CD4049s. Anyone got proof that buffered ones exist?

[brett]

Hi.
Like a lot of folk, I've used CD4049 hex inverters a few times.  And I've read MANY times to avoid the buffered ones.  But do buffered CD4049s actually exist?

Let me check the evidence:
1.  The TI range consists of the CD4049UBE and others.  People seem to assume that UB refers to "unbuffered".  However, there is no CD4049B in the TI range.  In a 16 pin DIP, they are all UBE.
2.  National make CD4049BCJ/BCL/BCN/BMJ.  Are these buffered?  No.  They are all identical chips to the TI CD4049UBE.  The schematic shows the same arrangement of 3 CMOS devices per inverter.
3.  Intersil make a CD4049BMS.  This is slightly confusing because the datasheet refers to it as a CD4049UBMS (similar to National's CD4049UBM), but again the schematic shows that it is identical to all of the chips above (i.e. unbuffered).
4.  Philips make the HEF4049BPs.  The datasheet doesn't include an internal schematic, so there mught be extra CMOS devices lurking within.  But I doubt it.  The specs are virtually identical to all of the devices above.  If there are buffers in these inverters, they don't work very well.

This is a big call, but I think the 4049 "B" for buffered myth is busted.

So how would the myth get started?  Well, these hex inverters are used as buffers.  They only need to be "unbuffered" hex inverters to be good buffers.  But they still make good pedals.  Also, the CD4050B is the complementary hex non-inverting buffer device.  It contains 5 CMOS devices per channel, and is a "stronger" buffer than the 4049.  Maybe someone thought that the 4049B shared some of these characteristics of the 4050B?  Lastly, it might seem natural that there is a difference between a 4049B and a 4049UBE.  But I strongly suspect it is just that TI is the only manufacturer who used "UBE" as their suffix.

Any other ideas or information out there?

[gez]

Yeah they exist I've got some.

[gez]

Whoops, no that's no the case (sorry Brett).  I once ordered buffered but got UB...hmmm.  I'm sure I had some once though...who knows.

[TELEFUNKON]

http://pdf1.alldatasheet.com/datasheet-pdf/view/102930/ETC/IW4069BE.html

see page 4, expanded view: 3 serial inverters... ("hard")

fairchild had them, similarly, too, IIRC

[MartyMart]

I also have a few, I only know because they "dont work" in any of the classic cmos circuits.
I'll look 'em up when I have a 'mo.

MM.

[TELEFUNKON]

"buffered" buffer (see fig.3 : 3 serial inverters...)
http://pdf1.alldatasheet.com/datasheet-pdf/view/17715/PHILIPS/HEF4049B.html
"unbuffered" buffer (see 1st. schem,page 5)
http://pdf1.alldatasheet.com/datasheet-pdf/view/8175/NSC/CD4049UBC.html

[Peter Snowberg]

This is a big call, but I think the 4049 "B" for buffered myth is busted.

Sorry, but no. No myth. Why would there be any myth? 

I started playing with 4xxx CMOS parts in the 1970s. At that time there were still alot of "A" versions in supply channels. These were the first CMOS parts and they were unbuffered. These were very quickly replaced by the "B" buffered series which provide improved ability to drive downstream logic. Quickly, almost every CMOS part that wasn't relying on analog capability was improved with a buffer stage.

You'll notice that all the logic functions that are purely digital have a "B" in the part these days.

1.  The TI range consists of the CD4049UBE and others.  People seem to assume that UB refers to "unbuffered".  However, there is no CD4049B in the TI range.  In a 16 pin DIP, they are all UBE.

Today there isn't nearly the need for discrete inverters in the logic world. The most popular uses for these parts now are places where the analog capability gets used. That's why the "4049UB" parts survived. The 4049B is still quite popular, but now TI calls it the 74HC4049. http://focus.ti.com/lit/ds/symlink/cd74hc4049.pdf If you look at the datasheet you'll see the block diagram for the 74HC4049 is made from three inverters in series.

