stomp compressor : Help for newbie

[gazza]

Hi,

I'm looking to build a simple stomp-box compressor for guitar and bass.

I'll be using vero stripboard, so I'd prefer as simple a design as will suit so that i'm less likely to make a mistake in designing the layout.

I'm thinking of making john hollis's 'flatline' compressor which uses a single dual opamp and a led & ldr.   This is because it seems a little bit more flexible than an orange squeezer and less complex to build than the mxr dyna-comp/ ross compressors.  

so a few questions before I start !

Is it true that the optical link will mean a cleaner signal rather than the slightly dirtier electronic links as in the orange squeezer ?

Would it be possible to do what the EH black finger does and have a switchable lamp or led ???    I like the idea of being able to choose between the slower lamp response or the faster led response.  

Will this compressor have a lot of that typical high frequency filtering effect because its optical ?    should I be looking to try and add a filter to boost the highs ?

Also, would I need a boost stage to avoid any volume drops when switching ??

hope people can help or point me in the correct direction.  I know next to nothing about compressors.  

thanks !

Gareth

(PS i'll be wanting to use a strat with single coils, a semi-solid with humbuckers & flat wounds, and a bass with jazz type pickups)

[David]

So far, I've only built a Flatline on breadboard.  My impression of it was that it was smooth, quick-acting and had terrific sustain.  It also seemed to have a pretty flat frequency response.  I don't recall noticing any coloration at all.  

I don't think you would have a choice of light sources with the Flatline.  One of the op-amps is set up as a rectifier to drive the LED.  Driving a lamp is not my field of expertise.  I do know that if you use the right parts it WILL work.  Does it sound better than an Orange Squeezer or a MXR?  Mark Hammer or Constantin Necrasov might be able to answer this better than I can.  I do plan to at least breadboard an Orange Squeezer since I just took delivery of a BUNCH of FETs!

ROG simulators, here I come!!   :twisted:  :twisted:  :twisted:

[Mark Hammer]

In principle, optical compressors will have the following advantages:

1) LDRs reduce some of the envelope ripple that can often create the impression of distortion in non-optical compressors.

2) LDRs have, for all intents and purposes, NO overloading or distortion due to overloading or signal level.

I haven't made the Flatline, but I've made both the CLM6000-based compressors from the Anderton books and they sound nice and clean.

Their chief disadvantages are that :

a) The light source that the LDR depends on can (that's CAN, not WILL) have higher current requirements than a FET or OTA might.

b) It is far easier to slow down the response time of a FET/OTA to eliminate envelope ripple than it is to speed up an LDR to accommodate different sorts of input signals.

c) There is often more component-to-component variation in LDRs than in FETs/OTAs, hence a slightly lower chance of satisfying first-time build with optical.

Having said all of that, it's a big world out there, and somebody somewhere may have fabulous access to a wide variety of inexpensive LDRs but no easy access to reasonably priced FETs.  As well, if the intent is to power by wall current anyways, then current consumption is a nonissue.  And if the ultimate purpose is something intended for use as a guitar or bass compressor ONLY, with no notion whatsoever of having to handle, say, cymbals or snare in a pinch, then the added flexibility of non-optical is also a nonissue.

The OS is probably an easier and more trouble-free build, has a wealth of user comments, mods, and build reports to draw on, and sounds just fine.   Worth making if only to remember the late Dan Armstrong (see obit in this month's Guitar Player) with fondness and respect.

I posted some mods and an accompanying schem over at generalguitargadgets, and others have been as pleased with them as I was.  Take a look and see what you think.

Ultimately, though, it's 6 of one and a half dozen of the other.  Both will work fine for you.

[gazza]

So far, I've only built a Flatline on breadboard.  My impression of it was that it was smooth, quick-acting and had terrific sustain.  It also seemed to have a pretty flat frequency response.  I don't recall noticing any coloration at all.  

hmm,  I may have picked the wrong type of compressor.  It sounds like the flatline is very much a sustainer.   (maybe the sustain control was a hint !)

maybe I should build an orange squeezer for a more mellow effect.

OR

Maybe I could have a 'gentle' setting by having a very big cap across the LED and leaving the sustain setting low ???  ( the cap seems to be a 100 microfarad one anyway, which is fairly big) or reducing the intensity of the LED, or putting a resistor in series withthe LDR ?

