How to make an affordable 'Vein Finder', for use during venipuncture

FeaturedContest Winner
Picture of How to make an affordable 'Vein Finder', for use during venipuncture
One of the hardest parts of taking blood can be finding a suitable vein. Some patients are 'difficult sticks'; their veins are either very small, and/or deep, preventing health professionals from finding a site easily and quickly. Repetitive needle sticks are painful for the patient and may also lower their confidence in the ability of the phlebotomist performing the stick.

Many companies now market 'Vein Finder' products, but these products can cost over $5,000 (depending on quality and utility). That price tag is often outside the means of smaller clinics or facilities. This is a guide intended for hobbyists and DIY problem solvers -- not staff of a large hospital looking for high-end tech to aid in routine venipuncture. 

This tool works by using near-infrared wavelength LEDs to illuminate the flesh at the site. The veins will appear as dark bands because they are more absorbent of this spectrum of light than the surrounding tissue. It is similar in principle to holding your hand over a flashlight (something we all did as kids).

Estimated Time to Complete: 3-6 hours (depending on your plastic cutting and soldering ability)
Estimated Cost: ~15-50$ depending on shipping/availability of free samples/if you need to buy solder, spare wiring, misc.

Below is a quick video of it working, moving it up and down my forearm. It's hard to document how clearly the veins appear because of my camera's minimum focal distance, but this ought to give you an idea (imagine no blur!):

Step 1: Required materials and tools

Picture of Required materials and tools
Soldering Iron
Rotary tool, with attachments suitable for cutting through plastic
Safety Goggles
Respirator/dust mask
Needle Nose Pliers x2 (I found these useful)

Blank Circuit board
Rocker Switch, On/Off (I used)
Resistors, 56 OHM, 250mW, 1% (I used) x30
LED 3mm, 20mA, 1.9V, λ 628nm (I used) x30
Plastic Enclosure with battery contacts and PCB screw holes (I used OKW enclosures' Part# A9072129, with Part#s A9190002 and A0304031)
Friendly Plastic, Black/Black (I bought mine here)
Wire (less than 1' needed)
AA batteries x2

1-40 of 66Next »
danbemp (author) 2 years ago
Very excited to have this project be featured as a finalist in the UP! contest. Thank you everyone for your continued support and feedback on this project -- It was fun to build, and I hope to keep writing instructables that give others even better ideas.

akshay.d211 month ago

Hi danbemp !

I tried making this but it does not work it seems. I used a normal Red LED for it and powered them on a 9V battery. I used ledcalc to use 82 Ohm resistance. RED LED's are 1.5v Forward Voltage and 20 mA.

Thanks. :)

Kind regards from India.

Great work danbemp!

Before deciding the the wavelength of the LEDs, did you consider the transmission, absorption and reflection properties of the different skin layers? Is 628nm based on a real study or estimation? I just want to be able to chose the right LEDs before I order them :)

clarify the weld layout of resistors and LEDs

ToddR15 months ago

I was trying to build this with a 920nm infrared from radioshack. I have learned that the LEDs I bought from radioshack probably have a wide beam and hence you need to enclose them in black goop on the sides. All I see is a blur on an IR camera. There are places online where you can purchase narrow beam LEDs of 10-20 degrees to focus the light. I intend to pursue that.

bharatk27 months ago

plz give me detailed explanation about it...plzzz which components can we use

bharatk27 months ago

waw itz amazing idea....i like it verry much....i want to add more specifications in it....

it is very nice..

Deano45928 months ago

Brilliant idea and exactly what I was looking for; I simply can't justify spending the amounts of money they're asking of a Veinlite. ...with that being said, I've spent a considerable amount of time and money on making incarnations of it below trying different things. My experiences are, use good quality LEDs, my first attempt has 22x10MCD 635nm LEDs and an open slot at the top. Passable, but only in a dark, dark room. My second and third attempts I'm not sure of the exact wavelengths but they're BRIGHT. My advice is to NOT open up the top to be crescent; exactly like you were saying, it lets too much natural light in and you lose the vein as soon as you turn it parallel to insert the cannula. I haven't figured out exactly how to get around this yet for my fourth (and final) attempt (final because I've just finished the third and I'm tired and frustrated that what was my best attempt, I ruined by opening it up to be a crescent) but I think it's just the way it has to be; leave it closed, find the vein and watch it like a hawk when you remove the light. Or you could mark it, and THEN swab it with alcohol??

