Copper beads as thermocycling "block" - Was: [DIYbio] What do open source PCRs need?

145 views
Skip to first unread message

Mac Cowell

unread,
Feb 11, 2016, 3:19:14 PM2/11/16
to diy...@googlegroups.com
The well block that heats the pcr tubes seems a little tricky to fab. 

Inline image 1

I think the wells have conical sides with a spherical tip at the bottom. I measured the wells in the cad file in the ninjapcr repo - the side walls slope about 8.8 degrees from the normal, and the well ends 0.45" from the surface of the block in a flat bottom that has a diameter of about .075" (~2/32").

The closest tooling I can find from mcmaster to cut something like that is with this $21 Tapered High-Speed Steel End Mill (10 Degree Taper, 3/32" Tip Diameter, 1/2" Length of Cut) for the conical section, and this $19 Ball-End High-Speed Steel Two-Flute End Mill (1/8" Mill Diameter, 3/8" Shank Diameter, 3/8" Length of Cut). $40 of tooling and not quite right.

Besides getting a batch of this milled by a machine shop, what are some other options for heating the tubes besides hot air or liquid?

Lab Armor sells these metal beads designed to replace liquid water baths for heating and cooling, but their opening price is $100+.


Perhaps we could build a box with peltiers + heat pumps on the top and bottom and perhaps forced air on the sides, and fill it with these metal beads. We could pack the pcr tubes inside the beads in the box and thermocycle the whole thing. It might not be fast, but perhaps it would be smooth and steady.

Spent a little time hunting around for cheaper alternatives with the same idea...

Found a listing on ebay for 1lb of (supposed) aluminum granules for $21 + shipping.
Inline image 1



And a 100-pack of 1/32" diameter aluminum spheres on mcmaster for $12 (I think 5-10 packs would be needed? Not sure about sphere packing... are spheres optimal shape because the allow for the highest net density? or worst b/c they minimize surface contact with one another and thus thermal conduction?).



McMaster also sells Aluminum oxide grit as a media for abrasive blasting, $28 for 10 lbs of 18 mesh grit. I guess it would pack and conduct heat much differently than the smooth spheres (worse?), but it's much cheaper.
Inline image 2




Lastly, I found an online retailer that supplies metals for jewelers and casters. They are selling 1 troy oz of ".999 pure" copper casting grain for $3. 16 tubes might need 10-20 times that mass.

Inline image 3

Some quick googling suggest that aggregates of metals like these conduct heat much worse than solids... perhaps only 10-30%? as thermally conductive as it would be if it were solid. That said, pure copper is apparently 1.5-2x as thermally conductive as pure aluminum .

So in conclusion, perhaps $60 of copper beads surrounding tubes would perform "within range" of a milled aluminum block - and without requiring a finicky heated lid.

Just some fun ideas!

Mac


Inline image 2

On Mon, Feb 8, 2016 at 11:32 PM Cathal (Phone) <cathal...@cathalgarvey.me> wrote:
I am very impressed with ninjapcr;

* Cheaper design
* Uses stock Arduino firmware!
* Has ninja in the name


On 9 February 2016 01:42:36 GMT+00:00, Mac Cowell <m...@diybio.org> wrote:
Just saw another open pcr project - "ninjapcr". OpenPCR fork. 


NinjaPCR_500.jpg
On Tue, Feb 2, 2016 at 6:48 PM Nathan McCorkle <nmz...@gmail.com> wrote:
On Tue, Feb 2, 2016 at 6:05 PM, John Griessen <jo...@industromatic.com> wrote:
a fast ramping air PCR thermocycler that needs no lid because it also
shakes the contents, (which also contributes to ramp rate performance)?

Now you're encouraging/helping phase-change of liquid-to-gas... it would be worth checking for prior-art. Also shearing of the DNA, people might look at you cross-eyed if you so much as drop their sample tube, or mix it too fast with your pipette tip. You're going to have to confidently demonstrate you're not adding new variables to a protocol pipeline.

--
-- You received this message because you are subscribed to the Google Groups DIYbio group. To post to this group, send email to diy...@googlegroups.com. To unsubscribe from this group, send email to diybio+un...@googlegroups.com. For more options, visit this group at https://groups.google.com/d/forum/diybio?hl=en
Learn more at www.diybio.org
---
You received this message because you are subscribed to the Google Groups "DIYbio" group.
To unsubscribe from this group and stop receiving emails from it, send an email to diybio+un...@googlegroups.com.
To post to this group, send email to diy...@googlegroups.com.
Visit this group at https://groups.google.com/group/diybio.
To view this discussion on the web visit https://groups.google.com/d/msgid/diybio/CA%2B82U9LaXvTwzaxsdG4dbfB0ttW_by-xccwG5_FCU4xYKUeT6w%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.