OR

can I set the flatline up to be fairly subtly by simply turning the sustain control down ?.

as I said before, I'm a total newbie with compressors.  

Gareth

[Mark Hammer]

Both devices are capable of identical functioning and are instances of the same CLASS of device.

Limiters are devices that do nothing to the signal UNLESS it exceeds some predetermined level, at which point they reduce it in a manner dictated by how the controls are set.  Think of it like a compressor confined to only the loudest portions.

Compressors are essentially boosters and apply a fixed amount of gain.  As the input level increases though, the output level is adjusted downwards so that a relatively constant output signal level is produced.  It's like having an intelligent clean booster pedal that can say "Oh, I guess you didn't want THAT much boost" and turns itself down a bit, and a bit more when needed.  The impression that is created is similar to what you get IF the instrument was capable of longer sustain, but neither device actually increases sustain, regardless of what the manufacturers call them.

There are two ways to do this.  The Flatline sticks an LDR in the feedback loop of the op-amp.  As the resistance of the LDR is decreased in response to input signal, the gain of the op-amp is reduced.  In the OS, a slightly different logic is applied.  Here, the gain of the op-amp remains constant.  Instead, the FET and the 82k resistor that precedes it behave like a pot/volume-control that turns down the level of the input signal which eventually goes to the op-amp.  Although the amount of gain doesn't change, it is applied to a lower/attenuated signal, so the output is reduced.

Stock, the OS tends to have a lighter touch (i.e., less squish) and be valued for that.  But you can tweak component values on either to produce variations between light restraint and heavy squish.

[gazza]

Both devices are capable of identical functioning and are instances of the same CLASS of device.

Stock, the OS tends to have a lighter touch (i.e., less squish) and be valued for that.  But you can tweak component values on either to produce variations between light restraint and heavy squish.

Thanks for all the info Mark ! It's making much more sense now.  The OS circuit was more confusing to get my head around.  

If I want to mellow out the flatline compressor, is it simply a question of trying to reduce the gain available ?  

I'm not sure what type of LDR was expected with the circuit, I have a wide choice of different resistances to choose from, an order of magnitude on the minimum (~20k vs ~200k) and a lot on th edark resistance (   500k to 3Meg).  Can anyone help me here ?

is there any point in trying to play around with the capacitor across the led ?

I guess the answer is to just build the thing and find out !    

thanks for the advice.

Gareth.

PS the circuit is at John Hollis's web site here :

http://www.hollis.co.uk/john/flatline.jpg

[Mark Hammer]

Having things make sense to people always adds sparkle to my day, so we're off to a good start.

The cap in parallel with the LED does play a role in setting the response time of the LED and consequently the response of the LDR.

There is actually a lot of room to play around with LDRs here.  Bear in mind that in this instance (let's leave the LDR out of it for the moment), the gain of the op-amp is determined by the ratio of the feedback resistor (220k) plus the resistance to ground (10k fixed + 100k pot), to the resistance to ground.  So, with the pot set to zero ohms, the gain is (220k+10k)/10k = 23.  With the pot set to max resistance, the gain of that stage is equal to (220k+10k+100k)/(10k+100k) = 330k/110k = 3.

The extent to which the gain of that stage is adjusted by the parallel LDR would depend on the value of the LDR *and* the value of the other components.  For example, if the LDR is 3meg under dark perfectly conditions, then sticking it in parallel with 220k will make the effective feedback resistance equal to (1/220k)+(1/3meg) = 1/R, or about 205k, making the default maximum gain equal to 215k/10k or about 21.5.  Not much different.

If a big power chord causes that LDR to drop to a resistance of 100k, the gain will drop considerably.  Parallel resistances of 220k and 100k will get you an effective feedback resistance of about 69k, which will drop the gain down to a little less than 8.  Of course, if the LDR were to drop its resistance down to, say, 10k or less under conditions of maximum LED brightness, then the gain would be reduced even further.

Note that since the gain depends on BOTH the feedback resistance and the resistance to ground, once the effective feedback reistance becomes equal to or less than the combined value of the pot and fixed resistor, you'll have a gain of less than 2 and only a bit more than 1.

The upshot of this is that, depending on the parameters of the LDR, you can tinker with the values of the feedback resistor and the resistance to ground to achieve different ranges of max and min gain.  

