Forssell Opto Compressor

[soundguy]

It would be worthwhile to keep the audio path discrete...

dave

 

[hitchhiker]

I agree Soundguy,
myself I'll be into the Forssell amps but I thought there will also be a cheaper version for trying it out etc.

cheers
Lance

 

[tommytones]

Ya, we should have the 2 options here. Just like the output. The circuits are relatively simple so the component count is low therefore accomodating 2 different input and output circuits is possible.

 

[Svart]

My vote is for both, However I still think something similar to the GSSL i/o is simple and effective. it's even modular enough to likely be cut n' pasted from that pcb to this one in my opinion. just my 2cents..

 

[hitchhiker]

I'd vote for the gssl i/o if it works.(and the discrete i/o too, I mean to say both)

cheers,
Lance

 

[hitchhiker]

hi buttachunk,

I just tried a tl071 (which I had forgot was jfet input) in the makeup and again dropped the voltage via resistors. in this case 3k9 gets 16.4v.

the amp is distorting a bit at lowest setting and gets much worse as turned up. the lowest setting is also a bit too much makeup.

Is using resistors to drop the sup v creating the problem.Is this a bad idea?

I'm also noticing that whatever makeup chip is used it effects the GR meter as it is turned up.I don't know if I just didn't see it before or what?

cheers,
Lance

 

[hitchhiker]

I have a couple of 15v zeners. Would I connect the pos end to the pin 4 on the neg supply and the neg end of the other to pin 7?
I only have two so I don't want to screw up.

As far as the interaction of the makeup setting goes , it is also happening with the opa604 when no resistors are used on the supply v.

 

[Svart]

I'd like to keep this one simple as I would likely be using it on many channels, so maybe have a different version later with the iron? I'd rather see this first batch as simple as possible so that we can all play and add things as we work on them.

 

[hitchhiker]

hi buttachunk,

that created a cool smoke show on the current limit resistors....?

with the oter stuff hooked up to the reg +- 24v there must be some
problem?

I decided to go back to the 604 at 24v for everything and check that
GR meter sensitvity to gain makeup.

It's still there. Where should I measure the current?

Lance

 

[Svart]

my guess is that it wasn't a high enough wattage resistor or the correct resistence.. try 5.6k 1w, that always seems to be a good start for me..

 

[hitchhiker]

isn't this going to change the voltage for the rails and therefore the sidchain and meter section too?

Lance

 

[tommytones]

Okay, I'm getting a good idea now. I'm gonna take a stand and declare usage of Jakob's input stage from the ssl clone. I'll do it (try anyhow) in such a way that you can jumper the signal path if you don't want to use the input buffer. As for the discrete buffer, I'll just use an API schem I have which is very similar to Jakob's circuit minus a couple of components. I think this should work out just fine with Fred's amps. Either way, very simple and proven.

Stay tuned for the next updated layout. Should be tomorrow at some time. I'll let y'all know.

 

[DrFrankencopter]

The resistor is there to force a particular current draw from the circuit, which gets the zener into a range where it's dynamic resistance is lower.

Zener current is given by (Vin - Vzener)/R , so if R is real small you have lots of current flowing through, and power = I^2R, so it doesn't take long before you get past the limit of a 1/2 watt resistor. Try a 300 ohm resistor, that should only be at around 0.3 Watt for a 10 volt drop.

Cheers,

Kris

PS: What value did you use? 10 ohms is way too small...100 ohms is 1 Watt for 10Volts drop.

 

[hitchhiker]

The resistors were 20r 2W and they are toasted but the zeners are OK I think.

Does this kind of circuit leave the voltage of the rails that feed it at their
same +-24v?

 

[DrFrankencopter]

Yup, do the math and you'll see that you put 2.5 Watts into those resistors.

Will it leave the rails at 24V...that depends on how stiff the rails are. If you draw too much current, then you might sag the 24v rails, but if it's a well regulated supply it shouldn't be a problem.

Cheers,

Kris

 

[hitchhiker]

thanks for the lesson Kris , I'll give 300r a try .This is a good thing to learn. (Vin-Vzener)/R and power=I^2R

cheers,
Lance

 

[hitchhiker]

hey buttachunk, the smoke has cleared! No one got hurt.

I put in 330r 1/2 W and tried again........and no smoke this time.

the voltages all checked out and then I auditioned the tl071. I prefer it to the opa604 . there is some interaction still, when I turn up the makeup
the meter reads more GR. It's not as bad as with the 604 though.

It's a good size power tranny and the filtering is 4700u caps before reg,
220u & .1u after reg. I've used this supply at +-28v for quad8 ca137's
lots with no prob.No filter caps near the chips.

the sound with the Forssell opto is very quiet really.

I measured current accros the opto resistors ,aprox .40 ma. I think.Yes it changed when I turned the makeup control.

Where else should I measure?

cheers,
Lance

 

[hitchhiker]

right-o, I'll try it in the morning .

cheers,
Lance

 

[hitchhiker]

buttachunk,

I tested with 22u on input and still same.

added 100u on output ,so now same as your original with caps makeup
amp and still the gain adj effects the GR meter .

I double checked my bread board wiring and found no errors?

I'm gonna look again for errors.

Lance

 

[hitchhiker]

hi buttachunk,

yes the artwork isn't clear but I did figure out what you wanted. All cap and r values are exact same as you have specified.

only changes I've made is on Freds side chain because I have no .33u caps so I used .22u

My next closest available choice today for non polar would be up to 1u

Lance