MOSFET Biasing with 5V

[audioguru]

The tiny DMP3010 Mosfet is perfect for your requirement, even if it has minimum or maximum spec's.

Guess what? The curves on datasheets are for "typical" devices that you cannot buy. The text in a datasheet shows the minimum to maximum range of specs. The saturation graph for a TIP125 shows 1.26V at an output of 3A and an input of 12mA but the text says a maximum saturation voltage of 2V.
1) Where does 12mA with a pull down from 5V to 2.0V (the maximum Vbe sat is about 3V) come from?
2) Will you accept an output saturation voltage of 2V or will you buy and test many TIP125 Mosfets and hope to fine a better one?

 

[Pommie]

Guess what? The curves on datasheets are for "typical" devices that you cannot buy. The text in a datasheet shows the minimum to maximum range of specs. The saturation graph for a TIP125 shows 1.26V at an output of 3A and an input of 12mA but the text says a maximum saturation voltage of 2V.

Can you suggest an alternative to "reading the datasheet"? How do you decide that a certain device will do the job?

Mike.

 

[audioguru]

Can you suggest an alternative to "reading the datasheet"? How do you decide that a certain device will do the job?

Mike.

The spec's on a datasheet are the manufacturer's guarantee. Buy name-brand parts from a reputable recommended distributor instead of a cheap copy or manufacturer's reject from "you know where". Then do not exceed the datasheet spec's.

 

[Suraj143]

Hi, I did a further test on TIP125.
I disabled the multiplex & load all 48 LEDs continuously.The saturation is around 0.8V (Better than I thought). But the transistor is getting too hot.
But when I enable multiplex, the transistor is pretty cool.

1) Is this normal?
2)TIP125 has some inbuilt pullup resistors from Base To Emitter.But in my PCB I have another 10K pull-up for Base to Emitter.Can I remove that (See my picture)?

 

[ronsimpson]

I have another 10K pull-up for Base to Emitter.Can I remove that

Yes.

I disabled the multiplex & load all 48 LEDs continuously.The saturation is around 0.8V (Better than I thought). But the transistor is getting too hot.
But when I enable multiplex, the transistor is pretty cool.

I don't understand. When you disable multiplexing 48 LEDs are on. When you are multiplexing how many LEDs are on, at one time?

 

[audioguru]

You do not say the current of each LED and also you do not say what (the voltage and current) that drives the base of the Darlington.

I guess each LED has 20mA then 48 of them draw 960mA. Since your saturation voltage is low at 0.8V then the heating is 960mA x 0.8V= 0.77W. The datasheet shows that its thermal resistance from chip to ambient is 62.5 degrees C/W, then if your ambient temperature is 25 degrees C room temperature, the chip will be safe at only 87.5 degrees C which is far from its max allowed chip temperature of 150 degrees C.

Why did you say the load is 3.2A in your first post? Will the TIP125 ever drive 160 LEDs?

 

[audioguru]

I notice that you have a series 100 ohm base resistor. How did you calculate such a low resistance that might overload whatever you have driving it?

The base-emitter voltage of the TIP125 is shown as typically 1.5V when its collector current is 1A.
If whatever drives it goes down to 0.5V and the supply is 5.0V then the base current is (5V - 1.5V - 0.5V)/100 ohms= 30mA. Since the Darlington is shown to saturate well when its collector/base current is 250 times then the collector current can be 30mA x 250= 7.5A. Your load is not 7.5a so the value of the 100 ohms resistor is much too low.
I think your PIC is driving it and is overloaded.

 

[Suraj143]

Thanks Audioguru wonderfull information.

I like to drive my LEDs at 4mA average current.So when multiplexing I need to supply 4mA X 16 times = 64 mA per LED to get a descent brightness .Then a row driver will need to supply 64mA X 48 =3A.

Meanwhile I just checking by loading 48LEDs continuosly without multiplexing how the TIP performs .It is too hot when continuously loaded but LEDs brighter.But when multiplexing its so cooler but LEDs are dim.

