[sdiy] DIY Mixer - please help!

Fri Sep 4 14:46:52 CEST 2009

Jerry,
that sounds very interesting. Do you think you'd be able to whip up an
absolutely minimal schematic that contains just all the components for
the op amp and compares the approaches with the diodes, and without
them? It's hard for a newbie like me to picture what's going on in the
design you mention.

Thanks a lot
Damian

On Fri, Sep 4, 2009 at 2:44 AM, Jerry Gray-Eskue<jerryge at cableone.net> wrote:
> <<How would an op amp/vca need to be designed in order to gradually set
> on distortion as the signal nears rail voltage? And 'controlled'
> distortion before that - while keeping noise levels at a minimum until
> the 'controlled' distortion happens, and making the thing as linear as
> possible until that point?>>
>
> OK, it may be a bit tricky, but basically you want the op amp signal path
> the work normally when you are not near the rails, which is what they do
> anyway, so that part is easy. The trick is to add in a "Soft Clip" as you
> approach the rails.
>
> One way to get a soft clip is to add opposite facing Diodes parallel in the
> feed back loop of a gain stage to reduce gain as they turn on, the problem
> is they start conduction way too early, around .4V and are fully on at .5V
> to .9V typically about .6 Volts.
>
> However you might use Back to Back Zener Diodes in the feed back loop,
> controlling the "Clip Limit" by the Zener voltage. An additional benefit is
> that while one Zener is in Breakdown the other is in forward conduction
> giving a longer soft clip region.
>
> The other approach would be to use the opposite facing Diodes approach, but
> use Strings of diodes to get the turn on voltage you want, unfortunately
> this is about 16 diodes in a string for 10 volts. LEDs have a greater
> forward voltage drop, about 1.2 but still would require 8 in a string.
>
> The Hot ticket may be to use the Back to Back Zeners and if the clip is not
> soft enough start dropping the Zener voltage and adding stacked diodes to
> put more "Curve" in the clip.
>
> - Jerry
>
> -----Original Message-----
> From: synth-diy-bounces at dropmix.xs4all.nl
> [mailto:synth-diy-bounces at dropmix.xs4all.nl]On Behalf Of cheater cheater
> Sent: Thursday, September 03, 2009 1:32 PM
> To: synth-diy
> Subject: Re: [sdiy] DIY Mixer - please help!
>
>
> Jerry,
> You're right, tubes are definitely something to look at.
>
> A question regarding solid state now:
>
> How would an op amp/vca need to be designed in order to gradually set
> on distortion as the signal nears rail voltage? And 'controlled'
> distortion before that - while keeping noise levels at a minimum until
> the 'controlled' distortion happens, and making the thing as linear as
> possible until that point?
>
> I guess an important thing is to characterize the type of distortion
> that we'd want to introduce once it does happen. I guess the only word
> I can give is 'musical' because I simply don't know how to dress it in
> words better. Something that doesn't modify the mid band much, and
> crushes the highs in favor of keeping bass in, seems like a good plan.
> This would mean that the mid field where the interesting stuff happens
> stays unscathed. The highs that have all the transients that come
> through from the compressors, limiters, and plain old instrument
> transients, can be too much and often need to be clipped anyways. So
> that's the first thing that should go. They can sometimes be very high
> in level, especially without the engineer knowing about it because of
> imperfect monitoring conditions.
>
> I am not sure how to give a VCA character like this without
> introducing an equalization stage before and after the amplifier. What
> are some ways to do that?
>
> Cheers
>
> D.
>
> On Thu, Sep 3, 2009 at 5:01 PM, Jerry Gray-Eskue<jerryge at cableone.net>
> wrote:
>>
>> This is all just for Info, Personally I think a pure op amp setup would
> work
>> great, and If you want Tube distortion, just run the output through a Tube
>> preamp, or power amp.
>>
>> The hybrid thing sounds excellent too, the extra headroom would be nice.
>>
>> << whether they can replace the vintage parts
>> they are meant to be compatible with, due to their subjective
>> performance in the circuits they would be in (what is sparsely and
>> incorrectly called 'sonic quality').>>
>>
>> I have seen a lot of "subjective" reviews that say they are good, If you
>> care to look at a few you can check these out.
