Phototubes enable op amps to achieve true zero output

Infineon / Mitsubishi / Fuji / Semikron / Eupec / IXYS

Phototubes enable op amps to achieve true zero output

Posted Date: 2024-01-30

For example, a typical rail-to-rail output op amp cannot provide true zero voltage, it can only guarantee that its output is at least a few millivolts, while the resolution of a high-resolution DAC can reach tens of microvolts. The application requires true zero output, hence the problem.

Of course, some negative supply source is needed to increase the "headroom" to near zero. (I use the word "margin" because we are not dealing with positive supply from above, but with supply from below. A better word would be "margin.")

The Cuk configuration circuit is intentionally used again, just like the old circuit in the EDN, but the output voltage is only about -1 V and the output current is lower (less than 2 mA).

While exploring alternatives, the idea of ​​using photocells instead of any ordinary voltage converter arose. The result is the circuit shown in Figure 1.

Figure 1 uses photodiodes instead of voltage converters to help provide true zero voltage to the output of the op amp buffer of a high-resolution single-supply DAC.

The solution is comparable in size to circuits based on the Cuk configuration, but is less efficient. But since the excess power is no more than 0.1 W, this probably doesn't matter.

Such a solution has important advantages:

This is much simpler.

It produces very low electrical noise - which is significant when you're dealing with low analog signals. (In this circuit, even without output capacitor C1, the output noise is less than 1 mV.)

Any overvoltage on its output (which a Cuk converter might produce if there were any problems with feedback) is eliminated.

One should also note the perfect isolation level, although this is not important in our case.

Since the external contour of the device is determined by the photocell, a microphotocell AM-1417 (Toshiba) was used. It measures only 34 x 14 x 2 mm, has 4 sections (so 4 LEDs each), and produces about 3 V without any load.

The 4 LEDs are very common bright red series LEDs (L-513HURC, 1800 mcd at 15° angle) since silicon photocells are efficient in this spectral region.

Red is also better suited for +5 V supplies because their low forward voltage makes it very simple to double the efficiency by stacking them in pairs and delivering the same current through both.

This circuit produces 490…520 mV with a 2k load and 20 mA flowing through the LED. This is more than enough for many micropower operational amplifiers such as the AD8603/AD8607.

The output voltage of the photocell can be changed by changing the current flowing through the LED.

The photocell is a current source rather than a voltage source, so capacitor C1 is needed to reduce the output impedance of the circuit. If the negative voltage disappears for some reason, diode D1 enables a path to sink the current and protect the electrolytic capacitor.

As I mentioned, the output power is more than enough for a precision micropower op amp such as the AD8603. If you need more power, you can use a higher LED current, a more efficient LED/photocell pair, or simply connect more of these circuits in parallel.

#Phototubes #enable #amps #achieve #true #output