Somehow my LEDs are dim.I' m going to replace the column driving ULNs with a low saturation BJT.

For Row driving part I need to do with commonly available parts.Those special mosfets arent available here.

Any improvement to make it more brightness...!!!

 

[ronsimpson]

My head is slow. Please help me understand.
LEDs are on 1/16 in time. So there must be 16 TIP125 transistors.
You say 48LEDs so there must be 48 column drivers.
???
When you stop multiplexing the LED average current goes from 4 to 64mA. They must be hot and almost dead.
When you stop, the average transistor watts goes from 1/16w to about 1w.

 

[ronsimpson]

The saturation is around 0.8V

I' m going to replace the column driving ULNs with a low saturation BJT.

Darlingtons are not low saturation. (0.8v) Two transistor in a NOT-Derlington configuration should be more like 0.2V to 0.3V. I am thinking about both your choice of TIP125 and the column driver.
I would use low gate voltage MOSFETs but …..

LED matrix ... does not need drivers. In color.

 

[Suraj143]

My head is slow. Please help me understand.
LEDs are on 1/16 in time. So there must be 16 TIP125 transistors.
You say 48LEDs so there must be 48 column drivers.
???
When you stop multiplexing the LED average current goes from 4 to 64mA. They must be hot and almost dead.
When you stop, the average transistor watts goes from 1/16w to about 1w.

Exactly.

I have 16 rows so there is 16 TIP125 .

For coulmns I have 48 columns, so there is 6 ULN2803 (8X6=48) & 48 column resistors.

 

[audioguru]

I think you fried and ruined your LEDs by turning off the multiplexing and operating them at 64mA for more than a couple of milli-seconds.

This thread is confusing since its schematic is in your other thread. I marked it with what we are talking about. I see that the LEDs are operating with 22mA pulses, not 64mA but I used "typical" specs which could produce higher or lower currents and voltages if the devices have minimum or maximum specs.

 

[Suraj143]

Hi, Audioguru

You are correct.Now I see why my LEDs are dim.The above diagram is my very first design & I have added a ULN to drive the row transistor.At first I have added B772 transistor.It got heat.So I redesign the row to drive a Darlington directly from PIC pin.Now no more ULN in row side.But in any case I can add a ULN to row side.

I have three choices.

1) Add a logic level FET to row side - This cannot be done due to unavailable parts.
2) Keep the ULN in row side & add a high power transistor like TIP42 & drive it from ULN.Also replace column driving ULN2803 with 2N4401.
3)Drive rows directly from a darlington TIP125 from directly PIC pin via a 330R resistor, Also replace column driving ULN2803 with 2N4401.

Which will be better to make LEDs brighter?

Thanks

 

[audioguru]

Since you cannot buy ordinary common parts then maybe you should make something more simple.

 

[Pommie]

Can you run your matrix at a higher voltage? E.G. use a 9v supply and a 78L05 to power the pic.

Mike.

 

[Suraj143]

Can you run your matrix at a higher voltage? E.G. use a 9v supply and a 78L05 to power the pic.

Mike.

Hi I could re-arrange my PCB.I can supply 12V not 9V.I can power the PiCs and shift registers with 7805.

 

[ronsimpson]

Do you have current limiting resisters with the LEDs?

 

[Suraj143]

Do you have current limiting resisters with the LEDs?

Nop. Only a single resistor per column.

 

[Suraj143]

Does that NDP6020P Mofet is a good part for my Row problem?

 

[ronsimpson]

NDP6020P

I can't remember, I think you are using a 5V computer. With 5V G-S the FET can do 20A with 0.8V D-S, 10A=0.4V, 5A=0.2V
I think the part needs 4 to 5 volts of drive. Works at 3v at lower current.

Only a single resistor per column.

If you want more current then change the resistor. Changing the supply voltage puts more voltage across the resistor which equals more current. Changing the resistor also changes the current.