Accurate Homebrew Blue
About Those BLUE LEDs... October 5 2018
As you can see on this page I have tried both pure blue and warm white
LEDs over the years. I have come to prefer warm white for various
reasons, but one that I cite often is that they are "easier on the
eyes". Knowing that this sounds subjective, I decided to do a little
research to see if there is any scientific basis for this assertion.
This is what I found. Copyrighted images are used under Fair Use Copyright provisions and remain the property of their owners.
Image source: https://en.wikipedia.org/wiki/Color_vision
The chart above shows that human color vision is not
very sensitive in the blue region of the spectrum. As we go up in
frequency toward mid-band the sensitivity increases markedly and then
drops off again as we approach red. The blue, green, yellow, and red
lines represent the center frequency of common LEDs. Note that by the
time we hit blue the sensitivity is down to 20%.
Here we see the spectrum of a common blue LED. It is a very narrow
banded emission centered around a frequency of about 480 nM with
virtually no output throughout the rest of the band.
Image source: https://www.digikey.com/en/articles/techzone/2013/apr/defining-the-color-characteristics-of-white-leds
The above image shows the color spectrum of a
"white" LED at various color temperatures. In this case "warm white" is
defined as a Correlated Color Temperature
of 2600K-3700K. This shows significant optical power across a much larger area of the visible spectrum.
And finally, a side-by-side comparison of various different types of lighting.
So now that you have seen more than you really wanted to know on a flippin' radio
page, what does it mean? My take away from all this info is that pure
blue LEDs simply make the eyes work harder even if they have high
output. It is harder to distinguish fine detail and the optical power
has to be significantly higher to produce the same relative brightness
as warmer and more broad-banded light sources.
July 21 2012, Oct. 5 2018
Although I did for a while quite like the "ice blue" look of the TR-7 with blue LEDs,
the monochromatic light from the LEDs rendered the red dB-over-S9
numbers in the S-Meter black. A post from another ham on the Drakelist reflector alerted me to a "warm white frosted" LED pilot lamp replacement from Coin Taker, a supplier of parts for pinball machines. Click HERE
to see it. They were cheap enough ($0.69 each, as of this writing), so I
ordered a bag of 10. The cost was well under $10, including shipping,
and they arrived promptly.
The LEDs are intended as #44/47 replacements, requiring 6.3 VAC @ 40 mA.
They are non-polarized, which means that they probably have an internal
bridge rectifier. As such, they can easily run from a higher DC voltage
with a series dropping resistor. I initially chose to use a 180 Ohm 1/2 Watt
resistor in series with each pilot lamp socket but this seemed to make the lamp
life unreasonably short. Around 300 Ohms seems to be a workable value with the
newer and more efficient lamps (2018). Also added was a strip of bright
white plastic diffuser material cut from an empty shampoo bottle behind
each lamp, to brighten and even out the illumination. (Note: Drake
4-Line and earlier rigs have a white background behind the lamps which,
according to K4OAH, yellows over time and alters the color of the
display illumination). Below are some pictures of the results; it seems
to be very difficult to accurately capture the color on camera,
sometimes they look a bit on the green side, sometimes a little overly
blue, but you get the general idea. The red is back! The overall
coloring seems much closer to Drake original, which should make the
purists happier as well.
Note the shadow on the right from the lamp base, which is normal with the stock lamps.
Still a hot spot in the middle, but not as much as with the homebrew lamps. There's that shadow on the right again!
The TR-7 next to the RV-7, which still has the homebrew "ice blue" lamp. Hotspots are exaggerated by the camera.
Full front shot of the TR-7/RV-7.
Smile for your closeup, Mr. Meter.
You can compare the coloring with the Homebrew Blue photos below to get
an idea of the difference. Although there are various opinions of what
"Drake original" blue is, the general consensus among restorers is that
Lee Filters #172 "Lagoon Blue" gel is very close to the original. You can see
a sample of this color HERE. It is also roughly equivalent to Rosco E-Colour 172, which can be seen HERE.
The background color of this page is generated by the Hex color codes
for the Lee 172 swatch on the Lee web page. Rosco E-Colour 172 is noticably different from Lee 172, as can
be seen below (generated by
Rosco 172 Hex codes) and the page
background color. This may just be differences in the way the colors
were rendered for internet presentation, but is probably close to the
actual gel color.
Update 10/29/2014! Thanks to Dave, KD2E, who has provided information
to the effect that Roscolene #853 is the ORIGINAL Drake filter material.
I have added Roscolene #853 right below Rosco E-172. Thanks, Dave!
|Rosco E-Colour #172 Lagoon Blue
To further add to the mix of possibilities, Roscolux #67 is referenced as a substitute for Roscolene #853 and similar colors
|Roscolene #853 Middle Blue
And also recommended by Rosco is E-Color #724 Ocean Blue (Thanks to Stan KM4HQE for the info)
|Roscolux #67 Light Sky Blue
Joe KC9LAD recommends Roscolene #854 Steel Blue as being identical to NOS Drake filter material.
|Rosco E-Color #724 Ocean Blue
|Roscolene #854 Steel Blue
DISCLAIMER: How accurately these colors render on your
computer screen depends on the accuracy of the manufacturer's digital
samples and the calibration of your monitor. How they look in your radio
will depend on the number of layers used, the color of the background,
and the color temperature of the illumination source.
