Author Topic: Inverts and Negative Voltage (Read 10018 times)

stampedeboss · « **on:** March 29, 2015, 06:42:45 pm »

Working on an LED Driver using PT4115, WS2811S and an inverter. On my breadboard I used a Renesas HD74LS04) and everything worked fine. For those that don't know, the WS2811S is a LED 3 channeol controller chip. It uses PWM on the negative side of the LED circuit.

Now for the question, I've tried to change to another inverter without sucess, I've tried a Toshiba TC7WU04FU and 74HC3GU04 without success. I don't understand why,it seems that the inverters don't react to the -4.9v output by the WS2811. Can someone help me better understand how inverters work? I thought I understood but I guess I don't. I also having issues determining the differences between the HD74LS04 and the others.

Any insights?

Zero999 · « **Reply #1 on:** March 29, 2015, 06:58:45 pm »

Perhaps it's because it uses TTL levels?

Try the 74HCT3G04
http://www.nxp.com/documents/data_sheet/74HC_HCT3G04.pdf

stampedeboss · « **Reply #2 on:** March 29, 2015, 07:04:50 pm »

I tried a 74HC3GU04 without success, what the difference?

mikerj · « **Reply #3 on:** March 29, 2015, 07:17:18 pm »

The WS2811S uses a single positive supply, why do you think you are you getting -4.9v out of it?

A schematic of your circuit would greatly help.

IanB · « **Reply #4 on:** March 29, 2015, 07:18:33 pm »

Inverters contain logical NOT gates. They turn a logic 1 into a logic 0 and vice versa. Not sure what you are trying to do here with mention of "-4.9 V"? That doesn't seem like a logic level (logic levels are typically 0 V = false, 5 V = true).

SeanB · « **Reply #5 on:** March 29, 2015, 07:24:22 pm »

I am going to guess the LS04 is happy with a floating input as being a TTL one level, the rest are not, they expect a valid logic one level, and if the output is not actively driven to 5V you need a resistor to pull it up to 5V. A value between 1k to 4k7 will work, depending on the capacitance and what is also attached to the pin. If it is a long string of LED's then 330R to 560R will work, but if it is just the gate alone being driven direct the 4k7 will work.

TTL treats any voltage above 0.8V as logic one, and real TTL provides a pull up current by it's design. CMOS gates need it to be over 2V4 ( or 3V for some CMOS) and do not source any current from inputs to pull it up if it is an open collector ( or drain) output, they need the resistor to 5V or an active pull up to work.

Zero999 · « **Reply #6 on:** March 29, 2015, 07:27:59 pm »

Quote from: stampedeboss on March 29, 2015, 07:04:50 pm

I tried a 74HC3GU04 without success, what the difference?

Read the datasheets.
http://www.nxp.com/documents/data_sheet/74HC3GU04_Q100.pdf
http://www.nxp.com/documents/data_sheet/74HC_HCT3G04.pdf

stampedeboss · « **Reply #7 on:** March 29, 2015, 07:49:15 pm »

Thank you that helped.

As you can see from the attached schematic, the inverter sits between the ws2811s and the dim input on the PT4115B. 3 PT4115B to be specific, one for each color. The PT4115 uses DC voltage or a wide range of pulsed dimming. Applying a voltage of 0.3V or lower to the DIM pin turns the output off and switches the device into a low current standby state.

Supposely, the WS2811 outputs PWM, I've really struggled with understanding the WS2811 output since from my reading the datasheet focuses on the communications protocol with the LED controller.

I see the TTL vs CMOS differences, I don't see where the HS74LS04 is a TTL device since it worked.

stampedeboss · « **Reply #8 on:** March 29, 2015, 07:53:41 pm »

Quote from: IanB on March 29, 2015, 07:18:33 pm

Inverters contain logical NOT gates. They turn a logic 1 into a logic 0 and vice versa. Not sure what you are trying to do here with mention of "-4.9 V"? That doesn't seem like a logic level (logic levels are typically 0 V = false, 5 V = true).

Because the only changes reflected in state are when the volt meter is reversed, positive to negative. Then I see a shift from near 0 to 4.9v. Wire + to + shows no changes, reads the same 19ma regardless of on or off.

That may just be my lack of understating of negative voltages.

Zero999 · « **Reply #9 on:** March 29, 2015, 07:56:58 pm »

No, the voltage never goes below the negative rail. It's just the operation is reversed so whenever the output of the WS2811 is 5V the output from the inverter will be 0V and vice versa.

As SeanB said, TTL inputs are happy left floating and just need to be pulled low to activate. The CMOS inputs need to be tied to either VDD or VSS, not left floating.

The WS2811 has open drain/collector outputs so would've worked with the TTL LS inverter but not the CMOS inverter. To use it with the CMOS inverter, just connect each output of the WS2811 to the same positive supply voltage as the CMOS inverter, via a 10k resistor.

SeanB · « **Reply #10 on:** March 29, 2015, 07:59:00 pm »

You need pull up resistors between the outputs od IC2 on pins 3,2 and 1. Any value from 1k to 10k will do, pulling the inputs up to 5V. That way the CMOS inverters will work exactly the same as the 74LS TTL ones.

stampedeboss · « **Reply #11 on:** March 29, 2015, 08:04:52 pm »

First of all THANK YOU, I've been trying to figure this out for days. This all now makes some sense.

