Introduction: Coffee Cup - PCR Thermocycler Costing Under 350$

Picture of Coffee Cup - PCR Thermocycler Costing Under 350$

This Instructable will describe how to make a PCR Thermocycler for Field or Teaching applications, from scratch, using a commercial temperature controller, a temperature probe, a PC cooling fan and basic shop tools.

The unit performs reliably, is menu driven for programming various PCR (Polymerase Chain Reaction) "recipes" (so you don't have to use a computer but there is free downloadable software that will let you do this and we provide the PIN diagram for making the cable). It cycles temperatures at 0.15 degree centigrade per minute. The difference in actual temperature (Process variable) from "control" temperature (Set variable) varies about 5 degree centigrade. That means we set the temperature at 5 degrees higher than the temperature we want to achieve. This is due to the bias in the thermocouple probe.

In a Nutshell: We use a piece of aluminum rod, drill a 1/4" hole in the center of one end for a cartridge heater and another hole near the rim of the same end for a thermocouple (temperature sensor). Additional holes in the same end of the rod are for sample tubes (0.2 ml eppendorf tubes).

This "heating block" is positioned on a small bracket over a PC cooling fan. Both are placed inside of a metal coffee cup. A commercial temperature controller ( available from Omega Engineering) is used to control the heating and cooling cycles. End of story.

It reaches and holds (soaks) at various temperatures and can also be used for "In situ" Hybridization protocols. It can be programmed from the face plate of the controller or connected to a computer by a serial cable. The control software is a free download.

Project Team:

R. Siderits
A. Marcus
F. Ivalde
A. Andoh
M. Swift
O. Ouattara
W. Lecorchick
S. Singh

Step 1: What Is a Thermocycler?

Picture of What Is a Thermocycler?

What is a Thermocycler?

A thermocycler raises and lowers temperature. For molecular or DNA applications it needs to heat up to about 90-95 degrees centigrade. DNA which usually exists in double strands in a helix (like a spiral staircase) and at about 92 degrees centigrade the strands will break and fall apart.
The solution that the DNA sample is in has extra "bases" that will connect to the broken rungs of the ladder to make two new ladders. To do this part we need to cool down the sample to the point where "bases" can connect to the separated DNA strands. If you cycle through this temperature range from 20-30 times you can make millions of copies of the original DNA.

So in summary, a thermocycler can make a million copies of one original piece of DNA. These copies can then be used for other types of research. The prototype that we build can take 6 sample tubes (0.2 ml eppendorf tubes). There is room for one blank, the temperature probe and a central cartridge heater.

Learning sites that are excellent:

- DNA Extraction Online Tutorial
- Home DNA extarction of DNA (Instructables)
- PCR Song (must hear)
- PCR Protocols

Step 2: Completed Version of the Coffee Cup Thermocycler:

Picture of Completed Version of the Coffee Cup Thermocycler:

Here is the final version showing a metal coffee cup that contains the heating block and a PC cooling fan. The lower half is a coffee "can" that contains the temperature controller. If your wondering why a coffee can and coffee cup, that's all we had to work with.

Parts List and links to sources:

6061 Aluminum Rod 1.5" round stock about 5" in length
Cartridge heater 1" length, 1/4" diam 80W
Temperature controller CN8282-R1R2C2
Instruction manual
Free Software for serial connection to computer
40 mm PC 110 volt cooling fan

Other materials:
- 4 foot extension cord, three prong, to cut off and use as power cord.
- Metal coffee cup and coffee can (your on your own here)
- Hand drill
- Tap for putting threads in the hole for the temperature probe #7 drill and 1/4-20 Tap.
- Basic tools (hammer, screw driver, pliers, wire cutters etc...)

Step 3: Make the Heating Head: Machinist and Hobbyist Approaches:

Picture of Make the Heating Head: Machinist and Hobbyist Approaches:

Make the Heating head. See following images.

This aluminum rod is used to make the heating head.

A center hole 1/4" will be for the cartridge heater.

7 holes are drilled on the same end of the rod for the 6 sample tubes and one hole for the
thermocouple temperature probe. The hole for the temperature probe is made with a
#7 drill and a 1/4-20 tap to thread the 3/4" deep hole.

