GSSL with switchable Turbo and Super SideChain boards

Hi,

I realize my first gssl step by step and I'm pretty happy with the result...it sounds good and even more since I plugged the Turbo board from expat !

All went well until I try to wire the two SideChains boards that I made from this schematic : http://sounddiy.free.fr/images/SSL/Doc/Sidechain/Sidechain_Sch.pdf

I have read many post on this forum and read and read again the explanations written in the document of expat audio to connect the 2 SSC boards with the turbo in switchable mode but no way...SSC does not work  :'(

You will find the design of my actuel wiring in attachments. Hope you can tell me where I'm wrong 

Regards.
Vince

Welcome Vince,
please excuse my edit to your drawing.
(This would better be placed at the 'The Lab' section in the related gssl-, turbo- or sidechainfilter help thread.)
Good luck

Harpo for "President" !

peace
r2d2

Hi Harpo,
I have post my question to the lab section.
How can I remove this post now ?

vince said:

Hi Harpo,
I have post my question to the lab section.
How can I remove this post now ?

Click to expand...

Must be a different question as your initial question should have been answered already by my 'edit to your drawing' when you click on this hyperlink.
Just let this post die.

at r2d2: please edit this nonsense out.

Thanks a lot Harpo,

sorry but I did not notice changes in the design at the first read... :-[

I'll modify my wiring asap.

One more question: I noticed that you had specified R7 + RV1 = 12K. Does that mean I have to change the R7 resistor (actually 1K) for a higher value ? Pretty sure the answer is yes...

Thanks again.
Vince

For usual you want unity gain behaviour for the passband in this spot, so R7+RV1=R6=12K. Unfortunately there is no perfect part and trimmers come with tolerances up to +/-20%, so this 10K labled trimmer might only be a real world 8K trimmer. Using a maybe 8K2 for R7 and your RV1 dialed in to a little less than center position should work on paper, if the trimmers wiper (doubtfully) never loses contact (this R7+RV1 is the opamps feedback resistor and when interrupted/failing will leave you with an opamp running open loop, only limited by its supply rail voltages). Using a single 12K fixed resistor instead of the R7+RV1 series string will give you less trouble and a spare 10K trimmer that might come handy for another project.

Thanks for these explanations Harpo.

I did the wiring as you described on the drawing and this time it works but not really as expected.

- When the toggle switch is on the OFF position, the threshold does not work anymore but makeup gain is still active. I assumed that the OFF position must only bypass the SSC. Am I wrong ? (my gssl also include a True Bypass....don't know if it matters)
- All the others positions on the toggle switch seem to make it sound similar, included the external (even if I have nothing plugged). I attach my own pcb design based on Greg's drawing http://sounddiy.free.fr/images/SSL/Doc/Sidechain/Sidechain_Sch.pdf. I may have made a mistake and I do not see it.
- The threshold seems to compress later than before (gssl with turbo only)

For the moment, GSSL with turbo and no SSC boards sound better.

Thanks again.

vince said:

- When the toggle switch is on the OFF position, the threshold does not work anymore

Click to expand...

Which toggle switch? GSSL/Turbo mode switch from previous wiring drawing with uncommon switch pole not at the center pin (exchange wires to pins1/2 for a common toggle switch pinout)? GSSL-bypass switch? Auto-cowbell-mute switch ;D?

but makeup gain is still active.

Click to expand...

So you fitted a link at control-pcb instead of the optional other half of the DPDT or rotary bypass switch between makeup pot wiper and 'E' to disconnect makeup gain.

I assumed that the OFF position must only bypass the SSC. Am I wrong ? (my gssl also include a True Bypass....don't know if it matters)

Click to expand...

Doesn't bypass, but with previously described parts changes you'll have two inverting opamp stages at unity gain in series, giving non-inverting unity gain for your SCF bypass setting. Your hardwire bypass has nothing to do with this.

- All the others positions on the toggle switch seem to make it sound similar, included the external (even if I have nothing plugged). I attach my own pcb design based on Greg's drawing http://sounddiy.free.fr/images/SSL/Doc/Sidechain/Sidechain_Sch.pdf. I may have made a mistake and I do not see it.

Click to expand...

