74hc595 noise problem ?

[omid_juve]

74hc595 noise

Hi
in our project we connecting 25 , 74hc595 to each other and shifting data on them
but suddenly we have some undesired result with the output of the 74hc595
i want to know what kind of noise consideration should i take into account for
connecting these numers of 74hc595 to each other ?
plz help me
Thanks

 

[xorcise]

74hc595 problem

Yes, there is a limit to how many devices you can drive, but with CMOS it can be quite a lot.

There is a matter of distance from the driver source for noise problems, but firstly did you decouple the power pins of the '595's with a 0.1ufd cap? Even a 0.01ufd can work.

 

[omid_juve]

i have problem in 74hc595

yes we use a .1uf cap on pin vcc to gnd i want to know if we should use any cap
on DS,SHIFT CLOCK or LATCH CLOCK ?

 

[Old Nick]

74hc595 spec

yes we use a .1uf cap on pin vcc to gnd i want to know if we should use any cap
on DS,SHIFT CLOCK or LATCH CLOCK ?

No, just decouple the power supplies. Are you building this on strip-board or a pcb? For better decoupling use a 47u 10u and 0.1uF caps at the power supply where it enters your board, and use 0.1uF cap at each power pin on the chips.

This does not sound like a noise problem though.

If you upload a schematic it may be possible to spot the problem. It is, as xorsise wrote, most probably a driving issue. Check your fan-out, check that you aren't mixing cmos with ttl. It may be that you need to place a few buffers in your circuit.

 

[omid_juve]

74hc595 constant current

we use them on a pcb and we have some sn74hc595 & m74hc595b1 that are mixed
in this design.
and if u want to know more about our project i can say that we construct a 16*200
led display and we use 74hc595 for putting the data on uln2003a and after the uln2003a it is connected to the columns .

 

[Old Nick]

cascading 74hc595

What is driving the inputs?
How are the chips connected?
what frequency are they working at (I assume it's very low, so probably isn't an issue).
It is imposible to give any more advice on what is wrong with your circuit (debug) without seeing at least the schematic. And exactly what the problem is, and where it appears. What do you mean by 'suddenly we have some undesired result with the output', did this used to work and then failed?

 

[omid_juve]

pic16f627a transient problem

please check the schematic of this design.
and about the undesired result i can say that when we have all the text printed on our
led display some of the dots undesirably blinking and when we use some cap on the power rails of the 74hc595 about the 97% of our problem is solved but still we have
a little blinking leds on our display. and so i sure about the noise effect on our led display.
thanks for all of your instruction until now
i still wait for your answer.

 

[PiCS]

74hc595 circuit

Put on last ic 74hc595(U25) pin 12(latchclk) to gnd 27pF and 1k resistor

 

[blueroomelectronics]

pic hc595

Have you calculated how much power you're going to need to light up that 16x200 array?
You'll be lighting upto 200 LEDs at any given moment, and with 1/16 duty you'll need to put a fair amount of current through each of them say 50ma so they're bright enough is 50ma x 200 = 10A

As for your circuit what's the purpose of the transistors on QH? Use a ULN2803 as its a hex driver.

What are you using as a row driver? It's gotta pass 10Amps

 

[omid_juve]

74hc595 pwm

why should i use 27pF and 1k resistor on pin 12(latchclk) to gnd on last ic
can u explain it for me ?

 

[syrax]

74hc595 debug

i think it is a vaiant to wire 2-3 MCU to speed up the sistem and one host to cotrol the slvaes

 

[x_zoli]

74hc595 for led displays

Hi, there are a few solutions to the problem:

1. You need decoupling for the HC595s, that means that you have to connect some 0.1u and 10u capacitors between Vdd and Vss. This is because of the high transient currents that occur when the outputs are switched, that can lead to erroneous outputs or even resetting the shift register.

2. Especially if you are using high current switched circuitry (like a bunch of multiplexed LEDs), there is a possibility that there are RF emissions, especially if you are using common emitter connections for the power transistors. These emissions can cause glitches on the digital lines that will trigger false clocks or latches of your shift register. This is the cause that you need to connect the RC as instructed above. You can leave out the resistor or even use a lower value, like around 50ohms. I also use the same config for HC595 in an industrial design.

3. The third, and I think the most important of your problems is the connection of the signals to the registers. You should be aware that all the inputs of the HC595s have some equivalent capacities that aren't that insignificant. When you connect 25 shift registers in parallel, and i mean the clock signals and the latch signals, you can end up driving a large capacitor with a single microcontroller pin. This can be possible, but at very small speeds because you literally have to charge the capacitors, and that takes some time. You could enhance your design by connecting some non-inverter buffers (HC245, for example), on the clock and latch lines, connecting together the input for 5 buffers and each buffer output should drive no more than 5 shift register inputs.

Hope it will work!

Best regards

 

[mojy_1]

hc595 simulation

you can use D flip flop(74ls74) for driving the last section of your circuit.

 

[xorcise]

74hc595 single input

There should be no issue to drive 25 74HC595's from a single PIC output. I have seen specs that show fanouts of 60 to 80 CMOS devices on a single driver. I believe Philips has spec'd a possible fanout of 30,000 for one of their CMOS IC's.

As blueroomelectronics said, check the effects of loading on your power supply. That will certainly create problems with lost/incorrect bits.

And I don't know if the question was asked, but, are you breadboarded or are you soldered to a PCB?

 

[omid_juve]

74hc595 computer control

it is soldered to the pcb

 

[blueroomelectronics]

driver 74hc595

it is soldered to the pcb

You built a PCB before you tested it? Always build a prototype of something as complex as the circuit you described here and never just use a simulation.

What made you think it would work? How many amps is your supply? And yes you MUST use 0.1uf caps near the power pins of ALL your ICs.

 

[syrax]

noise 74hc595

i think wneh data shifting with 16 led in column ~ 320mA i meen power comsumption is no to much

 

[Myself]

hc595 circuit

Hi Guys,

Why don't you consider using the IC named tpic6b595 as your LED driver. It consists of 74hc595 type shift register as well as power transistors.

I am also uploading the data sheet of this IC for you to have a look.

By using this you save space and money in your desings.

Take care.

Myself

 

[omid_juve]

74595 scan pwm

our project is cosisted of 10, 16*40 led display module
every module has its own row driver and 5 modules of them have 74hc595 and uln2003 as a column driver
and we test the 16*80 module on a breadboard and we don`t have any problem with any thing so we decided to construct the pcb for the whole project . and about the tpic6b595 and uln2803 the cause that we don`t use them is the price .
and the power supply that we use now for testing the
whole modules is the computer power supply
and the whole current that our led display consumes is about 4A with all the leds on .

 

[omid_juve]

hc595,trigger

our project is cosisted of 10, 16*40 led display module
every module has its own row driver and 5 modules of them have 74hc595 and uln2003 as a column driver
and we test the 16*80 module on a breadboard and we don`t have any problem with any thing so we decided to construct the pcb for the whole project . and about the tpic6b595 and uln2803 the cause that we don`t use them is the price .
and the power supply that we use now for testing the
whole modules is the computer power supply
and the whole current that our led display consumes is about 4A with all the leds on .