M74HC4094 DS missing STROBE waveform in Timing chart

[waclawek.jan]

In M74HC4094  datasheet - DocID1972 Rev 5 - Figure 4. Timing chart, STROBE waveform is missing:

 

JW

[waclawek.jan]

The truth table is flawed, too.

JW

[Peter BENSCH]

Thanks for the hint, Jan!

In the data sheet, Figure 4 is indeed incomplete; when rearranging the SGS document, the curve at Strobe was probably lost and is located exactly where you assumed it to be.

However, the truth table is actually (almost) correct (i.e. your red markings are wrong), other known manufacturers have only swapped the order of the lines. The only real error in the data sheet of the M74HC4094: line 5 of the truth table (i.e. between your red markings) states HXX as the input condition at OE/ST/DI, which should be HHH.

I myself also find the numbering of the outputs unfortunate (Q1...Q8), where every programmer starts at 0 - but the first data sheet came out in the early 90s of the last millennium, so please forgive us for that.

Regards
/Peter

@Lina_DABASINSKAITE let's see if we can find someone to correct this really old document.

[Lina_DABASINSKAITE]

Thanks @Peter BENSCH for looping me in. I will inform responsible colleagues to have a look a this. 

[waclawek.jan]

Hi @Peter BENSCH ,

The timing marks (vertical lines) in timing diagram indicate also rearranged waveforms, as the clock-related ones are too short to reach the clock waveform, OTOH the strobe-related ones are too long and reach up to the clock waveform... :)

In the truth table I believe I have the red marks correct, as the last two lines apparently aim at showing the function of the QS' output which changes upon *falling clock edge* (contrary to all other outputs) - see last two columns (QS and QS').

And, as at the falling clock edge neither of the "regular" shift registers output don't change, it's then indeed HXX for OE/ST/SI in the second-to-last row, as neither strobe nor shift-input can influence the outputs. at that moment.

I looked into RCA's DS, and yes, they have it wrong in this detail, too. Others (except ST - haven't looked at "original" SGS DS yet) just blindly copy/pasted.

JW

[waclawek.jan]

Just to show that how RCA was not infallible, two pictures from the same databook:  the left-hand picture is from the overview page 22, right-hand from the particular "datasheet" page 279:

And one more, from a 1975 RCA databook:

 

JW

[waclawek.jan]

This is a page from 1978 SGS-ATES COS/MOS B-series databook (I don't know where do I have it from):

It's a verbatim copy of the RCA databook, so they probably licence-built the 4000 COS/MOS line.

In other words, sometime later somebody in SGS or SGS/Thomson or STM redrew the diagrams, giving actual thought to the signals (fixing the old and introducing a fresh error :) ).

JW

 

[Peter BENSCH]

Phew, yes, there are plenty of mistakes in all the documents.

On page 22, the RCA book has the typical error of a manually repainted drawing (as can still be found today with manually typed information like phone numbers). The resulting book by Harris contains exactly the same error in the functional diagrams on its page 1-22.

The drawings on the top right of the RCA book, page 278 (page 279 in the pdf), which you have reproduced here on the right, are indeed correct. At the bottom of that page you will find the physical realisation of the latch with four FETs and two inverters, which have been replaced by a state-controlled D-latch in ‘more modern’ books. Here you can see that /Q of the last shift register stage is inverted three times and then output on pin 9 as QS, while Q of the same register is latched and then appears double inverted on pin 10 as QS’.

The ST data sheet also makes a blunder here and correctly indicates pin 9 on the left in fig 1, but not in the IEC symbol on the right, where pin number 8 is indicated for QS - which is GND. Sigh...

When I have some more time, I'll take another look at the diagrams, so far I've only compared them with more recent documents from other manufacturers.

I don't know if there's any point in cleaning up this mess of copied information of copied information, people have been getting on with it for 40 years, haven't they? 

/Peter

BTW: you used a smiley - weren't you the biggest opponent of that in forum software? ;)

[waclawek.jan]

> The ST data sheet also makes a blunder here and correctly indicates pin 9 on the left in fig 1, but not in the IEC symbol on the right, where pin number 8 is indicated for QS - which is GND. Sigh...

Nice catch!

---

This appears to be completely correct. Maybe result of the transition from 4000 to 74HC. SGS-Thomson High Speed CMOS, 1990:

So it was probably the redraw at rev.3 which moved the clock edge marks and omitted the strobe signal.

JW

 

PS. > BTW: you used a smiley - weren't you the biggest opponent of that in forum software? 

I write them manually as normal. If I'm in the mood, I replace the hyphen by a different one so that forum software won't convert to icon; this time I did not. Back when loading the thousand of jpegs slowed down the forum significantly, I simply blocked them at my side, so after posting I don't see those where I don't use the non-hyphen-hypen.