Forssell Opto Compressor

[hitchhiker]

hi buttachunk,

the neg leg of c3 is grounded.

with the makeup opamp in, I get changes in current to the SC opamps when turning the gain makeup knob.

without the makeup opamp I get no change in current when adjusting the gain make up knob.

cheers,
Lance

 

[FredForssell]

with the makeup opamp in, I get changes in current to the SC opamps when turning the gain makeup knob.

without the makeup opamp I get no change in current when adjusting the gain make up knob.

Have you checked to make sure that the gain makeup opamp isn't going into oscillation?

Also check for oscillation on the side-chain opamps. You may need to adjust the compensation cap values.

 

[hitchhiker]

I don't know how to check for oscillation, let alone know what values I might adj with.
I'm a rookie and over my head trying to solve this.

cheers,
lance

 

[hitchhiker]

Buttachunk,

Tried the cap tests and no change. I also measured the voltage accross the C3 terminals (without the cap in place) and it reads +.002v with no change if I adj gain knob.That is +.002 volts to the amp side. So the C3
cap grnd would be correct to grnd .
I also tried a bipolar at C3 but it gave the same results.

Without C3 in there is adequate makeup but no increase with gain knob.

Maybe Fred's right about oscillation. I have no scope just an analog tester,
so if there is a ball park compensation I can try , I'll do so?

I thought maybe it was related to a very low output impedence of my console inserts so I tried audio straight out of my soundcard but it acts the same.

cheers'
Lance

 

[hitchhiker]

hi Buttachunk,

put in .01u for C4. no change.

Dropped into the side chain circuit 100p caps accross pins 6&7 of the first opamp (1&2 is already bypassed & I tried a .01u here too) also 1&2 and 6&7 of the second opamp .

result is everything behaves the same.

How are you for the parts for this side chain? Do you need a couple of these optos to breadboard and test this circuit?

cheers,
Lance

 

[hitchhiker]

cool,

Is perkinelmer still the only place to buy these?

Lance

 

[Svart]

we are using the Vactec VTL 5C1 for the Opto right?

http://www.hoffmanamps.com/

has these for 10$ us. I don't know if that is a good price but i don't know too many places that carry these..

 

[Svart]

even better.. :green:

 

[mnats]

Even slightly cheaper from Allied:

Mfr.'s Part #: VTL5C1
Allied Stock #: 980-0700
Manufacturer: PERKINELMER
Description: OptoIsolator, Analog
Current Page #: 1199*
Previous Page #: 504*

Qty. In Stk.: 761

U / M:
1 Ea.
Min. Order Qty.: 1

Request Lead Time

1-49 $4.070 Ea.
50-99 $3.660 Ea.
100-149 $3.330 Ea.
150-249 $3.050 Ea.
250+ $2.760 Ea.

 

[hitchhiker]

mnats ,
beat me to it . thats where I got mine.

hey buttachunk,
When I dug up my invoice I learned that mine are VTL5C9 not 5C1
and that maybe is why I'm having problems with the makeup/GR interacting. oops!

Input
Current
(mA)

VTL5C1 = 40 ma

VTL5C9 = 2 ma

What do you think?

Lance

 

[hitchhiker]

hi buttachunk,

yes it will be interesting, I havn't compared all of the differences but
when I was measuring current I never found anything above .5 ma.

The transformer is a Hammond rated at 50VA.

cheers,
Lance

 

[hitchhiker]

hi buttachunk,

I looked over the PDF data sheets at Perkinelmer and I'd say the vtl5c9 is
going to be hyper sensitive to current change in Fred's SC compared to
his specified vtl5c1.

so your makeup design is very likely going to work as you expected with the correct optos in the SC.

cheers,
Lance

 

[Svart]

yo Tommytones, how're the PCBs coming my friend?

 

[mnats]

Looks like hitchhiker's version with Fred's sidechain and butta's makeup is the only one breadboarded and working yet...I made a layout myself of this combo so I can try it out soon:

Redrawn Schematic (29kB PDF)
Checkplot (28kB PDF)
Bottom Layer (20kB PDF)

All the pots can be pc mount or wired from the board. The Vactrol VTL5C1 is there instead of the Silonex proposed earlier. I plan to make a single-sided layout for Fred's makeup stage once someone has built and tested this too.

Hope this helps in the meantime...

 

[FredForssell]

I just looked at the schematic at mnat' link and have a couple of comments that you can take with a large a grain of salt as you feel necessary.

I would give some further thought to th gain make-up section of this design. My concern is based on the fact that as shown, this stage (U3) is a variable gain amplifier with a gain range of appox 100 (40 dB). My personal feeling is to avoid using variable gain amplifiers where ever possible because, imo, changing the feedback factor changes causes the sonic quality of the amplifier to change as well. This means that the sound of the circuit is, in part, a function of the gain at which it is set. That's too many variable for me. YMMV. It is possible that there could be stability issues if the circuit as drawn if not tested carefully upon completion for oscillation and compensated correctly.

Personally, I would change VR6 to a fixed 10k value, change R12 to appox 2k (for a fixed gain of about 15 dB), and decrease the value of C16 to about 47 uF.

For level control I recommend a 100k log pot in parallel with the LDR cell of PR1. This will cause a very minor amount of attenuation working against the 6k series R (R8+R9). I'd then reduce R14 to about 300 ohms.

Also, as drawn, the amplifer is operating with 100% DC feedback and I do not think that you need to worry about using an output coupling cap. Also I am guessing that the electrolytic coupling caps are 22uF not 22 nF as shown. I would use a far smaller value on the input side and eliminate the output cap.

The above changes reflect my personal design philosophy of try to design sonically neutral equipment. If you wish to have your compressor contribute some sonic flavor, you will probably want to ignore the above statements. None the less, I think it is something to think about before the circuit boards are made. While you can use shorting jumper in place of capacitors, changing the way the level is control in the gain make-up amplifier is less easy to modify once a board is built.

Something to think about anyway.

 

[hitchhiker]

hi mnats,

I noticed in your re-drawn schemo that the makeup amp INPUT comes off of conn 2 where as in buttachunks schemo it's input is off of conn1.

Which way should it be ? Fred didn't seem to have issues with the way
you have drawn it.


hey buttachunk.... first thing I'm gonna try is your design moved over to take input from conn 2.....it may stop that interaction thing.


Then I will try Fred's suggestions for changes in the makeup as I too would like to hear this compressor as neutral as possible .



cheers,
Lance

 

[hitchhiker]

hi buttachunk,

I just tried out taking the input to the makeup amp from conn2 with no caps in or out and there was no effect on the sidechain and meter circuit.

but.... the wrong optos now are too touchy to dial in at all.The SC went from smooth tuning to useless. I guess the effect of the makeup amp input from conn1 changed the currents of the SC to be right for the 5C9
optos. It really worked very well and was easy to adjust to however much
ratio and threshold you wanted. a cool accident !

cheers,
Lance

 

[StephenGiles]

Can anyone suggest what alterations the circuit would need for VTL5C4/2 or NSL 32 optos? These are all I have.
Thanks
Stephen

 

[hitchhiker]

buttachunk,

I'll order some VTL5C1 later this week, after I add some things to my list.

Also gotta hear it as it's meant to be heard. hey hey :thumb:

cheers,
Lance

 

[Gustav]

Ill be more than happy to make these boards. If you come up with something double sided that wouldnt be a problem (I just scanned thru the pages and thought I saw some mention of that)

I could also start out by making just a few boards for testing if that is neccesary.

Got this one downloaded optopcb_B.zip. Is that of any interest on its own, or are you implementing that part into your design ?

Gustav