Daisy-chaining 7 IC's

Bend component leads if they are long enough to reach another component.

Just use the strips themselves for low-current signals.

Use the strips but run a bare (thick) wire as well and solder it all along the strip for high-current.

Use wire for any connections that don't fit the above.

BTW, the best way to cut strips is a 1/8th" bit in a pin vice. Just drill enough to cut the track.


Rob

Annnnnd scratch that. Broke again. Just the 5th chip. Since the 5th is on a different breadboard I had to use longer wires and that seems to have killed it again. Before it was connected directly but I was gunna run out of wire and space to plug stuff in. Maybe if I tore off the little +V and GND rails on the sides of the breadboards and used those..

I can't adjust the SPI clock -- I think the library does it at some point by its self. I decided to try "bitbanging" instead SPI just to see if it'd work (It says it's slower), but that seems to be too slow for the TLC's to accept.

Man, these things are /really/ sensitive..

but that seems to be too slow for the TLC's to accept.

SPI doesn't normally have a minimum speed.

Did you try using shiftOut()?

I think the library does it at some point by its self.

Why do you say that? Looking at the source code I can see nowhere that is changes the speed except in the setClockDivider() function.

EDIT: You're not constantly calling SPI.begin() and SPI.end() are you?


Rob

Didn't try shiftOut(), and I really have no idea anymore. My head's fuzzy, I've been at this for weeks now so I'm starting to skip over simple stuff (see hz/mhz confusion) and it's four in the morning so that ain't helping much.

I'll work on it some more later. I just gotta figure out some way to get the chips to stay in the breadboard now. Seems like taking two of the power rail bus type things off the breadboards and using them is working. Maybe. I hope.

Sounds like its time to firm up the assembly.
Get some perfboard, some wirewrap sockets, and wire up a real assembly.
Wirewrapping is nice, goes quick, easy to correct if you miswire something or need to make a change.
This board only took an evening to build up for example.
You can see the '328 & crystal/caps that make up an arduino, and some shiift registers above it.
On the left is a MIKROE483 FTDI module for USB interface to a PC.
Add your TLC5940s and LEDs next to them, be an easy wiring job.

Where do I get that ribbon cable? I've been looking all over for that and the best I can find is discontinued IDE cables at Radioshack that I'm running out of. x.x

I think that particular cable was from here

crimp on connector

If you need wider connectors, you can get them here, I bought 14s & 16s recently.
http://www.king-cart.com/phoenixent/product=RECEPTACLES-SOCKETS+IDC+.1SP+%2526+2MM+SP/exact_match=exact
Can be found at digikey, mouser, pain in the butt locating them.

Blah. That bus idea ain't working. Is there a device that exists who's only purpose in life is to accept a signal and repeat it again to get rid of line noise? Given what I have on hand and my abilities, I really don't see a way to avoid using one.

OK Crossroads, you have got me convinced to try wire-wrapping for prototypes ( after wasting 2 hours today finding an intermittent fault between tracks of a vero ( perf ) board ).

My reluctance has been due to my having worked on TVs in the 60s with first tagstrip wiring, then wire-wrapping, and eventually printed circuit boards !

I still recall a billboard advert in England proudly stating that Murphy ( I think ) was the only hand wired TV made !!!

Go for it Boffin - I bet it would take care of Nofew's problems too.

I am going to look into it tomorrow, after seeing some of your boards !

What is wire wrapping, exactly? I'm entirely self-taught here and pretty much everything up to now has been done by guess-and-check. School didn't even touch on this stuff so I'm sorry if I'm asking a lot of really basic questions.

Wirewrap sockets are mounted on a card, then wire is wrapped around a pin and connected to another pin.
I use socket strips like these,
http://www.king-cart.com/phoenixent/product=SOCKETS+WIRE+WRAP+DIP+%2526+SIP/exact_match=exact
or sometimes sockets if I have them the right size. Usually just lengths of strips tho.
Solder down the 2 ends, maybe a pin in the middle if its a longer length.
I use the Modified hand wirewrap tool, has a built in stripper in the handle. Slide the wire thru ~3/4", pull it down to catch the cutter, and slide it out.
Have a little pair of wire cutters to cut the wire to length, and a small pair of needle nose pliers to hold the wire as I start twisting to keep it from getting tight around a corner when spacing is tight enough that I can't just press a finger on the wire.
End result is nice neat board, very secure connections, quick to assemble, easy to make changes or fix mistakes.
I used to buy rolls of 100' 30 guage wire here & there, recently bought four 500' rolls to take advantage of the economy of scale pricing.
Thought I had lost my wirewrap tool also (had it since early 80's), so I bought one each both on this page - and then found my original, so now I have 3.

Interesting, owo. Might have to go do that soon..

I have good news and bad news. The good news is, it's not the wires! Yay!

The not-so-good news is I found out by connecting the chip directly to the Arduino. Using the same wires that control the first four chips just fine, I put them to the fifth chip, and it didn't work.

I know from experience that any chip I put there won't work, so the breadboard must be at fault. I have no idea how or why -- I used a multimeter to test all the strips and they work fine. I didn't have to stick the probes in far (they didn't even fit into the holes), so it doesn't make much sense to me.. Though the chips do pop up, maybe it has to do with that even if they're stuck down.

Regardless, point is the issue can be tracked down to me having a faulty breadboard. Is there any way to fix this or should I just grab a new one?

Have you checked continuity of the +5v and ground buses on the breadboard? On some breadboards, they have a break half way along.

Been there , done that..

in my case, ( early on, when I rushed into things ) I had the ground return for all the LED driver chips daisy chained, so that there was a voltage drop when I connected ( in my case ) more than 3 displays. This voltage drop confused the data levels to the chips and gave odd results.

Just running a separate wire linking all the ground pins of all the driver chips solved it.

Hope it helps you....

Breadboards aren't the most reliable things even when they are good. If you have bad one I'd bin it, look how much time it's cost you so far. But try another one first, just in case :slight_smile:


Rob

point is the issue can be tracked down to me having a faulty breadboard.

I would rethink your approach. Rather than building a giant thing the first time, you may want to experiment with one module and then chain them together. This approach breaks down the tasks to more manageable levels and allows you to build up your board gradually.

@dc42: Yes, I have. They run all the way.

@boffin1: I'm doing what you said (the second one, when it worked) -- Have been from the start, but thanks for at least telling me not to try daisy-chaining them.

@Graynomad: I tried another, now it's being even weirder. The chip works. I know this because it's responsible for eight green and eight blue LED's, and the eight green ones work flawlessely. However, the eight blue ones have gone nuts. They're very dim (normally they're the brightest on the display) and only turn on fully once in a blue moon (excuse the pun). However, if I turn only a few LED's on at a time, they work. I don't know what to think -- The PSU is more than capable of running every LED at once (it could run six of these displays at the same time), they're grounded, wires are in just fine.. I don't know what can be causing this. I even added a 6th IC to see if those blue ones wouild mess up too, and it just gets weirder. The next 11 do, the the last 5 are fine. I'm thinking about scrapping the project entirely and designing a PCB instead.

@dhenry: I've been doing that.