4
\$\begingroup\$

What are the requirements when selecting a diode for a Baker clamp circuit?

  1. How does the reverse leakage current of the diode affect the circuit? Can a diode with relatively high reverse leakage current (e.g. in mA range) turn the transistor on spontaneously?

  2. In case of single Schottky Baker clamp, is there a scenario where lower Vf is not desirable, or do we always want the diode Vf to be as low as possible?

enter image description here

enter image description here

\$\endgroup\$
1
  • 2
    \$\begingroup\$ As you have probably found, you can't select these characteristics completely independently. A diode of a given type (eg. Schottky) with relatively low Vf will typically have relatively high leakage (especially at high temperature). \$\endgroup\$ – Spehro Pefhany Jan 2 '19 at 19:27
2
\$\begingroup\$
  1. Yes, high reverse leakage might lead to the switch always being partially on, even when you try to turn it off. High reverse leakage is kind of like putting a (nonlinear) resistor in its place. Then you can see it's like a collector-base feedback biasing scheme that will could slightly bias the transistor into active mode.

  2. If you are switching at very high speeds, even small Schottky diodes might not work as well as a fast switching diode since you are essentially putting another (small) capacitor parallel to Cbc. Then a 1n4148 might actually work better in some cases, even if Vbc gets more forward biased. You might need to add a speed up capacitor as well in that case.

This question assumes done knowledge of Baker clamps, so for readers that aren't familiar: when a BJT is in saturation (switching), Vbc can become forward biased. When that happens, charge builds up in the base that takes a long time to dissipate (100s of ns). So if you want to switch fast, you can limit that charge build up by making sure Vbc never goes into forward bias (put a diode between the b and c known as a Baker clamp). To work best, Vf of that diode should be as small as possible. Additionally, you can limit the amount of charge stored by keeping Ib as small as possible or by adding a capacitor in parallel to Rb to "pull out" the excess charge when the switch goes off. That is known as a speed up capacitor.

\$\endgroup\$
2
  • \$\begingroup\$ A '4218? I looked that up because I didn't recognise it, and apparently that's a zener diode--are you sure you didn't mean a '4148? And I'd think a schottky would be a good option for high switching speeds, given the lack of reverse recovery time in schottky diodes. At least use a FRED. \$\endgroup\$ – Hearth Mar 21 at 20:31
  • \$\begingroup\$ Ha! Typo on my phone, thanks I'll fix it. I meant 4148. I just made (for fun/HAM) a 12V mosfet driver driven by a 2.7 V square wave at 7 MHz. I used a Baker clamp + 10nF speed up capacitor with a bat54 (Schottky) and compared to a 1n4148. I thought the bat54 would work better, but the 1n4148 made slightly better squares. Both worked well. At less than 1 MHz, the bat54 could work without the speed up capacitor, but the 1N4148 still needed it. So I think the 4148 works better for a narrow use case at high frequencies when also using a speed up capacitor. \$\endgroup\$ – KD9PDP Mar 22 at 3:49

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.