Step 4: Making the Coffee Cup Part of the Thermocycler:

Picture of Making the Coffee Cup Part of the Thermocycler:

Make the Coffee Cup part of the Coffee Cup PCR system.

We used a fly cutter to make the hole in the bottom of the coffee cup (mug).

Four smaller holes are for the feet that attach to the mounting holes on the fan that is
inside of the mug.

Mark the holes with the fan sitting on the outside of the mug.

Step 5: Wire the Controller:

Picture of Wire the Controller:

Wire the controller:

Read the warning in the manual that comes with the temperature controller!

Instruction manual

See the shop notes pages below that explain various aspects of the wiring.

Step 6: Performance:

Picture of Performance:

Performance evaluation:

The serial cable is connected to the controller.
The free downloaded software form is used to write a recipe for the following tests.
The data is "logged" to data files that were imported into excel for graphing.

Test:: PASS

Pass - Hold at 50 degree centigrade up to 10 hours.

Pass - Hold at 92 degrees centigrade for 30 minutes.

Pass - Run menu (recipe) to go from RT (room temp) to 55 over 4 minutes; then
Loop 25 times: from 55 to 72 where you soak for 2 minutes, then raise temperature to
92 where you soak for 1 minute; then at end of the cycles hold at 55 degree
centigrade indefinitely.

Step 7: Free Software

Picture of Free Software

Screen grabs from the free software which lets you control the unit over a
serial cable.

Free Software for serial connection to computer

Downloading the free Software from the Omega Engineering website lets
you control the "temperature controller" and quickly write complex "recipes" to
store into the unit and run at any time.

You don't need to use the serial cable since the unit can be programmed to
run recipes from the front panel, but it is a quick way
to test and optimize the temperature parameters while capturing the data to
a log file that you can import into excel for graphing.

The software also has a graph function but it is somewhat limited.

Step 8: Where Do We Go From Here:

Picture of Where Do We Go From Here:

Where are we going from here?

- Swap out heating heads for Chromogenic In Situ Hybridization, slide holder
- Two sample heating block using ducted RC hobby fan
- Vented "Tray block" with 20 sample deck
- USB powered Micro PCR platform
- USB control using "Phidget" interface and incorporate two Peltier Junctions
- Micro-controller based (Parallax) controller for computer free 9 volt version
- Internet accessible interface using IOBridge

- Make a single sample "Pen based" version that runs off of a USB port.

Step 9: Disclaimer (fair Warning):

Picture of Disclaimer (fair Warning):
Follow ALL reasonable safety guidelines and actually
wear safety equipment related, but not limited to, the following:

1) Electrical currents that you should not play with.
2) Power-tools that you never read the instructions for.
3) Any tools that you never learned to use properly.


4) Anything that heats up, pinches, cuts, squeezes, flies off, carries electric current,
causes other types of traumatic, caustic, thermal injury or biological contamination.

5) This is a project for scientific teaching or field research and is NOT to be used
for diagnostic testing!

We are sharing our experience, NOT TELLING YOU WHAT TO DO.

If you choose to try this then - it is at your own risk!

  • No really, we're not kidding about this.

Step 10: Project Overview:

Picture of Project Overview:

Facility: RWJUHH Experimental Pathology Division and
Center for Parabiotics Research

Section: Prototype Applications - Research Projects

Application: Molecular Diagnosis and Medical Cytology

Technique: Rapid Prototyping for field research

Title: "Coffee-Cup" PCR Thermocylcer for field or teaching applications costing less than 350$


LuisA197 (author)2016-09-07

great job!

moshsparrow (author)2015-01-28

Im sorry any idea for how to make a QPCR please thanks great Job

DrRadium (author)moshsparrow2015-11-13

One of the most straightforward ways to do quantitative PCR in a DIY environment is competitive PCR. It can be done as RT-PCR as well. Basically, a series of reactions are done with a different amount of a competitor DNA (or RNA for RT-PCR). The competitor needs to use the same primers and be similar in sequence to the target but easily distinguished on a gel. Usually it is the target with a deletion or an addition. The competition acts as an internal control to get around differing reaction efficiencies. Regardless of the efficiencies between reaction tubes, when the ratio of target and competitor PCR products are equal, the ratio of starting templates was also equal. Since you know how much competitor was added, you know how much target template was present.

