Is a "Compander" Chip Worth it in a BBD Chorus?

[rthryhorysak]

I've been looking at BBD designs and the Cool Audio V571 is something that I'm looking at adding. Basically, can all the filtering before that is typically used be omitted (my guess is not completely), or rather heavily simplified by removing the pre/de-emphasis filtering, and raising cutoff frequency for aliasing filtering.

If BOTH have to be used, it doesn't seem worth adding, but then again I've never tried it, so I'm wondering what the performance improvement is with the chip, and if the extra parts and $2 dollar chip can be justified.

https://cabintechglobal.com/pdf/COOLAUDIO_V571_DATASHEET.pdf

[StephenGiles]

It depends how you are going to use the chorus - in a band situation, probably not worth the hassle as there would be so much other noise to battle with anyway, but for recording or impressing your girlfriend (!!!) the achievable noise reduction would be well worth the cost of additional parts.

[Transmogrifox]

...so I'm wondering what the performance improvement is with the chip

It all depends how it is applied.  Both pre-emphasis/de-emphasis and companding are used to improve signal/noise ratio, so using both together yields better SNR than either one by itself.  To quantify the improvement you need to understand the sources of noise and its bandwidth, characterize your signal level (dynamic range).  Then you can determine how much noise reduction you get from pre/de-emphasis filters,  and if/how much you need to compress the dynamic range to get to the desired SNR. 

If you don't have the technical training/knowledge to calculate that, you can just build it and use your ears to determine whether the noise is tolerable.

Pre/post emphasis filtering is the most simple and cheap.  Many commercial stompbox designs stop there (I really can't think of many guitar-oriented stompboxes that use companding--rack-mount gear, yes).

As StephenGiles states, it depends how it's going to be used. 

Of course, since companding isn't a perfectly invertible effect it will add its own characteristic quality to the tone, which may turn out to be something you're after.

In the end, a $2 chip for a 1-off pedal isn't a really big deal unless you're so strapped for cash you can't afford to occasionally have a coke.

If you intent to sell a lot of this thing, then you probably should crunch the numbers yourself to see if the improvement in SNR is marketable. 

[PRR]

> it doesn't seem worth adding

The '571 chips were hot for telephone systems in the 1980s. They also got used in noise-sensitive audio. Especially BBD systems which are barely tolerable naked. Telephony changed and the mass market for '571 disappeared; so did production. BUT the '571 was SO useful in small audio that COOLaudio made the effort to put it back IN production.

Seems to me it is "worth it", for more than a few people.

[Mark Hammer]

BBDs have limited headroom, so the utility of including a compander will depend partly on how hot a signal you plan on feeding your chorus.  For instance, if your intent is to "punish" your amp with a chorussed overdrive, then using a compander will offer the BBD some protection against undesirable clipping.  You'll want the amp's tone to be a result of its own clipping, as dictated by the overdrive, and not the result of overloading the BBD.

That said, many chorus pedals omit companding unless they anticipate clock noise being a serious issue.  Case in point is the venerable Boss DC-2 chorus, whose use of companding is in service of reducing the noise of two counterswept clocks that extend an open invitation to heterodyning.   The noise of ONE clock is often manageable with simple pre-emphasis/de-emphasis.

[Scruffie]

From my experience, no, if you want a chorus with a long delay such as the EHX Polychorus which is almost a slapback chorus at some settings then yes it's absolutely useful (although without going to a BBD with a lot more stages your filter cut offs are still going to be quite low) but once you mix the clean signal in with the chorus, the slight raise in cut off of your filters just doesn't really amount to much change in tone over say, the CE-2 or Small clone.

You can get away with simplifying the input anti-aliasing filtering a bit though on a standard chorus with one but with a 16-pin chip plus extra parts needed, it's not really simplified, plus you may find in more 'real time' than a delay pedal that you don't like the companding action which does have an attack and release.

The smart money i'd say is on the EHX clone theory, which uses a J-FET for a wider range of sweep over a diode as a variable resistor but at a higher clock frequency so the filtering doesn't have to be as heavy to keep the noise down.

[EBK]

The smart money i'd say is on the EHX clone theory, which uses a J-FET for a wider range of sweep over a diode as a variable resistor but at a higher clock frequency so the filtering doesn't have to be as heavy to keep the noise down.

Having built two Clone Theory pedals now, while standing on Scruffie's giant shoulders, so to speak, I agree.   

[Scruffie]

The smart money i'd say is on the EHX clone theory, which uses a J-FET for a wider range of sweep over a diode as a variable resistor but at a higher clock frequency so the filtering doesn't have to be as heavy to keep the noise down.

Having built two Clone Theory pedals now, while standing on Scruffie's giant shoulders, so to speak, I agree.   

They're fairly average but I certainly appreciate the compliment!

I would also add, as Stephen has alluded to before when discussing flangers, that a slight amount of noise can actually be pleasing to the ear and I think is a part of what we deem 'analog' sounding. Considering most vintage recording equipment was going to introduce it to some degree and that amplification will too, it's not an insane theory, as long as you're not hearing the noise as a repetitive swoosh that then becomes far too obvious.

