LED Sequencer - NPN or PNP transistors ?

Thread Starter

01-0077

Joined Jan 12, 2021
30
My apologies everyone, I was reading from scribbled notes when I quoted the current draw of each led bank. The current draw is in fact 231mA per bank. Looking through the side of the lens it appears that each bank has 12 white leds in parallel with a current limiting resistor.
 

Thread Starter

01-0077

Joined Jan 12, 2021
30
So you actually prefer to use 1 inverter chip, 5 resistors and 5 transistors, over the one chip solution?

Now don't get me wrong, I'm not trying to convince you of anything, especially since this board seems more like a competition then anything else sometimes.

I'm just curious.
I hadn't discounted your advice, I am simply overwhelmed. This is my first attempt at designing a circuit and as you can tell I am on about an 89 degree learning curve. As the discussion was around BJT's and MOSFETS I just ran with that. I don't know what a non-inverting high side load driver is, let alone the circuitry required to drive it, but I am all ears.
 

Thread Starter

01-0077

Joined Jan 12, 2021
30
Would anybody like to share the math behind calculating the base resistor, given collector current of 231mA and an appropriate PNP transistor (2N4403 as suggested by Audioguru?).

Does this calculation change if using a darlington PNP?
 

Thread Starter

01-0077

Joined Jan 12, 2021
30
First pass at the circuit I described in post #12, an all-MOS solution. The inverters can be any CMOS inverting device with at least 5 sections, such as the CD4069.

This is an edit of something from another thread, so ignore the LED specifics, MOSFET part number, etc. and use what you have / can get / whatever.

The U2F input can be tied to anything, just not left floating. Also, the circuit needs two power supply decoupling capacitors.

ak
View attachment 230489
Thanks AK, this looks great but I have no idea of how to pick an appropriate p-channel mosfet, any suggestions?
 

Audioguru again

Joined Oct 21, 2019
6,672
Since this thread changed the LED current 3 times and talks about using a bipolar transistor, a Mosfet or a darlington then I am leaving this totally confusing thread.
 

Thread Starter

01-0077

Joined Jan 12, 2021
30
Since this thread changed the LED current 3 times and talks about using a bipolar transistor, a Mosfet or a darlington then I am leaving this totally confusing thread.
Fair enough, I did screw up on the LED current and I apologised for that. I have been thrown various options regarding the switching side of the circuit and am only trying to understand these options. In no way did I intend to upset you and I do appreciate the assistance and advice you have offered.
 

ElectricSpidey

Joined Dec 2, 2017
2,757
I hadn't discounted your advice, I am simply overwhelmed. This is my first attempt at designing a circuit and as you can tell I am on about an 89 degree learning curve. As the discussion was around BJT's and MOSFETS I just ran with that. I don't know what a non-inverting high side load driver is, let alone the circuitry required to drive it, but I am all ears.
There is no circuitry needed to drive it, that is why I suggested it.

A non-inverting high side driver means when the input is high the output is also high and when the input is low the output is "off" not low.

You simply connect the outputs from the 4017 directly to the inputs of the drivers and connect the LED modules directly to the outputs.

One chip contains 8 channels, you would be using 5 channels, and tying the other 3 inputs to either ground or positive, in fact you can simply leave the unused channels open.

CE020 MIC2981 Jun05 (microchip.com)
 

AnalogKid

Joined Aug 1, 2013
10,986
Maybe it's time to reboot.

Being stuck with common-cathode LEDs, and using the CD4017, reduces the large number of circuit options way down. But there still are enough that there is no one "best". The driver transistors (or multi-section device) can operate as either a saturated switch (a limiting condition of a common (emitter or source) amplifier, or an (emitter or source) follower, a non-saturating current amplifier. So there are four possible solutions, a 2x2 grid of bipolar-or-MOSFET and switch-or-follower. Wait -- *plus* the side issue of device packaging, since some of the options are available in integrated packages in addition to discrete parts. I can argue both the superiority and inferiority of each option, but it's not always about me.

