Maker Pro
Maker Pro

SMPS's with cascoded opto feedback

eem2am

Aug 3, 2009
429
Joined
Aug 3, 2009
Messages
429
Please Note these two equivalent offline flyback simulations.
-one has cascoded opto feedback , and the other does not.
Note how the non-cascode version has 2V of overshoot on vout at start-up. Due to the lack of a cascode, this overshoot is virtually impossible to get rid of whilst keeping sufficient gain and phase margin. The casc0ded version has no overshoot, and achieving this was simple.
Cascodeing is extremely cheap to do...why does everyone not do it?

Schematics and LTspice simulations of offline flybacks attached. (one cascode and one no-cascode, but they are the same power level, vin, vout, np/ns, Fsw.)
 

Attachments

  • Flyback _Cascode opto.pdf
    18.2 KB · Views: 171
  • Flyback _opto.pdf
    17.6 KB · Views: 164
  • Flyback _Cascode opto.txt
    8.4 KB · Views: 120
  • Flyback _opto.txt
    7.8 KB · Views: 130
Last edited:

jpanhalt

Nov 12, 2013
426
Joined
Nov 12, 2013
Messages
426
And, I thought it was just because not everyone builds SMPS.

Just be thankful that we are not discussing why university EE programs in the UK don't include courses in analog design.

John
 

eem2am

Aug 3, 2009
429
Joined
Aug 3, 2009
Messages
429
the cascode is slightly more expensive, but a sot23 bjt and a few resistors is a small price to pay for a more robust power supply.

Is the cascoding of the opto feedback no different to using a common collector connection of the opto transistor?

Common collector connection attached, as well as its LT spice simulation
(Comon emitter connection of opto also attached for good measure)
 

Attachments

  • Flyback _opto Common collector.pdf
    17.8 KB · Views: 143
  • Flyback _opto Common emitter.pdf
    17.3 KB · Views: 131
  • Flyback _opto_CC.txt
    8.1 KB · Views: 106
  • Flyback _opto_CE.txt
    7.5 KB · Views: 116
Last edited by a moderator:

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
Can you post images rather than pdfs?

If you want me to put some effort into your question lectures then at least make it easy for me.

Can you also let me know if this is a real question, or are you just suggesting that nobody can design SMPS power supplies?
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
Also, apart from the inversion of the output signal, explain the practical difference between common emitter vs common collector for an optocoupler.
 

eem2am

Aug 3, 2009
429
Joined
Aug 3, 2009
Messages
429
Thanks, I believe that if we can find an answer to this question then it is amazing, because this is not written anywhere, even the Basso book doesn't cover it. (Basso pgs 288-308 covers these connection methods, but doesn't say which is more conducive to achieving stability)
You can see the opto transistor and see its either in CC or CE....the "VREF1" and "VREF" nodes are a 5V rail, provided by the LT1243 controller.

The schematic and simulation attached, show the differences in the optocoupler connection being either common emitter or common collector
"
here is the schematic in JPEG
29uyvr5.jpg

..It is also attached in pdf



Here is the question again....
Regarding opto-isolated flyback SMPS’s, which opto-coupler connection method is more conducive to facilitating a higher bandwidth in the feedback loop whilst maintaining stability?….is it the "common emitter" connection of the opto, or the "common collecter" connection?

An LTspice simulation showing both connection methods is attached. In the simulator, the “common emitter" connection” version is far harder to make stable.

LT1243 datasheet...(its basically same as UC3843)
http://cds.linear.com/docs/en/datasheet/1241fa.pdf
 

Attachments

  • Flyback _opto_CE_CC compare.pdf
    27.7 KB · Views: 130
  • Flyback _opto_CE_CC compare.txt
    16.1 KB · Views: 210
Last edited:

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
There is no common emitter or common collector in anything you have shown.

edit: ask yourself "what is it common with"?
 

eem2am

Aug 3, 2009
429
Joined
Aug 3, 2009
Messages
429
"VREF" and "VREF1" are the supply rail provided by the LT1243.... this is what they are commoned with.
Its the opto transistor, inside the CNY17 optocoupler that is connected either CC or CE.

Pages 16, 17 18 of the following state that common collector is better...

..it states how "The common-collector configuration eliminates the miller
effect of the output transistor’s collector-to-base capacitance"
and generally increases achievable loop bandwidth.
http://cds.linear.com/docs/en/datasheet/4430fc.pdf
 

Arouse1973

Adam
Dec 18, 2013
5,178
Joined
Dec 18, 2013
Messages
5,178
"VREF" and "VREF1" are the supply rail provided by the LT1243.... this is what they are commoned with.
Its the opto transistor, inside the CNY17 optocoupler that is connected either CC or CE.

