Self-destructing 14-pin DIP

[GK]

I only got up to the scene where Tom Selleck rescues Kirstie Alley from the sentry robot. The film was just too daft and well before then it was obvious that the blonde was never going to get her norks out.

The detonating DIP was a highlight though:

http://youtu.be/rzX6wxeneaY?t=1080

 

[MK14]

It looks like, he shorted out multiple pins of the IC, while probing it.
It may have damaged the IC, hence causing it to overheat. Potentially making gases build up inside the chip, and hence the mini-explosion.

I haven't watched the rest of the documentary, so may have missed other information.

 

[Alex Eisenhut]

Looks like latch up. Let the nerds clean up the mess, I'm chatting up Cynthia Rhodes instead.

[station240]

"See that red stripe it's not a regular commercial chip"

lol you can see the TI logo, and they didn't even sand the numbers off.
Not entirely clear, but the bottom line seems to read 74LS...

[MK14]

"See that red stripe it's not a regular commercial chip"

lol you can see the TI logo, and they didn't even sand the numbers off.
Not entirely clear, but the bottom line seems to read 74LS...

You need to put your glasses on, or get your prescription updated.

It seems to be a CD4011AE !
Made in Malaysia.

Feel free to zoom in on the clearer picture, below:

[Kleinstein]

A latch-up can do quite some damage to a chip - especially in older computer HW with a powerful supply. I have seen the port drivers (7406 or similar AFAIR)  in 2 printers to have a blown away top. The likely cause was connecting the printer while still powered and than latch-up triggered by ESD.

[GK]

It seems to be a CD4011AE !

OMG! Just think of the depth of that chips capacity for being re-programmed with sinister AI! It will override the CPU! Terror!

 

[MK14]

A latch-up can do quite some damage to a chip - especially in older computer HW with a powerful supply. I have seen the port drivers (7406 or similar AFAIR)  in 2 printers to have a blown away top. The likely cause was connecting the printer while still powered and than latch-up triggered by ESD.

Latchup is very possible.
The old/original CD4011AE, A = unprotected inputs (i.e. lacking ESD protection diodes, if I remember correctly)

You just had to look at them funny, for them to instantly destroy themselves.
Jokes aside, I remember they would break, way too easily.
TTL was considerably more robust, and the 40xxB types (B), were more reliable.

[MK14]

It seems to be a CD4011AE !

OMG! Just think of the depth of that chips capacity for being re-programmed with sinister AI! It will override the CPU! Terror!

 

If I remember correctly, with the unprotected A type (40xxA etc), you could actually make a number of projects out of them, by using their analogue mode.
E.g. A set of hex inverters (A), configured to be a radio (which was in Electronics magazines of that time era). Using the analogue mode as amplifiers.
(I could be mis-remembering, and maybe it was the buffers, non-inverters, I can't remember, too long ago).

[bd139]

This reminds me of when I was doing A level tech many many years ago here. My friend was sitting next to me and suddenly there was smoke. Lots of it. Came from a 555. I don’t know how he had done it but it’s amazing how much magic smoke there is in a 555.

About ten years ago I came across a whole bunch of IC tubes, chucked out from local college, twenty of which contained various 555 derivatives. Could I get that much smoke out? Hell no. Just a few wisps. Disappointed. Sent about 15 ICs to their grave trying to find out how he did it.

[GK]

A latch-up can do quite some damage to a chip

Yes, the scene was totally representative of reality.

[Kleinstein]

The smoke and flames in the video look more like a battery of old tantalum cap going off all the way, or a little more.  Those little bastards can sometimes burn a hole right through an FR4 board. So maybe 100 tantalum caps to get the fireworks as shown. Normally one would consider a mixture of Ta and MnO2 powder a dangerous combination. So there is quite some energy stored in those little caps.

[Cyberdragon]

Everytime stuff (vehicals or buildings) gets damaged in a movie random stuff always shoots sparks and flames (using little charges that Bigclivedotcom has disected and even done skits with, or they're cheap and animated). But it's always out of wierd places like consoles behind screens or buttons, random walls or ceilings, or wires fall off the ceiling and spark. Wouldn't damage be localized or maybe reach up to the nearest fuse/breaker with the controls protected? Or has anyone actually seen protection circuits go nuclear and result in movie style fireworks?

[MK14]

At around 8:50, things get a tiny bit smokey.

