Shunt-Widerstand (Rsh) der Photodiode im Null-Bias-Modus

The Naked Truth About Designing for Zero Bias

Look, designing precision measurement equipment is hard enough without your detector’s noise ruining the party. If you are building a high-end pulse oximeter, an X-ray CT scanner, or an in-vitro diagnostic reader, you probably already know that zero bias is teh way to go.

But here is the thing that nobody really explains in standard electronics classes: the actual hero (or absolute villain) of your circuit is the photodiode shunt resistance.

I see designers all the time slapping a massive reverse bias on their diodes because they think they need “speed.” Spoiler alert: for low-light, low-frequency medical applications, you probably dont. Applying a reverse bias just injects nasty shot noise and dark current into a system that cant afford it. You want to operate in photovoltaic mode. And in photovoltaic mode, Nebenschlusswiderstand dictates almost everything about your noise floor.

Most generic detector manufacturers are selling you garbage when it comes to true zero-bias performance. They quote you a “typical” spec on page four of a datasheet, but when you actually plug it into your transimpedance amplifier, your signal-to-noise ratio goes straight out the window.

Let’s break down what Nebenschlusswiderstand actually is, why it makes or breaks a precision medical circuit, and how you can stop your op-amp from amplifying pure noise.

What is Photodiode Shunt Resistance (Rsh), Really?

If you look at the equivalent circuit of a photodiode, it’s not just a perfect little current source. You’ve got a current source in parallel with a junction capacitance (Cj) and a resistor. That parallel resistor is your Nebenschlusswiderstand (usually written as Rsh).

In plain English? It’s the resistance of the zero-biased photodiode junction. If you look at the current-voltage (I-V) curve of a photodiode, the Nebenschlusswiderstand is literally the slope of the curve right at the origin, where V = 0.

In a perfectly ideal, physically impossible universe, a photodiode would have an infinite Nebenschlusswiderstand. No leakage, no problems. In the real world, actual values range from tens of Ohms (for some really terrible or specific infrared materials) to thousands of Mega-ohms (Giga-ohms) for high-quality silicon.

Why do medical equipment designers care so much about this single parameter? Because when you are operating at 0V (photovoltaic mode), you don’t have dark current from an applied bias. The dominant source of noise in your detector is the thermal noise generated by that exact Nebenschlusswiderstand . If your Rsh is low, your noise is high. It is what it is. Thermodynamics doesn’t care about your project deadlines.

The Big Debate: Photovoltaic Mode vs. Photoconductive Mode

Let’s get controversial for a second. The industry is obsessed with photoconductive mode (applying a reverse bias). Yes, reverse biasing widens the depletion region, drops the junction capacitance, and makes the diode super fast. Great for fiber optics and telecom.

But for precision medical measurements? It’s usually a terrible idea.

When you apply a reverse bias, you create a steady dark current. That dark current generates shot noise. Shot noise is proportional to the DC current flowing through the device. If you are trying to detect a weak fluorescence signal from a blood sample, that shot noise will bury your signal completely.

Instead, we use photovoltaic mode (unbiased, 0V). In this mode, dark current is essentially zero. The photocurrent varies linearly with the incident light, and the temperature stability is way better. But without dark current, what causes the noise?

You guessed it. The photodiode shunt resistance.

In photovoltaic mode, the thermal noise of the Nebenschlusswiderstand becomes the absolute dominant current noise in the system. If you skimp on your detector and buy one with a low Rsh, you are just shooting your SNR in the foot before the signal even reaches the amplifier.

The Mathematics of Noise (Without the Academic Jargon)

I promised to keep this practical, but we need to look at the actual math to understand the pain of a bad Rsh. Don’t worry, I won’t use unreadable formatting.

The thermal noise (often called Johnson noise) generated by the Nebenschlusswiderstand is a current noise. You can calculate it using this standard formula:

Ij = √( 4 * k * T * B / Rsh )

Let’s look at the pieces of this puzzle:

• Ij is the thermal noise current (in Amps RMS).
• k is Boltzmanns constant (1.38 x 10^-23 J/K).
• T is the absolute temperature in Kelvin (room temp is roughly 298 K).
• B is your noise measurement bandwidth in Hz.
• Rsh is the photodiode shunt resistance in Ohms.

Look at where Rsh sits in that formula. It’s in the denominator.

Wenn Ihr Nebenschlusswiderstand goes up, your thermal noise current goes down. It’s an inverse square root relationship. If you want to cut your detector noise in half, you need a photodiode with four times the Nebenschlusswiderstand.

