Open Collector Output Boost

[ADWSystems]

I was drawing a schematic with a LM393 and NPN transistor to boost the output current. Originally I had a pull-up resistor on the '393 output and a base resistor on the NPN, but when I drew the schematic I forgot the base resistor which got me to thinking if I really needed a base resistor. What are the pros and cons to including and omitting the base resistor? The pros I can see is one less resistor and a simpler circuit. I don't see any cons, does anyone else?

View attachment 63761

 

[ronsimpson]

looks OK.

 

[ericgibbs]

I was drawing a schematic with a LM393 and NPN transistor to boost the output current. Originally I had a pull-up resistor on the '393 output and a base resistor on the NPN, but when I drew the schematic I forgot the base resistor which got me to thinking if I really needed a base resistor. What are the pros and cons to including and omitting the base resistor? The pros I can see is one less resistor and a simpler circuit. I don't see any cons, does anyone else?

View attachment 63761

hi
Do you plan to have any positive/hysteresis feedback from the output.?

If so without the base resistor the output of the LM393 will only swing from close to 0V and approx 0.7v, this small change means that a low value of hysteresis resistor is required.

I usually split the LM393 pull up into two resistors [ie make one the base resistor], this means the LM393 output will swing from 0v to Vs/2, which gives an increased hysteresis range.

 

[ADWSystems]

There is hystersis.

What is the pro to splitting the resistors versus not?

 

[Mikebits]

There is hystersis.

What is the pro to splitting the resistors versus not?

Eric explained it.

I usually split the LM393 pull up into two resistors [ie make one the base resistor], this means the LM393 output will swing from 0v to Vs/2, which gives an increased hysteresis range.

 

[ADWSystems]

I'm sorry requoting something doesn't clarify it.

With and without the base resistor, the output is swings 0 to Vs/2. Using formula provided in the On Semi datasheet, Rl (pull-up resistor) has minimal effect on the hysteresis. Dropping from 10K to 5K, raised the upper threshold 40mV. Is there another effect of the base resistor I'm missing.

 

[ericgibbs]

I'm sorry requoting something doesn't clarify it.

With and without the base resistor, the output is swings 0 to Vs/2. Using formula provided in the On Semi datasheet, Rl (pull-up resistor) has minimal effect on the hysteresis. Dropping from 10K to 5K, raised the upper threshold 40mV. Is there another effect of the base resistor I'm missing.

hi ADW,
I will run some simulations in LTS to show the point I am making, will post tomorrow.

Others disagree with what I have found in practice, that connecting the LM393 output directly to a transistor base does effective the reliable operation of the Off point hysteresis, the On point is unaffected.

E,

 

[ccurtis]

I'm sorry requoting something doesn't clarify it.

With and without the base resistor, the output is swings 0 to Vs/2. Using formula provided in the On Semi datasheet, Rl (pull-up resistor) has minimal effect on the hysteresis. Dropping from 10K to 5K, raised the upper threshold 40mV. Is there another effect of the base resistor I'm missing.

The output of the comparator as you have shown it cannot rise to Vs/2 as you state. It will rise to no higher than the Vbe of the transistor it is driving, as Ericgibbs stated. The pull-up resistor has little effect on hysteresis, true, but Vbe limiting the output of the comparator to no more than 0.7 volts has a great effect of how much hysteresis can be achieved by adding a positive feedback resistor.

Edit: Without studying the idea in detail, going further on the hysteresis topic, it seems to me that the collector of the buffer transistor could be connected to the - input of the comparator through a resistor to provide overall positive feedback (the buffer transistor provides the first 180 deg inversion), assuming that whatever might be connected to the collector does not also limit the maximum voltage on the collector too much.