Criteria for selecting switching transistors

[boylesg]

RE: this circuit



What are the selection criteria if I were to replace the transistors in this circuit with others.

E.G. BC327 => BD140, Darlington => 2SD2348 (horizontal deflection power transistor from a tv)

It seems with the above replacements I can get a maximum of 1.1mA through the base of the BD140 with no base resistor and a max of about 40mA through the base of the 2SD2348. That is a long way off the apparent saturation level of 1.23A or so.

The result seems to be that the 2SD2348 generates a lot of heat when I replace the car headlight bulb (5W) with a flyback transformer from a tv. I take it the 2SD2348 is not fully open and hence has a high power output then the low resistance flyback transformer draws a high current.

I do have other assorted tv transistors that I can try. Including another darlington with specs not far off the BU941P used in the above kit. BU941P was not available in multisim so I have selected an available Darlington for the above circuit spec.

- - - Updated - - -

I decided to try a 2SA1015 in place of the BD140 and discovered I can get much more current through the collector - of the order of 0.3A ignoring the fact that the current is pulsed and the real value is no doubt greater than this.

So there is presumeably an issue of transistor impedance here that I need to take into consideration. I.E. The BD140 has a much greater impedance than the 2SA1015, and I guess the BC327, and hence the NE555 is not able to push as much current through its base and hence I can only get about 40mA through the BD140 collector.

The BD140 specs make no mention of resistance but the 2SA1015 specs do (Base intrinsic resistance rbb’ VCB 10 V, VCB=10 V, I =1 mA, f=30 MHz 30R)

So is there some way of determining that a BD140 is not all that compatible with a NE555 from the specs or general knowledge of transistors?

 

[FvM]

There are too much wrong assumptions in your post to correct it in a few words, e.g. the complete "base impedance" point. In a first order, the said low power transistor differ by current gain, rated current and power. Base I/V characteristics are almost similar and don't need to be considered within the accuracy requirements of the present problem.

 

[boylesg]

There are too much wrong assumptions in your post to correct it in a few words, e.g. the complete "base impedance" point. In a first order, the said low power transistor differ by current gain, rated current and power. Base I/V characteristics are almost similar and don't need to be considered within the accuracy requirements of the present problem.

Well can you explain briefly why the base current for BD140 is so much smaller than 2SA1015 in this application - about 3 mA average pulsed compared to roughly 0.3mA average pulsed. What is causing this difference?That would be start for further reading at least.

 

[godfreyl]

Why do you think the base current of the BD140 is so low? Did you measure it somehow?

I suspect there is a mistake. The base current can not be so low unless the transistor is broken.

 

[boylesg]

Why do you think the base current of the BD140 is so low? Did you measure it somehow?

I suspect there is a mistake. The base current can not be so low unless the transistor is broken.

I broke the circuit and inserted the multimeter probes. I undertsand that the reading I am getting is not going to be the real value since the current is pulsed and my multimeter is not sophisticated enough to read peak current and all that. But there never the less seems to be a difference of 10 fold between the two transistors.

 

[FvM]

But there never the less seems to be a difference of 10 fold between the two transistors.

If the transistor isn't broken, you may have mixed up the terminals.

Most engineers will assume a constant Vbe of 0.7 V for the base current calculation. BC327 has about half the rated current of BD140. As expectable, the typical Vbe values for identical Ic/Vce operation points are slightly lower (by a few 10 mV) for BD140.

Using BD140 may be reasonable for Ic > 0.3 A due to the higher current gain in this range, in other words in combination with non-darlington output transistors.

 

[D.A.(Tony)Stewart]

The schematic clock resembles the 555 timer but without the p/n, which some vendors rate the output at 200mA source @12V, but only sink 5mA @5V

1. What is the voltage, Vcc for the 555? 5V? that would explain a part of your 1.1 mA sink current problem,, the other parts are ; design choice of sink vs source, measurement error {inductance and duty cycle} and supply voltage.

2. Choices of current gain for switches are depending on you to read the specs around rated current and voltage when saturated.

3. Not all vendors use the same specs unless JEDEC qualified specs are used or follow an agreed industry standard. There were/are at least 16 different vendors since 1971.

4. Current gain for saturated switches depends on hFE and V-I operating point but drops as current rises as a rule from 300 max for low current to 5 for humungous switches. Darlingtons are a special case where cascaded current gain is offset by increased voltage drop.

5. I suggest you source at 12V and use a "common collector" design instead for pre-driver then "common emitter" to drive the transformer. ( the "name" comes from the pin in common with in and out currents.



6. MOSFETS come in 4 types and offer better performance if adequate drive. p/n type and enhancement/depletion mode types. But for fast pulses the input capacitance takes a bit more current during transitions.

7. Automotive Design Engineers need dozens of switch but must meet to tough requirements like reversed battery 24V operation , 100V load dump surges and more, so they have nicely packaged solutions called SSR's or solid state switches with Hi or Lo side drivers.