Page 1 of 3

John Hollis - Flatline Compressor (Layout + Transfer)

Posted: 23 Sep 2007, 05:10
by markm

Posted: 12 Oct 2007, 01:13
by oldrocker
The Flatline is another John Hollis great design. I use mine all the time and it really makes my guitar stand out in the band. Especially clean undistorted guitar chords.

Posted: 13 Oct 2007, 14:32
by modman
Credit where credit is due - I linked to the schematic on his website:
http://www.hollis.co.uk/john/circuits.html

What a great website, will add it to the links.
And thanks MarkM for the layout! Try to add a link to schematic you used - I find it impossible to debug without schemo's.

Posted: 13 Oct 2007, 17:08
by markm
modman wrote:And thanks MarkM for the layout! Try to add a link to schematic you used - I find it impossible to debug without schemo's.
Yeah, I've been trying to add schems with my posts and it's a slow process as I'm adding them as I come across things in my files which to be quite honest.......are a damn mess! :roll: :x
I'll get there though!
That's why I'm glad were able to edit here, makes it so much easier!

Posted: 13 Oct 2007, 18:01
by modman
That's no problem at all Mark, until I learn to make layout I'll add them when I can google them. Just don't want to link to a wrong version.

thanks for all your work!

Posted: 06 Nov 2007, 07:25
by seniorLoco
a true gem indeed !! 8)

Posted: 06 Nov 2007, 12:57
by briggs
I built one of these but modified the envelope section to allow control over the attack and decay - sounds great now. Quite unique and capable of a huge "squish" 8) I'll put the scheme up soon. Just need to get round to drawing them nice and neatly :?

Posted: 06 Nov 2007, 19:36
by bajaman
I'll put the scheme up soon
I would love to see - even a rough drawing - I have been working on a good optical compressor idea for some time now - John's design looks interesting ( as do all his other designs) , but I can see some shortcomings in them all :wink:
Cheers
bajaman

Posted: 06 Nov 2007, 22:57
by destro
Yeah, I would like to see that schematic as well. I am doing a parts order for it and the voodoo vibe right now!

Posted: 07 Nov 2007, 02:45
by briggs
OK. I'll try and get a photo of my scheme up. It now looks nothing like the Hollis Flatline. As usual I've morphed it into a completly new design without even thinking about it!

It uses two dual op amps (two for compression block and two for te follower section) but it has controls for gain, sensitivity, attack, decay and intensity. The intensity control will probably be removed from the final version, I need to tweak this current version to get it's best sounds - at the moment it's super versatile but needs a few tweaks here and there!

Posted: 07 Nov 2007, 02:50
by bajaman
I have been working on a good optical compressor - almost got the Demeter Compulator sussed with help from pz and others, so I am curious to see your efforts here Fred :wink:
Cheers
bajaman

Posted: 07 Nov 2007, 02:58
by seniorLoco
i dont see the need of an 'attack' or 'decay' on the simple JH flatline !
The attack pretty much depends on how hard you hit the strings and can't care less on the decay ..... just keeping it simple 8)

Posted: 07 Nov 2007, 12:11
by briggs
just keeping it simple Cool
It's not in my nature - you can easily sub in your prefered values and just leave it. I like the option to have some crazy "swell" type compression.

Posted: 07 Nov 2007, 12:23
by briggs
Image

As you can see it looks nothing like the flatline anymore. This is an old scheme, some of the values have be tweaked but until I look at my breadboard I can't remember which ones! It can do all sorts - mild compression to full out "swell" squish 8)

Posted: 07 Nov 2007, 12:24
by briggs
Oh, btw. The follower section is that from the Meatball envelope follower - nothing new :wink:

Posted: 07 Nov 2007, 13:31
by analogguru
sorry about my critics, but on the first view I found the following:

You are using a 4,7p Mica in parallel with the 2M2 of the first opamp - a normal ceramic wouild do it too...
especially with an unnecessary 2µ2 input coupling electrolytic.
With 1M input impedance you only need 47n fotr a flat response down to 20 Hz. So a 100n film-cap would do a better job.
(1M and) 2M2 in the feedback path is a high value with adds resistor noise exceedingthe opamp noise, better lower the values and add an input buffer.
Output coupling of 1µF with a 10k load (neglecting the 4k7 restistor to virtual ground) acts like a high-pass filter. For linear response i would choose for this value 10µ.

analogguru

Posted: 07 Nov 2007, 14:06
by briggs
Thanks for the input AG. Much appreciated from a guy of your knowledge 8)

Yeah, I've changed the value of the resistors on the input op amp to 220k and 470k, this also lowers the gain down to around 2.13. The lower input impedance lowers the noise right down.

The mica was used because it was the only 4.7p cap I had, I suppose you could use a ceramic with very little impact on sound. I couldn't be bothered with hitching up an input buffer 8) I'm too lazy :wink:

On the schematic (I know it's hard to read and looks like 2.2uF) the input cap is 2.2nF - rolls of a fair bit of bass, a little too muchinfact, I'm now using a 15nF input cap, rolls off at 48Hz - keeps the low end tight. I also lowered the output cap - 470nF. Rolls of Bass @ 55Hz this again was to keep the low end tight. For flat response (e.g no bass roll off of any kind) use input cap of 33nF and an output cap of 10uF).

With those mods it's quite a nice comp. Different to the regular offerings :)

Posted: 07 Nov 2007, 14:21
by analogguru
with the 1uF cap it rolled off bass at 15Hz; below what humans can hear) this again was to keep the low end tight. For flat response (e.g no bass roll off of any kind) use input cap of 33nF and an output cap of 1uF).
sorry, but not with the 10k sensitivity-pot at the output.

analogguru

Posted: 07 Nov 2007, 14:32
by briggs
What does it roll off at? In my calcs I get 15Hz using this formula.

Frequency Roll off = 1 / (2pi x R x C)?

Posted: 07 Nov 2007, 15:23
by analogguru
What does it roll off at? In my calcs I get 15Hz using this formula.

Frequency Roll off = 1 / (2pi x R x C)?
Have a loook at the
Capacitor Reactance Table

As you can see, a cap with 1µF has a reactance of 8 kOhm at 20 Hz. With a load of 10k (potentiometer only) you have a voltage divider 8k/10k that meens approx. half signal or a loss of 6 dB at 20Hz. A loss output voltage of 0,7 would be 3 dB which is normally used for corner frequencies.

Since many stages can be cascaded, I am always calculating for a loss of only - 1,5 db. My "easy-remember"-formula is:

"A load of 220k requires 220n coupling capacitor for no loss."

That means load is 470k half cap-value necessary: 100n
load of 100k = double value = 470n
10k leads to 4µ7.....

easy..... hmmm ? :wink:

analogguru