siderits (author)2013-03-10

Sorry it took so long to get back to you. The RT-PCR version works, is cheap, modular and it uses ONLY USB power. I 3D printed the enclosure but the principles can be applied without a custom housing. Here is an overview of parts.
First you need to heat up a DNA sample to 100 degrees centigrade. You might think that would be impossible but I used two minco thermofoil adhesive backed 1 inch circular heaters wired to two USB ports. Now you need to have a switch turn them on and off so I used a dual solid state relay connected to a phidgets: 8 analog in/8digital in/8digital out board. Now you need to sense the temperature so that you can turn the heaters on and off at the right time. I used a phidgets temperature sensor with a J type thermocouple. OK so now you have the temperature side down, how about the light side. I used a photoresister to monitor the light level in the sample chamber and brought that into the phidget board on another analog channel.
No you only need two more things. A light source the emits at a blue light (I used an ultra bright LED at 480-490nm on the second of the dual Solid State Relays) to excite the SyberGreen when it binds to double stranded DNA and a band pass filter to only let a narrow band of green light through to the photoresistor. The band pass filter was from Fisher scientific. I wrote the interface to sense temp and light and to control the relays for heating with LABVIEW.

Long story short, I've got an essentially disposable RT-PCR platform that works on USB power only and costs less than 300$ total cost. You can stick it onto the back of any standard Notebook computer. This enable field research at a fraction of the cost of a small RT-PCR platform. For what its worth, you can run it from a remote as well.

szilva (author)2013-02-22

Hi, thank you indeed for the coffee cup PCR, is there any chance to see any further details about this latest development you were mentioning in March? We are planning to build a simple PCR right now, but we would go for a qPCR (instructions permitting :) if possible.

mayan12 (author)2012-01-23

Late to this party. Your post is a bit old, did your plans for the USB power source ever come to fruition?

siderits (author)mayan122012-01-24

Yes. Amazingly, YES.

I was told (often) that you can't heat something to 100 degrees centigrade (denaturation temp for DNA) with a USB (4.5 Volts and 100 mA). Turns out its really easy. We used two adhesive backed 1" diameter Minco Thermofoils on separate USB ports. A quick program in Labview to monitor temperature from a thermal probe, a styrofoam thermal chamber to prevent heat loss and a USB Phidget relay switch to control the heaters made it work to commercial nearly specs. We can control, soak, cycle and ramp with no problem.

We then started on the Real Time component using a 1$ "teal" LED and a UV LED to excite Sybr green and a cheap 1 cm (25$) band pass filter from Edmond Scientific to select out the emission wavelength for "Sybr greener". A 50 cent photoresistor should give us the ability to detect amplification in real time for under 200$.

That's where we're at now.

knakker (author)siderits2012-03-22

Wow very cool! What band pass filter are you talking about, I cant find it. :( Would love to check that in our lab too. :)

siderits (author)knakker2012-03-23


Thanks for the kind words; We just printed our our chamber on an UP3D! printer (see attached image) . It cost 11 cents. The front and back panels hold the adhesive backed Minco Thermofoils (as the USB powered heat source). One set of wires in the image is the pressure fit Aqua LED with the perfect wavelength to excite Sybr green, we also now have a second LED in UV range. The other side holds the Band Pass filter in front of the photoresister or photodiode. Its really easy to swap out LEDs and bandpass filters to fit your needs. All controlled with USB "Phidgets" (SSR, 222 controller and Temperature sensor) and interfaced quickly with Labview graphical programming platform. Good weekend project. Its essentially disposable, or you could have several in a row, etc... As a matter of fact you could actually run it off of a solar panel and control it with a 9 volt battery powered microcontroller, even remote the data over wireless (or just log it).
We're now tweaking aspect of the insultation, when time permits.

We've considered using a sinmple LED display on a cheap Single Board Computer (SBC) also from Phidgets. Most of the "solar" work was done by the 10$ medical centrifuge team (see one of our other Instructables for that one. Great idea, used empty bullet casings as buckets and celophane for virtual tubes. Anyway, here are the filters we used.

Edmund Scientific Filter INT 535nm 11.80 mm diameter item number 43071, ordered a second type with a slightly different wavelength pass-through (532 nm item number 43070). We wanted to make sure that we dealt with any secondary peak issues.

