Making use of micro-current source LM334
Since a magnitude of the current prior Xiangnong 1uA, precision> 10 bit current source, but searching through ADI, Ti, as well as ST, the result is only minimal 1mA current source chip - is the LM334 series (also 134 234, the difference It is the work of different temperatures). But I read the manual chip and found it to be the current resistance of the output circuit according to access, and then I thought, I change its resistance, you can not change the output current up? But my concern is, 1uA 1mA relatively large gap between the actual results may not be so ideal.
but! ! ! !
In a review of my blog's found a 'senior' is worked out, according to his description, the results seem pretty good:
the original link
through my calculations, he's the best precision can reach 16.6bit, even able to there are also long-term guarantee of 10.9bit, precision is enough.
Here are some specific programs:
First, we need to find their datasheet Quguan net, I saw two of Ti and ADI, I found two programs presented are not the same, but the formula is the same.
[Manual]
First introduce the ADI:
its encapsulation information, we can clearly see that this thing is in fact three pin, an input voltage V +, output a voltage V-, to a need to set the resistance Rset.
The basic principle of the equivalent circuit shown in Figure 1 LM134. With respect to the negative input terminal V- pin, A1 is applied a reference voltage of 64mV. A1 drive to the Q2, R to the pin is held in 64mV, regardless of the value of RSET. The ratio of the transistors Q1 and Q2 is 17: 1, so that the current flowing from the V- pin is always the total current flowing into pin V + 1/18. The total current referred ISET, is equal to:
At this point we do the math, we need a Rset = 67mV / 1uA = 67k resistor on OK. But things really that simple? Although I do not do this, but apparently directly on intuition tells me it is certainly not. As to why not, we can look at the operation of the older generation who, on the air-conditioned room is warm, but also cool, the purpose is to measure a temperature coefficient. This temperature coefficient that is come from? We read the manual carefully, you will find that the output current of the chip (the ISET) temperature coefficient of about 0.33% / ° C, so that the current variation due to temperature rise is caused by (0.4) (0.33) = 0.132%. In doing so precision current source can not ignore this problem. Fortunately, ADI later in the manual gives us a typical zero-drift circuit:
Next Ti see us, I think this formula described in more detail:
