I searched, could not narrow down the results enough, so I'm asking directly. My apologies if this is a commonly asked question...
I'm slowly putting together a spud headphone amp using a triode-wired 12HL7 pentode per channel. It looks like best performance will be achieved with plate current of >35mA per tube. I currently plan to use a 1000uF electrolytic cap bypassing the 47R cathode bias resistor to get about 1.8V bias. But I'd much rather use a red or green LED to get that voltage.
Unfortunately, all the through-hole LEDs I've found are rated for a maximum 20mA to 30mA continuous current. Are there >100mA rated LEDs with wire leads? All I can find are fancy SMD-2 types.
If the answer is that I have to parallel LEDs, what would be the most effective way to make sure they share current equally? Something like a 10R resistor in series with each?
Or... Is it better to use some kind of Schottky or rectifier diodes that can handle higher current? I could see putting three Schottky rectifiers in series to get about 1.8V forward voltage drop. Has anybody tried that?
Well, I answered my own question by using Mouser's search more effectively.
I found some 100mA rated through hole LEDs. This should work:
151053RS03000 Wurth Electronics | Mouser
--
I'm slowly putting together a spud headphone amp using a triode-wired 12HL7 pentode per channel. It looks like best performance will be achieved with plate current of >35mA per tube. I currently plan to use a 1000uF electrolytic cap bypassing the 47R cathode bias resistor to get about 1.8V bias. But I'd much rather use a red or green LED to get that voltage.
Unfortunately, all the through-hole LEDs I've found are rated for a maximum 20mA to 30mA continuous current. Are there >100mA rated LEDs with wire leads? All I can find are fancy SMD-2 types.
If the answer is that I have to parallel LEDs, what would be the most effective way to make sure they share current equally? Something like a 10R resistor in series with each?
Or... Is it better to use some kind of Schottky or rectifier diodes that can handle higher current? I could see putting three Schottky rectifiers in series to get about 1.8V forward voltage drop. Has anybody tried that?
Well, I answered my own question by using Mouser's search more effectively.
I found some 100mA rated through hole LEDs. This should work:
151053RS03000 Wurth Electronics | Mouser
--
Check out the LEDs at Marlin Jones LED & LED Displays | MPJA.COM
I posted impedance data for the 10W versions -- The Red Light District - another PP EL84 amp -- they are pretty inexpensive and can be heatsinked right onto the chassis.
I posted impedance data for the 10W versions -- The Red Light District - another PP EL84 amp -- they are pretty inexpensive and can be heatsinked right onto the chassis.
Thanks for this. Great idea.
I'm thinking the 10W versions good for 1A current would be complete overkill for this particular application. What about the 1W LED SMD like this red one?
Red 1W LED SMD Diode 50lm 10 Pack | MPJA.COM
2.0V to 2.2V forward V drop is what I'm looking for.
350mA max current rating. That's 7X more than I'll be pulling through them.
10-pack for less the three bucks.
Any problems you see with that one? Is the package impossible to work with in a DIY tube amp?
--
I'm thinking the 10W versions good for 1A current would be complete overkill for this particular application. What about the 1W LED SMD like this red one?
Red 1W LED SMD Diode 50lm 10 Pack | MPJA.COM
2.0V to 2.2V forward V drop is what I'm looking for.
350mA max current rating. That's 7X more than I'll be pulling through them.
10-pack for less the three bucks.
Any problems you see with that one? Is the package impossible to work with in a DIY tube amp?
--
Last edited:
Looks like these 1W LEDs are commonly sold mounted on a heat sink. MPJA has them naked. I suspect they're not good for 350mA forward current if they're not on a heat sink.
The heat sink for these looks like this:
They're cheap enough. They look like they're just pieces of aluminum. I suspect heat sinking to the chassis would work well enough.
--
The heat sink for these looks like this:
They're cheap enough. They look like they're just pieces of aluminum. I suspect heat sinking to the chassis would work well enough.
--
That's just the PCB, not the heatsink itself, they are meant to be then bolted down to a heatsink. Running them naked even at 50mA will not be a great idea... Although at ~50mA, I think the chassis will be plenty. If you want low impedance look to lumiLEDs/luxeon Philips/osram, CREE, and other name brands. The longevity on the Chinese/no-name LEDs is poor, even at low current in my experience (aquarium/hydroponics lighting) the last few years. The name brand stuff will have a lower voltage drop, and a lower dynamic impedance.