KhiemHa (author)2010-11-28

Hi there, I'm recreating this exact (or as close as i can get to it) Thermo-cycler and was wondering how to go about cutting the aluminum rod for the Heating Head, it doesnt seem like i can do this without a milling machine; this is a big problem for me because I'm not sure if any schools in my area or Home Depots have this machine for me to use.

any advice on how to go about this?

siderits (author)KhiemHa2010-11-29

No problem. You don't need a milling machine or a lathe. I just got a little carried away. The most improtant thing is that the heat radiates out from the center and the block cools from the outside surface. That's what gives this thing such good performance characteristics. The holes for the sample tubes should be at the same distance from the heating source and the cooling surfaces.

Just use a hack saw and a hand drill.

You can use a square piece of stock with holes drilled at a distance equally placed from the center heating cartridge. Then cut "fins" to enhance cooling and approximate a round shape.

You can also use the hack saw to take the corners off the square block. You might even consider scribing an octagon on the surface and cutting a square block down to an octagon shape to drill the holes into.

I used a round shape so that temperature would diffuse radially and at a constant rate while the outer surfaces are cooled by the PC fan at the bottom of the mug.

Here's another idea, cast a heating block using JB weld compound or even use the JB weld to create a custom shape from some scrap material.

Another solution would be to pick up some aluminum tubing (to hold the sample tubes) then set short segments of the aluminum tubes in a JB weld "form" to create a custom heating block.

When you get ready for the programming step using the PC let me know I'll send you a few examples. The performance is surprisingly good on this thing.

If you get stuck let me know and I'll help. By the way this is perfect for using the Quick Extract mRNA or DNA extraction kits

and yes you can make this into a real time PCR platform without much trouble.

Best regards;


KhiemHa (author)siderits2010-12-11

Hi there again, so after a long waiting period, my county school system gave the A-OK to purchase the temperature controller so it should be here hopefully by the 13th! I'm going to install the software on my laptop today so i was wondering if you could help me out with explaining some of the wiring when the controller arrives.

I solved my heating head issues because the auto-tech teacher in my school had a milling machine and agreed to cut it for me.

However, when i purchased the fan my budget made me jump at the cheapest fan available which came back to be so tiny (it can't be more then 1.5 inches wide and long)

I'm currently wondering if a longer cooling period could suffice in its usage or whether i should be reading the companies return/exchange policy right now also i have the cartridge heater which fits nicely into the heating head, but has 2 wires at the end causing me to wonder how exact am i to wire this. i do not have the controller yet but will inform you immediately when it comes.

siderits (author)KhiemHa2010-12-19

I turned another heating head on the lathe, if you want it, I'll send it to you.
If so send address to

I'd be delighted to help you get going with the program and would be glad to send you my "recipes" (that's what the program calls it's stored macros).

You will also be able to program the controller directly from the faceplate but logging the data to a text file and then importing it to an excel worksheet is fun.

matchett808 (author)2010-03-18

How is the software coming? I had another idea using (and please laugh) ncurses and bash? lol (hooked up just to a bog standard serial

KenCan (author)2010-02-22

 Historical Moment:  A fellow grad student Adam Richman built a homemade PCR/thermocycler machine in ~1990. In fact, he sold several to top labs when there were only 1 or 2 other thermocyclers on the market.  He used coffee cup heating elements, washing machine water inlet/outlet valves, an acrylic tub made from layered sheets and a thermal sensor.  The only expensive parts were a TRS-80 to control it, some kind of a/d converter and a cheap stirplate for underneath to keep a stirrer going.

It should be easy to have a pc-controlled version as mentioned above.

chun2003lee (author)2009-11-23

Could you re-design this, to allow for PC control, wihout the need for the temprature regulator.

Ie the PC takes direct temp measurements and switches on the heater and fan to control the thermocycler directly.   Hopefully this should reduce the cost dramatically while not affecting the accuracy of the device.. can you think of a way to do this, or is it imposible?.

siderits (author)chun2003lee2009-11-24

That turned out to be easier than I thought using the Phidget modules (  You can even do it with a simple Scratch script ( a picoboard and the lego WeDo robotics kit.  We're working on a USB powered version now.  We may use the USB soldering iron concept for the heater or a silicone rubber strip heater.

Thanks for the comment!

chun2003lee (author)2009-10-22

sorry to rain on your parade, im not hear to make enemies either, but your design has one flaw!, Its to expensive.

