Welcome to Solid State Guitar Amp Forum | DIY Guitar Amplifiers. Please login or sign up.

March 28, 2024, 11:24:03 AM

Login with username, password and session length

Recent Posts

 

enhanced orange squeezer type compressor

Started by rring, September 01, 2012, 06:51:04 PM

Previous topic - Next topic

rring

I thought someone might also be interested in my compressor design. It originated from looking at the Orange squeezer, but has more features, namely, totally discrete signal chain, gain control via a H11F1M opto FET and uses a zetex ZXCT1041 current sensor as a ideal full wave precision rectifier. I also have a simple op amp implementation that works fine(for rectifier) but the ZXCT1041 is really great for converting signal level to control voltage- it works reliably up 50Khz or so. The circuit provides control for compression level and out level. Its quiet, has no critical adjustments and uses little power.

J M Fahey

Wow!!! A *totally*new project!!   :dbtu:
Congratulations  :tu:.
Post some MP3  ;)

Roly

As JM said - Wow!  I've worked around broadcast compressor/limiters, and years ago was on the edges of a design for such a product, and I tip my hat to you - much respect.   :dbtu:

Perhaps you could give us a bit of a run down on the IC's you have used and why, and the design problems you encountered and how you solved them, so others can get some insight into the development of this deceptively simple circuit.
If you say theory and practice don't agree you haven't applied enough theory.

rring

I am glad someone likes it (besides me)!

I had some basic goals, discrete signal chain, low noise and nice compression control and not very complicated.
the first problem.. what to use for voltage controlled resistor? The led/photoresistor works fine but you have little control of attack and decay. So I chose the optofet which is linear with respect to current(not voltage) that is why the driver to the optofet is current feedback. It is a current controlled resistor. This part works great with signal swings less than 1 volt, so fortunately here the signal swing across it is  low.

The next issue was generating the control signal. Here you could just used a diode like the Orange Squeezer, but it has offset error and ripple because it is half wave. This creates distortion and limits how fast the decay can be. I chose to use this nifty zetex current monitor IC. It senses difference between the two top terminals(positve or negative) but always give the difference as a positive voltage. This allows it to act like a perfect fullwave recitfier- with very few parts. This IC would be great for any kind of envelope sensing in the audio range.

The op amp I use is special in that it is rail to rail input and output but also can have the input swing way above and below the rails without
phase shifting or slew overloading, originally per its data sheet, I used it as the fullwave recifier and it worked fine. Ultimately, I used the ZXCT1041. Other op amps could be used. In this circuit, it is used as a diode drop compensated buffer for my diode peak detector and as the volage/current converter for the the optofet.

I will post some sound samples this evening.

rring

Here is a compressor demo: Its recorded on a DR1 portable recorder about 3 feet away from my battery amp from an earlier post. It starts with no compressor, then compressor, alternating a couple more times and ending with the compressor. The effect is subtle and the compression level is set about 1/2  of max.

Roly

Quote from: rring on September 02, 2012, 12:54:53 PM
I chose to use this nifty zetex current monitor IC. It senses difference between the two top terminals(positve or negative) but always give the difference as a positive voltage. This allows it to act like a perfect fullwave recitfier- with very few parts. This IC would be great for any kind of envelope sensing in the audio range.

Technising at its best - I  <3) this; and as an old tech, may I take my hat off to you for this particular insight.  It's all good, but this is one of the few great recognitions that happen from time to time.   :dbtu: tres  8)

I've been thinking about a new bench audio oscillator, and this gives me some new ideas for tight level control, many thanks.  :tu:
If you say theory and practice don't agree you haven't applied enough theory.

rring

I'm glad its useful to someone! I used the part for its original purpose and saw an app note describing alternate applications. The part could be used for all sorts of envelope filters or even a high side octave. The only drag is its only availabe as a sot23-5 surface mount part. So for those who feel more comfortable with through-hole parts, may not want to use it.

J M Fahey

Can't you post a couple pictures?
This and the other projects.
I'm intrigued how you mixed SMT and through hole parts.
Or is it all SMT?
Thanks.

rring

Here are some pictures.... I mainly build with surface mount because the parts end up being cheaper most of the time and I can get the boards really dense. I find smt easy enough to hand solder but you need a good set of tweezers and a nice soldering station!

J M Fahey

Mmmmhhhhhh , real PRO, congratulations. :dbtu: :dbtu:
One non-electronic doubt: your PCBs look professionally made (not like the ones we home make), don't they become too expensive in small quantities?
I see the parts IDs are silkscreened so it's not the affordable Express PCB offer.
If you have some convenient PCB supplier tips, we are all ears, that's one stumbling block for homebuilders.
And again:  :dbtu: :dbtu: :dbtu:

rring

Yes they become expensive. Sometimes I break down and pay for them myself but I am fortunate in that I do circuit design for a living and I can often piggy back my experiments and home brew on to other boards I am prototyping for my work. I can just cut them up into the appropriate sub sections. I really try to make my hobby designs as simple  as possible so they can be perf built for someone who does not want to invest in a PCB.

J M Fahey

Oh !! *now* I understand.  8)
I make amps commercially and know the actual costs and minimum quantities involved, that's why I am amazed.
I silkscreen and process my own PCBs but the minimum practical run is 10 or 20 of anything.
And preferrably 50/100.
For prototypes (or stamp size miniboards) I became quite proficient on Thermal Transfer, but results always betray the kitchen table cradle.
Positive photosensitive boards yes, provide quite Pro looking results, but of course still lack parts silkscreening.
Congratulations again.   :dbtu:

jfetter