IMAG0717.jpgIMAG0718 (1).jpgIMAG0719.jpg
Can we buy one online?
imjbt2311 months ago

what can i use as an alternative for the friendly plastic ?

jrashata1 year ago

Can i use LED 630 nm?

twolf81 year ago

Seems like this might be easier to build using common bus resistor network SIPs inplace of the individual resistors.

balubaid1 year ago

hi brother, I am surprised at the great project that helps patients Varjua that the information sent by the most beautiful on the project in order to apply it in the national

hi brother, I am surprised at the great project that helps patients Varjua that the information sent by the most beautiful on the project in order to apply it in the national

my email (oaoab2010@hotmail.com)

This video is very cute to me. Always I want such project would be completed with great skill. I hope, only experts who are very experienced to this sector should research and apply in real works. To see more related topic: click here: http://www.whichisthebetter.com Hope to well.
This video is very cute to me. Always I want such project would be completed with great skill. I hope, only experts who are very experienced to this sector should research and apply in real works. Hope to well.
Raphango1 year ago
Whoa man! Great idea!
HBSkirmit2 years ago
A great project and good instructions. I have just one suggestion to simplify. Rather than use thirty 56 ohm resistors, just use one with wattage large enough to handle the total current of all 30 leds. Again, great job !
danbemp (author)  HBSkirmit2 years ago
I'm an *extremely* novice 'electrical engineer', so you may be onto something here. However, I remember that when I was doing research for this project it was mentioned many times that it's better to use multiple resistors in parallel with the LEDs. Something about how diodes aren't very good at sharing current. When I saw that, I knew there could be a possibility of some of the LEDs getting very hot -- and when you're designing a device to be used on people, safety always trumps simplicity.
Hello - I'm an actual electrical engineer - and I can answer this. The reason you *must* distribute the resistors is because the LEDs all have slight different forward drops, and have very strong voltage/current relationships, thus if you put all the diodes in parallel, the one with the lowest turn-on voltage will turn on first (or turn on 'hardest') and hog all the current. It's actually called 'current hogging' in our business. Whenever you have devices with non-linear current vs. voltage and varying threshold voltages, you distribute resistors just like you have them. In this use, we sometimes call them "ballasting resistors", since their job is to balance the current drawn by each nonlinear load (the LED in this case). I see the one-resistor suggestion come up all the time in forums - and actually several times in the comments to your project - but don't do it. It is a poor design practice. If you are lucky, one or a few LEDs will hog and overheat, if you are unlucky, the lowest turn-on device will smoke, followed by the next lowest, etc. until they are all toast. I hope this helps someone!
realety1 year ago
Fantastic!! May I make one suggestion to the build? Cutting the case at the top end so that you have an open slot, making it look more like a tuning fork. With that it would be more like the commercial units and it would be able to be used during the venous access and eliminate the "eye off the ball" that would take place when removing the light later.
fuzzynurse2 years ago
I love this! When I have time, I might just have to do this. When I work NICU and pediatrics, we have a light that does the same thing but it shines through the hand/foot/arm so it's only useful on kids up to about 2 years old. I doubt it's very expensive because we have 2 or 3. This, though, would be great for adults who, surprisingly, are often much whinier about getting stuck. However, have you looked into the legality of using this in a professional setting? I wonder how much of the cost of the commercial devices has to do with needing FDA (or whoever) approval. Anyway, you're now one of my heroes!
danbemp (author)  fuzzynurse2 years ago
Thank you so much! Please do build one, I'd love to see another one in action.

Good call about the FDA -- I did some research and after a quick brush up on my legalese I think this device falls under this category: http://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfcfr/CFRSearch.cfm?fr=880.6350 , meeting the exception described here http://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfcfr/CFRSearch.cfm?fr=880.9 . Plus, since it's not a commercially marketed device in this instance, it's build and use at your own 'risk'.
Long wavelength red LEDs (both 660 and 730 nm) can be obtained for reasonable prices here:


The design of the device would have to be modified a bit to accommodate the different form-factor of the package, but at 1W each, you wouldn't need many of them.
There is a LOT of wasted energy with your multiple resistor design. I'd venture to say that you are probably burning more in the resistors than the LEDs themselves. It's just going to wear out the batteries faster. As long as the single resistor is sized properly, heat isn't an issue. You can find some forward voltage calculators online, especially helpful for the entry level. I like http://ledcalc.com/
danbemp (author)  zach.stoltenberg2 years ago
Your link is telling me to put 24 resistors in parallel!
fred27 danbemp2 years ago
No. The resistors go in series. That way they get the same current through them. However there's a voltage drop off around 2v across each resistor so you could only do this with a higher voltage battery. If battery life isn't a big deal for you then I wouldn't worry about it.
LED's can have a little tollerence in the voltage required to burn. Lets say you have 2 LED's which you bought in a bundle, then the first LED can require 3.05V to burn, but the second one might need 3.10V to burn. Using only resistor will reduce the lifetime of the LED's becouse they are burning at a differend voltage then required.
Giving each LED it's own resistor will increase the lifetime of the LED's.