After reading the article i looked closely at the cn8200 controller and the thermocouple, the combined cost of these devices comes to 700-800 dollars.
for a total sample size of 5 samples it not worth it plus not ever one has a machine shop at home either.

Good work, but not really practical or economically worth wile.

siderits (author)chun2003lee2009-10-23

A little rain keeps the air fresh.  That's one of the things that makes Instructables great, Honest exchanges about ideas! 
No enemies here.

Here's a breakdown of the low cost and higher sample number version.
We just wanted to demonstrate some control features and get some data about performance in a radial heat diffusion design.
The basic cn8200 cost me 285$ (new from Omega electronics), you don't need the one that connects to the serial port on the computer.  The basic unit is programmable from the keypad. I used the serial connect just to prove that the performance was as good as a commercial unit and to make fast changes to the program.  You can get both cheaper on Ebay and remember to try 

You can also paste a silicone rubber heater to the bottom of an aluminum block that has many more sample wells (say 20).  The radial heat diffusion our design will be revisited for the USB powered version.  Turns out a USB powered Soldering iron with or without a 9V battery assist works just as well as the cartridge heater.  Can you imagine a PCR system that runs on 2 USB ports!  Anyway, the silicon rubber coated heater is 3x4 inches, is very thin and it will draw about 2-3 amps.  It will heat a sample block up to 100 degrees centigrade in about 90 seconds at 6V.  Just use silicon cement not thermal paste to stick it onto the block so that the silicon coated heater and the aluminum sample block stay together.  The principle is the same.  The thermocouple is about 20$.  If you want to stay under 200$ then I'd suggest using the Phidget USB controller.  The on-board rela can switch a higher rated relay but consider adding a diode to the input terminals to "reduce inductive sparking" 

By the way we're about to post a 4 slide CISH IHC heated water bath that will hold slides at either 50 C or 98 C for 10 hours using a 20$ temperature controller (from MTM temperature controller ).  We are also trying a high temp limit controller but we needed to CNC our own PCB -  It will use the silicone coated heater as an example.

By the way, we're still working on the RTPCR part of this project using ultrabright LED, to excite the sybr green, a photoresistor and two band pass filters.  We're using the Scratch program and a PicoBoard to monitor temperature with a thermistor bead and to take timed light readings.  This would mean that anyone with a picoboard could run the RTPCR program from the internet.  You would just need to get the temperature cycling with any kind of controller i.e. phidget or a completely USB powered version.  One USB to power the fan, one for the heater and one for the PicoBoard.  If i need robotics I'll use the Lego Wedo robotics servos the are controlled by the Scratch Program.  You can even control switches with the servos, which would mean that the entire project could be run by a Scratch program that might take less than an hour to write.

Best regards


ErikaJavert (author)2009-09-23

Dear Dr. Siderits, Are you in current need of a lab assistant? Or possibly a wife?

siderits (author)ErikaJavert2009-09-23

Thanks so much for your comments. The RT-PCR module is coming along really well. Turns out you can do it with some new cheap ultrabright LED's, the lego WEDO robotics kit, a CDS (cadmium sulfide) photoresistor and the Scratch Pico Board. You don't even need a PMT (photo multiplier tube). Really all very cheap and replaceable parts. The trick now is to run the RT-PCR module from only a USB (4.5 V at 100 ma). We think that possible as well so it should run on a notebook. We use Scratch as the programming language so writing software for MAC-PC-Linux-Unix platforms is no problem. By the way, enthusiasm is the only requirement we have for membership on any of our research teams (and of course we're all wedded to science).

killianzim (author)2009-07-14

This is incredible! Our lab is about to purchase a $10,000 thermocycler, which makes your price tag all the more impressive.

siderits (author)killianzim2009-07-14

We are working on one that uses a battery powered (for the USB part) cigarette lighter for the heating element. We'll use a hobby fan to cool it and control from a microcontroller board. Should be under 100$ but will only take 2-4 samples.

snowdruid (author)siderits2009-08-08

why don't you use a peltier element? by changing the polarity you can change from cooling to heating cycle but I guess that would be a little bit to power hungry for a portable application...