>>
>> My point is Nothing can give you "Tube sound and distortion" like a tube
>> can. There are a lot of active discussion groups on this very subject, and
>> it gets down to which current production tube they like best and think is
>> the Same or Better than the original tubes.
>>
>> http://www.guitarcenter.com/Accessories,Page-1.gc?src=tube&o=6&ipp=15
>>
>> http://www.tubedepot.com/12ax7reviews.html
>>
>> http://www.tubedepot.com/6l6reviews.html
>>
>> http://www.tubedepot.com/el84reviews.html
>>
>>
>> <<And I presume tube circuits are
>> going to be just amazingly difficult to manufacture on a scale larger
>> than two units every five years.>>
>>
>> Not really, its just another type of electronic device, Look how many Tube
>> devices are available in large quantity here:
>>
>>
> http://www.guitarcenter.com/Guitar-Amplifiers-Guitar-Amplifiers-and-Effects.
>> gc
>>
>>
>> - Jerry
>>
>> -----Original Message-----
>> From: synth-diy-bounces at dropmix.xs4all.nl
>> [mailto:synth-diy-bounces at dropmix.xs4all.nl]On Behalf Of cheater cheater
>> Sent: Thursday, September 03, 2009 10:26 AM
>> To: synth-diy
>> Subject: Re: [sdiy] DIY Mixer - please help!
>>
>>
>> Jerry,
>> tube production is of course going on, but it's nowhere near the
>> output that existed 'back in the day'. There are also questions on the
>> quality of those tubes and whether they can replace the vintage parts
>> they are meant to be compatible with, due to their subjective
>> performance in the circuits they would be in (what is sparsely and
>> incorrectly called 'sonic quality'). And I presume tube circuits are
>> going to be just amazingly difficult to manufacture on a scale larger
>> than two units every five years.
>>
>> Ken stone's tube vca is indeed interesting but I'm not sure if it's
>> the same kind of thing that would go in the summing bus of a mixer. It
>> is, however, a start of *some* sort, definitely.
>>
>> D.
>>
>> On Thu, Sep 3, 2009 at 3:21 PM, Jerry Gray-Eskue<jerryge at cableone.net>
>> wrote:
>>>  <<The tube manufacture
>>> market has mostly folded. And then he won't give you the best ones
>>> anyways :)
>>> - if this design is to be used by a lot of people, you can easily see
>>> how resource problems would start happening with the insufficient
>>> (negative) worldwide supply of tubes that we are experiencing>>
>>>
>>>
>>> The Boutique guitar amplifier market has keep tube production going,
> Tubes
>>> are far from obsolete discontinued museum pieces.
>>>
>>> There are a lot of tubes in current production in
>>> Russia (Sovtek http://www.tubedepot.com/sovtek.html ) and
>>> China (Sino http://www.tubedepot.com/sino.html) and several other
>>> manufactures, some of these are just re labeling the China/Russia tubes,
>>> some are not - some nice stuff and a good selection of tubes for audio
>> use.
>>>
>>> http://thetubestore.com/
>>>
>>>
> http://www.vacuumtubes.net/?source=google&gclid=COnhmc3EhpwCFRmbnAod8GqW_g
>>>
>>> http://www.tubedepot.com/?gclid=CKLamfnGhpwCFRaenAod2Svd-g
>>>
>>> <<- tubes seem to require impedance matching transformers, unless I'm
>>> wrong.>>
>>>
>>> Hammond makes these transformers and are still in business, you can buy
>> them
>>> from Mouser and other distributors.
>>>
>>>
>>>
>>>
>>> -----Original Message-----
>>> From: synth-diy-bounces at dropmix.xs4all.nl
>>> [mailto:synth-diy-bounces at dropmix.xs4all.nl]On Behalf Of cheater cheater
>>> Sent: Thursday, September 03, 2009 8:58 AM
>>> To: synth-diy
>>> Subject: Re: [sdiy] DIY Mixer - please help!
