Aaargggggggghhhhh Can't get any current from ny TEG

Submitted by ITSadmin on

Printer-friendly versionPDF version

Forums: 

Hi all,

I'm new to this fourm and having a major problem with my TEM. The max rated current is 2.1Amps and max voltage is 2.19Volts with a THot of 50 deg celcius. and the max temperature difference across the hot and cold plates is 70 deg celcius.

I setup a test to see what currents I can get from the TEM as shown in the diagram below; the hot side is exposed to the light from a 60W light bulb and the cold side is attached to a coper pipe of 120mm in length and 40mm diameter. After 15 minuites the air on the hot side was 62.7deg celcius and the temperature of the copper pipe was 24.1 deg celcius giving a 38.6deg celcius temperature difference acros the hot and cold plates. The thing is I have the TEG powering an LED, and an ammeter connected in series with the LED so I can get the circuit current and with a temperature differnece across the hot and cold plates of 38.6 deg ceclius I'm getting zero current :? ?!?!?!?

What am I doing wrong? With this temperature differnece I would expect currents of at least half the max current as I generate a temp difference of more than half the max temperature differene!

I have contact melcor to ask if I'm doing something wrong, but that was a week ago and they have still not got back to me and I can't proceed with my project until this is sorted.

Has anyone ever used a TEM as a generator for the temperature range I am using? Is it possible to generate a decent current with Thot being about 60 deg celcius as I thought it's all about the temperature difference of the two plates?

I would be greateful if anyone out there can help as I don't know what I can do to generate a decent current at this temperature ( want about 0.9-1.0Amps )

Re:Aaargggggggghhhhh Can't get any current from ny TEG

Hi,

Vmax and Imax are not useful for the TEM in generation! They are only cooling parameters.
With a single module with that DeltaT is impossible to achieve that current.

I think you need bigger DT, and the max output power at a specified DT depends on the ratio RL/Rint, where Rint is the internal resistance of the module (at the mean temperature), and RL is the load resistance (your led). The max output power is achievable only if that ratio is 1.

Moreno

Re: Aaargggggggghhhhh Can't get any current from ny TEG

Dear "Aaargggggggghhhhh",

Nice figure by the way.

As someone else has already noted, the cooler parameters don't (directly) tell you much about the performance as a power generator.

First, just some definitions.

Vmax and Imax are the voltage and current which give maximum cooler power (at zero dT, which would require a perfect heat sink to achieve!). At Vmax, Imax you get Qmax of cooling power.

dTmax is the maximum temperature difference you can achieve. None of these things have anything to do with power generation, but they all tell you stuff about the module. And from them, you can estimate power generation properties.

So, in your case you have applied heat to acheive 39 K of temperature difference. Check to see that the module produces an open circuit voltage. Bigger dT, bigger voltage.

Using the magic of thermoelectrics, we can estimate the open circuit voltage (only roughly) from Vmax:

V_open = Vmax * dT/300
~ 2.19 * 39/300 ~ 0.28 V

That "300" comes from 300 K for room temperature, the typical condition for determining Vmax. Call it magic <g>.

We can also estimate the internal resistance of the module:

R_int = (Imax/Vmax) * (1 - dTmax/300)
~ (2.1/2.19) * (1 - 70/300) ~ 0.73 ohms

Now if you short circuit the module you should get:

I_short = V_open/R_int
~ 0.38 A

Check your setup, without the LED in the circuit. My guess? Your LED needs a larger voltage than 0.28 to 'turn on'. So it is behaving as a high resistance device, hence no current will flow.

Let us know how you get on.

- Cronin

[quote="Anonymous"]Hi all,

I'm new to this fourm and having a major problem with my TEM. The max rated current is 2.1Amps and max voltage is 2.19Volts with a THot of 50 deg celcius. and the max temperature difference across the hot and cold plates is 70 deg celcius.

Aaaaargggggghhhhh

Also make sure you're forward biasing the LED. Even if you have sufficient voltage to turn on the LED in the forward bias direction, if you have the opposite polarity, you're not going to get any light output.

Re:Aaargggggggghhhhh Can't get any current from ny TEG

Dear Cronin,

thanks for these "magic formula".

Could you tell me where you have found them, or give me a reference for a specific article?

Is there a big approssimation in them or not?
I've tried to put some data from a manufacturer, but the calculated internal resistance is lower than the actual resistance measured at 300K. (0.52 vs 0.96 ohm)

Thanks

Moreno

Re:Aaargggggggghhhhh Can't get any current from ny TEG

Moreno,

I made them up <g>.

Seriously, I just used standard expressions for Vmax and Imax which relate to the temperatures and the devices effective seebeck coefficient and internal resistance. Once you have those, you can back out the rest. The main approximation I suppose is that the resistance and seebeck coefficient are independent of temperature. They are not, of course, but still this should be pretty close.

As for a basic reference, see

S. W. Angrist, Direct Energy Conversion, 1982.

I have the fourth edition, and the expressions for Vmax & Imax are on page 151. But all the magic is in there.

Harrumpf. I just realized I goofed in those expressions. dT=0 for the Qmax, Vmax, Imax condition so the expressions I gave above are wrong. Dopey me!

Hang on and I'll give better ones in a sec. Not good. Not good.

[quote="Anonymous"]Dear Cronin,

thanks for these "magic formula".

Could you tell me where you have found them, or give me a reference for a specific article?

Is there a big approssimation in them or not?
I've tried to put some data from a manufacturer, but the calculated internal resistance is lower than the actual resistance measured at 300K. (0.52 vs 0.96 ohm)

Thanks

Moreno[/quote]

Re:Aaargggggggghhhhh Can't get any current from ny TEG

Dear Moreno,

See my post in a new thread with more "magic formulas". They may be good magic or bad magic. We'll see.

My corrected expression for R_int will bring you your calclated internal resistance value by about 30%. The rest of the difference with your measured number may be due to generating an internal temperature difference when you measure the resistance. The internal dT will give an apparently larger R_int.

You might think you can elimate this effect by just using a smaller current when measuring R_int. But that won't work. As soon as you turn on the current, the temperatures start changing. And even if the dT is small, so is the voltage (becaue the current was small). To get the right value for R_int, what you need to do is measure V before dT has developed. That can be tricky if you aren't set up for it.

- Cronin

[quote="Anonymous"]Dear Cronin,

thanks for these "magic formula".

Could you tell me where you have found them, or give me a reference for a specific article?

Is there a big approssimation in them or not?
I've tried to put some data from a manufacturer, but the calculated internal resistance is lower than the actual resistance measured at 300K. (0.52 vs 0.96 ohm)

Thanks

Moreno[/quote]

RE: &quot;Aaargggggghhhhh&quot;

Hello,

I have some results based on the suggestion from Cronin, but I'm afraid they are disappointing.

Firstly, as Croning suggested, I watched the open circuit voltage across the TEG as the temperature at the hot side increased, the open circuit voltage did increase.

I then applied heat on the hotside, as shown in the diagram of the original postt, to generate a temperature difference across my TEG of 38.4 degrees Celcius. The recorded open circuit voltage of the TEG was only 0.069V; a power of ten out from Cronin's calculation ( Eq 2 of Magic Formulas thread; V_open = Dt/300 * Vmax)!!!!

The specification of the TEG states that the module resistance is 0.85Ohms with a hot side temp of 25 deg Celcius and 0.95Ohms at a hotside temperature of 50 deg Celcius, since my hotside temp is approximately 67.1deg Celcius I used the basic equation y = ax + c to get the module resistance at 67.1 deg Celcius, the resistance I got was 1.0184Ohms.

THEORETICAL RESULTS: V_open - 38.4/300*2.19 = 0.28V
I_short = V_open / R_int = 0.28/1.0184
= 0.27Amps

PRACTICAL RESULTS: V_open @ deltaTemp =38.4deg celcius: 0.069V
this would mean that the short circuit current would:-
Short Circuit current = V_open/R_int
= 0.069/1.0184
= 0.067A
Can anyone explain why the practical open circuit voltage is of the order 10 out???? :?

Re:Aaargggggggghhhhh Can't get any current from ny TEG

As you are measuring low voltages (and not only electrical power), it could be that you haven't 38.4°C on the ceramics of the module. Take in account the spreading of the heat on the heatsink. Pheraps the copper pipe under the module is hotter than the thermocouple point on it.
Have you put the thermal interface material well to reduce the parassite DT between heatsink and module's ceramics?

Hi.

Moreno
moreno.buffolo@peltech.com

Re:Aaargggggggghhhhh Can't get any current from ny TEG

When you say thermal interface material, are you talking about the aluminium foil? If you are, I decided to remove it as I couldn't get the temperature of the air on the hot side to reach above 50 deg celcius ( don't know why as I managed it before) so I removed the foil and painted the area around the TEM black, managed to get the temperature above 70deg celcius: one point my digital thermometer could not display any more readings as the temperature exceeded it's max.

As for querying whether I do indeed have 38.4 deg celcius (how did you get the degree symbol) then I thought if the copper pipe was in contact with the cold side of the TEM, then leaving it long enough the copper pipe would become in thermal equilibrium with the ceramic plate. I am assuming the copper is quite faily sheilded from the heat source as when the ambient temperature of the hot side was greater than 60.0 deg Celcius, the copper pipe was still reading at about 25.0 deg celcius, it was 22.8 deg Celcius before the heat source was applied!

Do you think I should paint the TEM black, as this will increase the heat flow to the hot side, as I have often thought that the plate being white is not an ideal colour as I want to use the TEM as a generator; using the solar energy as the heat source ( eventually ) . So if I painted the hot side black then it would absorb heat better, wouldn't it?

Anyway, you're probably right; even though I have an ambient temperature on the hot side of say 60 deg celcius, it doesn't mean I have a hot side temperature of 60 degrees celcius as what is important in TEM is the heat flow into the hot side, isn't it?

What I will do is get a good heat sink and redo the test.

Watch this space,

TGen.

Re:Aaargggggggghhhhh Can't get any current from ny TEG

My doubt was that the temperature of TEM cold side could be high, so reducing the DT.
But it could be the hot side too low.
The black painting of the modul's hot side or (I think better) of the aluminum foils is a good idea. For thermal interface I mean thermal grease or similar to applicate on the sides of the modules.

Do you know N, the thermocouple number of the module?
If yes the open voltage Vopen must be:

Vopen= 2*N*alfa*DT where alfa is near 200*10^-6 [V/K] for your temperatures (aproximated) and DT is your DT.

from your data you have a Melcor CP 0.8-17-06 ( see http://www.melcor.com/CP_Series/CP0.8-17-06L.PDF )

so N=17.. and Vopen = 2*17*200*10^-6 *38.4 = 0.261V (like the first magic of Croning :) )

I think that in Melcor website they have a software to simulate their models, even in generation.