[wraper]

Latchup is very possible.
The old/original CD4011AE, A = unprotected inputs (i.e. lacking ESD protection diodes, if I remember correctly)

CD4000A series do have ESD protection and are not original series. Original series are CD4000 without suffix, they are slower and accept lower max voltage but even those do have ESD protection diodes.

if I remember correctly

You don't.

[MK14]

Latchup is very possible.
The old/original CD4011AE, A = unprotected inputs (i.e. lacking ESD protection diodes, if I remember correctly)

CD4000A series do have ESD protection and are not original series. Original series are CD4000 without suffix, they are slower and accept lower max voltage but even those do have ESD protection diodes.

if I remember correctly

You don't.

You're right, I've got mixed up about it.

Internet searches, seem to indicate that the 4000A series was still a fairly old version of the CMOS logic, and had a number of weaknesses, which were resolved by moving to the newer/better 4000B series of CMOS logic.

E.g. http://www.nutsvolts.com/magazine/article/understanding_digital_logic_ics_part_4

THE ‘4000A’-SERIES OF ICS
The initial 1972 range of digital ICs was known as the ‘4000A’ series; it used the basic type of CMOS inverter shown in Figure 1, but incorporated extensive diode-resistor ‘clamping’ networks to protect its MOSFETs against damage from static charges, etc. Thus, a complete A-series inverter stage took the basic form shown in Figure 3.

Commercial testing of the early A-series range of CMOS devices quickly revealed a number of design problems.

Hence the move to 4000B, which was what was probably confusing me.

I vaguely remember, that although the early devices, claimed to have ESD protection, it didn't work very well, until later series came out.
But I can't remember if that was 4000A or 4000B series CMOS, I was thinking of (with somewhat working ESD protection), sorry.

[MK14]

What I was really trying to refer to was, making a radio out of early 4000 CMOS logic, using its "analogue" mode.

E.g.



From this old thread:

http://www.antiqueradios.com/forums/viewtopic.php?f=2&t=247725

Post subject: Re: Electronic Experts - Need Advice on CMOS CD4066BPostPosted: Apr Wed 02, 2014 4:13 pm
Member

Joined: May Tue 30, 2006 4:46 pm
Posts: 9424
Location: Santa Rosa, CA   
Years ago, I bought a Dick Smith crystal set at the Wellington, NZ store.

This simple circuit uses a CD4007 as a chain of audio amplifiers:

Image

The stages shown as inverters are really just single transistors.

A novel use of "logic" but it works!

Rich

Ironically, someone else, from 2002, also can't quite remember about the "unbuffered" CMOS logic things.

My vague recollection is when the 4000 series cmos came out in the 70s there were some unbuffered types.... The logic inverters of those could be used to make amplifiers by putting a resistor between the input and output so that it would bias itself at 1/2 the supply rail.

I wonder if this is not a 4066B but a earlier version. Quick search on this didn't turn up anything.

But, there is or was also a 4016 switch that is not buffered. Could this be that instead?

Someplace I have a copy Don Lancaster's "CMOS Cookbook" from the late 70s that talks about this sort of stuff.

[wraper]

4000B series have unbuffered parts as well. http://www.ti.com/lit/an/scha004/scha004.pdf They have 4000UB suffix, you can but them easily: https://eu.mouser.com/Search/Refine.aspx?Keyword=CD4001UB

[MK14]

4000B series have unbuffered parts as well. http://www.ti.com/lit/an/scha004/scha004.pdf They have 4000UB suffix, you can but them easily: https://eu.mouser.com/Search/Refine.aspx?Keyword=CD4001UB

I'd maybe have to try it, to find out (4000 vs A vs B types). But I'm NOT really too bothered at the moment, so won't be progressing it, now. But MAYBE in the future ?

You seem to be talking about the "digital" usage ones. Fair enough.

But what I was really referring to, was using them as "analogue devices". To make radios (amplifiers) and things. I.e. linear mode.

It goes OUTSIDE what the datasheet mentions/intends, I know. It comes from electronics magazine circuits, from a very long time ago (1970s/1980s).

I suspect (but would have to try it to be certain), but it needs the older, original 4000 or 4000A circuitry, to work well/best.
The 4000B, unbuffered or buffered, would not be so good, and/or would NOT work in that mode of operation (see link in earlier post, from nuts&volts). Since their design seems to have been improved, which removes that capability.

As regards static/ESD damage.
I can't remember if it was the 4000 or 4000A series (my best guess is it was 4000A's, because I seem to remember wanting the fully unprotected 4000 series (but that could be anytime from just about first availability to much, much later), but they were largely unavailable, then, even though we are talking about when 4000 CMOS were first or soon available to the general public, give or take a number of years, as I don't like to mention my exact age).