--
Sent from my Android device with K-9 Mail. Please excuse my brevity.

--
-- You received this message because you are subscribed to the Google Groups DIYbio group. To post to this group, send email to diy...@googlegroups.com. To unsubscribe from this group, send email to diybio+un...@googlegroups.com. For more options, visit this group at https://groups.google.com/d/forum/diybio?hl=en
Learn more at www.diybio.org
---
You received this message because you are subscribed to the Google Groups "DIYbio" group.
To unsubscribe from this group and stop receiving emails from it, send an email to diybio+un...@googlegroups.com.
To post to this group, send email to diy...@googlegroups.com.
Visit this group at https://groups.google.com/group/diybio.
To view this discussion on the web visit https://groups.google.com/d/msgid/diybio/B6CCA208-6B66-4930-976F-07F272249B1B%40cathalgarvey.me.
For more options, visit https://groups.google.com/d/optout.

John Griessen

unread,
Feb 11, 2016, 3:38:45 PM2/11/16
to diy...@googlegroups.com
On 02/11/2016 02:18 PM, Mac Cowell wrote:
> googling suggest that aggregates of metals like these conduct heat much worse than solids... perhaps only 10-30%? as thermally
> conductive as it would be if it were solid. That said, pure copper is apparently 1.5-2x as thermally conductive as pure aluminum
> <https://en.wikipedia.org/wiki/List_of_thermal_conductivities>.
>
> So in conclusion, perhaps $60 of copper beads surrounding tubes would perform "within range" of a milled aluminum block - and
> without requiring a finicky heated lid.
>
> Just some fun ideas!
>
> Mac

The beads or grit all have air space that is insulating, so they won't conduct well, and they have mass, so the low conduction and
mass inertia makes for an extra long heat up exponential curve.

Actively stirred hot air, then opening a door will beat all these plus metal milled blocks too.

Zinc is easy to melt and inexpensive. You could make up a plate with pins in it of your desired shape of vials and cast it. Just
for fun. Lead is even easier, and easier to find laying around since we still use it for car batteries and tire weights. Not PC
though.

Copper is not easy to cast -- pours like molasses/honey, (yes, I've done it), and you're at glowing yellow orange temperature.
Not fun unless you like metal a lot.

Simon Quellen Field

unread,
Feb 11, 2016, 3:42:46 PM2/11/16
to diybio
It is easy to cast tin. Melts at 450 F.
It melts on the kitchen stove (use an old cast iron skillet).
The mold can be made from clay or plaster of Paris, or you can sand cast it.
Place the plastic tubes upside-down in a baking pan and cover them in silicone rubber.
When the rubber hardens, it is a negative mold for the plaster of Paris.

You can also use lead-free solder (melting point 422 to 430).
Zinc will melt on a cheap electric "fifth burner" stove (so the wife won't complain, and you can do it outdoors without a hood). The molten zinc will then dissolve aluminum, and you can cast with the alloy that results. You can melt US pennies as your zinc source.

This email has been sent from a virus-free computer protected by Avast.
www.avast.com

-----
Get a free science project every week! "http://scitoys.com/newsletter.html"


Mac Cowell

unread,
Feb 11, 2016, 10:04:38 PM2/11/16
to diy...@googlegroups.com

---------- Forwarded message ---------
From: Mac Cowell <m...@diybio.org>
Date: Thu, Feb 11, 2016 at 2:32 PM
Subject: Re: Copper beads as thermocycling "block" - Was: [DIYbio] What do open source PCRs need?
To: John Griessen <jo...@industromatic.com>


According to wolfram alpha, copper is 6 times more thermally conductive than tin (400 W/mK vs 67 W/mK).

That's pretty big. Now I want to run the experiment. 

We use a solid tin, and a solid aluminum thermocycling block vs copper beads packed all around the tubes. Even if the beads lose 80% of their thermal conductivity due to airspace and poor contact, they will still be conducting (theoretically) 0.20 * 400 W/mK = 80 W/mK, which is more than the solid tin.

Additionally, I would like to see what effect forcing air through the copper matrix has on the cooling rate...

I'm not a physicist, anyone else want to explain how this is really going to work?

On Thu, Feb 11, 2016 at 1:48 PM John Griessen <jo...@industromatic.com> wrote:
If you do any casting with plaster of paris you need to dry it completely before pouring hot metal, or the steam coming out can
splash the hot metal at you.  (never experienced that and never want to -- always poured dry molds)  Ambient conditions can get
plaster looking dry, but not dry enough -- needs oven time after days of air drying, or more oven time from start.