>>>
>>>
>>> Jerry,
>>> there are several problems with this:
>>> - tubes are not as easy to predict. You can't get new, 'great' tubes
>>> nowadays unless you buy them from someone who regularly pulls them off
>>> ebay and has a warehouse of ten thousand of them. The tube manufacture
>>> market has mostly folded. And then he won't give you the best ones
>>> anyways :)
>>> - if this design is to be used by a lot of people, you can easily see
>>> how resource problems would start happening with the insufficient
>>> (negative) worldwide supply of tubes that we are experiencing
>>> - solid state is, I guess, easier for people to DIY.. and I guess we
>>> have more experts for this here!
>>> - tubes seem to require impedance matching transformers, unless I'm
> wrong.
>>>
>>> However a tube-based design could be very interesting sometime in the
>>> future. There is no denying that this would be the next step after a
>>> 'clean and nice' solid state design is finished - a tube-based summing
>>> bus would be the obvious choice for something that complements a
>>> solid-state summer.
>>>
>>> Thanks
>>>
>>> D.
>>>
>>> On Thu, Sep 3, 2009 at 2:39 PM, Jerry Gray-Eskue<jerryge at cableone.net>
>>> wrote:
>>>>
>>>>
>>>> Given that, and assuming you would like the "tube" type distortion I
>> would
>>>> say that you should go ahead and use a tube(s) for the summing buss,
> this
>>>> does leave you with the issues of scaling the output down for recording,
>>>> headphones, or other lower voltage applications, but it is all easily
>>>> "doable" and could make a very nice setup. The other advantage is this
>>>> system could drive speakers directly and would be a cool setup for live
>>>> performances.
>>>>
>>>> If you want to max out the tube effects, you could use standard tube
>>> preamp
>>>> designs for your inputs and dump use of op amps altogether.
>>>>
>>>> - Jerry
>>>>
>>>>
>>>>
>>>> -----Original Message-----
>>>> From: synth-diy-bounces at dropmix.xs4all.nl
>>>> [mailto:synth-diy-bounces at dropmix.xs4all.nl]On Behalf Of cheater cheater
>>>> Sent: Thursday, September 03, 2009 8:29 AM
>>>> To: synth-diy
>>>> Subject: Re: [sdiy] DIY Mixer - please help!
>>>>
>>>>
>>>> I think Seb has mentioned a good point here, Jerry.
>>>>
>>>> What seems to be a very important thing in analogue electronics
>>>> (especially mixers) is not *whether* they distort. There will always
>>>> be distortion and you can't escape it. You might even want to
>>>> introduce it by overloading channels - that happens quite often.
>>>>
>>>> I think the important thing is *how* they distort.
>>>>
>>>> So when designing the amplifier in the summing stage, it could be very
>>>> useful to have a lot of headroom, but to be able to dial in a little
>>>> bit of distortion by going higher.. and then the higher you go the
>>>> more distortion you get. I guess this requires some sort of
>>>> information on what sort of circuits distort the best.
>>>>
>>>> Also I'm not very happy with using divide-down/resistors/whatevers in
>>>> the summing bus input. The first problem is that it changes the way
>>>> the mixer works depending on how many channels are in use (and the
>>>> additional complexity of added switches/relays/etc is not worth it).
>>>> The other problem is, as I understand it, that such approaches either
>>>> require an active element (which adds distortion) or a passive network
>>>> (which can easily create problems in impedance matching)
>>>>
>>>> For example Manley seem to be using solid-state strips, but the
>>>> summing bus is a tube based circuit, which can easily run on high rail
>>>> voltage, since tubes like high voltages, and also distorts nicely,
>>>> since tubes, well, distort nicely. But I think that a similar approach
>>>> can be had with a discrete design based on solid state circuits, I
>>>> might be wrong of course (probably am).
>>>>
>>>> So the question is - how do you make a summing stage that can work on
>>>> high voltages that reach, say up to 75V without distortion, and then
>>>> progressively starts being less and less linear, behaving like a nice
>>>> soft clipper? I think it would be nice if the behavior when you're
>>>> near the rails were similar to what I described.
>>>>
>>>> I think it would also be nice to be able to dial in the symmetry
>>>> yourself by inputting a DC bias - instead of the circuit changing
>>>> symmetry in itself. For example I sometimes see this used for clipping
>>>> off 'spikes' on just one side of a signal. I understand that one trick
>>>> is that if you leave the spikes only in the 'negative' part of the
>>>> signal, they would happen when the tweeter 'pulls' and not when it
>>>> 'pushes', giving you a mellower high-end.
