Low-Leakage Capacitors #1

New Issue

molly · 2025-12-29T19:04:42-05:00

molly commented

2025-12-29 19:04:42 -05:00

In Ray Wilson's original design, he uses a 0u01F Polystyrene (or Polycarbonate) capacitor in the sample & hold part of the circuit. Polystyrene capacitors are now impossible to find, and polycarbonate are pretty difficult to find.

This issue is opened so that I can remember to someday find a low-leakage capacitor to use in-place of the originals. I've settled on the KP1830310061 from Vishay Beyschlag/Draloric/BC Components for the first batch of boards assembled, however I am not confident at all that these will work particularly well, especially since I have done all of 5 minutes of research to say they'll work in the first place.

Purpose of Low-Leakage Capacitors in this circuit

The reason Ray used low-leakage capacitors was to avoid voltage drift (or sag) while holding a note down. Without a sample & hold circuit to grab the CV voltage when a note is pressed and store it, parts of the circuit would begin to heat up and thus drift dramatically (you might start with middle-C but after a couple of seconds end up with a very very different note!) When a note is pressed the CV voltage is sampled into the low-leakage cap, then a JFET opens disconnecting the keyboard voltage from the CV output circuit. The CV output circuit will continue to produce whatever voltage is stored in the capacitor until either the note is released (closing the JFET and preparing the circuit to sample another voltage) or the capacitor is discharged.

This is the important part...

The capacitor storing the CV voltage will slowly discharge through parasitic leakages. This is just a fact of analog electronics and is unavoidable. The longer the note is held, the more this capacitor discharges and thus the more the output CV drifts too. Too great a drift and you end up with the original problem, the note you wanted to play starts okay but after a few seconds ends up at a very different note. So, this sample & hold capacitor needs to be as low-leakage as possible for the stability (and usability) of the synthesizer.

The Plan Forward

I'll use this issue to track what capacitors I try and how well they work. When I've decided on a capacitor that works well enough I'll formalize it by changing the footprint on the PCB and updating the list of materials. I'll also post little thoughts or any extra updates as I think of them too.

In Ray Wilson's original design, he uses a 0u01F Polystyrene (or Polycarbonate) capacitor in the sample & hold part of the circuit. Polystyrene capacitors are now impossible to find, and polycarbonate are pretty difficult to find. This issue is opened so that I can remember to someday find a low-leakage capacitor to use in-place of the originals. I've settled on the KP1830310061 from Vishay Beyschlag/Draloric/BC Components for the first batch of boards assembled, however I am not confident at all that these will work particularly well, especially since I have done all of 5 minutes of research to say they'll work in the first place. ## Purpose of Low-Leakage Capacitors in this circuit The reason Ray used low-leakage capacitors was to avoid voltage drift (or sag) while holding a note down. Without a sample & hold circuit to grab the CV voltage when a note is pressed and store it, parts of the circuit would begin to heat up and thus drift dramatically (you might start with middle-C but after a couple of seconds end up with a very very different note!) When a note is pressed the CV voltage is sampled into the low-leakage cap, then a JFET opens disconnecting the keyboard voltage from the CV output circuit. The CV output circuit will continue to produce whatever voltage is stored in the capacitor until either the note is released (closing the JFET and preparing the circuit to sample another voltage) or the capacitor is discharged. This is the important part... The capacitor storing the CV voltage will slowly discharge through parasitic leakages. This is just a fact of analog electronics and is unavoidable. The longer the note is held, the more this capacitor discharges and thus the more the output CV drifts too. Too great a drift and you end up with the original problem, the note you wanted to play starts okay but after a few seconds ends up at a very different note. So, this sample & hold capacitor needs to be as low-leakage as possible for the stability (and usability) of the synthesizer. ## The Plan Forward I'll use this issue to track what capacitors I try and how well they work. When I've decided on a capacitor that works well enough I'll formalize it by changing the footprint on the PCB and updating the list of materials. I'll also post little thoughts or any extra updates as I think of them too.

molly commented

2025-12-29 19:08:37 -05:00

Ray Wilson's Original Capacitor

Here's the performance of Ray Wilson's original capacitor as documented on his website.

Third C up pressed: Ideal Voltage 2.000V
Keydown 2.0038V
30 seconds later 1.9972V
30 seconds later 1.9897V
30 seconds later 1.9819V
30 seconds later 1.9748V
30 seconds later 1.9672V
30 seconds later 1.9598V

He also gives us the following information:

Sample & Hold Performance - Output Voltage Droop
The control voltage output of any analog based keyboard that uses a sample and hold which stores a voltage on a capacitor will eventually droop as any leakage path to the capacitor will eventually cause it to do so. These numbers represent what I believe is typical performance for this keyboard controller circuit when used with a .01uF polystyrene cap for C19 and the LF444CN for IC U3. It is important to clean ionic soils from the PC board using flux remover especially in the vicinity of these components as well as the 2N5457 sample switch transistors Q2 and Q3.

Future Testing

For future testing, try to use as good of a multimeter as I can (I don't think I have anything that can do 4 decimals of precision, perhaps I'll have to borrow something from work) and also document what components are being used for C19, U3, Q2, and Q3.

# Ray Wilson's Original Capacitor Here's the performance of Ray Wilson's original capacitor as documented on his website. Third C up pressed: Ideal Voltage 2.000V Keydown 2.0038V 30 seconds later 1.9972V 30 seconds later 1.9897V 30 seconds later 1.9819V 30 seconds later 1.9748V 30 seconds later 1.9672V 30 seconds later 1.9598V He also gives us the following information: > Sample & Hold Performance - Output Voltage Droop > The control voltage output of any analog based keyboard that uses a sample and hold which stores a voltage on a capacitor will eventually droop as any leakage path to the capacitor will eventually cause it to do so. These numbers represent what I believe is typical performance for this keyboard controller circuit when used with a .01uF polystyrene cap for C19 and the LF444CN for IC U3. It is important to clean ionic soils from the PC board using flux remover especially in the vicinity of these components as well as the 2N5457 sample switch transistors Q2 and Q3. ## Future Testing For future testing, try to use as good of a multimeter as I can (I don't think I have anything that can do 4 decimals of precision, perhaps I'll have to borrow something from work) and also document what components are being used for C19, U3, Q2, and Q3.

Sign in to join this conversation.

1 Participants

Notifications

Due Date

No due date set.

Dependencies

No dependencies set.

Reference: ZOE/UTIL-002_AdvancedKeys#1