I've been working with some BJT transistor constant current circuits and of these 4 basic designs I like the Zener option. Very simple to calculate and, to me, a pretty straight forward design. Are there any real benefits or pitfalls of using basic diodes or Zeners over a standard resistor voltage divider base bias? I also like the JFET options or using LM317s for higher currents...

 

I like this one from Bob Pease.
Low dropout voltage, and temperature compensated, but initial accuracy is a bit wayward.

 

The zener version is wasting 5V across the sense resistor. Only good for very low currents.

 

I like this one from Bob Pease.

I've looked at several current mirror options but haven't really worked with them yet. I'll give this one a try!

 

I've looked at several current mirror options but haven't really worked with them yet. I'll give this one a try!

It's not a current mirror, it's the same as the usual two-transistor current source (which you haven't included in your original set) with the addition of the schottky diode.
The two-transistor current source has the base of the drive transistor as its voltage reference, but of course it drifts with temperature. The schottky subtracts a voltage from that reference, and its voltage drifts with temperature by the same amount, so you have a reference which is temperature-stable.

 

Hmmm, OK... Except for the additional R1 connection to the base of Q2 and the Schottky also at its base it looked like a "mirror" circuit I've seen to me. Anyway, I'll check it out!

 

Hmmm, OK... Except for the additional R1 connection to the base of Q2 and the Schottky also at its base it looked like a "mirror" circuit I've seen to me. Anyway, I'll check it out!

Maybe I'll remember where I saw it. I bet Bob Pease wrote a good explanation (it would almost certainly be an entertaining explanation)

 

The Bob Pease circuit also has better current regulation with a change is load resistance (stiffer) due to the added feedback gain from the second transistor.

Below is the LTspice sim showing the difference in current regulation with a change in the collector load resistance for the 1-transistor and the 2-transistor circuits.
Depending upon the application that difference may or may not be important.

 

it looked like a "mirror" circuit I've seen to me.

Look again.
A "mirror" circuit has one base shorted to its collector as below:

 

I bet Bob Pease wrote a good explanation (it would almost certainly be an entertaining explanation)

LOL, I'm sure it was. I've read some of his articles and have his book on analog circuits.

 

I guess I misinterpreted. Alan Wolke W2AEW showed this mirror and then another similar circuit that I thought was also a mirror. It's the second circuit that is very similar to Bob's. I assume the SMD BAT or BAS can be replaced by a through hole 1N58xx Schottky?

 

With resistors only, there is no temperature compensation for the transistor's Vbe.

The two-transistor version (that you do not show) has much better regulation.

I'm a big fan of the LM317 approach. My garage goes from -10 F to +100 F, and an LED driver I have out there just sits there like a stone.

ak

 

Does that include the LM317L in the TO-92 package? Seems like a good 2 "transitor" approach? Seems like none of the responses here are in favor of the diode or Zener approach other than Bob's design with the Schottky.

 

Does that include the LM317L in the TO-92 package? Seems like a good 2 "transitor" approach? Seems like none of the responses here are in favor of the diode or Zener approach other than Bob's design with the Schottky.

I have used the two diode circuit.

 

Is there any real advantage to it over the standard resistor voltage divider base bias?

 

Is there any real advantage to it over the standard resistor voltage divider base bias?

Yes, it is not sensitive to supply voltage.

 

Does that include the LM317L in the TO-92 package? Seems like a good 2 "transitor" approach?

The LM317L constant-current circuit is a single device with one resistor.
It does have a slightly higher total voltage drop then the transistor approach.

Below is the LTspice sim of a LM317 constant-current circuit for the input voltage going from 0V to 10V with different values of load resistor:
That circuit has a minimum voltage drop of about 3V with a stiff output (current varies very little with a change in load resistance once the constant-current value is reached).

 

Does that include the LM317L in the TO-92 package?

Yes. The temperature stability is about the same. Note that the L version in a TO-92 package can dissipate very little heat.

ak

 

Look again.
A "mirror" circuit has one base shorted to its collector as below:

View attachment 304190

That's just a basic current mirror. More complex current mirrors may or may not have a transistor in a direct diode connection. Even pretty simple enhancements to improve performance in some regard often insert other things in the way, such as a resistor in the base path of the diode-connected transistor to reduce sensitivity to transistor beta or a resistor in the collector path of that transistor to better match the nominal Vce of the output transistor.

 

Is there any real advantage to it over the standard resistor voltage divider base bias?

As with most things, there are both advantages and disadvantages. It comes down to what is important.

Are you willing to pay more (more parts, more real estate on your PCB, etc) in order to have your current source be stiffer, or more accurate, or less sensitive to temperature variations, or (whatever other aspect the more complex circuit offers)? If so, then there's a real advantage.

If the basic circuit works well enough for what you are using it for, then there is no advantage to using a more complex circuit. It may perform better, but that better performance does not translate into any real advantage if you don't benefit from it -- all you will have done is paid more for a circuit than you needed to.