This is why, when you are measuring ultra-low light levels in something like a pulse detector or an analytical spectrometer, you cannot accept a photodiode with an Rsh of 10 MΩ. You need hundreds of Mega-ohms, or preferably Giga-ohms.

The Op-Amp Trap: How Shunt Resistance Ruins Amplifiers

Here is the part that bites almost every junior engineer I’ve ever worked with. The photodiode doesn’t exist in a vacuum. You connect it to a transimpedance amplifier (TIA) to turn that tiny photocurrent into a usable voltage.

A standard TIA has an operational amplifier and a feedback resistor (Rf). If you are measuring tiny signals, your Rf is going to be massive—maybe 100 MΩ or even 1 GΩ, just to get enough gain.

But op-amps aren’t perfect. They have their own internal input voltage noise. And how much of that op-amp voltage noise shows up at your output? It depends on the noise gain of the circuit.

For low frequencies, the noise gain (NG) of a TIA is approximated by:

Noise Gain = 1 + ( Rf / Rsh )

Do you see the trap?

If you use a cheap photodiode with a Nebenschlusswiderstand of 5 MΩ, and you need a feedback resistor of 500 MΩ for your signal, your noise gain is:
1 + (500 / 5) = 101.

Your circuit is taking the innate voltage noise of your expensive, “low-noise” op-amp and multiplying it by 101! Your output will look like a fuzzy caterpillar on the oscilloscope.

Now, imagine you sourced a premium detector from a specialized custom photodiode manufacturer like BeePhoton. You drop in a diode with an Rsh of 2 GΩ.
Your new noise gain is:
1 + (500 / 2000) = 1.25.

You just reduced your voltage noise amplification by almost 100x just by picking a detector with a proper photodiode shunt resistance. This is the secret sauce of medical equipment design.

Temperature Dependence: The “Rule of 6”

So you tested your circuit on your lab bench at 22°C and everything looks great. Then your device goes into a hospital environment, sits next to a warm power supply, the internal enclosure temp hits 35°C, and suddenly the machine fails calibration. What happened?

Shunt-Widerstand is highly temperature-dependent. It doesn’t just sit there. As the temperature goes up, the diffusion current increases, and your Nebenschlusswiderstand drops like a rock.

The industry rule of thumb is this: Shunt-Widerstand approximately halves for every 6°C increase in temperature. (Conversely, it doubles for every 6°C decrease).

If you start with an Rsh of 500 MΩ at 20°C:

• At 26°C, it drops to 250 MΩ.
• At 32°C, it drops to 125 MΩ.
• At 38°C, it’s down to 62.5 MΩ.

Your thermal noise is creeping up, and your TIA noise gain is exploding because the denominator (Rsh) is shrinking. This is exactly why medical devices need heavy thermal management, or better yet, detectors that start with such an astronomically high Nebenschlusswiderstand that even after a thermal derating, they still perform flawlessly.

Material Matters: Silicon vs. InGaAs vs. Germanium

Not all semiconductors are created equal. The material you choose fundamentally limits the Nebenschlusswiderstand you can achieve.

If you are working in the visible to near-infrared spectrum (like 400nm to 1100nm), Silicon (Si) is king. The bandgap of silicon allows for exceptionally high Rsh values, which is why Si PIN photodiodes are the default choice for medical imaging and life sciences.

Once you push into the longer infrared wavelengths, you have to use materials with smaller bandgaps, like InGaAs or Germanium. A smaller bandgap means more thermally generated carriers at room temp, which drastically lowers the Nebenschlusswiderstand.

Here is a quick reference table I put together based on typical room-temperature values. (Keep in mind, active area size also changes these numbers—larger active areas have lower Rsh).

Photodiode Material	Typical Shunt Resistance (Rsh)	Primary Application	Noise Profile in PV Mode
Silicon (Si PIN)	100 MΩ to > 5 GΩ	Medical CT, Pulse Oximetry, Spectroscopy	Excellent / Extremely Low
InGaAs	1 MΩ to 50 MΩ	Telecom, NIR Spectroscopy	Mäßig
Germanium (Ge)	1 kΩ to 100 kΩ	Older IR detection, Power meters	Poor / High Noise

As you can see, if you are designing life-saving equipment, you want Silicon whenever the wavelength allows it.

Real-World Case Study: Rescuing an IVD Fluorescence Reader

I want to share an anonymized story from a client we worked with recently. They were designing an In-Vitro Diagnostic (IVD) fluorescence reader. The concept was brilliant, but their prototype was failing miserably. The fluorescence signal from their assays was so incredibly weak that it was getting entirely lost in the baseline noise of their detector board.