Pages 16, 17 18 of the following state that common collector is better...

..it states how "The common-collector configuration eliminates the miller
effect of the output transistor’s collector-to-base capacitance"
and generally increases achievable loop bandwidth.
http://cds.linear.com/docs/en/datasheet/4430fc.pdf

I think your right, an emitter follower configuration will be faster because the miller capacitance for a common emitter version is CM = Cbc*(AV+1). So because the common emitter version has voltage gain greater than 1 this will magnify the capacitance and the dominant miller capacitance will take effect. The emitter follower version has a voltage gain of less than 1 so the miller effect is removed and a capacitance slightly less than Cbc dominates but for best results it is recommended to have a base pull down resistor to achieve much better results than an open base in for emitter follower configuration.

Thanks
Adam
 

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
How can a transistor whose base is not connected to anything be said to be operated in common emitter or emitter follower configuration?
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
Someone has to explain to me how common emitter and common collector can apply to an optocoupler where there is no common connection (or more obviously, no connection to the base).

To my mind, the transistor will behave *exactly* the same whether the emitter is grounded and there is a load resistor connected to the collector or vice versa. What is happening at the base emitter junction is completely independent of the potential at the emitter.

Sure, an emitter follower has a gain of less than 1, but that requires that the output affects the emitter potential with respect to the base. That's simply not happening here.

The only difference will be that one configuration will have a faster rise time and the other a faster fall time. But this is simply because turning on the transistor causes a rise in one configuration and a fall in the other.

There may be other differences if there is a significant load on the output, and whether this is a source or sink, but then the differences are not due to the configuration but to the load.

This suggests that miller capacitance is a minor factor (bottom of page 1), and also suggests that you need a base resistor (so that you really do have a common emitter or common collector configuration) to be able to manipulate things to deal with this effect.

Without the base resistor, all you're doing is observing the effect of junction capacitance on switching speed and observing that you can switch on faster than you can switch off.

From page 10:
Another limitation of the Miller effect is that it only really becomes apparent when the turn-off mechanism is by means of an active discharge through a base to ground capacitor. In the case where the base is left virtually without any discharge path, the effects of “Miller the killer” are not noticeable...
I would encourage you to read page 7 which indicates that a base resistor is required should you wish to have anything other than a "common emitter"

The next step in switching speed improvement is to move from a standard 4 pin part to a standard 6 pin coupler. This type of coupler gives the user access to the base of the phototransistor and allows us to modify the standard common emitter switching circuit discussed above...
Maybe I'm missing something, and if I am, I'll be quite happy to be educated.

edit: and while I spend ages writing a tome, Kris gets to the point in one line.
 
Last edited:

Arouse1973

Adam
Dec 18, 2013
5,178
Joined
Dec 18, 2013
Messages
5,178
Someone has to explain to me how common emitter and common collector can apply to an optocoupler where there is no common connection (or more obviously, no connection to the base).

To my mind, the transistor will behave *exactly* the same whether the emitter is grounded and there is a load resistor connected to the collector or vice versa. What is happening at the base emitter junction is completely independent of the potential at the emitter.

Sure, an emitter follower has a gain of less than 1, but that requires that the output affects the emitter potential with respect to the base. That's simply not happening here.

The only difference will be that one configuration will have a faster rise time and the other a faster fall time. But this is simply because turning on the transistor causes a rise in one configuration and a fall in the other.

There may be other differences if there is a significant load on the output, and whether this is a source or sink, but then the differences are not due to the configuration but to the load.

This suggests that miller capacitance is a minor factor (bottom of page 1), and also suggests that you need a base resistor (so that you really do have a common emitter or common collector configuration) to be able to manipulate things to deal with this effect.

Without the base resistor, all you're doing is observing the effect of junction capacitance on switching speed and observing that you can switch on faster than you can switch off.

From page 10:
I would encourage you to read page 7 which indicates that a base resistor is required should you wish to have anything other than a "common emitter"

Maybe I'm missing something, and if I am, I'll be quite happy to be educated.

edit: and while I spend ages writing a tome, Kris gets to the point in one line.

Hi Steve
If you connect up an Opto coupler correctly I am pretty sure you can have common emitter or common collector. Well all of the manufacturers think you can. I am not saying the attached circuit is correct, the user was asking what is better common collector or common emitter. The base terminal of the opto coupler uses light through the air for connection rather than a wire. It's the same thing just a different way of getting the energy across in the form of Photons. As you know the photon is responsible for the energy transfer not the electrons of charge.
Thanks
Adam
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
Adam, you need to read the reference I gave.
 

