1
\$\begingroup\$

I'm salvaging parts from old computers, read the datasheet of those parts and try to understand the important things build stuff. While i now understand some crucial things about mosfets there is alot that i need to learn. Anyway i'm playing with the tlc5940 (16ch 12bit led driver) and i wanted to create a "animated" illumination for my room. I already created some led drivers with transistors , but also mosfets. Not logic level mosfets. To do so i needed to build a complex "Saturation" circuit involving other 2 transistors per Mosfet.

In my local electronics shop they on ly have higher voltage mosfets. Also on the websites i normally buy parts.

The salvaged mosfets are all N-type Logic level ... so perfect both for the raspberry and the Arduino.I can saturate them correctly, there is almost no voltage drop, they are really fast and they sink alot.

My question is:

Looking at the datasheet i could not find that big difference regarding each different mosfet. But they all have another number.I tought the amperage but some lower number mosfets have more amps.

05N03L  = 80 Ampere
07N03L  = 30 Ampere
15N03L  = 42 Ampere
32N03L  = 50 Ampere
46N03L
55N03L
60N03L

So what are those 2 numbers in front of the code?

N means N-type, L means logic level ... 03?

And most important question:

If i use different types of those mosfets with the tlc5940 could there be a SERIOUS problem or would that work nicely?

note: pullup 10k each ch, tlc with 6,8k resistor to let just sink enough to activate the mosfet, keep all channels under 3A.

Sample circuit using arduino pins

schematic

simulate this circuit – Schematic created using CircuitLab

And here using TLC5940NT

schematic

simulate this circuit

\$\endgroup\$
1
\$\begingroup\$

Just look up the data sheets...

The 60N03L is a 60 amp 30 volt device and the 55N03L is a 55 amp 30 volt device. I didn't look up any more because a trend seemed to be occuring.

Always check the data sheets though.

Having now checked a few more data sheets it appears that the trend is bucked. The "03" part of the name consistently seems to imply that the voltage rating is 30V BUT, for instance, the IPB05N03 device is an 80 amp device that can handle peaks (on period of 80 us) up to 160A: -

enter image description here

Of note is the performance when the gate drive signal is 3.0 volts (i.e. a lowish 3V3 logic drive level). With a 2A load you would probably say the volt drop is about 0.1 volts and therefore the conducting power dissipation figure is 200mW. If we looked at the 07N03 device the power dissipation is about the same but with a 3.2 volt gate drive.

You could probably run either of these devices without a heatsink and expect the device not to rise much above 70 degC and possibly only 50 degC with some decent flowing air around the device. To confirm this check the data sheet for the temperature coefficient of the package - it should be stated as X degrees per watt.

You should also note that the 05No3L won't be able to conduct anything more than 20A with a gate drive of 3.0V and, I suspect, if you looked at the data sheet for the 07N03L it will be maybe 30A at 3.2 volts but it'll get too hot without a heatsink. For instance, the 05N03L at 3v gate drive will probably drop about a volt at 20A and this means a power dissipation of 20 watts - definitely it will need a really good heat sink.

Regarding: -

If i use different types of those mosfets with the tlc5940 could there be a SERIOUS problem or would that work nicely?

The circuit you have shown is likely to work with a load of no more than 2A without a heatsink but you need to check what the PWM frequency is and how fast the driver can charge up the capacitor inside the gate of the transistor. Typically the 05N03L has a 2.5nF gate capacitance and if you fed it from a 100 ohm resistor, 5 x CR = 1.25 us. Clearly uit would be foolish to run this at a PWM frequency of 1MHz but 100kHz is looking OK but it's worth examining what the switching losses are likely to be: -

You could estimate that half way through switching the current is 1A with 6V across the FET - that's a power of 6 watts and lasts for about 1us. With a 100 kHz PWM this is dissipated twice in a 10us period therefore power due to switching is about 20% of 6 watts i.e. 1 watt.

If you had a 10 ohm resistor feeding the gate (and the driver could push the current into the device's gate), the charge time would be one tenth and switching losses at 100 kHz would be about 0.1 watts. Alternatively, running at 10 kHz may be a better option.

\$\endgroup\$
  • \$\begingroup\$ @cocco [For the 3rd time.] Comments are not for extended discussion; this conversation has been moved to chat. \$\endgroup\$ – Nick Alexeev Oct 31 '15 at 18:43
  • \$\begingroup\$ @Andy doesn't seem to like the SE chat either (with all due respect to Andy.) \$\endgroup\$ – Nick Alexeev Oct 31 '15 at 18:46
  • \$\begingroup\$ @Andyaka I don't mind the conversation. But you might do it in the chat as well. \$\endgroup\$ – Nick Alexeev Oct 31 '15 at 18:47
  • \$\begingroup\$ @Andyaka I sense you may have a misconception. Chat doesn't have to be concerted. It can be as slow and unconcerted as comments under a post. A separate chat room is created, your conversation is not interleaved with others. \$\endgroup\$ – Nick Alexeev Oct 31 '15 at 18:51
  • \$\begingroup\$ perfect now i would say. bigger mosfets handle more amps right? so just confirm that on a 40A it would work (3.3v @ 2A no heatsink under 70deg). btw the pwm freq on arduino is 480/960HZ i could use a 10k resitor prolly ... the tlc5940 should have no problems? there is no resistor stopping the flow \$\endgroup\$ – cocco Oct 31 '15 at 18:51

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.