Double check my soldering? SN74HC595N x 5

Hi there,

I'm working on connecting five SN74HC595N shift registers in series, and before I start moving forward I was wondering if someone would be able to tell me if I'm soldering them together correctly?

I'm trying to connect the pin 8 (GND) and Pin 13 (OE-) to the common ground I'm setting up on row 1. I'm using row 10 as a common 5v, connecting pin 16 from each of the registers to that row.

I hope this makes sense... I plan on connecting several LEDs to a single board and I wanted to make a mini "RAM stick" that I can use to control them. I haven't soldered anything but the 5v row and ground rows yet.

I've attached some images of my work, and a quick diagram I made up, for whoever wouldn't mind giving them a look. I'm new to soldering, so please don't judge me too harshly! Lol

Thank you!



You need a 0.1uF ceramic capacitor between the power and ground of every shift register chip. You can solder them on the back to make the leads from them as short as possible.

Ah mm.. It looks lika baking. That can be riscy, tinn floating where it should not be.
You could use a tiny, one strain, copper wire that You prepare for the soldering path. Put it in place and solder one end first and the other when the first soldering has cold down.
Or use isolated tiny copper wire wire to make soldering path. Wire wrap cable could work.
To each such board I would put one bulk capacitor, tantal, of some 10-47 micro Farad at the Connection Point of incoming Vcc. As close as possible to + of each circuit I would 4.7 micro Farad tantal.

No using a tantalum across a supply it a bad idea, they can fail short and start a fire. UL certification forbids such a thing.

Asking the wiring, especially the power supply connections, thinner is a bad idea.

Maybe 0.1 ceramics at each circuit is preferable. I don't know how spiky the 595 is. Such ceramics were mandatory for circuits like dynamic memorys that were rather nasty regarding power consumption spikes. I used tantalums for many years in the past and they didn't fail during 20 years. Anyway, use some bulk capacitor for the board nere the incoming Vcc.

Such ceramics were mandatory for circuits like dynamic memorys

They are compulsory for all digital ICs

Here is why:-
http://www.thebox.myzen.co.uk/Tutorial/De-coupling.html

Yes, G_M, that is what reality is. Ceramic capacitors are "faster" to deliver current then electrolytic ones. Different types of circuits have differently sharp current consumption spikes. A latching device like the 595 probaly produce current spikes so the ceramic cap is preferable. Also note the bulk capacitor for the board.
Your info about hazards using tantal electrolytic capacitors I have not heard of. They have served me well.
Regarding "baking soldering" I think that using a copper wire, some 1/2 mm thick or so would be a more safe way to keep the tinn away from swimming all over.

Thanks for all the feedback, I'll start tweaking my design!

Starting and testing things in a small scale, verifying that Your theories work out is really good. Make the project advance one step at the time.

Suggest you use good machine pin or dual beam sockets for the I.Cs.

Ceramic capacitors are "faster" to deliver current then electrolytic ones.

That is not true. Did you not read the link? It is all to do with the self resonant frequency of the capacitor caused by the inductance of a capacitor.

Yes. In case of any short or overload, using sockets for the 595 makes replacing a damaged 595 much more easy. I once filled a 19 inch rack with 19 boards, most of them selfbuilt using wire wrapping and sockets. Really a good suggestion for a long time working design.

@G_M
I'm sure Your facts are true. I refer to "good design" 40 years ago and it has served me well during 30 years of designing quite advanced digital electronics. How often does a designer step down to frequency analysiz of spikes? Maybe in failing space projects or military ones. Working in both those fields I haven't heared those questions.

I also miss a means of connecting whatever is to be connected to the output of the shift registers. No space to solder in even a basic pin header.

Railroader:
How often does a designer step down to frequency analysiz of spikes?

He dose it once at the beginning of his career and applies the results for the rest of his life.
Anyway it is not the frequency analysis of spikes, that is just silly, but the frequency response of capacitor types.

Maybe in failing space projects or military ones. Working in both those fields I haven't heared those questions.

I find it very hard to believe that. In fact I will go so far as to say that can not be true. You can not have 19 boards of digital logic work correctly or reliably without using ceramic capacitors.

Grumpy_Mike:
He dose it once at the beginning of his career and applies the results for the rest of his life.
Anyway it is not the frequency analysis of spikes, that is just silly, but the frequency response of capacitor types.

Most of us don't do this, ever. I for one haven't. I just listen to what people who actually did those things say, and apply their knowledge. Much easier that way. That applies for so many things in life, not just electronics. No use to re-analyse everything.

Anyway, 100 nF ceramic caps are really cheap, and it won't hurt adding them, so I'm simply sticking to the "IC needs cap" rule.

I just listen to what people who actually did those things say, and apply their knowledge......... so I'm simply sticking to the "IC needs cap" rule.

Nothing wrong in that.

Anyway, 100 nF ceramic caps are really cheap, and it won't hurt adding them, so I'm simply sticking to the "IC needs cap" rule.

Given how small 100nF capacitors are, and how essential, I wonder why no semiconductor manufacturer puts them inside the package for the chip. (Or maybe one tried and it didn't catch on).

You use to be able to get sockets with de-coupling capacitors integrated, however, different chips have different pinouts.

Caps have very large footprints compared to the i.c. die.

Given how small 100nF capacitors are, and how essential, I wonder why no semiconductor manufacturer puts them inside the package for the chip.

Most chips these days arn't that big. I am talking about surface mount parts. In the days when chips all came in DIL packages ceramic capacitors were quite large. Now surface mount capacitors can be very small and would fit in the old DIL packages but as DIL packages are in genral a think of the past it is not worth doing.

While 0.1uF is a standard value it is sometimes better to use smaller values especially if you have a lot of chips or want to reduce the cost. One place I worked when they wanted to trim the cost, they replaced all the decoupling chips with 27nF, saving about £0.03 on the bill of materials. Not much of a saving you might think but the production run was for 4 million units, so quite a saving in the end.