But, they would be very easily destroyed in my hands (NOT intending to break them on purpose, just with normal usage). Tending to not last very long, on bread boards. Supply voltage too big, static electricity and maybe input voltages, outside of the supply rails, were all suspects.

[Kleinstein]

Everytime stuff (vehicals or buildings) gets damaged in a movie random stuff always shoots sparks and flames (using little charges that Bigclivedotcom has disected and even done skits with, or they're cheap and animated). But it's always out of wierd places like consoles behind screens or buttons, random walls or ceilings, or wires fall off the ceiling and spark. Wouldn't damage be localized or maybe reach up to the nearest fuse/breaker with the controls protected? Or has anyone actually seen protection circuits go nuclear and result in movie style fireworks?

I have seen a case where the mains fuse (actually 2) failed to do it's job. So a simple short from an blown light bulb did some damage to the wiring and the switch.  So a 16 A LS switch failed and the 32 A or 50 A fuse and the next level fuse at some 100 or 150 A blew. Even just the plain NYM cables (3x1.5) turn brown from the outside. The fuse panel was surprisingly intact - just the one LS.

Part of the problem was that the outlet and fuse panel was very close to the distribution transformer (only some 4 m) and thus very low impedance - so more like an area you want CAT 4 tools.

[GK]

What I was really trying to refer to was, making a radio out of early 4000 CMOS logic, using its "analogue" mode.

E.g.

I recognised that schematic straight away, because I built it ~30 years ago. It was a Dick Smith kit from Smith's Funway into Electronics, Vol 2. The only thing I ever managed to tune in on it was the local AM radio broadcast station.

[MK14]

I recognise that schematic straight away, because I built it ~30 years ago. It was a Dick Smith kit from Smith's Funway into Electronics, Vol 2. The only thing I ever managed to tune in on it was the local AM radio broadcast station.

Did you use an original CD4007, or a later CD4007A or even CD4007B version ?
(Not remembering the answer now, is fine).

It won't surprise me, if such circuits are a very bad way of making such circuits. But, I'm interested, in using "funny" techniques, whereby a standard digital logic device, is used to make a radio.

I use to use ZN414 based, radio circuits. But with limited results (not brilliant), as far as I can remember.
But the fact that you can use an IC radio chip, in an apparent, standard small transistor like package, was interesting as well.
Plus doing the IC with only three legs, is cool as well. The final icing on the cake, is that it is happy to run on only 1.5 Volts. So a single AA/AAA battery, is an extra fun factor to it (ZN414), as well.

[wraper]

4000B series have unbuffered parts as well. http://www.ti.com/lit/an/scha004/scha004.pdf They have 4000UB suffix, you can but them easily: https://eu.mouser.com/Search/Refine.aspx?Keyword=CD4001UB

I'd maybe have to try it, to find out (4000 vs A vs B types). But I'm NOT really too bothered at the moment, so won't be progressing it, now. But MAYBE in the future ?

You seem to be talking about the "digital" usage ones. Fair enough.

But what I was really referring to, was using them as "analogue devices". To make radios (amplifiers) and things. I.e. linear mode.

You certainly can get CD4000UB to operate in "linear" mode.

[bd139]

4000 series is cool. Lots of "unspecified uses" for it. Thandar had some good ones. Front ends of DMM's using unbuffered 4000 series. Worth looking at the timebase on the SC110 scope as well - that had one as the sweep gen differential amplifier: https://k1.spdns.de/Vintage/Sinclair/Other%20Inventions/Other%20Electronics%20Products/Thandar%20SC110A%20Oscilloscope%20Service.pdf (page 5)

Then there's Bob Pease on it: https://www.electronicdesign.com/analog/whats-all-cd4007-stuff-anyhow

[MK14]

4000 series is cool. Lots of "unspecified uses" for it. Thandar had some good ones. Front ends of DMM's using unbuffered 4000 series. Worth looking at the timebase on the SC110 scope as well - that had one as the sweep gen differential amplifier: https://k1.spdns.de/Vintage/Sinclair/Other%20Inventions/Other%20Electronics%20Products/Thandar%20SC110A%20Oscilloscope%20Service.pdf (page 5)

Then there's Bob Pease on it: https://www.electronicdesign.com/analog/whats-all-cd4007-stuff-anyhow

Thanks, that is VERY interesting!