Also, it is a bit of work to solder the resistors, but where I buy them, resistors cost 8 eurocents per 5, so it isn't worth to leave em out for the price.

ps. Funny thing to try: use a 3V coincell (CR2032) and place a red and a green LED over it, what happens is that only one LED will burn becouse of the different voltages required. The same thing might happen with 30 paralel LED's, some might not even go on.
Oh, and about the lost power in the resistors:
Paralel resistors:

P=U*I = I*R*I = 30(number of series)*0.020^2(current squared*56(resistence) = 0.672W
Power = voltage * current
Voltage = Current*resistence

1 big resistor:
R=(30/56)^-1 (becouse 1/R = 1/R1 + 1/R2 + ... + 1/R30 = 30/R1 with R1=R2=...=R30)
R= 1.867ohm

P = 1.867(ohm) * (30*0.020)^2 = 0.672W

Both situations give the same power. With the extended lifetime of the LED's, I would recommend the 30 resistors as used in the Ible
alcurb2 years ago
Why does this remind me of a stud finder? It's very clever. Would you get better contrast with IR LEDs? If so, the device can be used in conjunction with a camera-equipped smartphone to see the IR light.
danbemp (author)  alcurb2 years ago
I read your comment this morning and I've been thinking about it while at work. On the one hand, I *do* think you'd get better contrast with IR light, however, using a camera to see the light has the drawback of not being able to easily mark the veins. You need to be able to do this to still know where the veins are when the device is taken away.

It's still an interesting idea, and a very creative solution. As an aside, I was thinking this may be how the accuvein tool works, with one additional feature. Basically, the accuvein may use IR light to see the contrast between flesh and veins, and then the computer in the device maps that image and displays it back onto the same patch of skin it's viewing. Just a hunch. Thanks for the comment.
iudom2 years ago
How to make an affordable 'Vein Finder', for use during venipuncture is a topic that needs thorough explanations as done in the University of Nigeria site http://www.unn.edu.ng
cstandley2 years ago
This is great. Have you tried it on a dog or cat? There are days I could sure use one!
danbemp (author)  cstandley2 years ago
I'm always amazed how Vets and Vet Techs are able to draw blood from dogs and cats -- seriously uncooperative patients. That having been said, I had tried it on my cat and she was not too happy about it. All the fur makes it difficult to use.
wizodd0 danbemp2 years ago
It's not unusual for a vet to shave the area...if only a drop or so is needed, the veins on the outer edges of the ears are easy to see by transmitted light.
we would spray the leg down with rubbing alcohol to draw blood. and you get a good person to position the animal
Mandebelle2 years ago
Awesome design! Will this help avoid veins a little deeper in the tissues? (I do weekly subcutaneous infusions of human immune globulin into the fatty tissue between veins. This requires inserting five 9mm needles at a 90 degree angle to the skin and taping them into place for 2.5 hours. Each infusion set has 6 needles, so if I accidentally hit a vein that spare needle comes in handy. If I hit a second vein though, I'm supposed to remove the needles from all 5 sites and start over again with a new infusion set. Not fun.) It would be fantastic if your device could help find deeper veins (up to 9mm below skin surface) and so avoid all those extra sticks.
danbemp (author)  Mandebelle2 years ago
Sounds like you're involved in a pretty cool project. Is it for research? I'm curious to hear more about it.

I think my device would be helpful in your situation but one cm is admittedly deep. In theory it should work. Where on the body are you doing these injections? Keep me posted if you decide to do the build.
muy bueno lo voy a copiar en mi profesión ne sera de mucha ayuda gracias sobre todo cuando el paciente esta hipo tenso
danbemp (author)  rubendariob20052 years ago
Muchas gracias. ¡Que tengas suerte y ya me contarás cómo te va!
1-40 of 66Next »