Totem pole gate driver with "extras"

Status
Not open for further replies.

Flyback

Well-Known Member
Hi,
All the totem pole gate drivers i find on the web dont have the "extras" that are needed to make this work well.
Eg, they dont have an upper Darlington for the upper txtor of the totem pole.
And thus they lack the added circuitry to sweep out minority carriers from the Darlington to make it switch faster.
They also lack the cctry added to sweep out the minority carriers from the lower PNP of the totem pole.
Also, the web offerings lack the diode used to reduce cross conduction.

As you know, the easiest way to do totem pole driver is to actuate the totem pole with a low side common emitter BJT to pull down the centre point....and a resistor to pull up the centre point..

But as discussed, the ones on the web are all pretty awful.
Do you know of any with the needed "extras"?
 
A true "Totem pole" output is all NPN and does not give a particularly high positive voltage. It originated in TTL logic ICs.



You are talking about a symmetric push-pull type circuit, Class B amp style.
They are very different in operation and overall characteristics and should not be confused.

A symmetric push pull with normal darlingtons gives a very reduced output swing and is inefficient power wise, due to the voltage drops across the power transistors when they are switched on. You can get better power efficiency using Sziklai pairs.

The only things normally used in high speed / high power circuits with darlingtons are relatively low value base-emitter resistors at each stage and reverse diodes across each base-emitter on each device.

When the drive signal switches, the devices that were previously on have the B-E junctions reverse biased equally, with the voltage divided via the diodes.

The drive circuit need an appropriate voltage offset between the base connections of the two halves to allow for the chain of B-E drops and control current appropriately to each, depending if it's on or off.

Things that use darlington configuration outputs like that in commercial high power gear often have auxiliary power supplies to allow the drive circuit voltages to exceed the high current power rails, so the devices can be driven nearer full saturation.

High speed PWM systems may have separate drive for the upper and lower halves, to allow a dead time and prevent shoot through.


There are numerous integrated gate drives that commercial manufacturers seem to find quite acceptable. Why are you trying to re-invent the wheel in the first place?
 
There are numerous integrated gate drives that commercial manufacturers seem to find quite acceptable. Why are you trying to re-invent the wheel in the first place?
Thanks, its a legacy product.....and we havent the money to redevelop it just yet, and it works, so just keeping it going for now....also, its discrete and AYK, chip shortage is a problem now , so we are not keen to move to an integrated solution just yet......having said that, its an opto based isolator with a common-emitter on the "high side" of the chip.....this yanks down the centre point of the said pushpull pair, and makes it switch....a pullup resistor switches on the upper Darlington of the pushpull-pair when the CE "BJT" turns OFF.
You are talking about a symmetric push-pull type circuit, Class B amp style.
Thanks, yes, quite right, lower is PNP, upper is Darlington.
When the drive signal switches, the devices that were previously on have the B-E junctions reverse biased equally, with the voltage divided via the diodes.
Thanks im trying to picture what you kindly say here, ....we just have a diode from base of Darlington to base of PNP...(cath to PNP base)
A symmetric push pull with normal darlingtons gives a very reduced output swing and is inefficient power wise, due to the voltage drops across the power transistors when they are switched on
Thanks, im simulating ours and getting about a volt across the Darlington when its ON.
High speed PWM systems may have separate drive for the upper and lower halves, to allow a dead time and prevent shoot through.
Thanks, im sure youre right though ours doesnt have that.
Thanks again, i'm sure your right , though ours just has the one high side rail for drive and main gate-drive
The only things normally used in high speed / high power circuits with darlingtons are relatively low value base-emitter resistors at each stage and reverse diodes across each base-emitter on each device.
Thanks again, though ours doesnt have these things.

I wish i could show the schem but am restricted......ours does have a BJT/Zener "thing" which stops it giving gate pulses when the high side gate drive rail falls too low.......there is then a 1MEG resistor from that circuit to the base of the PNP....and its this 1MEG that i cant work out what its for.......getting rid of it on the LTspice sim makes no real difference to the switching.....however, the 1MEG does keep current flowing in the diode between the base's when the Darlington is ON....so i wonder if its to make this diode turn on quicker?....ie, turn on quicker when the PNP gets turned ON?......Also, when the (upper) Darlington is ON, the base-to-base diode would have very low voltage across it if it werent for this 1MEG....so maybe this is it...ie, it keep the base of upper and lower switchs apart from each other when the upper switch (DARLINGTON) is ON?....this may lead to even less shoot-through i am wondering....or in fact, by keeping the base-to-base diode in conduction when the Darlington is ON....then maybe there is a minority carrier sweep-out path for the PNP?...thus the PNP turns ON and OFF quicker.?
 
Last edited:
Status
Not open for further replies.
Cookies are required to use this site. You must accept them to continue using the site. Learn more…