The rule of thumb is that when you have a pair of parallel resistors, their combined effective resistance is determined more by the smaller value, up to a certain point.   If the 220k resistor were 10k, then it would take whomping changes in the LDR value to change the effective parallel resistance.  What this means is that you can play off the different resistance values until you get some arrangement that lets whatever you can eke out of the LDR you have produce a reasonable change in gain.  Ultimately, you may find yourself with, say, a 10k pot and 1k fixed resistor to ground or maybe a 250k pot and 39k resistor to ground or whatever.

Please note as well that different LEDs produce different degrees of illumination for different input current.  Personally, I'm partial to the high brightness (>2000mcd) ones since they needless current to shine, hence allow longer battery life.  Of course, since the brightness of the LED is also determined by how much the default gain is (the LED gets its signal from that first stage), LED choice and gain stage requirements will interact.

The down side of this is that it may take a while for the final component values to "settle".  The up side is that you WILL be able to make it do what you want no matter WHAT LED or LDR you are stuck with.

Again, for all these reasons, my gut sense is that a first-timer may likely have less difficulty with making an OS even though in some respects the Flatline is a more "bulletproof" design by virtue of using an LDR.

[gazza]

Mark,

Thanks for the advice.  I can see that I'll  need a few different leds and ldrs in sockets and have to spend a bit of time setting the thing up.  

I saw a modded OS schematic somewhere with a full wave rectifier and all sort s of tweaks - but I can't find it again.  

Anyone know where it is ?

I'm tempted to try both circuits out now.

thanksl,

Gareth.

[gazza]

I saw a modded OS schematic somewhere with a full wave rectifier and all sort s of tweaks - but I can't find it again.  

Anyone know where it is ?
Gareth.

Doh !   found Mark Hammer's modified schem on GGG - how i missed it last time I don;t know. (its at http://www.generalguitargadgets.com/diagrams/orangesqueezer_marks_variable.gif  )

QUESTION !

how do I add the millenium bypass LED indicator circuit to it ?  I think there needs to be a farily high resistance to ground for the LED fet to turn on/off BUT  there is only R18 which is 10K.    

what can I add to sort this out ?

thanks,

Gareth.

[RLBJR65]

Can't help with the technical stuff but if your just getting into your build, there is a nice layout for this project at Geo http://www.geofex.com

Also you might want to look at Mark's expansion mod. I have not tried it out but it looks simple enough. http://hammer.ampage.org

[Mark Hammer]

Just a comment on converting the Flatline from compression to expansion.

Please note that the time constants for pleasing expansion are different than those for pleasing compression.  Where a compressor that sluggishly reduces gain and holds it down, in response to a suddent transient, can sound nice, an expandor will sound better if the control element (LDR in this instance) has a fast onset and fast offset.  In other words, you want it to add a little more zip to initial attack transients, then get the hell out of the way.

In that respect, simply lifting one end of the *same* LDR and reconnecting it to another part of the circuit, may certainly yield dynamic expansion, but not necessarily anything you might like.  It may well be advisable to have two different LDRs with different response times.

(Post script to original September note, added March 19:  Rick Lawrence indicated on another thread that he did the Flatline-to-Punchline conversion and reported that it worked quite satisfactorily and in a noticeable way.  Perhaps he got lucky with LDR choice, perhaps he has absolutely no taste  or is too damn busy working on the next box to notice the shortcomings of the last one :lol: , or perhaps the LDR response time matters much less than I think it does.  Given that few people will deliberately seek out an expansion circuit WITHOUT ever being interested in compression, any experimentation with flipping around the LDR location is a lowcost experiment.  You're in the neighbourhood anyways.  May as well drop in, right?)

[rubberlips]

someone wouldn't have a copy of John Hollis' flatliner schematic would they?
Tried the site and can't access it
I've got RG's layout, but it would be nice to see the schematic

Pete

[loki]

someone wouldn't have a copy of John Hollis' flatliner schematic would they?
Tried the site and can't access it
I've got RG's layout, but it would be nice to see the schematic

Pete

I have it, but don't remember the page where i found it....
if you give me your email i can send it to you
bye

[RickL]

Mark, your second surmise about my impression of the Punchline conversion is probably the most correct. I have never claimed to be a gourmet when it comes to effects, only a gourmand. If it makes some sort of noise and every knob has some sort of effect I'm happy.  :lol: