1
\$\begingroup\$

I am using this BC807-40 transistor.

My maximum collector current through the transistor is 200mA.

I want to calculate the delay time between the transistor input Vbe and output Ic.

Is there a way or formula to calculate the delay time theoretically? I couldn't find any delay specification mentioned in the datasheet.

Whereas, for a MOSFET which I have used in a different circuit, I was able to find the delay time:

pn

It is mentioned in the datasheet as turn on delay time and turn off delay time.

But this delay parameter is not available for all transistors.

Why don't the manufacturers of transistors mention this delay in their datasheet? Is it because transistors switch slower when compared to MOSFETs? Or which switches much faster between input and output?

Please explain.

\$\endgroup\$
3
  • 1
    \$\begingroup\$ BJTs can switch quite fast, compared to FETs. In fact, fast switching circuits are based on BJTs. See ECL. \$\endgroup\$
    – jonk
    May 21, 2020 at 8:54
  • \$\begingroup\$ Could you please explain a little more on how BJTs are much faster than MOSFETs in switching applications, please \$\endgroup\$
    – user220456
    May 23, 2020 at 5:35
  • \$\begingroup\$ Just look on the web about ECL as compared to FETs. It's "out there" to find. Or ask a different question. I was just letting you know your assumptions may need some added research, is all. \$\endgroup\$
    – jonk
    May 23, 2020 at 7:17

3 Answers 3

2
\$\begingroup\$

That's a general-purpose transistor and specified for slow switching and slow (AF= audio frequency) linear operation. You can find general-purpose transistors specified for faster switching.

enter image description here

The speed of a BJT can also be influenced a lot by the circuit it's in, in an emitter follower configuration usually it doesn't saturate and will switch a lot faster. You can add a Schottky diode clamp to prevent saturation at the expense of a couple hundred mV more "on" voltage. You can use a speed-up capacitor across the base resistor.

Note that the datasheet MOSFET speed is dependent on driving the gate with about 1.5A peak. If you drive it with a high value series resistor or a wimpy drive circuit it will be much slower due to the Miller and input capacitances.

\$\endgroup\$
9
  • \$\begingroup\$ Thank you for the answer. 1. How are you identifying that my selected transistor is specified for slow switching? \$\endgroup\$
    – user220456
    May 21, 2020 at 9:37
  • 2
    \$\begingroup\$ There is no specification for switching speed. ;-) \$\endgroup\$ May 21, 2020 at 9:37
  • \$\begingroup\$ Thank you. So, in your terms, what is your range of time when you say - slow switching? 500ms - 600ms? And fast switching transistor datasheet, which you shared, the "delay time" parameter is the equivalent parameter of the turn on time of the device? Like, if I give 0.7V between the base & emitter of the transistor & base current of 15mA, I will get the collector current 150mA flowing after 15ns max, right? \$\endgroup\$
    – user220456
    May 21, 2020 at 9:43
  • \$\begingroup\$ Where a microsecond or even two doesn't matter, so LEDs, relays, perhaps switching at a few kHz to a beeper. In the case you mention 35ns (delay + rise). Turn-off is much slower. \$\endgroup\$ May 21, 2020 at 9:47
  • \$\begingroup\$ Thank you for the answer. Could you also answer my 3 questions in my previous comment? I am not able to understand for which one you have replied and provided answer. \$\endgroup\$
    – user220456
    May 21, 2020 at 9:50
3
\$\begingroup\$

Stored charges, in a bipolar forward-biased collector-base junction because of saturation, may be un-characterized in transistors intended for fast non-saturated uses.

\$\endgroup\$
1
\$\begingroup\$

Why don't the manufacturers of transistors don't mention this delay in their datasheet?

Because it's not a relevant measure in many, if not most, use cases of transistors.

The MOSFET you've picked is specifically meant for switching, and the "Conditions" column nearly fully describes the measurement setup for measuring the delay, which, however, is also a realistic circuitry for actual operation of the MOSFET.

So, that number is not a characteristic of the MOSFET, but a characteristic of a specific circuit (including the MOSFET). I hope you see how this makes little sense for a transistor that might not be as optimized for a single use case as this MOSFET!

The BC807 you refer to (which used to share a datasheet with the BC327, back when it was built – more than 20 years ago!) is a self-claimed "general-purpose PNP". If you want a transistor for switching, it might or might not be your choice – I'd go with "not", since it's a rather small PNP. So, the switch timing is largely irrelevant to the transistor; if you want to know about that, you'll first need to define a test circuit (like the one described in the "Conditions" column of the BSP317B) and then realize that a BJT is not a FET and turn-on isn't limited by a gate capacitance (since there is no gate); you'd then try to describe "turned on" as voltage thresholds, and come to the conclusion that, in fact, the datasheet won't tell you how fast your semiconductor junction is able to react – but it does give you a 3dB sensitivity point at 80 MHz, which is probably more valuable if you're actually trying to use a PNP BJT.

In fact, not really sure how much info even is in these turn-on and -off times: most of this looks like you could calculate it yourself from the actual transistor parameters – after all, you need to charge or discharge the gate capacitance, and that has an exponential charge curve, like any capacitor, together with the couple "cross" capacitances from a standard MOSFET model given above in the same table. So, with the gate capacitance that will definitely be central part of most MOSFET datasheets and these "parasitics", and the gate impedance, most of these numbers should be simply inferable, and that line would be totally superfluous, technically, since redundant.

It's still useful to the engineer, of course.

\$\endgroup\$
6
  • \$\begingroup\$ Thank you for the answer. But, are you saying that MOSFET has limited applications other than switching which is why the manufacturers mention the delay times, whereas the BJT or the general purpose transistor can be used as an amplifier as well as a switch, so, the manufacturer will not be knowing in what application I am going to use my transistor as? 1. "The MOSFET you've picked is specifically meant for switching" - Please help to understand how are you identifying this Mosfet is for switching and are they many uses for MOSFET apart from switching? \$\endgroup\$
    – user220456
    May 21, 2020 at 9:36
  • \$\begingroup\$ This specific MOSFET is optimized for switching. Not all MOSFETs. This specific BJT is not optimized for switching. Not all BJTs. You pick devices based on what they are made for, not the other way around! \$\endgroup\$ May 21, 2020 at 10:27
  • \$\begingroup\$ Well, if the datasheet specifies switching times, it's very likely optimized for switching. Why else would it do that? \$\endgroup\$ May 21, 2020 at 10:28
  • \$\begingroup\$ Ok But if the datasheet doesn't specify the switching times, what would be the typical delay in switching or how to find it? \$\endgroup\$
    – user220456
    May 21, 2020 at 10:43
  • 1
    \$\begingroup\$ that really depends on the type of transistor, what you want to switch, how you define on and off, and, most importantly, on the configuration that you're using it in, which is why I stressed so much that the MOSFET datasheet also gave the configuration with which these times were achieved. The MOSFET itself doesn't have a switching time – the specified circuitry including the MOSFET has one. \$\endgroup\$ May 21, 2020 at 10:50

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.