While I'm at it, why doesn't the PWM out from the WS2811 effectively drive the DIM on the PT4115? The driver claims PWM control and the WS2811 shows PWM output.

I've tried that before heading down this Inverter track, I have also got it to work using 2n3904 + 4.7K pull ups between the WS2811 and PT4115.

SeanB · « **Reply #12 on:** March 29, 2015, 08:12:11 pm »

Same reason, no pull up resistor so it pulls low and stays low from board capacitance.

stampedeboss · « **Reply #13 on:** March 29, 2015, 08:13:59 pm »

So, in theory add the resistors and eliminate the inverter should work.......

SeanB · « **Reply #14 on:** March 29, 2015, 08:21:42 pm »

Yes, provided the connecting cable is not too long. Just remember that the output of the control IC and the step up converter are different, you will have full brightness from the LED's at zero brightness from the WS8211, and vice versa. Output is active low on the WS8211 and the PT4115 expects active high for PWM, so you might still want to have the inverter there to make the code easier.

stampedeboss · « **Reply #15 on:** April 01, 2015, 08:59:21 pm »

I'm back and even more confused.

Status: received a number of the 74HCT3G04 inverters and the circuit is still not working. I've even tried pull up/down resistors without success on both HC and HCT versions of the inverter.

With Pullup - Always on
With Pulldown - Always off
Inverter from WS2811, always on.

Measurements using a voltmeter on the WS2811 output pin show it 1.4v when that color is to be off, and .02v when its to be on. This is where I'm really confused. I understand the WS2811 as "low" side of the circuit, since it is attached to the negative side of the LED. If I attach my positive probe to a +5v source and the negative to the WS2811 output pin I get solid readings of 4.9v and near 0 with 4.9V being on the "on" state.

stampedeboss · « **Reply #16 on:** April 02, 2015, 05:00:20 pm »

No one?

AJ Reynolds

Zero999 · « **Reply #17 on:** April 02, 2015, 07:23:45 pm »

What value of pull-up resistor did you use?

How long is the cable?

Is that 1.4V measured with the pull-up resistor in place? If not then that's what you should expect. The LEDs are dropping 3.6V at the tiny current taken by the 10M input resistance of the meter.

stampedeboss · « **Reply #18 on:** April 02, 2015, 07:54:38 pm »

I tried 1k, 4.7k, 10k,100k On bread board.

1.4v no pullup, no cable on a pcb trace 320 mil wide less than 3/8 of an inch. No Led at the time of measurement.

This where I'm now confused. Data sheet shows max of .8 for low but typically of 1.2. On the other inverter which shows max .8 at .9 inverter flips. Shouldn't 1.4v with max of .8 switch to high?

Right now can't use the new inverter on breadboard because it's smd.

AJ Reynolds

Zero999 · « **Reply #19 on:** April 03, 2015, 09:03:38 am »

Please post a schematic.

stampedeboss · « **Reply #20 on:** April 03, 2015, 06:16:17 pm »

Currrent sch from eagle. From what I've read my voltage is negative-going and is discussed a little in one of the TI datasheets. Not sure how this effects the other inverters.

Zero999 · « **Reply #21 on:** April 03, 2015, 06:56:30 pm »

Where are the pull-up resistors?

stampedeboss · « **Reply #22 on:** April 03, 2015, 08:31:49 pm »

Sorry this design didn't have them. It is what worked on my breadboard using the hex inverter 74ls. When I got the boards and tried to use various smd 3 circuit inverters unsuccessfully. LED is always on.

I went back to the breadboard to test the pullup resistors. I'll update and repost.

AJ Reynolds

stampedeboss · « **Reply #23 on:** April 03, 2015, 08:39:54 pm »

Quote from: stampedeboss on April 03, 2015, 08:31:49 pm

Sorry this design didn't have them. It is what worked on my breadboard using the hex inverter 74ls. When I got the boards and tried to use various smd 3 circuit inverters unsuccessfully. LED is always on.

I went back to the breadboard to test the pullup resistors. I'll update and repost.

AJ Reynolds

I've read that the WS2811 can only be pulled to ground since its constant current, which backs my testing on the breadboard. Looks like I have a bunch of coasters, short of trying every available Inverter in TSS0P8 since I can't figure out why the dip version inverter works and the SMD ones do not I'm not sure what to do. As you can see from the layout there really isn't any room for resistors. Here's the schematic:

Zero999 · « **Reply #24 on:** April 03, 2015, 10:10:55 pm »

The drawing with the pull-up resistors should work.

Did you try soldering the pull-up resistors between the inverter inputs and +5V on the PCB?

I doubt you'll be able to find the old LS parts in SMT.

To understand the differences, you need to look at the internal schematics of the logic gates.

TTL

CMOS


EEVblog Main Site	EEVblog on Youtube	EEVblog on Twitter	EEVblog on Facebook	EEVblog on Odysee

Author Topic: Inverts and Negative Voltage (Read 10018 times)

Share me