The Original Homebrew Blue
I was preparing to make another set of blue LED pilot lamps for my
newly acquired R7A when I spotted an interesting new LED offering on
Radio Shack's web site: SKU#276-023,
a high-brightness blue LED in a 4-pin dip package. This package is also
offered in red, green, and white, but blue was my primary interest for
reasons that should be obvious. My mind immediately jumped to the
A trip to a nearby Radio Shack netted me several LEDs and a 276-150
experimenter's circuit board. I stopped at Ace Hardware on the way back
to pick up some cheap automotive lamps in the proper form factor to use
as base donors.
The LEDs are rated at 20 mA maximum forward current. I decided to use
two LEDs in series for each lamp, and calculated that a 390 Ohm, .25
Watt series resistor would be suitable. Measured current with two LEDs
fed from 13.8 Volts DC was just under 20 mA.
The bulbs of the pilot lamps were duly sacrificed in homage to the
Boatanchor gods, the bases cleaned, and the Radio Shack circuit board
was cut into small strips with a Dremel cutting wheel to fit the lamp
bases. The LEDs and current-limiting resistors were assembled onto the
circuit board, and the assembly soldered to the lamp base. It all fit
together quite nicely into a neat, compact package. Here are some
pictures of the Version 1 assembly (all pictures are clickable):
If you are wondering where the resistor is, it is down inside the lamp
base. The end can just be seen to the left of the left LED. This worked
out very well for the S-Meter, but the center of
illumination was too far away from the lamp base to work well for the PTO. In the second version, I
made the circuit board strip shorter and
trimmed the lamp base with my Dremel to be even with the end of the
lamp socket when inserted. This worked out better, although some
experimentation with the position of the lamp socket was still required
for the most pleasing illumination. In both versions, the circuit board
was secured to the lamp base with only the soldering to the ground
point of the LED string. Here is the Version 2 assembly:
The S-Meter illumination is very smooth and even, fading slightly at
the edges. PTO illumination has two slightly brighter spots, fading
gently to a fairly even blue at the edges of the window. Careful
adjustment of the lamp position results in what I think is a cool
"cathedral" effect, which I attempted to capture in the photos below.
The effect is much more attractive when seen firsthand, of course. I
lightly sanded the surface of the LEDs to "frost" them and (hopefully)
diffuse the light a little better, but I don't think this had a major
I took the full frontal pic above with the original PTO lamp still in
place to show the difference in the new lighting. The greenish tint of
the original is slightly exaggerated by the camera, but the difference
is striking nonetheless. Overall, I was quite happy with the results.
I'm sure that something
could be done to diffuse the light better, or perhaps more LEDs could
be stacked together to even the glow out. But I think this is pretty
darn good for a couple of hours at the bench. The LEDs are pretty
bright at this level; a more subdued blue can be
achieved by using a higher resistor value. I found that a value of 750
Ohms also works well but is not quite as bright. The only problem I had
I had to buy more LEDs so I could make up more lamps for my TR7 and
RV7, as the LED lamps I brewed up previously weren't as bright. At some
point in the future, I will investigate adapting these LEDs to the
Drake L7 as well.
One final note: as I was working on the second TR/RV7 combo, I
discovered that the lamp socket in the RV7 was wired backwards
compared to my two TR7s, R7A, and other RV7. The shell of the socket
was wired to +13 Volts, and
the center pin was wired to ground. This is undesirable for a couple of
reasons, one being that it is easy to accidentally ground out the shell
of the lamp socket if you are trying to fine-tune the lamp position
with power on. This will blow the 5 Amp internal fuse in the TR7
instantly! The other reason is, of course, I wanted all the LED lamp
assemblies to be the same for reasons of compatibility and uniformity.
This was easily corrected by unsoldering the wires from the lamp socket
and reversing them. This is probably an isolated fault, but it would be
wise to check the polarity of any lamp sockets before LED modules are
If this method isn't to your liking, or you you don't wish to build
your own, there are
also online vendors who manufacture and sell some very nice LED lamps
for Drake equipment. I do not make these for sale, this information is
provided so that you can have the fun of making your own.
Check out N9OO's cool blue LED lamp replacements at The Radio Lab Works
if you are interested in purchasing ready-made lamps.
Carey Lockhart, KC5GTT, also offers LED lamp replacements. You can see
Willi Rass, DF4NW, manufactured the blue LED digital readout shown in
my TR7 above. Check out his replacement LED readouts, in various
For a little bit different kind of blue, give a listen to my friend Laurie McClain's beautiful
cover of Kate Wolf's "Cornflower Blue" by clicking HERE.
page was created using the
all-in-one internet application suite
original material contained on this web site is ęCopyright 2008-2012 James
Shorney (except where noted)
Drake is a registered
trademark of R.L. Drake, L.L.C.
All other Trademarks and Copyrights are those of
Information contained in this page may be freely
for non-profit educational purposes, providing that the author and
source(s) are credited, except where prohibited by other Copyrights
here by reference.
This information MAY NOT be included in any work or compilation offered
for sale, or any eBay or other online sale or auction, except by
reference or hypertext link, without prior written consent by the