siderits (author)snowdruid2009-08-08

That's a great idea. I have a few thoughts on that 1) you need to use several since thermal shock would rip them apart (one for cooling and two in step for heating, 2) although they require low voltage 14V they need higher amps 3-5 I think and the temperature controller that I used would have needed external switches to handle it; 3) I could have gotten pre-made thermal units with cooling panels and fans with the peltier junction but that would add on cost. We're developing a USB switched Pen-PCR version now that may use either a battery assist USB soldering iron type approach or possibly a battery powered-propane lighter for the heating element with a "Scratch Pico Board" to monitor the light and temperature cycles. An LED will excite the fluorecence and two band pass filters will separate excitation from emission of the marker. We have the 100 ma restriction on an average USB port but I may use two or three ports separately. Cheers!

matchett808 (author)2009-07-16

On the software side of it....It would be really easy to put together a custom application to do this....

....user input ---->interpreter---->instructions written to CRON ------>comms carried out by small C programmes

just an idea for ya to ponder....

...Nice build btw! lol

siderits (author)matchett8082009-07-17

Thanks; I think your right, that's the way we need to go. I'll probably prototype it with a multithreaded Python application through a Phidget interface board. Total cost for the system would then be about 150 dollars. Still working on the light sensor to make it a Real Time (RT) PCR platform but am stuck on using a photoresistor or a photodiode to detect 0.1 LUX fluoresced light. We may use a battery operated or even a butane powered heating source controled by dc switch on the IO board.

beninjoseph (author)2009-07-14

great job was very impressed.... best of luck

siderits (author)beninjoseph2009-07-14

Thanks for the positive comments. We're woking on the RT part of the RT-PCR system in a coffee cup.

beninjoseph (author)2009-07-14

is there a way to make a microtome ? would be of help ....

siderits (author)beninjoseph2009-07-14

Funny you should mention that. We're working on a "coffee cup microtome" for under ten dollars and will upload later this year. There was one on the drawing boards a few years ago called a "Pocket microtome" with a manual screw drive that cut about 10 microns and used razor blades, but I don't think it ever went into production and it probably has all kinds of patent restrictions. I know that one of the scientific houses has a pretty expensive one.

Otterthanwho (author)2009-07-11

Just like your functional, economical centrifuge, this unit offers expanded testing options to facilities and organizations that are financially challenged but want to provide advanced diagnostic tests to their patients and communities. To generously offer and provide this kind of information free to the public simply shows the generosity of you and your team.

siderits (author)Otterthanwho2009-07-14

Thank you very much for your comments. The Experimental Pathology and Rapid Prototyping teams are a bright and fun group of people to work with. I'd like to offer thanks to Instructables for making this kind of information available for entire regions of the world that have restricted economies and health care infrastructures that are still shaky at best. The same goes for areas of the US that have few resources to devote to education of scientists in their formative years. We are now working on a USB version for under 200$ and an RT-PCR Photocell circuit for a photocell.

Murf (author)2009-07-13

From someone who works in the research field (well actually studying now... a PhD student) and knows how much a stock PCR machine from one of the major suppliers... that`s amazing! And great. One of the main deterrents for small organizations to pursue research is costs. I commend you on finding an incredibly cheap alternative.

siderits (author)Murf2009-07-14

Thanks for the comments. We're working on the USB version for 200$ using either the interface board and/or the Parallax microcontroller board (there seem to be two camps on the team). It looks like the Photocell circuit, an ultrabright LED and two bandpass filters will make a two sample RT-PCR system possible for under 200$. It would "graph" to an LED display or to a serial interface to free graphing software for the microcontroller board.

ironsmiter (author)2009-07-13

So, If i understand the picture/text from step one.... I merely need to carefully bake myself a couple times in the oven, in order to achieve low-cost, home-brew human cloning? YES! Come tomorrow, there should be Thousands of half-baked me's running around! Step 2 : Take over the world!

siderits (author)ironsmiter2009-07-14

Step three: Destroy all remaining ovens.

About This Instructable



More by siderits:Ilustrated CAD process for creating a 3D printed assistive device in ABS plasticRapid technique for Fabricating Disposable Thermal Chambers to be used in testing Low temperature USB powered Thermofoils and heating elementsPrototype configuration for a USB Powered Low cost; real time continuous radiation monitoring system for Hospital Surgical Pathology, Hematology, Biological waste management and Research facilities
Add instructable to: