Hi,
again I am trying to find a way to use my TEC to control temperature of an enclosure. It is 2 Ohm, rated for up to 6A (12V). I "know" TEC should be powered by smooth voltage, not PWM. So I was planning to use a power source powerful enough (maybe old laptop charger with 19V/60W) and use an inductor (and capacitor) to smooth driving PWM to something reasonably stable. Quickly I have found inductors rated for such huge currents are "expensive" and large. Also the inductance is pretty low. I have chosen a this coil but I don't know it is usable. It is rated for 8A and inductance is 0.6mH. But it is "common mode choke". If I understand it properly common mode chokes are used to filter only common mode noise - there are 2 coils on a single core (so it is a transformer?). Differential mode current (going in opposite direction through one and the other coil) should pass unhindered while common mode noise (current trying to flow the same direction in both coils) should be blocked. But in this scenario the differential current causes magnetic flux to cancel in the common core and there is no risk of saturation of the core. But what if I want to use it as a single coil for the buck converter? If the 8A rating truly means the core will saturate only with more than 8A what if I use it as a single coil (not connecting it to power and GND lines as intended but one after another) - will it make a 1.2mH coil with maximum 4A current?
Hi,
The reason the common mode choke is so much cheaper than a "normal" choke is the core does not support the full 6A.
Reason:
A common mode choke has two coils. They are wired such that one cancels the other.
when properly connected, one coil caries +6 Amps (or whatever current you have) and the other coil caries -6 Amps. So the core does not have to deal with the 6 amps.
This type of inductor is meant to add impedance to the current that is on both the +6 and -6 wires (aka the noise)
I was afraid of this. So this coil is unusable for buck converter, right? I just don't understand the 8A limit. The coil is 15mOhm, with 8A current it is dissipating "only" 8A8A15 mOhm ~ 1W. Is it so much power for the huge beast when small resistor is rated for 1/4W? Also another "proper" coil has 40mOhm, yet it is rated for 10A. So I hoped ...
BTW a laptop processor has TPD 45W. AFAIK processors use very low supply voltage - 1-2V IIRC. To dissipate 45W they need over 20A. How they get such large current? If I were designing a computer I would probably need HUGE coil. I don't remeber seeing one when I disassembled a dead laptop. Maybe there is some better way to get a lot of current?
I "know" TEC should be powered by smooth voltage, not PWM.
This is a fairly common misconception.
There are a few circumstances in which rapid, large temperature changes across the module can cause cumulative damage leading to element failure in the long term, but those are easily avoided.
Unless your particular situation is one of those, your efforts are probably misguided.
1/ What's a TEC?
2/ A laptop charger will provide a pretty smooth voltage - perhaps 50mV of ripple.
Allan
TEC should mean (Peltier) ThermoElectric Cooler.
I am not interested in ripple voltage of laptop source. I think the Peltier does not need extra smooth voltage. But I wonder how laptop makes 20A@2V for the processor from 3A@19V source. Maybe there is some other approach than naive pulsing current via a single inductor?
Yes, you do want to run peltier devices with a fairly smooth voltage/current for good efficiency:
You want a buck converter with output voltage adjust input. The inductors and capacitor in the buck
converter do the work for you.
BTW a laptop processor has TPD 45W. AFAIK processors use very low supply voltage - 1-2V IIRC. To dissipate 45W they need over 20A. How they get such large current?
The higher the switching frequency, the smaller the coils and capacitors need to be - they only store energy
for half a switching cycle and size relates to energy storage, not the power. New ultra-fast switching GaN
(gallium nitride) power converters run at MHz rates and require even smaller inductors than the present
generation. GaN is potentially much better than silicon for power electronics: Johnson's figure of merit - Wikipedia
Currently GaN MOSFETs are niche, and require specialist driver chips and super-low inductance packaging,
I don't think they are easy components for hobby use alas.
GaN devices are widely used in RF power amplifiers at up to several GHz.
Allan