IC problems building The Great Destroyer (dwarfcraft devices)

[knutolai]

Thought I'd build The Great Destroyer from dwarfcraft devices as a simple sideproject. The circuit wants a MC14049BCP or a "buffered" (?) CD4049. Turns out is really picky about the IC as it does not work properly with any of the CD4049UBE I got from Tayda. By this I mean that the voltage starve doesn't sound right. It doesn't get any of the glitchy madness of the original.

Farnell lists the CD4049UBE as a replacement for the MC14049BCP so Im rather clueless on what I could use that would work. The MC14049BCP is out of production and I would like to find a available substitute before going to ebay.

Any suggestions?

Here's the layout:

[anchovie]

The "U" in CD4049UBE (or any other CD4*** chip) stands for "Unbuffered". MC14049BCP is a manufacturer-specific product code but the general chip family name is "4000 series" and the chip you want is a buffered 4049, so try and find a CD4049B without the U.

[knutolai]

wow why are these IC's so hard to come by? The CD4049BE isn't in stock at either Farnell, Mouser, ELFA (sweden), RS-Online, Digikey...

[anchovie]

Try HEF4049B - that's the product code for NXPs version of the chip.

[anchovie]

Stock no. 73-679-15 at ELFA.

[anchovie]

306-667 at RS.

[knutolai]

Thx! Found it even cheaper on Banzai. The code isn't specified, but they state it is the buffered variety.
http://www.banzaimusic.com/CD4049-buffered.html

[puzzlebunker]

Seems like CD4049 variant that's needed is no longer in production. I've tried ordering ones described as "buffered" and with the part code CD4049BE (as opposed to CD4049UBE) but what I was sent were chips with "UBE" stamped on.

Not even sure the stuff posted about "buffered" and "unbuffered" is correct because the 4049 itself is described as a buffer - what on earth would be meant by a "buffered" buffer and an "unbuffered" buffer?

Another thing to look out for is difference between inverting and non-inverting buffers. I'm wondering if this is where some confusion originated.

The Texas Instruments datasheet I've got for the CD4049 describes the inverting version as "CD4049UB" and the non-inverting as "CD4050B". So I'm wondering if the 4050B would work?

[iainpunk]

euhm, having both U and non-U 4000 and 14000 series in my collection, all 4049's are inverting
the difference is that the 'buffered' versions have 3 complementary stages stacked, while the 'un-buffered' versions have a single stage.

getting the 4050 to work is a little more hassle, since it can't bias itself like an inverting one can.

fun fact:
the CA3130 and CA3160 also have a CMOS complementary gain stage, but use it as their output stage. this makes them really nice sounding opamps when they are clipping to rails.

cheers

[anotherjim]

This could probably be redesigned to use a quad NAND or NOR gate chip which are all buffered and very much available.
Incidentally, that Tagboard layout doesn't terminate the unused inverters to stop them from misbehaving. Another one found does...
https://dirtboxlayouts.blogspot.com/2019/07/dwarfcraft-devices-great-destroyer.html

[Rob Strand]

I haven't traced the circuit but I suspect it should at least do something with an unbuffered device.
Maybe there's a problem.

All the unused inputs on the right side of the chip are floating.   Good idea to ground those.
[Jim already said that.]

Here's buffered vs unbuffered,
https://www.ti.com/lit/an/scha004/scha004.pdf

The  MC14049BCP is buffered and IvIark's comments seem to imply that's what is intended,
https://tagboardeffects.blogspot.com/2012/08/dwarftcraft-great-destroyer.html

A good deal of 4049 overdrives use the unbuffered devices.


Traced from IvIark's layout (not from original PCB),

[ElectricDruid]

wow why are these IC's so hard to come by? The CD4049BE isn't in stock at either Farnell, Mouser, ELFA (sweden), RS-Online, Digikey...

Probably because they were designed in the '70s and they're now pretty much obsolete. I grew up learning with 4000-series CMOS chips in the 1980's, but after that they've been superseded by any number of other logic families, and these days you wouldn't design a circuit that used them.

They were great chips and some are still around, but honestly? Use 74HC-series or whatever if you possibly can. I love 4000-CMOS for all the memories, but that was then, and this is now!

[zesak]

Why not CD4069UBE ?

[anotherjim]

The 4069UB would probably be similar to a 4049UB.

74HC4xxx parts are 7v max. Not unusable but the supply volts needs to be lowered.

[Rob Strand]

Traced from IvIark's layout (not from original PCB),

Is the wiring of that 22k pot correct?

To me it looks like it is doing nothing.    The idea is it should adjust the bias.

I suspect the wiper of the trimpot should connect to pin 5 of the IC and pin 6 of the IC should wire to pin 1 (ccw) of the trimpot.

What that does is it lets the biad voltage on the output of IC1b.

If this is done I suspect the pedal will work, perhaps not like the original but it will still work.

When you put in a 4049B device the first stage IC1c has a lot of gain.  It therefore has more chance of clipping.
When it clips it will produce a output at IC1b.

With a 4049UB the first stage gain might be lower and the output of IC1b might get stuck with no output.

If the 22k is use to set the gain then that brings up a conflict on how to set the DC bias on IC1b.

The last stage has a feedback pot for gain but at least it has an input cap which lets feedback set the bias.

[zesak]

Really funny, as i am finding out all the cd4049, cd4069 and comparator-based distortions,

that i built that ended up in my "real crap" pcb and parts bin, actually worked correctly.

What i did not realize back then (and after hearing the great destroyer youtube sound sample, how bad it is, exactly like my builds)

is that they were probably supposed to stutter and decay horribly, and generally sound farty and crappy by design.

No offense meant to those chasing the unusual distortion sounds, but for me it is just too rough and ugly, suggesting something has gone wrong.

[Rob Strand]

What i did not realize back then (and after hearing the great destroyer youtube sound sample, how bad it is, exactly like my builds)

is that they were probably supposed to stutter and decay horribly, and generally sound farty and crappy by design.

IMHO, having high gain 4049B's *and* DC coupling pin 5 to pin 6 is really asking for the stuttery .... farty etc sound.  The DC bias on IC1b is not well defined.   Some fluke-biased units might sound better than others and there might be a magic STARVE setting.

I suspect adding say a 1uF cap between pins 5 and pins 6 would help a lot.   That makes the second stage look more like the third stage.    It should bias better.  Positive of added cap to pin 6, same idea as C2.   With this change the Drive pot remains connected to pin 6 and the trimpot remains connected to pin 5.   The added cap goes in place of the small wire between the stages on my schematic.

The outputs of each stage should have a DC voltage which is very loosly mid-supply.   Definitely not really close to either the +rail (voltage on pin 1) or ground.

[anotherjim]

The Starve pot may have a lot to do with it. If by magic the inverter does bias midway, the shoot thru current will be at maximum and the supply pin voltage drops. This raises the open loop gain compared to the full 9v.
Another wrinkle in the 4049 is that the N channel is fatter than the P for the higher sink current of this types output so in a fight there is a tendency toward 0v.
However, If I connect x3 4069 inverters in series and fit a single overall feedback resistor it just doesn't bias properly.