JG

Mac Cowell

unread,
Feb 11, 2016, 10:18:04 PM2/11/16
to diy...@googlegroups.com
Here's the fermi question: if instead of the conventional aluminum block + heated lid system, we were to pack the sample tubes into a volume of copper particles, and in turn pack these into a box with surfaces that are a either conductive heater/coolers like those for the aluminum block in the conventional design or vents with fans that can blow air into or out of the box and through the copper matrix, then:

1) what distribution of the copper particles' shapes and masses is optimal for thermocycling the tubes? Spheres? Chain? Random chunks? Big, small, or a mixture?

2) if the thermocyclers have identical power supplies, which system is more efficient, and by how much? 1.2x, 2x, 10x, 100x... Whatever the answer I think it will be an interesting surprise.

Consider a hot air convection thermocycler: is it better to preheat the airstream away from the sample and maximize air speed, or is the reduced effectiveness of the slower airspeed that results from packing the heater around the samples offset from conductive heat transfer?

Mac

Simon Quellen Field

unread,
Feb 11, 2016, 10:45:57 PM2/11/16
to diybio
It occurred to me that I had been putting the premium on heat capacity instead of on thermal conductivity.

Graphite (especially pyrolytic graphite) has exceptional thermal conductivity, and is very easy to machine.

Cut from sheet steel a silhouette of the tubes you want to use, and chuck that in a drill press.
It will easily cut a hole in the graphite the exact shape of the tubes you want to place in the block.

You might even be able to use the graphite block as the heating element directly.
Just run current through it, and it heats up.
Blow cold air on it to cool it down -- it cools faster than metals, because it has less heat capacity.

And regular graphite is cheap [pyrolytic not so much :-) ].



-----
Get a free science project every week! "http://scitoys.com/newsletter.html"


--
-- You received this message because you are subscribed to the Google Groups DIYbio group. To post to this group, send email to diy...@googlegroups.com. To unsubscribe from this group, send email to diybio+un...@googlegroups.com. For more options, visit this group at https://groups.google.com/d/forum/diybio?hl=en
Learn more at www.diybio.org
---
You received this message because you are subscribed to the Google Groups "DIYbio" group.
To unsubscribe from this group and stop receiving emails from it, send an email to diybio+un...@googlegroups.com.
To post to this group, send email to diy...@googlegroups.com.
Visit this group at https://groups.google.com/group/diybio.

John Griessen

unread,
Feb 12, 2016, 11:03:45 AM2/12/16
to diy...@googlegroups.com
On 02/11/2016 09:17 PM, Mac Cowell wrote:
> Consider a hot air convection thermocycler: is it better to preheat the airstream away from the sample and maximize air speed, or
> is the reduced effectiveness of the slower airspeed that results from packing the heater around the samples offset from conductive
> heat transfer?

max air speed will give max heat transfer. Stirring by going past turbulence generating fins is needed to avoid laminar flow
where different layers can have different temperatures. Heater around the samples will definitely make the air slower, and
temperatures less even. Beads of metal will just be intermediate heat capacity creating heat flow lag.

Lee Nelson

unread,
Feb 12, 2016, 8:02:37 PM2/12/16
to diybio

You could use metal infused filament and 3D print the block.

https://blog.pinshape.com/3d-printing-metal/

You may be able to make a computer model of the heat transfer on a site called SimScale.

https://www.simscale.com/

Alex D

unread,
Feb 15, 2016, 8:57:15 AM2/15/16
to diy...@googlegroups.com
HI all, 

I dont really understand how Ti is not an option here? I have looked at the tables and Ti is stated 20x less then copper and roughly 10 time less conductive then Al ? I am not a specialist in this area so perhaps someone could clarify or had an experience with Ti? But I do camping sometimes and all of my cooking equipment is Grade A titanium, what I can add is when you boil a water in titanium cup and then pour water out of it, the cup itself goes from ~90C to room temperature in a matter of 3-4 seconds. Also I have used the titanium cups to boil water on the fire and on the induction oven , I havent tried copper cops but it boils water definitely twice faster then any Al cups ive had. 

anyone can clarify on that?

--
-- You received this message because you are subscribed to the Google Groups DIYbio group. To post to this group, send email to diy...@googlegroups.com. To unsubscribe from this group, send email to diybio+un...@googlegroups.com. For more options, visit this group at https://groups.google.com/d/forum/diybio?hl=en
Learn more at www.diybio.org
---
You received this message because you are subscribed to the Google Groups "DIYbio" group.
To unsubscribe from this group and stop receiving emails from it, send an email to diybio+un...@googlegroups.com.
To post to this group, send email to diy...@googlegroups.com.
Visit this group at https://groups.google.com/group/diybio.