I've enjoyed reading the Bob Pease, article. He had some very interesting comments to make about it.

The scope schematics is interesting as well. The low end Hitachi crt oscilloscope(s), used a number of 4000 series chips, in their 20 MHz scopes. Which surprised me, when I was trying to repair one.
They were especially used in its timebase, alt/chop modes and sweep start/running/stopped modes and stuff.
Probably to make the 1KHz prob cal output as well. 4011 Nand gate oscillators, if I remember correctly.

Since it was an ancient, supposedly all analogue scope, I was surprised to see the 4000 series chips.

[MK14]

You certainly can get CD4000UB to operate in "linear" mode.

You can, but the article (Nuts&Volts, link in earlier post of mine) seems to indicate, that they are less useful in analogue applications, than some of the older types. Because they still have the improvements (minus the buffering), which makes them better for digital purposes, but less so, if you want to use them in a linear fashion.
Which reduces the available gain, and hence they are potentially less useful.
Although Bob Pease, if he was still around, might disagree, and indicate improvements in other respects, despite the reduced gain.

[wraper]

You certainly can get CD4000UB to operate in "linear" mode.

You can, but the article (Nuts&Volts, link in earlier post of mine) seems to indicate, that they are less useful in analogue applications, than some of the older types. Because they still have the improvements (minus the buffering), which makes them better for digital purposes, but less so, if you want to use them in a linear fashion.
Which reduces the available gain, and hence they are potentially less useful.
Although Bob Pease, if he was still around, might disagree, and indicate improvements in other respects, despite the reduced gain.

There was no mention if they are better or worse for analog application.

[MK14]

You certainly can get CD4000UB to operate in "linear" mode.

You can, but the article (Nuts&Volts, link in earlier post of mine) seems to indicate, that they are less useful in analogue applications, than some of the older types. Because they still have the improvements (minus the buffering), which makes them better for digital purposes, but less so, if you want to use them in a linear fashion.
Which reduces the available gain, and hence they are potentially less useful.
Although Bob Pease, if he was still around, might disagree, and indicate improvements in other respects, despite the reduced gain.

There was no mention if they are better or worse for analog application.

Yes there does seem to be, but I could easily be wrong. But I think we should stop discussing it.

I remember from long ago, that the older/original chips were recommended for linear stuff, but the newer/later ones were not. I just can't remember the exact details, such as if it was 4000, 4000A, 4000B, unbuffered, buffered etc etc.

The linked to document, seems to me, with a quick look, to suggest some reasons why the analogue capabilities of the 4000 series, worsens/diminishes, as it gets more modern and improves, its digital abilities.

Anyway, I fully accept that I could be wrong, as I have only quickly read through the document, and can't clearly remember the details from many decades ago.

If you are REALLY concerned (I am, and might experiment in the future), it is probably best to just try experimenting with analogue circuits, using the 4000/A/B/UA/UB etc chips, rather than getting into long arguments on forums about it.

I find it a neat concept, because it is sort of a one input pin, one output pin, op-amp like solution.

Real op-amps are so cheap and available these days, there is little point in trying to use 4000 series devices (which are somewhat obsolete (or heading that way) as well). unless very interested in electronics and/or vintage stuff.

I wanted to build a radio using the CMOS 4000 inverters, at the time (many decades ago), but never actually did it. So, I guess it is part of my long electronics todo wish list.

[wraper]

Yes there is. But I think we should stop discussing it.

Excuse me, but I don't see how this could count as such:

One disadvantage of the B-series is that its propagation delays are larger than those of the old A-series. To counter this problem, a few new-generation devices are produced in an ‘unbuffered’ format (denoted by a ‘UB’ suffix), but incorporate all the other improvements of the B-series.

Typically, UB inverters have an AC gain of 23 dB at 10 volts, and are useful in several analog applications. Note that the bandwidth and propagation delays of a CMOS device vary with supply voltage and with capacitive output loading. Figure 10 lists the typical propagation delays of both UB and B-series inverters when used with supply values of 5V, 10V, and 15V when driving a 50 pF load.

[MK14]

I can probably find the article, I am trying to remember, from probably the 1970s online (many of the older articles get scanned into PDFs and can be found with some, possibly hard work). Which explains why the early 4000 series are good for analogue stuff, but the later ones are not.

It is possible that the 4000UB devices you are referring to, did NOT exist, at the time of the rather ancient article(s), I am thinking of. So, it could be that the 4000UB devices, are better, than some of the other 4000 series devices. But the article(s), I was thinking of, did not cover the 4000UB, as it did not exist then.

