Is the reference here to 'positive' on Cext (pin 10) correct?
Since pin 11 connects via Rint or via the external timing resistor to Vcc, then surely the positive terminal of the external cap should be on pin 11, not 10? This would be important if the cap were an electrolytic one.

 

I agree the positive terminal of a polarized cap should point towards 11 (and VCC)

 

Thanks. I suppose it would be possible, albeit unusual, to use the Rext timing resistor to discharge Cext and use a pull-down current source to charge it, so the suggested configuration could work. It would be interesting to see a scope shot of the cap waveform to confirm what happens.

 

Cext on SN74121 is different from that on SN74123.
With SN74123 Cext is at ground potential and should be connected to GND.

With SN74121, Cext is not at ground potential. The datasheet is correct. The positive terminal of an electrolytic capacitor should be on Cext.

 

Thanks for the clarification.

 

SN74121N

Upper trace is Cext (10)
Lower trace is Rext/Cext (11)

 

Is the reference here to 'positive' on Cext (pin 10) correct?
Since pin 11 connects via Rint or via the external timing resistor to Vcc, then surely the positive terminal of the external cap should be on pin 11, not 10? This would be important if the cap were an electrolytic one.

It's correct. TI, et al. don't give sufficient detail for us to understand why.

I breadboarded the circuit and there's a negative going pulse of about -3.2V at pin 11. I was surprised that the waveform wasn't exponential.

 

Thanks for the helpful waveforms. As you say, TI et al are certainly mean with detail.
I'm trying to create my own Spice model for the IC (to help posters in other threads) and had to guess how the pulse generator works. My guess gives the right Q and _Q outputs, but not that negative-going pulse. So, back to the drawing-board .

 

As you say, TI et al are certainly mean with detail.

I haven't seen a datasheet from any manufacturer that gave enough detail. I think Fairchild Semi had an application note (AN-366?) that said using resistors larger than the recommended range could cause permanent damage. I tried with the 560k that was used in that thread. It didn't work and, luckily, didn't seem to damage the part.

 

Hello,

Here are both the datasheet and the application note from fairchild.

Bertus

 

Thanks for the info. Unfortunately it still needs guesswork for the modelling, but I think I've cracked it now. My pin 10 and pin 11 sim waveforms match pretty closely to MrChips' scope shots in post #6. How's this?

 

Looks good. But why are you doing this, may I ask? I am just curious.

 

I'm doing it to help posters in other threads (as I mentioned in post #8). Other folk have been searching for a Spice model compatible with LTspice, but no-one seems to have one.