# Darlington Power Dissipation Question

Evening all,

I'm using a VB325SP ignition coil driver chip in a design of mine - it's a dedicated IC from ST. Here's the datasheet -

http://www.datasheetcatalog.org/datasheet2/2/05cozo7wzapcwa4x6itr0eo8z63y.pdf

I've run into a few problems however... A couple of questions -

1) Regarding pin 8, named 'base darlington'. The typical application diagram on page 9 shows no connection to this, and the block diagram on page 1 shows it's connection - I can't quite seem to work out what this pin is for.

VCC is a logic voltage supply, so does this pin need connecting to the voltage source used to drive the coil, namely battery voltage, with a current limiting resistor to limit current to 150mA, as there is a maximum current input for that pin in absolute max ratings.

2) A thermally related question, how does one work out how much heat this IC will generate? There is the thermal resistance data, fine - but I can't work out how to generate how many Watts of heat I'm producing in the IC to use this in the first place? I know its not as simple as I^2*R as with a MOSFET, but there must be a way? Any clues?

I can't figure out what the BD pin is for (for monitoring the switch?), but you don't need to worry, the example application circuit shows its not required.

The Vcc must be supplied with a regulated 5V - anything above 7V will destroy the chip (abs max ratings). It also need standard logic decoupling of course.

Heat dissipation to a first approximation is about 1 to 1.5 times the current - the saturation voltage of the darlington times its current.

However since this is a high voltage switch there will be significant power dissipation on switching (switch off takes upto 25us) - if the output voltage is 100V at 6.5A then a smooth ramp-down could dissipate about 8mJ, or 8W at 1000Hz... Output voltage depends on the inductor being switched. Only at switch-off will there be a high voltage (inductive load).

The Vcc must be supplied with a regulated 5V - anything above 7V will destroy the chip (abs max ratings). It also need standard logic decoupling of course.

Yep, I've got Vcc at a regulated 5v with appropriate local decoupling etc.

Heat dissipation to a first approximation is about 1 to 1.5 times the current - the saturation voltage of the darlington times its current.

However since this is a high voltage switch there will be significant power dissipation on switching (switch off takes upto 25us) - if the output voltage is 100V at 6.5A then a smooth ramp-down could dissipate about 8mJ, or 8W at 1000Hz... Output voltage depends on the inductor being switched. Only at switch-off will there be a high voltage (inductive load).

Ah, I think something has just clicked with me - this is a sinking IC, not a sourcing IC - hence the lack of a battery voltage connection to the chip.

As for 'output voltage', are you referring to the voltage seen on the HVc pin when the load is switched off, i.e. the inductive spike?

That's the one - the actual voltage depends on the ignition coil primary and secondary and the spark-gaps.

MarkT: That's the one - the actual voltage depends on the ignition coil primary and secondary and the spark-gaps.

Of course - many thanks for your help!

Heat is heat it is the power controlled by the transistor and the load dictates the power drawn.. So in this case you need the pulse width and the duty cycle to determine the total load power drawn by the IC. Does it have a thermal pad under it.. I wasn't looking for that when I rad the schematic ST was using for some of the values listed on the data sheet. Were I you I would look for some app notes on the product. Might help a great deal.

Bob

Docedison:
Were I you I would look for some app notes on the product. Might help a great deal.

Bob

This was my first train of thought - but after looking it would appear ST are very hush hush about this product range… But yes, it’s in the PowerSO10 package so the entire underneath of the package is one big thermal pad, that doubles as the high voltage pin.

EDIT: So - doing a quick analysis, assuming a 4ms PW and a 33% duty cycle. If thermal dissipation during the switch being on is for example, 6W - the actual component of thermal dissipation due to the switch being on is 6W * 0.33 = 1.98W. Thus taking the total power dissipation to 1.98W + switching dissipation?

So, looks like I need to do some more reading on how to calculation dissipation due to switching…