Simon Quellen Field

unread,
Feb 15, 2016, 11:07:37 AM2/15/16
to diybio
Titanium has a heat capacity of 0.125 calories per gram.
It thus sits between aluminum with 0.22 and copper with 0.092.

The thermal conductivity of titanium is about 12, on a scale where copper is 223, and aluminum is 118.

As titanium has a density of 4.43, to copper's 8.96 and aluminum's 2.7, and titanium is stronger than copper, I suspect that a titanium cup weighs less than one made from copper, and perhaps also one made from aluminum.

Boiling water in an aluminum cup might take a tiny amount more time than in a titanium cup, due to the slight difference in heat capacity and mass, but it will be dwarfed by the heat capacity and mass of the water, and any difference may be difficult to measure.

The low thermal conductivity of titanium is a problem, since it will heat unevenly. A copper bottomed pot will heat more evenly (which is why you find copper bottomed pots for sale), and aluminum is not far behind. But you would not want to scramble eggs in a titanium skillet over a fire -- you'd have to be very careful not to scorch them. Cast iron skillets, known for their low thermal conductivity (you can hold the iron handle while the rest of the skillet is in the fire), still have three times the conductivity of titanium, and use their mass to make up for the low conductivity, heating slowly enough that the heat has time to even out.

For a thermocycler, you want something with low heat capacity, low density, and high thermal conductivity. Graphite would be very hard to beat. Aluminum would beat titanium because of the dismal thermal conductivity of the latter.

The price of the titanium used in camping gear has fallen to about $4 a kilogram due to new cheaper manufacturing processes. Aluminum is $1.47 per kilogram, and is easier to machine, and much more thermally conductive, and less dense (2.7 g/cc to titanium's 4.5). These make it much preferable as a heat sink material, and also preferable in a thermocycler.

-----
Get a free science project every week! "http://scitoys.com/newsletter.html"


John Griessen

unread,
Feb 15, 2016, 11:47:30 AM2/15/16
to diy...@googlegroups.com
On 02/15/2016 07:57 AM, Alex D wrote:
> t it boils water definitely twice faster then any Al cups ive had.
>
> anyone can clarify on that?

It is probably because of the extreme thinness of the strong Ti sheet the pan is made of.
So you can probably see the flame pattern in the boiling on the pans bottom -- very little heat spreading, mostly flowing through
the thin sheet and transferring to the liquid.

Pan thinness of Ti pans is usually about 1/2 that of aluminum that I have seen. And conductivity is 10X different Ti less than
Al. The thinness is not all of it. The fire is much hotter than boiling is another thing. Even though the conductivity of a
half as thick ti pan might be 5X less than aluminum, there is still plenty of temperature difference to get heat flowing to
the boiling water. Maybe the formation of hot spots helps speed the boil? Hot spots would be causing early boiling that stirs
the water well and speeds heat up. Could the thicker aluminum pan be so much more even that laminar layers of hot/warm/cool water
form from the bottom up in the pan without stirring much for a several minutes, while the Ti pan is already stirring like mad
because of one spot boiling early?

I've seen a neato product for camping called a turbo rocket boil something that advantages Titanium. It has insulating sleeve
around a shell separated by corrugated metal from an inner pot such that the flow of propane flames goes along the sides of the
boiling pot effectively. That pot boiler has a vertical shape, compact, protected from dings by having the two layers. I bet its
inner pot is really thin and that is its trick.

Simon Quellen Field

unread,
Feb 15, 2016, 3:35:19 PM2/15/16
to diybio
The time to notice a spot that is boiling might be short.
I can insert my soldering iron into water and get to a boil very quickly.
The time to get all of the water to 99 degrees Celsius should be about the same in aluminum or titanium, for quantities of 100 ml or more.


-----
Get a free science project every week! "http://scitoys.com/newsletter.html"


--
-- You received this message because you are subscribed to the Google Groups DIYbio group. To post to this group, send email to diy...@googlegroups.com. To unsubscribe from this group, send email to diybio+un...@googlegroups.com. For more options, visit this group at https://groups.google.com/d/forum/diybio?hl=en
Learn more at www.diybio.org
--- You received this message because you are subscribed to the Google Groups "DIYbio" group.
To unsubscribe from this group and stop receiving emails from it, send an email to diybio+un...@googlegroups.com.
To post to this group, send email to diy...@googlegroups.com.
Visit this group at https://groups.google.com/group/diybio.