LEDsupply sells good stuff, free shipping too for small packages, I buy all my LED constant current drivers (aquarium reef lighting builds) from them. They are a little more money than other suppliers, but they use nicer MCPCBs than most anyone else for better thermal performance ( not that it matters for your application, really) but the free shipping offsets the increase in cost from the better PCBs.
LEDs - Cree, Luxeon, Nichia and 5mm LEDs
LEDsupply sells good stuff, free shipping too for small packages, I buy all my LED constant current drivers (aquarium reef lighting builds) from them. They are a little more money than other suppliers, but they use nicer MCPCBs than most anyone else for better thermal performance ( not that it matters for your application, really) but the free shipping offsets the increase in cost from the better PCBs.
LEDs - Cree, Luxeon, Nichia and 5mm LEDs
Last edited:
Thanks, that helps. Now I understand what's going on mechanically. I had it all wrong.
The little LED itself gets soldered onto the PCB, and that PCB is bolted to the heat sink (or metal chassis). I think I got it now.
So I'd want the 1-up red 625nm version of this?
Cree XLamp XP-E2 Color LED
The little LED itself gets soldered onto the PCB, and that PCB is bolted to the heat sink (or metal chassis). I think I got it now.
So I'd want the 1-up red 625nm version of this?
Cree XLamp XP-E2 Color LED
Last edited:
Like this one?
CSD01060A Wolfspeed / Cree | Mouser
At 0.05A I guess I'd need two in series to get 1.8V forward drop. Certainly easier to implement than the 1W LEDs.
I couldn't find a lower current or lower voltage version than this 600V 1A rated part. 50mA should just barely tickle it.
One thing I've noticed about LED's -- since they are a consummable item even name-brand devices go EOL-- that's how they get to some of the parts houses who cater to us DIYr's.
UV LED's are replacing mercury emitters -- this will be a DIY-saster for those incautious types who don't read the warnings.
UV LED's are replacing mercury emitters -- this will be a DIY-saster for those incautious types who don't read the warnings.
I've used Vishay Telux LED's at 50 mA for bias in a headphone amplifier. I used 3 in series for bias of about 6.3 volts. I used the yellow ones as they had lower impedance.
https://www.vishay.com/docs/83138/tlwr7600.pdf
https://www.vishay.com/docs/83138/tlwr7600.pdf
I think I understand. I should install them right next to each other, so that they're all at about the same temperature, right? How about if I install them so that their plastic bodies are all touching each other?
Thanks.
Also, thanks kodabmx and tikiroo for the suggestions. That Vishay 70mA LED looks perfect. This is a 'use up the junkbox parts' kind of project, so I'm going to see if I can get away with paralleling some of the many red LEDs I have in my parts bin.
Last edited:
I know I'm really late to the party, but what about using
The 1500 PIV part has a 2.0 V drop at 500 mA; based on the conventional diode equation, I'd expect its forward voltage to be around 1.8 or so at 50 mA.
Maybe not as “pretty” as a LED. But dâhmned durable, essentially indestructible.
_______
I once too wondered about using "LEDs in parallel" and the problem (perceived) of fair current sharing between them. Then my co-worker gave me a little string of tiny blue LEDs attached to a 4.5 volt (3 AA cells in series) battery as a Christmas Cubicle Happiness gift.
As an electrical engineer type, I was immediately struck by the fact that ALL of the 4 dozen LEDs were hooked exactly in parallel and shared one simple ballast resistor to limit current. The implication was that the particular LEDs were remarkably good at fairly sharing current for a particular quiescent forward voltage.
Its funny how such serendipitous findings fall into your lap. Secretly, well after Christmas, I decided to test the theory that current sharing would continue to be accurate as while not changing out the ballast resistor, LEDs were removed from the parallel chain.
Sure enough, when I clipped the string in half, putting half as many LEDs on the ballast+battery configuration, the remaining bulbs glowed noticably brighter. Having … in my back pocket (ahem) … a 4½ digit multimeter, I carefully measured the VF forward voltage of the remaining 2 dozen-in-parallel chain.
The original 48 had a VF of 3.31 V. The 24 had VF = 3.39 V. And they all glowed uniformly brighter. No wusses. So, I cut it in half again. 3.41 VF. Then one of the lights got dim, and burned out.
_______
So, taking Wavebourn's Advice, feel free to just put LEDs from the same lot — and tested for VF — in parallel. Its especially "nice" when they all glow the same brightness. Proves sharing is working, sort of.