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
If you connect up an Opto coupler correctly I am pretty sure you can have common emitter or common collector.
You can connect it with the emitter to 0V and the collector through a resistor to V+ or you can swap the transistor and the resistor over; this will determine whether the voltage goes high or low when the optocoupler is energised, but the transistor doesn't (can't possibly) behave any differently if the base isn't connected to anything!

It cannot meaningfully be described as either common emitter or emitter follower topology.
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
Consider Vce. Will that ever be different if a resistor is connected between emitter and ground vs collector and Vcc.

I suggest that with no connection to the base, the answer must be no. The only difference is a 180 degree phase shift and a level shift.

edit: Read section 2 of this: http://www.renesas.com/products/opto/technology/usage/index.jsp

Miller effect will play no part in differentiating the behaviour of an opto in "CC" or "CE" configuration.
 
Last edited:

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
That explains it nicely.

If a document refers to the collector output topology as "common emitter" and the emitter output topology as "common collector" or "emitter follower", it's because the author hasn't realised that those names they aren't appropriate, or the author is using them as (misleading) shorthand for collector output and emitter output, because they are easily recognised.

This will also be true when an external base-emitter resistor is used to speed up the turn-off of the phototransistor.
 

Arouse1973

Adam
Dec 18, 2013
5,178
Joined
Dec 18, 2013
Messages
5,178
Consider Vce. Will that ever be different if a resistor is connected between emitter and ground vs collector and Vcc.

I suggest that with no connection to the base, the answer must be no. The only difference is a 180 degree phase shift and a level shift.

edit: Read section 2 of this: http://www.renesas.com/products/opto/technology/usage/index.jsp

Miller effect will play no part in differentiating the behaviour of an opto in "CC" or "CE" configuration.

Yes Steve
I did mention the use of a resistor to speed up the device and you will then see a marked change in the response of the rise time between the two. Is phase shift correct how you have used it? Phase shift is a shift in time between two signals. Should this be signal inversion.

Regarding your link which part of it is not technically correct with regard to base current coming from the collector. This is incorrect, the base current comes from photons hitting the base collector junction causing a photo current (base current from light). Apart from having a base junction optimised for this purpose they are the same as normal BJTs. You can even cut the top of a 2N2222 and use this as a photo transistor.

With regards to CB CE and CC I think we will have to beg to differ on this. The way I look at the different configurations is which ever terminal of the transistor is not the input or output has to be the common one.

This is then quite simple to work out even if the text book diagrams that are used show negative voltages where you would normally see a positive voltages in normal conventional circuits. This is quite of the case in Physics books

This can be quite non intuitive at first glance and take some working out when we have been used to circuits being connected up the other way.
This has been a very interesting discussion.
Thanks
Adam
 

eem2am

Aug 3, 2009
429
Joined
Aug 3, 2009
Messages
429
ok thanks, ill read these but we are pitting linear.com of my #7 post against your Vishay and Renesas....we'll see which of these has it wrong soon
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
Is phase shift correct how you have used it?

Yes

Phase shift is a shift in time between two signals. Should this be signal inversion.
Phase shift can be due to a time shift, but 180 degrees of phase shift is a simple inversion and applies to this signal.

Regarding your link which part of it is not technically correct with regard to base current coming from the collector. This is incorrect, the base current comes from photons hitting the base collector junction causing a photo current (base current from light). Apart from having a base junction optimised for this purpose they are the same as normal BJTs. You can even cut the top of a 2N2222 and use this as a photo transistor.
I'm not sure what you are suggesting here. Are you saying that in a phototransistor the emitter current is different from the collector current?

edit: Let's be quite pedantic. Without a base connection a phototransistor is essentially operating as a field effect device. The photons are causing a charge to accumulate and recombination is discharging it.

With regards to CB CE and CC I think we will have to beg to differ on this. The way I look at the different configurations is which ever terminal of the transistor is not the input or output has to be the common one.

Ah, so you *do* think that a 2 terminal device can have a different current in each of its 2 leads.

This is then quite simple to work out even if the text book diagrams that are used show negative voltages where you would normally see a positive voltages in normal conventional circuits. This is quite of the case in Physics books
I will now sit back and watch you explain how the voltage drop across a phototransistor is different depending on whether a resistor is connected in series with the collector or the emitter.

Or even better how, with the base open, that the collector current can differ from the emitter current.

Perhaps you could also venture to tell us whether we should place resistors in series with the anode or the cathode of LEDs and which provides the better performance.

This can be quite non intuitive at first glance and take some working out when we have been used to circuits being connected up the other way.
I agree. And I suspect that when you've shown me how it works, I'll be more educated on the subject.
 
Last edited:
Top