Help with 74LS170 4x4 Register

[JHarmon]

I've traced a problem in a more complex circuit down to the 74LS170 registers (Datasheet).

I swear I have tried almost everything with this chip. I have tried 3 different orders of the chip, so I very highly doubt that the chip is faulty.

I made a separate circuit on a different breadboard, and just tried to write to the chip and read the word back. Apparently I am misreading the datasheet, because I cannot even do that.

I have the addresses set to 0,0. I have Read enable at 0V and i have a double-pole toggle connected to the write enable (which I toggle back and fourth when I want to write data... violently, of course, because I am becoming frustrated). I have the D1-D4 pins set to 5V, 0V, 5V, 0V, and the Outputs are all connected to LED's. So, yeah... Am I understanding the function of this chip? Or am I doing something horribly wrong?

Yes, I have double, triple, quadruple checked the pin numbers.
Yes, the LED's are going the right direction.
Yes, it is plugged in.
So confused...

 

[gorgon]

First of all, working with a LSTTL chip, never connect 5V directly to inputs. Always use a pullup of some kohms, like 2k2 or 4k7 to Vcc (5V). No problem connecting 0V directly.
Try to change that and see if the chip starts working.

When toggling inputs with a switch, I suppose you know that each toggle makes a lot of pulses, due to contact bouncing. This should not be a problem for you, but will be if you use the switch to clock something, like a counter.

TOK

 

[JHarmon]

Well, I added the pullups (thanks for the tip, that might explain why a lot of these more complex 74LS chips aren't working correctly)... Also, I technically have the toggle connected to a monostable circuit with a 555 timer, just to make sure

Anyway, I still can't read/write a word... Rather frustrating since a project I'm working on revolves around these chips...

Checked it with a meter: still nothing... (I'm stumped, so I'm just trying things now!)

 

[gorgon]

Sketch a schematic of how you have connected the chip, and explain in detail how you operate it.
You say you connect LEDs, how do you do that? I suppose you know the outputs will sink more current than it will source?

BTW You are sure the chip is in working order?

TOK

 

[JHarmon]

I uploaded a simple little pinout.

Now when I flip the toggle around a little, it should write 0,1,0,1 to the flip flops, and then my data-outs should go to 0V, 5V, 0V, 5V, correct?

I just have the LED's connected to ground (correctly), but I checked it with a meter, as well, and none of the outputs went to 5V...

I have checked pretty much everything with a meter, including the toggle, to make sure that it is going to 0V.

 

[JHarmon]

Also, I have tried several chips, from 3 separate orders and manufacturers (don't ask, I counted them incorrectly in my schematics a couple of times ), so I'm pretty certain that at least one of them was working.

 

[gorgon]

Ok, Just had a look at the datasheet. You need to add pullup resistors on all the outputs, like 2k2. The outputs are open collector and will only pull to 0V, If you connect your LEDs via a resistor from Vcc to the output, you will see an inverted output. Light when '0' and off when '1'.

TOK

 

[KrisBlueNZ]

gorgon, where did you get this advice that LS inputs shouldn't be connected directly to VCC? I haven't heard it... AFAIK it's fine to do that. What is the reasoning?

 

[gorgon]

gorgon, where did you get this advice that LS inputs shouldn't be connected directly to VCC? I haven't heard it... AFAIK it's fine to do that. What is the reasoning?

It is a standard warning in all TTL and LSTTL design books. The reason is the construction of the input transistors. If you hardwire the input to Vcc you will backfeed the input transistor and a corruption of the logic and increased power consumption may occur. If you look at the signal drawings in the datasheet it's referenced to 3V not 5V. This is due to the input current all TTL devices has.

A normal TTL pullup is in the 1k region, and the input high current is 400uA. For the LSTTL the current is lower and you may increase the resistance a bit, depending on the noise level, but I've seen 2k2 - 4k7 as a normal range.
The sink current into the input is much higher, like 1.6mA. This is why a pulldown is not very practical in LS/TTL circuits, and almost never used.

TOK

 

[KrisBlueNZ]

Interesting. Does it cause some sort of latch-up? I'll google some application notes.

 

[JHarmon]

You are a lifesaver haha

It has passed 3 AM here, and I can finally sleep. I thought I would never get it to work!

Thanks!