I am interested in digging up the old articles, anyway. I will have a look and come back to this thread, if I come across them.

[Zero999]

The buffered chips will probably oscillate, if used in a low gain application because they're actually three gates in series, which not only have a high gain, but lots of phase shift. Connecting the input to the output, will make a ring oscillator, unless the feedback node is connected to 0V with a low impedance, compared to the feedback resistor, at AC.

[MK14]

Excuse me, but I don't see how this could count as such:

With extensive research, and while having much fun and bringing back fond, past memories while doing it. I managed to find one tiny paragraph, explaining some of what I was trying to properly remember.

Summary:
It DOES really need to be the ORIGINAL 4000 series, i.e. No A or B types. The 4000UB type isn't mentioned for that year, as far as I can tell, so it may not have been around then, or I didn't look hard enough.

The explanation, is because the original 4000 series, had just a single transistor complementary pair. Which, hence allowed reliable/stable analogue/linear circuits to be constructed.

Coincidentally, the original 4000 series, did not have the ESD protection, which was what I was somewhat remembering, in relation to this issue.

It seems to be saying that later types (4000A and if available in the applicable type then, 4000B), are too unstable, to use in (at least the circuit it gives), due to the extra transistors. Which cause instability.

There were other information articles/projects/examples, explaining about it. But they would take considerably longer to find.

Please refer to bottom of page 26, in the following link.

Source:
https://www.americanradiohistory.com/Archive-Electronics-Today/ETI-Circuits-No-2-1978.pdf

I don't really definitively know if the (presumably) later 4000UB series, would be suitable, in those circuits. So you could be right (or not).

Does it really matter, since a modern, low cost, much better in most measurable ways, op-amp, would probably make a much better solution.

The original 4000 series, are probably getting increasingly harder to get, these days. I seem to remember, them being very difficult to get, even forty years ago (at least in the UK), as they were rapidly replaced by the 4000A and 4000B devices.

[MK14]

The buffered chips will probably oscillate, if used in a low gain application because they're actually three gates in series, which not only have a high gain, but lots of phase shift. Connecting the input to the output, will make a ring oscillator, unless the feedback node is connected to 0V with a low impedance, compared to the feedback resistor, at AC.

Probably exactly right!

That seems to be similar to what the mini-article is saying (implying/hinting etc), linked to above. I.e. Instability.

[IanMacdonald]

The supply voltage is important for good analog performance. Too low and you will have a dead band in which neither FET conducts. Too high and the conductance will overlap causing a high quiescent current.

[MK14]

The supply voltage is important for good analog performance. Too low and you will have a dead band in which neither FET conducts. Too high and the conductance will overlap causing a high quiescent current.

That makes a lot of sense. I think it is similar to what Bob Pease's (in the article he wrote, details earlier in this thread) concerns over the excessive sensitivity to changes in supply voltage, which these types of circuit, has.

Because, with normal CMOS (even the latest modern stuff), and when using it digitally. You are usually/often warned, to NOT allow any inputs to float at around half of the supply voltage and/or have too slow a signal transition, from low to high, or high to low (and hence it would be at the half way voltage for some of the time).
Because it can cause excessive current/power consumption. Essentially because it can cause the top and bottom, MOSFETs to conduct at the same time (shoot-through), hence the excessive current/power use.

Hence the "Too high and the conductance will overlap causing a high quiescent current", part of your explanation.

So, by keeping the supply voltage low enough (but while still being above the minimum supply voltage you just mentioned and probably more to safely be in the type of "linear mode" we are discussing, so the MOSFET gates get enough voltage to begin to turn on etc), it avoids/minimizes the possible excessive current consumption.

[MK14]

Just in case anyone is interested.
The magazines I was aiming for, was the UK edition of ETI (Electronics Today International), around the 1975 to 1978 time period (narrowed down by my estimates, of when I read the article(s) ). I encountered other electronics magazines, but feel ETI makes the most sense.

My gut feeling was before 1975 (4000 CMOS seems not so available in the UK, to hobbyists/me of that era) would be too early, and after the end of 1978, things (as regards 4000 CMOS, linear/analogue use), had moved on too much (i.e. it would probably be mostly/always op-amps and stuff, then, in magazine circuits).
But it could have been outside of those years.

A large number of those magazines (in PDF form), are available here:
https://www.americanradiohistory.com/ETI_Magazine.htm