018 Location: Southampton, Hampshire, UK. Posts: 1,061	Re: LED indicator for A DIY PSU Thanks for all the replies. It appears I do have a superbright red led from an unused Velleman sound to light kit. As opposed to the old LED, which drops 1.8V, the Velleman one only drops 1.6V and lights up a lot brighter just using the test leads of my multimeter. I will proceed to limit the current to 2 mA or thereabouts and fix it to my PSU cabinet.

11th Oct 2019, 11:02 pm	#8
kalee20 Dekatron Join Date: Feb 2007 Location: Lynton, N. Devon, UK. Posts: 7,077	Re: LED indicator for A DIY PSU What is the LM317T regulator feeding? If it's something which is guaranteed to take rather more than 20mA, you can just connect the LED (with a lower series resistor) across the LM317. You save 20mA of current, and the LM317 has an easier life too!

11th Oct 2019, 11:44 pm	#9
Jolly 7 Octode Join Date: Feb 2018 Location: Southampton, Hampshire, UK. Posts: 1,061	Re: LED indicator for A DIY PSU The regulator outputs 1.2 to 25V and will be used as a bench PSU. It doesn't light up at all until the output is 4V, using a 20K resistor. I chose this value as it allows the LED to glow using 1.7mA at optimum brightness from the 37V feed. The other issue I'm worried about is injecting LED noise into AM radio circuits, although I could be wrong.

12th Oct 2019, 12:03 am	#10
kalee20 Dekatron Join Date: Feb 2007 Location: Lynton, N. Devon, UK. Posts: 7,077	Re: LED indicator for A DIY PSU OK if the LM317 output is variable, forget my suggestion. However, LED noise is likely to be small: I have never had any problems. More likely to get trouble with LM317 noise (which I HAVE seen).

12th Oct 2019, 12:19 am	#11
broadgage Nonode Join Date: Jan 2010 Location: North Somerset, UK. Posts: 2,129	Re: LED indicator for A DIY PSU Simple LEDs seldom produce significant noise, they are used in all sorts of sensitive equipment without ill effect. LED lamps as increasingly used for lighting, CAN produce problematic noise, but this originates from the switched mode driver circuit, not the actual LEDs.

12th Oct 2019, 12:46 am	#12
Jolly 7 Octode Join Date: Feb 2018 Location: Southampton, Hampshire, UK. Posts: 1,061	Re: LED indicator for A DIY PSU OK, thanks for explaining. In relation to RF noise, does the stepdown transformer itself determine how much RF is being generated ? For example, I use the older heavier transformers as opposed to the really light modern ones found in almost all power supplies nowadays. If I were to use the latter type with a bridge rectifier instead of an SMPS, would I get a comparable low RF noise output ?

12th Oct 2019, 1:46 am	#14
ajgriff Nonode Join Date: May 2004 Location: Halifax, West Yorkshire, UK. Posts: 2,587	Re: LED indicator for A DIY PSU Be radical and use a neon with a series resistor across the mains after your on/off switch. So much nicer visually and no brightness variation. Alan