Common base PNP where collector and emitter currents differ significantly?

[ocset]

¬¬Hello,

In the attached 6W Buckboost LED driver schematic, the high-side current sense is referred down to ground via the common base connected PNP  (FMMT560Q).

We are worried about   situations in which the collector and emitter currents of the PNP would start to differ  significantly. (eg, at  low [eg sub 1mA]   collector currents, or at high temperature, where  emitter-base leakage current may make collector and emitter currents differ.)

Can you confirm, in this schematic,  that  collector and emitter currents are always likely to  remain close in magnitude?


FMMT560Q datasheet
https://www.diodes.com/assets/Datasheets/FMMT560Q.pdf

LTspice simulation also attached.

[T3sl4co1l]

More or less by definition, the ratio (alpha) gets bad around saturation.

Tim

[jmelson]

Somebody has been smoking too much <something>.  That circuit is INSANE!  two op-amps, a switching power supply controller, several assorted transistors, 4 voltage sources, and on and on, for 6 W?

There are single chips that will do this very nicely, possibly with an additional resistor or two to set the parameters.

Jon

[T3sl4co1l]

Somebody has been smoking too much <something>.  That circuit is INSANE!  two op-amps, a switching power supply controller, several assorted transistors, 4 voltage sources, and on and on, for 6 W?

There are single chips that will do this very nicely, possibly with an additional resistor or two to set the parameters.

Jon

You must be new here, look at his other posts too

Tim

[ocset]

Somebody has been smoking too much <something>.  That circuit is INSANE!  two op-amps, a switching power supply controller, several assorted transistors, 4 voltage sources, and on and on, for 6 W?
There are single chips that will do this very nicely, possibly with an additional resistor or two to set the parameters.

Thanks, i should have said, Vin is from 18v to 137vdc....where it must work at 6w
But it must also work at 1.2w when vin = 8-14vdc.
And must work at 3.6w when vin = 14-18vdc.

More or less by definition, the ratio (alpha) gets bad around saturation.

Thanks, yes , i can see i have forgotten to add a base resistor, as in page 12, figure 3 of the following datasheet....

http://www.analog.com/en/products/power-management/led-driver-ic/buck-boost-led-drivers/lt3756.html

[BrianHG]

Try just using a lower voltage LED string, IE leds with some in parallel instead of just all serial, with either 1 buck converter (lower efficiency at high voltages) or 2 buck converters, with the higher one switching straight on when the voltage gets too low,  better efficiency at high voltage.

[Zero999]

Somebody has been smoking too much <something>.  That circuit is INSANE!  two op-amps, a switching power supply controller, several assorted transistors, 4 voltage sources, and on and on, for 6 W?
There are single chips that will do this very nicely, possibly with an additional resistor or two to set the parameters.

Thanks, i should have said, Vin is from 18v to 137vdc....where it must work at 6w
But it must also work at 1.2w when vin = 8-14vdc.
And must work at 3.6w when vin = 14-18vdc.

More or less by definition, the ratio (alpha) gets bad around saturation.

Thanks, yes , i can see i have forgotten to add a base resistor, as in page 12, figure 3 of the following datasheet....

http://www.analog.com/en/products/power-management/led-driver-ic/buck-boost-led-drivers/lt3756.html

You should have stated the input voltage range in the first post.

I've noticed, you've posted quite a few questions relating to an 18V to 137V LED driver. It's certainly possible. I've seen commercial switched mode power supplies which can work off this kind of voltage range.  For example, the Rigol DS1054 oscilloscope, works from 36VDC to 400VDC, range of 1:11¹/₉ even though it's only designed for 100VAC to 250VAC. Perhaps you can get the schematic for the PSU in the  Rigol DS1054 and work from that?

[ocset]

I mean, the PNP in the schematic of the top post  cannot saturate because its Base-Collector junction cannot get forward biased in that schematic and setup.

It seems weird though that such a simple  method of referring a sense signal down from the high side is not more widely used? I mean, its just basically a cheap PNP transistor as shown in the top post. It makes you search for some hidden gremlin in  this PNP method.

[Zero999]

I mean, the PNP in the schematic of the top post  cannot saturate because its Base-Collector junction cannot get forward biased in that schematic and setup.

It seems weird though that such a simple  method of referring a sense signal down from the high side is not more widely used? I mean, its just basically a cheap PNP transistor as shown in the top post. It makes you search for some hidden gremlin in  this PNP method.

Yes, the transistor can't saturate.

This seems odd: I'd expect the beta to be higher at higher temperatures and lower currents, down to a point of course.

[ocset]

Thanks yes, i wonder how low the collector current has to go before the Hfe starts  getting really low?

[jbb]

Oh treez, you don't disappoint  .

At risk of threadjacking, how about using a flyback converter to drive the LEDs? It'll probably be a little less efficient, but you get some advantages:

• use of turns ratio to optimise the number of LEDs
• buck-boost voltage gain characteristic to match the LEDs to the wide input voltage
• the LED current shunt resistor can be referenced to 0V
• a tertiary winding can be added to make some bias voltage (remembering that the LED string voltage is sort of constant)
• you can get one flyback controller chip to do most of the work.  I would suggest you consider some kind of over voltage protection circuit to prevent destruction if an LED goes open circuit
• assuming constant frequency and DCM operation, setting a peak primary current can give you a well-behaved input current limit

Alternatively, there are some options for pampas in the sensing system:

• LT makes some differential amplifiers which have pretty good CMRR and are rated for 250V common mode input
• LT and others also make current sense amplifiers which do the same thing as your U1 and Q2, but might be a bit smaller