Germanium transistor leakage rises with temperature, which can move a pedal’s DC bias even when every resistor and control stays unchanged. Old stock does not escape the physics simply because it spent decades in a drawer.
NOS means new old stock, not electrically identical stock. Two transistors carrying the same part number can differ enough in gain, leakage, and temperature behavior to make one land neatly in a circuit while another pushes the same circuit toward weak output, gating, excess noise, or an awkward bias point.
A quick hFE reading does not always tell the whole story. If a tester does not separate leakage current from transistor gain properly, some of the current it reports as amplification can actually be current flowing when the base is not being driven in the normal way.
Test conditions matter too. Transistor gain changes with collector current, so a number measured at one operating point may not describe what the same device does inside a pedal running at a much smaller current. Good sorting therefore looks at the transistor under conditions reasonably close to the circuit that will use it rather than treating one display number as a final grade.
Leakage deserves its own measurement. A device with attractive gain but excessive leakage may pull a sensitive bias network away from its intended operating point, while a lower-leakage part with less spectacular gain can behave far more predictably once installed.
Simple vintage fuzz circuits make transistor selection especially obvious because their bias can depend heavily on the actual devices fitted. More stabilized designs can use emitter resistance, negative feedback, adjustable collector resistance, or other bias arrangements to absorb a larger spread in transistor gain and leakage.
This is why a leaky germanium transistor is not automatically useless. The same part can be troublesome in one circuit and perfectly workable in another whose bias network gives the leakage somewhere less destructive to go or provides enough adjustment to put the operating point back where the designer wants it.
The limited run behind the selected NOS 2N1175 transistor batch makes more sense through this lens. Finding a box of old parts is only the first step when a production pedal still needs repeatable gain, sensible noise, acceptable leakage, and a bias point that behaves like the circuit was designed around it.
Selection also affects manufacturing yield. A builder may buy far more transistors than the number of pedals planned because some devices fall outside the useful window for that particular design, even though those rejected parts may still function normally in a different application.
Touching a transistor during measurement can even warm it enough to make a sensitive leakage reading drift. Careful sorting gives the device time to settle and keeps measurement conditions consistent; otherwise, two readings taken minutes apart can appear to describe two different parts.
Circuit designers have several ways to handle the problem. A manual bias control lets the player or builder correct the operating point, while a more stabilized bias network can reduce how strongly transistor variation reaches the rest of the circuit.
More elaborate designs regulate temperature itself. A temperature-regulated germanium bias system can monitor the transistor’s DC condition and control a heating element so the device operates around a deliberately chosen thermal point instead of following every change in the room.
Thermal control cannot make every germanium transistor identical because gain spread, leakage, noise, and other device parameters still differ from sample to sample. It can remove one moving variable, which makes the remaining selection and bias work easier to reproduce from pedal to pedal.
Screening therefore becomes part of the circuit design rather than a ceremonial hunt for rare cans. Builders who measure leakage, gain, bias behavior, and temperature drift are not polishing vintage mythology. They are reducing the number of unknowns before a decades-old semiconductor is asked to behave consistently in a modern product.
NOS means new old stock, not electrically identical stock. Two transistors carrying the same part number can differ enough in gain, leakage, and temperature behavior to make one land neatly in a circuit while another pushes the same circuit toward weak output, gating, excess noise, or an awkward bias point.
NOS describes inventory, not electrical consistency
Vintage transistor data sheets allowed ranges because manufacturers never expected every device leaving a production line to behave exactly alike. Germanium adds another nuisance for pedal builders because collector leakage can be large enough to matter in simple high-impedance bias networks, especially in circuits that were designed around a small handful of parts.A quick hFE reading does not always tell the whole story. If a tester does not separate leakage current from transistor gain properly, some of the current it reports as amplification can actually be current flowing when the base is not being driven in the normal way.
Test conditions matter too. Transistor gain changes with collector current, so a number measured at one operating point may not describe what the same device does inside a pedal running at a much smaller current. Good sorting therefore looks at the transistor under conditions reasonably close to the circuit that will use it rather than treating one display number as a final grade.
Leakage deserves its own measurement. A device with attractive gain but excessive leakage may pull a sensitive bias network away from its intended operating point, while a lower-leakage part with less spectacular gain can behave far more predictably once installed.
Bias turns a transistor into part of a circuit
Bias sets the transistor’s idle operating condition before the guitar signal arrives. Move that point, and you change how much room the waveform has to swing, where clipping begins, how symmetrical it becomes, and whether the stage can pass a clean signal at all.Simple vintage fuzz circuits make transistor selection especially obvious because their bias can depend heavily on the actual devices fitted. More stabilized designs can use emitter resistance, negative feedback, adjustable collector resistance, or other bias arrangements to absorb a larger spread in transistor gain and leakage.
This is why a leaky germanium transistor is not automatically useless. The same part can be troublesome in one circuit and perfectly workable in another whose bias network gives the leakage somewhere less destructive to go or provides enough adjustment to put the operating point back where the designer wants it.
The limited run behind the selected NOS 2N1175 transistor batch makes more sense through this lens. Finding a box of old parts is only the first step when a production pedal still needs repeatable gain, sensible noise, acceptable leakage, and a bias point that behaves like the circuit was designed around it.
Selection also affects manufacturing yield. A builder may buy far more transistors than the number of pedals planned because some devices fall outside the useful window for that particular design, even though those rejected parts may still function normally in a different application.
Temperature exposes weak screening quickly
Germanium leakage is strongly temperature dependent, so a device measured cold on a bench can behave differently after warming inside an enclosure or sitting under hot stage lighting. The electrical shift can move the bias far enough to alter output level, clipping shape, sustain, and the point where a circuit begins sounding gated.Touching a transistor during measurement can even warm it enough to make a sensitive leakage reading drift. Careful sorting gives the device time to settle and keeps measurement conditions consistent; otherwise, two readings taken minutes apart can appear to describe two different parts.
Circuit designers have several ways to handle the problem. A manual bias control lets the player or builder correct the operating point, while a more stabilized bias network can reduce how strongly transistor variation reaches the rest of the circuit.
More elaborate designs regulate temperature itself. A temperature-regulated germanium bias system can monitor the transistor’s DC condition and control a heating element so the device operates around a deliberately chosen thermal point instead of following every change in the room.
Thermal control cannot make every germanium transistor identical because gain spread, leakage, noise, and other device parameters still differ from sample to sample. It can remove one moving variable, which makes the remaining selection and bias work easier to reproduce from pedal to pedal.
Screening therefore becomes part of the circuit design rather than a ceremonial hunt for rare cans. Builders who measure leakage, gain, bias behavior, and temperature drift are not polishing vintage mythology. They are reducing the number of unknowns before a decades-old semiconductor is asked to behave consistently in a modern product.