2.  National make CD4049BCJ/BCL/BCN/BMJ.  Are these buffered?  No.  They are all identical chips to the TI CD4049UBE.  The schematic shows the same arrangement of 3 CMOS devices per inverter.

Yes, they are buffered. Having three inverters in series is what makes it a buffered inverter.

http://focus.ti.com/lit/ds/symlink/cd4049ub.pdf

The 4049UB is clearly a single inverter in the TI datasheet. The 4050 uses two inverters in series and you'll notice it has the title 4050B.

PS: Don't trust block diagrams in datasheets to be actual circuit representations.


On suffixes; Don't bother with any letters past the "B" or "UB" because these are 100% vendor specific codes for things like package and  temperature range.

[Stephen]

TRUE though???

In the manufacturers part number you want a UB suffix in there

And a buffered 4049 will not work??

Why do they call them buffer inverters if they are not buffered?  Is this a cliche in electronics lingo??

What does buffer than mean??

[Sir H C]

PS: Don't trust block diagrams in datasheets to be actual circuit representations.

+1,000,000 they put them in there to throw off the competition.

[Peter Snowberg]

TRUE though???

In the manufacturers part number you want a UB suffix in there

And a buffered 4049 will not work??

Why do they call them buffer inverters if they are not buffered?  Is this a cliche in electronics lingo??

What does buffer than mean??

Yes, you want the "UB" in the suffix.

Buffered 4049s either work poorly or not at all as amplifiers.

Buffered inverters are buffered. A buffered inverter will have just a B on the end of the part number. An unbuffered inverter will either end the part number with "UB" or "A" (i.e. 4049UB, or 4049A).

A simple inverter uses a single stage. -> 4049UB
A buffered inverter will use three stages. -> 4049B
A non-inverting buffer will use either two or four stages. -> 4050B

What does the buffer mean? Transistors can only produce so much gain in either current or voltage. If you want more gain, you need to use multiple stages. The first stage in a buffered inverter uses very small transistors to keep input capacitance as low as possible. The third stage uses much larger transistors to make the output current higher. The second stage is similar to the first because it doesn't need to switch much power, it is there just to invert the signal being fed to the output buffer which is also an inverting stage.

PS: Don't trust block diagrams in datasheets to be actual circuit representations.

+1,000,000 they put them in there to throw off the competition.

I suspect the same thing. 

[Gilles C]

I can assure you they exist...

It seems that every time I buy some, they don't work in that kind of circuit, and I realise I bought the buffered ones. 

Gilles

[brett]

Hi.
Thanks for the feedback.  Myth NOT busted.

In summary, then:  The TI 4049UBE and National CD4049UBMJ and UBCJ are unbuffered.
The Philips HEF4049B, National CD4049BMJ/BCJ devices and 74C devices are buffered. The Intersil CD4049BMS is shown as unbuffered on the schematic, but seems likely to be buffered.

cheers

[Peter Snowberg]

In summary, then:  The TI 4049UBE and National CD4049UBMJ and UBCJ are unbuffered.
The Philips HEF4049B, National CD4049BMJ/BCJ devices and 74C devices are buffered. The Intersil CD4049BMS is shown as unbuffered on the schematic, but seems likely to be buffered.

Intersil made several 4049s including the CD4049BMS and CD4049UBMS. Every one of the searches I did for the BMS part turned up the datasheet for the UBMS part. These days Intersil makes 7 different 4049s (four of them specified for new designs) and they're all unbuffered.

There isn't perfect consistency between the part numbers from all the manufacturers, but the "U" seems to survive. The 74HC4049 is buffered, the 74HC04 is buffered, but then there is also the 74HCU04 which is unbuffered and very much like the 4049UB.

You can simplify it down to the presence or lack of a "U" in the part number for unbuffered.

Check out the general description on the National 4001/4011 datasheet:

They have equal source and sink current capabilities and conform to standard B series output drive.