William Beeson

unread,
Feb 15, 2016, 7:52:33 PM2/15/16
to DIYbio
Why couldn't a system like the following work:

Two large temperature reservoirs: Ice water and sous vide at 99C
A programmable valve that controls the proportion of flow from each source
A peristaltic pump that flows at >100 ml/min
3D printed and insulated incubation chamber (10 mL working volume)

The temperature reservoirs could actually be hooked up to many "pcr" machines and all you need is the arduino controlled valve.  The "low" cost PCR machines are in the $600-700 range, but they can only handle a small number of tubes.  An advantage of this approach is that it would be scalable to much larger numbers of reactions -- which is useful if you want to do larger scale genotyping.  The tubing, valves, and peristaltic pump aren't very expensive.  Sous vide immersion heater is less than $200.  I know the sous vide can hold temperature accurately to +/- 0.5 C and ice saturated water is very constant at ~0C.  So long as the peristaltic pump and valves work reliably it should be very easy to calculate the mixture rate to get a given temperature so long as the tubing and reaction chamber is adequately insulated.


Simon Quellen Field

unread,
Feb 16, 2016, 1:13:07 PM2/16/16
to diybio
Keep one pot boiling (don't bother to even monitor the temperature).
Keep another pot full of cold water from the tap (don't bother to cool it, or monitor the temperature).
Put a battery powered pump in each pot, controlled by a cheap Arduino Nano for $1.83.
Put a digital temperature sensor in one or more of the 96 tubes in the rack.
Pump boiling water into the bath the rack is sitting in.
Use PID control software to tell when to stop pumping hot water and start pumping cold water.
Now you have a simple fast ramp thermocycler that can only cycle (it can't hold an intermediate temperature).
To get intermediate temperatures, the PID control software will control the speed of each pump.
No valves needed (they are five times the cost of the pumps).


-----
Get a free science project every week! "http://scitoys.com/newsletter.html"


William Beeson

unread,
Feb 16, 2016, 6:45:24 PM2/16/16
to DIYbio
I like the idea of just directly controlling the flow rate of two pumps instead of requiring a switching valve.  If you can adjust the flow by just adjusting the voltage/current to the pump in a feedback loop with a temperature sensor that is probably the cheapest.  I was originally thinking of using a calculation to determine theoretically what ratio should be mixed together.  It is probably a lot easier to just control the circuit (apply more voltage to one pump and less to the other until T is reached).

I think the advantage to using something like a sous vide (homemade or bought on amazon for $180) is the safety aspect.  If you have boiling water that probably requires an external heat source and the potential that the water could boil over or evaporate entirely if left unattended.  If you hold the temperature just below boiling and keep the top covered you can maintain the high temperature indefinitely without a risk of running out of liquid or overheating.  I think the commercial sous vide also have auto-shutoff safety features.  I see the point about using cold tap water.  Advantage of ice water is then you can store the reactions overnight at low temperature if desired.  I think a good igloo cooler can keep ice overnight.

Could this really be built pretty easily?  I already have the sous vide and a MiniPCR (for direct comparison).  

Simon Quellen Field

unread,
Feb 16, 2016, 8:20:00 PM2/16/16
to diybio
The Sous Vide cooker is overkill. A cheap ($20) rice cooker will boil water, and automatically shuts off when the weight gets low (there's a spring-loaded switch under the pot that turns the power off). Plain water just boils, it doesn't "boil over", as there is nothing to make foam from.

If you want to store something cold overnight, put it in the fridge.

There are two groups of people interested in thermocyclers: those that want a compact, reliable, neat, clean, piece of laboratory equipment so they can get their real job done, and those of us who like building gadgets. I am sometimes in each category -- for example, I love Cathal's Dremelfuge, but when I found I needed a centrifuge, I just bought one (they're cheap) and I have a nice appliance sitting next to the microscope that is convenient and easy. The hot and cold water method will never be a convenient appliance sitting on a lab bench, but it will be fun to design and build, and work the kinks out of. Let's keep looking into ways to make the convenient appliance as well.


-----
Get a free science project every week! "http://scitoys.com/newsletter.html"


John Griessen

unread,
Feb 16, 2016, 8:23:20 PM2/16/16
to diy...@googlegroups.com
On 02/16/2016 05:45 PM, William Beeson wrote:
> Could this really be built pretty easily?

If those $4.31 pumps can take the hot temp. Why not just find out?
(Me, I still like air as the transfer medium)

Simon Quellen Field

unread,
Feb 16, 2016, 8:52:14 PM2/16/16
to diybio
I ordered some before I posted their price. I'll let you know.

I also like the hot air version, especially for the convenient lab appliance I spoke of.
Most blow driers have a switch that turns off the heater to blow cold air.
Instrumenting that switch might be interesting.
The last link I gave was to my AC PID controller project, which could be used to control the temperature of the blow dryer heater.