Sometimes, the "best" approach is not about the circuit:

Where are you located?

Is maximum brightness critical?
What access do you have to electronic component distributors?
Surface mount or through-hole?
Home project / one-off / prototype / production run?
Cost pressures or limitations?

ak
 

Thread Starter

01-0077

Joined Jan 12, 2021
30
Sometimes, the "best" approach is not about the circuit:

Where are you located?

Is maximum brightness critical?
What access do you have to electronic component distributors?
Surface mount or through-hole?
Home project / one-off / prototype / production run?
Cost pressures or limitations?

ak
Good point ak, to answer your questions I live on the west coast of Australia (Perth), brightness isn't critical but the brighter the better, 2 local distributors with limited component selection, through hole, home project, cheaper is good.

I contacted my local distributors they don't stock the MIC2981 and the p-channel mosfets they have are all rated in the 10's of amps. They do have NE555, CD4017, CD4049 and a small range of PNP transistors. I can order online but at present the freight cost and delivery times are crazy. Given this, maybe the circuit you posted at post#35 modified to use PNP's and base resistors is my best option?? If so, I have attached datasheets for the lower IC rated (500-1500mA) PNP transistors that are in stock at my local distributor, hopefully one of these will work.
 

Attachments

AnalogKid

Joined Aug 1, 2013
10,986
The original (and more common) part number for the source driver IC is UDN2982. See if your distributor can get it.

For discrete transistors, to keep the base current down in the CMOS range you'll need a Darlington type.

TIP115, 116, 117 - ?

Any power darlington with a current rating of at least 1 A. 2A is better, 5 A is overkill.

Likewise, any p-channel MOSFET with a current rating of 1 A or better. Depending on the distributor, these might be considered a signal part rather than a power part.

ak
 

Thread Starter

01-0077

Joined Jan 12, 2021
30
Hi AK,

Spoke to my local store, they have the BD682 darlington in stock but it's 4A. They cannot get the UDN2982 but may be able to source the TIP115 and p-channel mosfets in the 1 to 2A range, hopefully they will have an answer for me tomorrow. Normally they only do special orders for quantities in excess of 1000 units which doesn't leave me feeling overly confident.
 

Attachments

sarahMCML

Joined May 11, 2019
363
Hi AK,

Spoke to my local store, they have the BD682 darlington in stock but it's 4A. They cannot get the UDN2982 but may be able to source the TIP115 and p-channel mosfets in the 1 to 2A range, hopefully they will have an answer for me tomorrow. Normally they only do special orders for quantities in excess of 1000 units which doesn't leave me feeling overly confident.
Personally, I'd go for a CD 4049 or 4069 and some p-channel mosfets. By far the simplest!
 

Thread Starter

01-0077

Joined Jan 12, 2021
30
Ok, they can order me some p-channel mosfets and have given me a few options, hopefully one of these will work (datasheets attached). Let me know which one you would recommend (if any) and I will place the order.

AK, is it then just a case of substituting the selected mosfets and CD4049 into the circuit you posted at post #35? You also mentioned 2 decoupling capacitors. My original circuit had a 47uF in front of the CD4017, would I keep this and add another one in front of the inverter or something else?

Thanks again,
John.
 

Attachments

Thread Starter

01-0077

Joined Jan 12, 2021
30
I know but it's all I can get locally, 5x 200V cost $12 AUD, 5x 400V cost $18 AUD. If I order from Mouser, International shipping to Australia for 5 mosfets is $24 plus the cost of the mosfets.
 

Thread Starter

01-0077

Joined Jan 12, 2021
30
Not quite, Element14 (Farnell) are located in Sydney NSW (4000 klm's from me) and they charge $16.50 delivery. Next day delivery is only for people living in NSW. With state border lockdowns due to covid I would be lucky to see that order any sooner than 2 weeks.

If the IRF9530 (or any other mosfet) is more suitable then I am happy to order online.
 
Top