But also consider the Rectron device!
GoatGuy
The 1500 PIV part has a 2.0 V drop at 500 mA; based on the conventional diode equation, I'd expect its forward voltage to be around 1.8 or so at 50 mA.
Maybe not as “pretty” as a LED. But dâhmned durable, essentially indestructible.
_______
I once too wondered about using "LEDs in parallel" and the problem (perceived) of fair current sharing between them. Then my co-worker gave me a little string of tiny blue LEDs attached to a 4.5 volt (3 AA cells in series) battery as a Christmas Cubicle Happiness gift.
As an electrical engineer type, I was immediately struck by the fact that ALL of the 4 dozen LEDs were hooked exactly in parallel and shared one simple ballast resistor to limit current. The implication was that the particular LEDs were remarkably good at fairly sharing current for a particular quiescent forward voltage.
Its funny how such serendipitous findings fall into your lap. Secretly, well after Christmas, I decided to test the theory that current sharing would continue to be accurate as while not changing out the ballast resistor, LEDs were removed from the parallel chain.
Sure enough, when I clipped the string in half, putting half as many LEDs on the ballast+battery configuration, the remaining bulbs glowed noticably brighter. Having … in my back pocket (ahem) … a 4½ digit multimeter, I carefully measured the VF forward voltage of the remaining 2 dozen-in-parallel chain.
The original 48 had a VF of 3.31 V. The 24 had VF = 3.39 V. And they all glowed uniformly brighter. No wusses. So, I cut it in half again. 3.41 VF. Then one of the lights got dim, and burned out.
_______
So, taking Wavebourn's Advice, feel free to just put LEDs from the same lot — and tested for VF — in parallel. Its especially "nice" when they all glow the same brightness. Proves sharing is working, sort of.
But also consider the Rectron device!
GoatGuy
Thanks, I stored that Rectron R1500B info. I'll throw some of those on my next Mouser order.
Another possibility; Fairchild MV50154 red LED.
MV50154 Everlight | Mouser
Datasheet - https://www.mouser.com/datasheet/2/143/ds300031-40261.pdf
The sheet says the MV50154 has max continuous forward current of 100mA.
Max power dissipation 180mW. 2V*50mA = 100mW.
Vf quoted as 1.6V to 2.0V at 10mA. I guess I'd need to pick a matched pair. No biggie.
Now I have lots of options. Thanks for the ideas, everyone.
--
Another possibility; Fairchild MV50154 red LED.
MV50154 Everlight | Mouser
Datasheet - https://www.mouser.com/datasheet/2/143/ds300031-40261.pdf
The sheet says the MV50154 has max continuous forward current of 100mA.
Max power dissipation 180mW. 2V*50mA = 100mW.
Vf quoted as 1.6V to 2.0V at 10mA. I guess I'd need to pick a matched pair. No biggie.
Now I have lots of options. Thanks for the ideas, everyone.
--
As I understand it: the LED maker sorts his buns into "bins" of near-matched forward voltage. The flashlight maker uses all-same-bin LEDs in a flashlight.
If you are not buying flashlight-factory quantity you may not be able to get "binned" LEDs.
OTOH, LEDs are cheap as chips (potato) so you could buy a 10-pack and test Vf yourself, find two very similar.
Yes, you're right.
It looks like the absolute max peak current is 100mA, max continuous current is 30mA. Max power dissipation is 78mW. The sheet ends with a table of specs referenced to 20mA continuous current. Sort of like those games mfgs used to play with audio amplifier specs for 'peak music power' vs 'continuous power.' Confusing.
The Fairchild MV50154 Everlight LED looks like it should work. Max peak forward current = 1.0A, max continuous current = 100mA. Qty 10 for $3.33 USD is not too bad.
But I already have a bunch of LEDs to use up. I'll test a bunch for forward V drop before paralleling them up.
--
It looks like the absolute max peak current is 100mA, max continuous current is 30mA. Max power dissipation is 78mW. The sheet ends with a table of specs referenced to 20mA continuous current. Sort of like those games mfgs used to play with audio amplifier specs for 'peak music power' vs 'continuous power.' Confusing.
The Fairchild MV50154 Everlight LED looks like it should work. Max peak forward current = 1.0A, max continuous current = 100mA. Qty 10 for $3.33 USD is not too bad.
But I already have a bunch of LEDs to use up. I'll test a bunch for forward V drop before paralleling them up.
--
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.