going out on a limb and praying the schematic is correct

Ian Rogers

User Extraordinaire

Forum Supporter

Most Helpful Member

• Friday at 2:11 AM

• #21

I know I'm always late at these parties, but there is a serial chip that you may find a lot easier.
It has 32 segments and current control. Its a MAX 7219 and they can be connected as one.
So you will have 96 LEDs' with just three devices.

Just a thought.

 

Upvote 0 Downvote

MrDEB

Well-Known Member

• Friday at 4:54 AM

• #22

Looked at the data sheet. Looks promising.
THANKS

 

Upvote 0 Downvote

MrDEB

Well-Known Member

• Friday at 8:26 AM

• #23

I recall I went down this road and sevral members suggested the 74hc595 would be a better choice.
After looking over data sheet etc. I have to agree that the 74hc595 is a better choice.

 

Upvote 0 Downvote

T

tumbleweed

Well-Known Member

Most Helpful Member

• Friday at 9:22 AM

• #24

Don't ground the QH' outputs

As to the rest of it, whatever floats your boat. Why you have some CLK pins common and others not I don’t follow... you will have to deal with them differently

 

Upvote 0 Downvote

MrDEB

Well-Known Member

• Friday at 9:42 AM

• #25

The reason I grounded the OH outputs is I was doing some research into issues with using the 74hc595.
The article suggested grounding or taking unused pins high.
I need to review my schematic for misconnections etc.
IF I get to the point of having boards made will be happy to share. My plan is to have JLPBC assemble the front (all the LEDs)
Thanks for your guidance

 

Upvote 0 Downvote

MrDEB

Well-Known Member

• Friday at 9:47 AM

• #26

I checked the CLK pins and they are all connected?
oups found one after I highlighted the net
thanks

 

Upvote 0 Downvote

T

tumbleweed

Well-Known Member

Most Helpful Member

• Friday at 9:51 AM

• #27

Do not leave unused INPUTS floating/unconnected.
Unused outputs can be left unconnected.

You have two of the HC595 using separate CLK6 and CLK8 signals, while the rest all use CLK1.
I don't know if that's intended or not

 

Upvote 0 Downvote

MrDEB

Well-Known Member

• Friday at 10:00 AM

• #28

I corrected them
Will remove the ground from the OH pins

 

Upvote 0 Downvote

T

tumbleweed

Well-Known Member

Most Helpful Member

• Saturday at 7:20 AM

• #29

You may want to think really long and hard about how you have the 595s connected and what that does to the code.

There are a number of different options, each with different advantages and disadvantages.

I'm not sure what you're looking to accomplish with the 595s vs all your other led projects

 

Upvote 0 Downvote

be80be

Well-Known Member

• Saturday at 9:53 AM

• #30

https://g.co/gemini/share/00204c305957 Try it out
You could twinkle all you want

 

Upvote 0 Downvote

be80be

Well-Known Member

• Saturday at 12:12 PM

• #31

Here the link to where the AI tells all about using 74HC595

Gemini - 74HC595 Chaining and Bit Shifting

Created with Gemini

g.co

 

Upvote 0 Downvote

MrDEB

Well-Known Member

• Yesterday at 3:27 PM

• #32

Questions that maybe need to be addressed?


WHY 10 – 74HC595's I wanted to avoid the “curtain effect” with the LED's turning ON in sequence from the #1 LED to #80. If it hypothetically takes 2 seconds to cycle through each shift register, then 16 seconds from #1 to #80





WHY USE 74HC595 SHIFT REGESTERS?


They were suggested for a better project. Far from the first Christmas stars using 2 – LED's in parallel with 1- 500 ohm resistor.





IN MY LATEST SCHEMATIC, ALL THE CLOCK INPUTS ARE CONNECTED TO A COMMON PIC PIN. ALL THE LATCH PINS AND DATAINS ARE SEPERATE WHY?


IMO this should make programming easier

 

Upvote 0 Downvote

be80be

Well-Known Member

• Yesterday at 3:34 PM

• #33

2 seconds to cycle through each shift register where you read that at you could clock out all 90 bits faster then you could see it happen
less then 8 ms for 90 bits

 

Last edited: Yesterday at 3:47 PM

Upvote 0 Downvote

T

tumbleweed

Well-Known Member

Most Helpful Member

• Yesterday at 3:46 PM

• #34

MrDEB said:

IN MY LATEST SCHEMATIC, ALL THE CLOCK INPUTS ARE CONNECTED TO A COMMON PIC PIN. ALL THE LATCH PINS AND DATAINS ARE SEPERATE WHY?


IMO this should make programming easier

Click to expand...

If you think that then you haven't tried to write the code yet to do this.

As far as 2 seconds for each shift reg, that's only if you add all sorts of delays where they shouldn't be

 

Upvote 0 Downvote

be80be

Well-Known Member

• Yesterday at 3:57 PM

• #35

@ tumbleweed swordfish basic is now free are you still going to keep it going.
and the USB stuff works with the free swordfish basic

 

Upvote 0 Downvote

F

For The Popcorn

Well-Known Member

Most Helpful Member

• Yesterday at 3:59 PM

• #36

This is freakin' painful to watch. MrDEB has reached a new stratosphere of clueless.

 

Upvote 0 Downvote

Papabravo

Well-Known Member

Most Helpful Member

• Yesterday at 4:34 PM

• #37

And, how are those prayers for the schematic being correct working out for you?

 

Upvote 0 Downvote

Beau Schwabe

Active Member

• Yesterday at 5:13 PM

• #38

The 595 shift register can shift out at 100MHz ... that's faster than most processors could bit bang the data and definitely faster than an interpreted code processor running from BASIC could shift it out.

The 595 is not your bottleneck. IOW... you don't really even need delays in your code between clock pulses for the 595 to function just fine.

If it is taking 2 seconds to shift data out, then something is incredibly wrong in your setup.




Reference: https://www.digikey.com/en/products/detail/nexperia-usa-inc/74HC595PW-118/946653

 

Upvote 0 Downvote