dumbbell question about "UBE" IC's.....

[plexi12000]

does the "UB" stand for "unbuffered"?  thank you!

[GGBB]

does the "UB" stand for "unbuffered"?  thank you!

Can you point to an example?

[Mark Hammer]

For CMOS chips, yes it does.  I assume this is with relevance to one of the various hex invertors, like a 4049 or 4069?

[blackieNYC]

And IIRC- if you got the wrong 4069, which is hard to find now, there's a fun Devi Ever pedal. Great destroyer.

[plexi12000]

Yes...i was looking at a pedal layout and the dude says to use the unbuffered 4007.   Is that the same as the 4007ube?

Mark.....could you explain what a cmos is?   thank you

[robthequiet]

Datasheet: http://www.ti.com/lit/ds/symlink/cd4049ub.pdf

The UB is inverting, the B is non-inverting. The B has an extra stage to reinvert (unvert?) the signal, thus buffering in one sense. The E refers to the package type, as in PDIP, etc. I leave the finer points of which is better for what to the more expert cadre.

[R.G.]

Actually, the "UB" stands for "un-buffered", the "B" for buffered".

The confusion is that the 4049UB and 4050B are listed on the same datasheet. 4049 is a logical inverter, 4050 is noninverting. The 4050 is inherently buffered, as there is another gain stage in each of the six sections, making it "buffered". The 4049 inverter comes in both a 4049UB and a 4049B version, unbuffered and buffered respectively.

The 4049UB and 4069UB are useful for pedals and analog amplification, as their internal gain is small enough that they can be biased into the linear range. The B series inverters have so much gain that they're effectively impossible to get ot amplify - everything comes out as square waves, and they can't easily be biased to give amplification.

UB is what pedal makers want, usually. UB and B are not interchangeable for analog applications.

[GGBB]

Sorry Mark and RG - I don't think that's accurate (but it may have been in the past). Texas Instruments uses the "U" suffix on the CD4049 to indicate inverting. Thus the difference between CD4050B and CD4049UB. They are both "buffered" because this device is a buffer. That may or may not be the reason for the "B" suffix, but since there is no CD4050 or CD4049U, I would hesitate to believe that the "B" means buffered because that suggests there is such a thing as an non-buffered CD4049/50x which I don't believe is the case. It doesn't make sense to have a non-buffered buffer. The CD4069UB is the non-buffered inverter, which is maybe where the confusion about what "UB" stands for stems from. In this case, "U" still stands for inverting, so "B stands for - not "buffered" for sure.

So as to the original question, technically "UB" does not universally stand for "unbuffered " when used as an IC suffix, or even as a CMOS IC suffix. This is specifically true for the Texas Instruments CD4049 and CD4069 ICs and probably others. Other manufacturers may however use "UB" to stand for unbuffered - there are no rules to prevent it or enforce it - you must always read the manufacturer's data sheet.

As to the CD4007 "Dual Complimentary Pair Plus Inverter" - there is only one version of it I think (and I don't think it's buffered).

EDIT: Correcting CD4049B to CD4050B

[Mark Hammer]

As I have always understood it, the "buffering" was to render the output of a CMOS chip (40xx series) compatible with the input of a TTL chip (74xx series).  That sort of output does not work in the linar applications of things like invertor-based overdrives.  And, having made the mistake 35 years ago, I can vouch that an unbuffered hex invertor is needed for overdrives.

Now, what suffix/code individual manufacturers use to identify these different sorts of outputs is a whole other thing.

As an aside, I will note that there are several ostensibly TTL chips that purportedly work fine for overdrives.  I've seen the 74HCU04 used in Japanese overdrive circuits that are largely indistinguishable from the Red Llama/EPFM-TSF.

[reddesert]

This TI application report may be useful for more on the difference between buffered and unbuffered CMOS devices: http://www.ti.com/lit/an/scha004/scha004.pdf

In the CD4049UB / CD4050B datasheet posted above, http://www.ti.com/lit/ds/symlink/cd4049ub.pdf , look at Figure 1 and Figure 2 plotting voltage-out vs voltage-in. The rounded corners in Figure 1 for the CD4049UB vs the square corners in Figure 2 for teh CD4050B show that the CD4049UB is "unbuffered" by TI's definition, and in the sense that is desirable for use in analog effects. The CD4050B is not.

[anotherjim]

All 4049 invert. All 4050 do not.
Often they are on same data sheet.

Buffering doesn't improve output drive, just sharpens up the response to input level change due to higher overall gain. In fact, if the input changes too slowly, the buffer jitters badly, causing unexpected operation of the next device or logic element. Very unfortunate if that is an edge triggered flip-flop or counter.

TTL inputs require relatively high sink current to pull them low, but they pull up pretty well to high all by themselves. Both 4049 and 4050 have N-channel output transistors with a much lower on resistance than their complement P-channel. See in the data sheet that output sink current is much higher than source. The 4069UB inverter has standard CMOS outputs with similar source & sink current ability.

[blackieNYC]

(Re: the great destroyer. One could imagine that Devi got a bunch by mistake and was determined to make something out of them. You can get it to do crazy sub- intervals and also get it to give a variety of squeals as you adjust your guitar's volume control. A fairly straight up distortion too. I did not get mine by accident, but got it on fleabay. I don't know if it's worth actually going out and finding the chip, but if you're a lemons-to-lemonade alchemist you might appreciate having something to do with your B chip.   I happened to be messing with mine when this thread popped up.)

[PRR]

> All 4049 invert.

Agree. A 4049 which does not invert is NOT a 4049. It is a 4050.

Originally all CMOS was non-buffered. This stuff was very new. They had to keep the price down. Naked single-stage CMOS was good-enough for many many purposes. (Heck, it sure beats the M²L diode-logic we used before.)

While TTL interface is one issue, generally CMOS could be "improved" if buffered. As CPUs grew to 6-zillion MOS devices it was silly a 4049 only had a dozen. Buffering "B" increases this to 36, but that is now chickadee feed cost. So generally they moved to making only B devices.

Fussy low-performance guys like us want old-style un-buffered parts, now "UB".

Buffering classic CMOS is not one extra stage but two. 4049(original) and 4049(new UB) is one inverter (per path). 4049B is three inverters. 4050(original) is two stages, so buffered well enuff for TTL and hi-C uses.

[GGBB]

This TI application report may be useful for more on the difference between buffered and unbuffered CMOS devices: http://www.ti.com/lit/an/scha004/scha004.pdf

I stand corrected - thanks!