>>>>
>>>>
>>>>>> I believe this is to achieve high headroom. Unconfirmed information
>>>>>> tells me that most op amp designs will distort more the closer you get
>>>>>> to the maximum output voltage.
>>>>>
>>>>> I doubt this is the case except when you're almost at the rails. A more
>>>>> important limitation is simply the maximum supply voltages audio opamps
>>>> will
>>>>> tolerate, which for the NE5532 (very low distortion opamp if the pcb is
>>>>> designed correctly) is +- 22V max. Not that having "only" 27 dBu max
>>> level
>>>>> is a problem IMO.
>>>>
>>>> Thanks for your comment Antti. The NE5532 sounds interesting... how
>>>> does it behave near rail voltage?
>>>>
>>>> Not that I expect it not to distort.. but does it distort in a nice way?
>>>>
>>>>>> Do you think that a 741 will handle this sort of use?
>>>>>
>>>>> No. The comment about 741 was meant to show how poor some of the
>>>> electronics
>>>>> were in some respects.
>>>>
>>>> - this just shows what sort of rube I am when it comes to real-life
>>>> parts and choices like that. :^)
>>>>
>>>> Thanks
>>>> Damian
>>>>
>>>> On Thu, Sep 3, 2009 at 2:07 PM, Seb Francis<seb at burnit.co.uk> wrote:
>>>>> Antti Huovilainen wrote:
>>>>>>
>>>>>> On Thu, 3 Sep 2009, Seb Francis wrote:
>>>>>>
>>>>>>> I guess it's a question of whether you consider a mixing desk to be
>>>> 'part
>>>>>>> of the sound' or you simply want to mix your instruments, insert/send
>>> to
>>>>>>> effects, etc. and choose what colouration to add when desired.
>>>>>>
>>>>>> Sadly, these two options will rarely produce the same result. It's
> hard
>>>> or
>>>>>> impossible to reproduce the effect of dozens of small nonlinearities
>> and
>>>>>> phase shifts in a single effect box.
>>>>>>
>>>>>> Whether you want "warm" or "clean" sound is a different question of
>>>>>> course. Both are valid choices.
>>>>>>
>>>>>
>>>>> I think I like "warm" & "clean" :)
>>>>> i.e. Some warm sound sources, but with an overall clean mix so that any
>>>>> 'clean' sounds can really come through.
>>>>>
>>>>> And even with a 'clean' signal path, judicious use of EQ (both analog
>> and
>>>>> digital) can go a long way towards a 'warm' sound.
>>>>>
>>>>> Seb
>>>>>
>>>>>
>>>>>
>>>>> _______________________________________________
>>>>> Synth-diy mailing list
>>>>> Synth-diy at dropmix.xs4all.nl
>>>>> http://dropmix.xs4all.nl/mailman/listinfo/synth-diy
>>>>>
>>>> _______________________________________________
>>>> Synth-diy mailing list
>>>> Synth-diy at dropmix.xs4all.nl
>>>> http://dropmix.xs4all.nl/mailman/listinfo/synth-diy
>>>>
>>>>
>>> _______________________________________________
>>> Synth-diy mailing list
>>> Synth-diy at dropmix.xs4all.nl
>>> http://dropmix.xs4all.nl/mailman/listinfo/synth-diy
>>>
>>> _______________________________________________
>>> Synth-diy mailing list
>>> Synth-diy at dropmix.xs4all.nl
>>> http://dropmix.xs4all.nl/mailman/listinfo/synth-diy
>>>
>>
>> _______________________________________________
>> Synth-diy mailing list
>> Synth-diy at dropmix.xs4all.nl
>> http://dropmix.xs4all.nl/mailman/listinfo/synth-diy
>>
>> _______________________________________________
>> Synth-diy mailing list
>> Synth-diy at dropmix.xs4all.nl
>> http://dropmix.xs4all.nl/mailman/listinfo/synth-diy
>>
>
> _______________________________________________
> Synth-diy mailing list
> Synth-diy at dropmix.xs4all.nl
> http://dropmix.xs4all.nl/mailman/listinfo/synth-diy
>
>