Their lead engineer called me, practically freaking out. “The op-amp is noisy,” he said. “We need a better op-amp.”

I asked to see their schematic. They were using a top-tier electrometer-grade op-amp. The op-amp wasn’t the problem.

Then I looked at the datasheet for the generic, off-the-shelf photodiode they bought from a massive catalog distributor. The guaranteed minimum photodiode shunt resistance was listed as just 10 MΩ.

Because they needed a massive gain to see the fluorescence, their feedback resistor was 2 GΩ.
Their noise gain was 1 + (2000 / 10) = 201.
The op-amp’s baseline noise was being amplified 201 times! Plus, the thermal noise current from a 10 MΩ resistor at room temp was drowning out the actual photons hitting the silicon.

I told them to rip out the generic diode. We replaced it with one of our specialized detectors by having them explore our Si-PIN-Fotodiodes. The specific model we gave them was custom-passivated to guarantee a Nebenschlusswiderstand of over 3 Giga-ohms.

The math flipped instantly.
New noise gain: 1 + (2000 / 3000) = 1.66.

The baseline noise dropped by orders of magnitude. Suddenly, they could read fluorescence concentrations that were 50 times lower than their previous limit of detection. They passed their internal validation the next week.

This is what happens when you respect the physics of Nebenschlusswiderstand.

How We Actually Measure Shunt Resistance in the Lab

You might be wondering, “How do I know if the datasheet is lying to me?”

Honestly, the only way to be sure is to measure it yourself. But you cant just stick a standard multimeter across the diode pins. The voltage output from a standard DMM will forward bias or reverse bias the diode too much, completely altering the resistance you are trying to measure.

Shunt-Widerstand is defined strictly at 0V. To measure it experimentally, you have to apply a tiny, tiny voltage—typically exactly +10 mV, and then -10 mV.

You apply +10 mV, use an ultra-precise picoammeter (like a Keithley electrometer) to measure the current, and calculate R = V / I. Then you do the same at -10 mV to ensure symmetry and average it out.

Because the currents are in the picoamp range, any environmental noise, vibration (microphonics), or stray light will ruin the measurement. The diode has to be in absolute darkness, thermally stabilized, inside a Faraday cage.

If you don’t have the gear to do this, you have to rely on a vendor who actually tests their diodes instead of just printing theoretical numbers.

A Word on Active Area

There is a trade-off I have to mention. You can’t just demand a 10 GΩ Nebenschlusswiderstand on a detector the size of a dinner plate.

Shunt-Widerstand is inversely proportional to the active area of the photodiode. If you double the area of the silicon chip, you are doubling the volume where thermal carrier generation can occur. Your Nebenschlusswiderstand drops in half.

This means mechanical design and optical design go hand-in-hand with electronic noise. If you can use a lens to focus your light onto a smaller photodiode, you can use a diode with a smaller active area. A smaller active area means a higher Nebenschlusswiderstand, lower junction capacitance, and a massively improved signal-to-noise ratio.

Stop using 10x10mm detectors when a 3x3mm detector with a cheap plastic lens will give you 10 times less noise!

Why Quality Manufacturing Matters

Creating a photodiode with an exceptionally high Nebenschlusswiderstand isn’t just about buying good silicon wafers. It’s entirely about the fab process.

The edges of the P-N junction are where leakage current destroys the Rsh. If the passivation layer (the oxide coating that protects the raw silicon) is poorly grown, or if there are impurities introduced during the diffusion process, you get surface leakage. Surface leakage acts like a parasitic resistor in parallel with your diode, tanking your total Nebenschlusswiderstand.

When you source from a dedicated custom photodiode manufacturer like BeePhoton, you are getting silicon that has been processed with incredibly tight control over junction passivation and gettering. That’s how we push our photodiode shunt resistance into the Giga-ohm territory where generic brands just can’t compete.

---

Häufig gestellte Fragen (FAQ)

---

Let’s Fix Your Noise Problems Together

Navigating the physics of low-noise analog design is tough, and you shouldn’t have to guess whether your detector is going to perform or fail in the field. If you are struggling with poor SNR, excessive thermal noise, or op-amp voltage noise amplification, the problem might just be the chunk of silicon you are using.

Stop letting bad Nebenschlusswiderstand ruin your precision medical designs. We specialize in engineering detectors that push the absolute limits of low-noise, zero-bias performance.

Ready to upgrade your circuit? Head over to our contact page to get in touch with our engineering team, or just drop a message directly to info@photo-detector.com. Tell us about your noise floor, and we’ll show you exactly how a proper custom detector can fix it.