-----
Get a free science project every week! "http://scitoys.com/newsletter.html"


--
-- You received this message because you are subscribed to the Google Groups DIYbio group. To post to this group, send email to diy...@googlegroups.com. To unsubscribe from this group, send email to diybio+un...@googlegroups.com. For more options, visit this group at https://groups.google.com/d/forum/diybio?hl=en
Learn more at www.diybio.org
--- You received this message because you are subscribed to the Google Groups "DIYbio" group.
To unsubscribe from this group and stop receiving emails from it, send an email to diybio+un...@googlegroups.com.
To post to this group, send email to diy...@googlegroups.com.
Visit this group at https://groups.google.com/group/diybio.

Alex D

unread,
Feb 16, 2016, 9:10:22 PM2/16/16
to diy...@googlegroups.com
Hello Simon, 

thanks for the full review)) the reason I have mentioned Ti is from a camping experience. I have also tried Al  cups but they loose to Ti I would say by 20-30% in time for heating, but in cooling for some reason Ti cools within few seconds while Al or Cu takes much longer time. In PCR cooling takes probably the most time as i have noticed I am not sure if there is a relay to switch Peltier's polarity so the hot side turns into cool side but perhaps by placing a second pertier unit right under it will harvest all the heat from the one that is attached to the heat sink perhaps... 

I recently met a guy who works for AUTOCAD he mentioned that they are working on liquid cooling PCR where it is heated the normal way not sure which is it peltier based or by enduction, but he said that they are trying to design a water cooled heat sink similar to the computer liquid  cooling system so as he mentioned times could be cut from 3h to 30min not sure how is it possible if 30 cycles by itself are 30min but thats what I have heard. 

There is another thing GE works on is magnetic cooling which creates magnetic field to align molecules and force them to loose the energy, i believe they already have built the prototype of a new type of refrigerator   not sure how fast that would be but lack of any mechanical parts would definitely have a benefit as well as lack of peltier units which in my experience break quiet often. 

Simon Quellen Field

unread,
Feb 16, 2016, 10:04:26 PM2/16/16
to diybio
You said:
for some reason Ti cools within few seconds

I joke that my job is teaching physics to people with short attention spans.
The effect you noticed is called "heat capacity".
Titanium has low heat capacity. So does air. Water has very high heat capacity.

I think the reason people are using big aluminum blocks in thermocyclers is to even out the heat, so all of the tubes are heated and cooled equally.
An experiment with 96 tubes in a thermocycler would be suspect if the tubes were not all subjected to the same temperatures.
Thermal conductivity is important in evening out the temperatures, and titanium has low conductivity.
Again, I would prefer graphite if we determine that the temperatures are uneven.

Flowing water or flowing air would have to be carefully arranged so that turbulence and thorough mixing occur, so that all tubes are heated and cooled equally.

Peltier junctions are notoriously inefficient. Most of the benefit in computer chip cooling is due to the large heat sinks and air flow, not to the Peltier junction.
Putting electricity into a Peltier junction causes the system as a whole to heat up, as most of that electricity goes into heating the Peltier junction.
A little bit goes into pumping heat from one side to the other, if the hot side is properly cooled, either by a large heat sink and fan, or by a liquid cooling system.

That said, there are situations where we throw energy at the problem without caring a lot about the cost.
In data centers, the chips are not cooled by Peltier junctions. They are water cooled. Power is the main cost in a data center, and more power is spent on cooling than on powering the computers.

There are very few cooling solutions that can drop the temperature faster than dunking the hot thing in cold water.
Cold water is also cheap.
Adding a spray of cold water to the hot air thermocycler is something we have not talked about yet, but it seems easy to accomplish with a cheap water pump and an atomizer nozzle. Blow cold air past the spray and let the phase change from liquid to vapor soak up the heat in a hurry.
The most efficient refrigerators are all designed around a phase change.

The advantage of using the liquid cooling system from a gaming computer is that it is inexpensive, widely available, and pre-made.
We should look into that for the convenient appliance thermocycler.
They are also quiet, which has many advantages in a lab.

On the related subject of inexpensive lab equipment:
I've noticed that single-pot induction burners are inexpensive ($99), and can be set to a particular temperature like an oven, in the range of 160 to 430 Fahrenheit.
While I would not expect the temperature to be all that accurate or precise, many lab tasks don't require high precision or accuracy. Boiling water, for example.
And they are easy to clean. Some have built-in automatic timers.
One for $49 has five heat settings. This one for $60 has ten temperature settings (140 F to 460 F). This one has 52 temperature settings in 10 degree F increments from 100 F to 575 F, and is programmable to 100 hours of heating time or delayed start.
And speaking of titanium, this one has increments of 5 degrees F, and 94 temperature settings. Boils water in 60 seconds.


