Prompted by a comment in another thread, I've set about combining the two boards I've been working on modernizing their design, into a single board...

I'm very close to finishing the redesign, to the point I'm validating the parts' footprints on the board and checking on pricing.

Had a bit of a surprise when I priced one IC and want to see if the equivalent I've worked up will do the same thing.

The original circuit uses the 74LS348, which looks like it's >$6 a pop, while one of the 74LS368 ICs on the board had an unused section... if I understand the datasheets correctly, both should do the same thing, right?

The 3-pin header was originally wired to one deck of a 3-position rotary switch, which put the unit into three different modes: DCT, LMT-2, LMT-100 (the DCT wiper was originally wired to pin 1). The DCT was the piece of equipment that I said was rare when new and likely nonexistent today, so I took out the circuitry specific to that mode (the 4052 and analog source resistors), and then after bypassing the rotary switch to remove the rest of the DCT functionality, I was left with a 2-position switch, which I have then changed to a jumper shunt block (J5 or J15).

 

No they are not the same function.
74LS348 is a priority encoder - 3 inputs decoded into 8 outputs.
74LS368 is a buffer/line driver - 6 inputs buffered to the matching 6 outputs.
[edited to add the vital word 'not']

 

All the left circuit does is poll the status of the switch connected to J5 / J15 to determine what model is being tested.

That switch is set at the start of the test; if there is a mismatch between the configuration jumpers and the switch on J5 / J15, then the tester is to throw a 'ERROR 01: Model mismatch'.

I changed the design from a rotary switch to the jumper shunt shown.

Shunt on J5 3-2 = Pin 3 grounded = ED7 high, ED6 low, telling the fixture to expect an LMT-100 unit.
Shunt on J5 1-2 = Pin 2 grounded = ED7 low, ED6 high, telling the fixture to expect an LMT-2 unit.
(the original design, before I took out the DCT mode, would have provided for grounding of pin 1, setting ED7 and ED6 high)

Therefore, I was led to assume that since the 74LS368 is an INVERTING buffer, I made the apparently incorrect assumption that I could simulate the same thing using an otherwise idle IC on the board.

Guess I'll have to look at adding this circuit back.

 

74LS348 is a priority encoder - 3 inputs decoded into 8 outputs.

It is an EN-coder - 8 inputs encoded into 3 outputs.

74LS368 is a buffer/line driver - 6 inputs buffered to the matching 6 outputs.

It is an *inverting* driver, not a buffer.

ak

 

It is an EN-coder - 8 inputs encoded into 3 outputs.

It is an *inverting* driver, not a buffer.

ak

Are we agreed these two chips are not equivalent?

 

They are functionally equivalent in his circuit for the two cases when a jumper is installed, which was the question. When no jumper is installed, they are not. That may or may not be acceptable to his system.

CORRECTION - The no-jumper case is ok. See post #8.

ak

 

Very well. I guess I will look at restoring that IC to the circuit.

 

CORRECTION - it should work as shown.

Left circuit:
With no jumper installed, inputs 5 and 6 are both high. 4 and 7 already are high, so grounded input 3 is the highest priority input. For this case, both outputs A0 and A1 are low. I have to keep reminding myself that all inputs and outputs are negative-true.

Jumper 1-2 grounds input 5. This drives outputs A0 and A2 low. So J1-2 = ED6 goes high.

Jumper 2-3 grounds input 6. This drives outputs A1 and A2 low, so J2-3 = ED7 goes high.

Right circuit:
With no jumper installed, both inputs are high and both outputs are low. Check.

Jumper 1-2 grounds input 2, which drives output ED6 high. Check

Jumper 2-3 grounds input 1, which drives output ED7 high.

Looks good to me.

NOTE - Right schematic input pins 6 and 10 must be either GND or Vcc.

ak

 

I appreciate the followup, and I also appreciate the reminder that the unused inputs need to be grounded. Thank you.

 

Incidentally, another section of this same board had a set of 6 status input lines that could be pulled down by external means. I finally realized the other day what all 6 inputs are for: Four are for the status of the controlled relays, and the other two are for detecting the unit model by way of jumper wires in the plug.

The circuit for those two lines turned out to be a duplicate of the circuit that we reasoned through a couple posts up. The addressing for both circuit sections even line up - only one bit is different in the addressing logic. Perfect!

 

u happy me happy