Oh and regarding pre/de-emphasis, yes you can leave it out with a compander (see the Ibanez Rotary Chorus which basically does what you're asking about and uses filters IIRC that are near identical to the CE-2) but i'd still use it anyway, there's a good reason it became an industry standard for BBD pedals, it works.

[rthryhorysak]

> it doesn't seem worth adding

The '571 chips were hot for telephone systems in the 1980s. They also got used in noise-sensitive audio. Especially BBD systems which are barely tolerable naked. Telephony changed and the mass market for '571 disappeared; so did production. BUT the '571 was SO useful in small audio that COOLaudio made the effort to put it back IN production.

Seems to me it is "worth it", for more than a few people.

Sorry, should have given context. The idea behind the pedal is going full Moog and using VCOs, VCAs, and VCFs in the chain and LFO. So I already will have a lot of 13700s and op-amps so I was wondering if I have access to something like an massive sweep of an LFO that is way past chorusing them maybe it was something to look into. I'm still on the fence and have to breadboard the design. I can handle the basic calculations but the more complicated like clocking the BBD and such, it gets a bit over my head. I was just looking qualitative data as far as these circuits in BBD pedals.

[rthryhorysak]

Oh also, the parts aren't much about cost, more PCB space, the goal is a 1590BB. The previous design took up the whole enclosure with SMD parts so the parts count is already high.

[armdnrdy]

Coolaudio V571m

Some of the better choruses use companders...and if the companies that produced them didn't think it was well worth the cost....they would not be included in the design.
Remember...we as hobbyists may not be very concerned with the cost...but if a company that designs and builds guitar effects wants to remain solvent and turn a profit...every cent is factored in.

[Scruffie]

Ah but that's not the question, the question was can corners be cut elsewhere if you used one and the answer there is you can't have everything.

But they certainly can work nicely in choruses of course, just not at the expense of good design.

[armdnrdy]

Ah but that's not the question, the question was can corners be cut elsewhere if you used one and the answer there is you can't have everything.

But they certainly can work nicely in choruses of course, just not at the expense of good design.

I felt that enough people answered the first part of the two part question...that I would skip it.
My response was to the second part of the question:
"If BOTH have to be used, it doesn't seem worth adding, but then again I've never tried it, so I'm wondering what the performance improvement is with the chip, and if the extra parts and $2 dollar chip can be justified."

He was also concerned about real-estate so....I posted the SMD model number.

[StephenGiles]

Perhaps a decision needs to be made as to which corners to cut, and compromises to be made if a chorus is to be stuffed into a tiny box. It may be better for a no compromise, all bells and whistles, low noise, top performance BBD board to require a larger enclosure, even to rack size.

[ElectricDruid]

I agree with Stephen. You need to decide what compromises you're going to make. If you want the best-possible quality from 1970's tech like BBDs, you probably need to throw everything at it, good anti-alias filtering, good reconstruction filtering and clock noise attenuation, and companding to reduce background hiss. It's not either/or with filtering or companding; they do different things.

Either you go for best quality and have a larger circuit, or you make a few compromises and have a smaller, simpler circuit. If it has to be small and sophisticated, do it with surface-mount parts. There are examples of all these approaches in stomp box history - some chorus pedals use the commanding for lower noise, and others go for lower cost or greater simplicity and leave it out.

HTH,
Tom

[StephenGiles]

Didn't Craig Anderton design a noise reduction circuit for recording purposes back in the 1970s/80s period? Would this be of any use in BBD circuits?

[Mark Hammer]

Probably worth noting that while companders were only occasionally found in chorus pedals and flangers, they were pretty much standard in analog delays.  Why?  Simply because, when aiming to squeeze most delay time from a single 4096-stage BBD (as was standard at the time), the clock rate one would have to use would pose much greater noise risk than the faster clock rates used  for chorus and flanging.  So, the means used to control noise needed to be more aggressive.

Consider that a 1024-stage BBD would be used to produce a maximum of 20msec of delay for most chorus units, while 4096-stage BBDs would be used to aim for 300-350msec maximum delay much of the time.  Four times as many stages that would have to aim for a maximum delay of 80msec in order to be using the same clock rates as a typical chorus.  With the goal of 4 times as long a delay, the clockrate wouldneed to be dropped by a factor of 4, where the clocknoise becomes MUCH more audible.

[anotherjim]

I think the least you can do is use emphasis, since suitable gain stages are usually going to be there anyway.
As per Boss CE-2...
https://www.electrosmash.com/boss-ce-2-analysis

[ElectricDruid]

Absolutely agree with Mark. Flangers and chorus are the "easy" BBD applications, in that you can keep the clock relatively high - indeed for flangers, you're mostly trying to get it as high as you can. Even then, some of the better ones put a 571 in - improves S/N, helps squash the clock noise in the expander stage.
Analog delay is a process of trying to recover something useful from mush. 10KHz clock rates are not unheard of, and that's clearly audible, and means you need plenty of filtering way down around the 3KHz mark. Note that although Nyquist reckons that signals can be accurately reconstructed at half the the original sample rate, he wasn't dealing with BBDs. BBD data sheets usually reckon on Fs/3 rather than Fs/2 as the maximum signal frequency. That's a practical rather than theoretical figure!

Tom

[Mark Hammer]

Analog delay is a process of trying to recover something useful from mush.

That quote should be preserved in the Smithsonian!