-----
Get a free science project every week! "http://scitoys.com/newsletter.html"


William Beeson

unread,
Feb 16, 2016, 11:55:39 PM2/16/16
to DIYbio
The induction burners are great from the safety perspective.  I skew a lot of my interest towards a community lab space where people with limited experience will work.  The biggest hazards in a biotech lab are burns, acids/bases, and cuts from broken glass.  That's a great price for an induction top.  When I was inside lowes and bestbuy the induction stove tops were several hundred dollars.  

I really like the flowing water system because when you have a closed loop and a high enough flow rate the water bath temperature will equilibrate very quickly (probably faster than even commercial PCR machines).  All we need is some simple software to control the temperature cycling and we would have a scalable, safe system that works for $200-300.  In my mind the big limitation of the other lower cost PCR machines on the market is they only accomodate 8-16 sample and still cost $600+ each.  The heating/cooling source in the proposed water recirculation PCR machine could be shared among many units consisting of the reaction chamber and the arduino controlled pumps -- reducing the cost per additional PCR machine to less than $50..   

Alex D

unread,
Feb 17, 2016, 2:00:52 AM2/17/16
to diy...@googlegroups.com
induction heater could be built way cheaper, but if something will spray the cold mist over the heat sink would it force the heat sink loose more heat? perhaps something that evaporates in far lower temperature then water even like alcohol , but then it has to be sprayed at very low temperature and cooled after it evaporated?  perhaps something like tosol or engine coolant or freon could be pumped 

Simon Quellen Field

unread,
Feb 17, 2016, 10:16:26 AM2/17/16
to diybio
Water is best for many reasons.
It has a much higher heat capacity than alcohol.
It is cheap, and the lab has a ready supply of it.
It is non-toxic, and easy to clean up in case of a spill.
It is odorless.
When sprayed into a stream of cold air from a blow dryer, it will be quite cool enough.

All that said, let's first try just the cold air. It will probably be quite enough by itself.
We can add the spray later if it is needed.

-----
Get a free science project every week! "http://scitoys.com/newsletter.html"


Bryan Jones

unread,
Feb 17, 2016, 11:27:57 AM2/17/16
to diybio
The downside to a water based system is that there are more moving parts. I'd imagine that pumps and valves tend to have a shorter lifespan than solid state parts like a peltier device. Water also tends to grow things if it's not changed regularly or the parts are not cleaned.

Simon Quellen Field

unread,
Feb 17, 2016, 11:52:03 AM2/17/16
to diybio
Quite true.
The two $4.31 pumps may have to be replaced yearly.
We should consider them consumables.
Given that half the time in operation, the water will be near boiling, I am less worried about things growing.
And most of the time, or at least at night, the device will be dry.
We could always add a little bleach or hydrogen peroxide to the water.

-----
Get a free science project every week! "http://scitoys.com/newsletter.html"


William Beeson

unread,
Feb 18, 2016, 1:32:38 PM2/18/16
to DIYbio
As long as the parts are low cost and the assembly is simple I think it's generally better to go the low cost route than the more reliable part route. You just need to make sure the system is designed so you can quickly recognize it's performance is degraded. The parts need to have a low probability of failure during any given analysis/operation, but multi year reliability shouldn't be needed or expected for most DIY. The concept of the pumps being a consumable is right on.

Nathan McCorkle

unread,
Dec 19, 2017, 8:29:11 PM12/19/17
to diybio
Exploring the limits of ultrafast polymerase chain reaction using liquid for thermal heat exchange: A proof of principle
Appl. Phys. Lett. 97, 264101 (2011); https://doi.org/10.1063/1.3530452

http://aip.scitation.org/doi/full/10.1063/1.3530452

"""
We explored the limits of speed by using liquid for thermal exchange rather than metal as in traditional devices, and by testing different polymerases. In a clinical setting, our system equaled or surpassed state-of-the-art devices for accuracy in amplifying DNA/RNA of avian influenza, cytomegalovirus, and human immunodeficiency virus. Using Thermococcus kodakaraensis polymerase and optimizing both electrical and chemical systems, we obtained an accurate, 35 cycle amplification of an 85-base pair fragment of E. coli O157:H7 Shiga toxin gene in as little as 94.1 s, a significant improvement over a typical 1 h PCR amplification.
"""

On Tue, Feb 16, 2016 at 5:51 PM, Simon Quellen Field <sfi...@scitoys.com> wrote:
I ordered some before I posted their price. I'll let you know.

I also like the hot air version, especially for the convenient lab appliance I spoke of.
Most blow driers have a switch that turns off the heater to blow cold air.
Instrumenting that switch might be interesting.
The last link I gave was to my AC PID controller project, which could be used to control the temperature of the blow dryer heater.


-----
Get a free science project every week! "http://scitoys.com/newsletter.html"


On Tue, Feb 16, 2016 at 5:23 PM, John Griessen <jo...@industromatic.com> wrote:
On 02/16/2016 05:45 PM, William Beeson wrote:
Could this really be built pretty easily?

If those $4.31 pumps can take the hot temp.  Why not just find out?
(Me, I still like air as the transfer medium)


--
-- You received this message because you are subscribed to the Google Groups DIYbio group. To post to this group, send email to diy...@googlegroups.com. To unsubscribe from this group, send email to diybio+unsubscribe@googlegroups.com. For more options, visit this group at https://groups.google.com/d/forum/diybio?hl=en

Learn more at www.diybio.org
--- You received this message because you are subscribed to the Google Groups "DIYbio" group.
To unsubscribe from this group and stop receiving emails from it, send an email to diybio+unsubscribe@googlegroups.com.

To post to this group, send email to diy...@googlegroups.com.
Visit this group at https://groups.google.com/group/diybio.

--
-- You received this message because you are subscribed to the Google Groups DIYbio group. To post to this group, send email to diy...@googlegroups.com. To unsubscribe from this group, send email to diybio+unsubscribe@googlegroups.com. For more options, visit this group at https://groups.google.com/d/forum/diybio?hl=en

Learn more at www.diybio.org
---
You received this message because you are subscribed to the Google Groups "DIYbio" group.
To unsubscribe from this group and stop receiving emails from it, send an email to diybio+unsubscribe@googlegroups.com.

To post to this group, send email to diy...@googlegroups.com.
Visit this group at https://groups.google.com/group/diybio.

For more options, visit https://groups.google.com/d/optout.



--
-Nathan

John Griessen

unread,
Dec 20, 2017, 2:58:08 PM12/20/17
to diy...@googlegroups.com
On 12/19/2017 07:28 PM, Nathan McCorkle wrote:
> Exploring the limits of ultrafast polymerase chain reaction using liquid for thermal heat exchange: A proof of principle
> Appl. Phys. Lett. 97, 264101 (2011); https://doi.org/10.1063/1.3530452

And here's the juicy details of heat transfer:


"Our thermal cycler, Fig. 1, used a liquid interface sand-
wiched between two high-powered Peltier devices,
(Marlow XLT2393, Dallas, TX), that worked in tandem to cycle the
temperature rapidly. They rejected waste heat to a pair of
fan-cooled heat sinks. These high efficiency heat sinks were
sized to allow for increased cooling rates due to their ability
to bias the heat exchange toward lower temperatures, opti-
mizing power usage at less than 400 W. A phenolic gasket
held 2.3 ml of thermal transfer medium, (gallium eutectic),
between the two face surfaces of the Peltiers, making direct
thermal contact between them and the sample capillary tube.
The liquid allowed for excellent thermal contact to the glass
capillary tube. The small volume needed to make thermal
contact required less than 200 J to raise the medium’s tem-
perature from 60 to 95 °C. In a typical thermal cycler, the
thermal resistance between the junction of the solid metal
block and the sample tube is a dominant factor in reducing
the ability to transfer heat in and out of the sample, resulting
in slow thermal cycling."

See how they key on, "required less than 200 J"?

Losing the air gap between solid blocks with thermal mass
is a speed up.

My plan to do air PCR with movable doors will allow quick
change from heating to cooling without need to switch the direction
or even on/off state of a Peltier cooler, so any thermal mass
built into the cooler that slows switching between heating and cooling
will have no slowing effect using air and movable doors.

The airpcr air will be the only medium of transfer of heat between hot nichrome
heater element wires and sample tubes. It will be moving fast to
give complete stirring for even temperature, and quick heat transfer.

Still busy with electroporation, but stay tuned.

--
John Griessen
industromatic.com Austin TX building lab gear for biologists

Jonathan Cline

unread,
Dec 23, 2017, 4:28:26 PM12/23/17
to DIYbio

Silicone mold for tubes mounted directly on automotive heater core ($28 *) for heat transfer (**) with fan ( > $10) mounted below to blow through the radiator fins.  Minimum component cost at least $35 for air design.


* 1980-1997 Ford F250 Heater Core - APDI 9010205.  Note that no new mechanical design will get cheaper or more efficient than this. 

** However the lab market has already moved away/rejected air driven pcr so not sure why rehashing that outdated design is considered productive.  No it won't be made better.  Also, no lab tech wants to deal with gunk on outside of tubes regardless of whether it optimizes heat transfer or for some other purpose.


-- 
## Jonathan Cline
## jcl...@ieee.org
## Mobile: +1-805-617-0223
########################

Reply all
Reply to author
Forward
0 new messages