I'd route it less space consuming, but no obvious pcb errors spotted (except L-R input shorted, but wouldn't matter with 2 SCF boards connected seperately). Compare signal level between L-in (does signal arrive here) and U1-pin1. Should read the same, but inverted polarity (assuming R1 and R2 both 20K in parallel by the shorted L-R-input or only R1 10K fitted from previous wiring drawing). With SCF rotary in position 1 for bypass (you have set the rotary end-stop washer to pos.6 after turning the rotary full CCW?) and R7+RV1=12K, signal level between L-in and U1-pin7 should measure the same with polarity back to non inverting. Case not and opamp measures +/-15V supply rail voltages at pins 4/8, NE5532 might be broken.

Hi Harpo,

I'll try to clarify because I said a lot of stupid things  :-X

- When the toggle switch is on the OFF position, the threshold does not work anymore

Click to expand...

It's about the SSC switch (Off, 60HZ, ..., Ext). But it wasn't for the OFF position but the external position (it's seem I have wired my lorlin switch up side down). So it's normal. Sorry for the confusion.

but makeup gain is still active.

Click to expand...

Due to the previous mistake, it's OK.

Doesn't bypass, but with previously described parts changes you'll have two inverting opamp stages at unity gain in series, giving non-inverting unity gain for your SCF bypass setting.

Click to expand...

I think I understand. In OFF position, no filterring but the unity gain become non-inverting.

Your hardwire bypass has nothing to do with this.

Click to expand...

OK

(except L-R input shorted, but wouldn't matter with 2 SCF boards connected seperately).

Click to expand...

The 10K was a shotcut for the drawing I did. In real life, the two 20K are shorted.

Compare signal level between L-in (does signal arrive here) and U1-pin1.

Click to expand...

Sorry for this dumb question but how can I measure this ? If it's a voltage measure, I have nothing between L-in and ground.

Thanks again for your help. I believe....it'll work!

vince said:

Sorry for this dumb question but how can I measure this ? If it's a voltage measure, I have nothing between L-in and ground.

Click to expand...

Feed your powered on GSSL input with a maybe 1Vrms 500Hz audio/sinewave signal. Measure AC voltage (1Vrms) at the left side of your lifted 47Ks or at the balanced line receivers NE5534 pin6. Same voltage should arrive at both your L-in terminals at your SCF boards, case not you missed the wire in between. Continue 'With SCF rotary in position 1 for bypass ...' as previously written.

Sorry Harpo if I don't understand as quick as I should...I try to do my best

Feed your powered on GSSL input with a maybe 1Vrms 500Hz audio/sinewave signal.

Click to expand...

Do I need a special equipment to do this ?

Measure AC voltage (1Vrms) at the left side of your lifted 47Ks

Click to expand...

By lifted you mean removed ? (English and me !!)
The AC measure will be done between this point and the ground. That's it ?

or at the balanced line receivers NE5534 pin6.

Click to expand...

NE5534 on the GSSL board ?

- I checked my resistances values and R6 is closer to 11K than 12K. Is this critical?
- is that a capacitor may be dead without it being visible (inflated) ?

Hope you will not loose your patience and Kick some...

vince said:

Feed your powered on GSSL input with a maybe 1Vrms 500Hz audio/sinewave signal.

Click to expand...

Do I need a special equipment to do this ?

Click to expand...

No. Amplitude and frequency is not critical. Feed it with maybe a single steady note from a keyboard or organ or use testfrequency wavefiles from wavelab/soundforge/... 

Measure AC voltage (1Vrms) at the left side of your lifted 47Ks

Click to expand...

By lifted you mean removed ? (English and me !!)
The AC measure will be done between this point and the ground. That's it ?

Click to expand...

Yes and yes (from your drawing you spelled it LIFT).

or at the balanced line receivers NE5534 pin6.

Click to expand...

NE5534 on the GSSL board ?

Click to expand...

Yes (or at the lefthand side of the removed 47K resistors).

- I checked my resistances values and R6 is closer to 11K than 12K. Is this critical?
- is that a capacitor may be dead without it being visible (inflated) ?

Click to expand...

not critical, but a little less level (0dB in -> 0.76dB out)
dead cap probably not.

Hi Harpo,

thanks for these explanations. I join my results in attachment. (500Hz Sine with TestGenerator)

Hope you will understand what's wrong 

Just to be sure. The voltage measure between Pin6 and Gnd for the 2 NE5534 on the GSSL board are those at the bottom (left hand side) and not those just under the two 47K resistors (cause for these I have no voltage on Pin6) ?

Regards.

Your meter seems a little insensitive. Nothing obviously wrong with your numbers. When switched to External mode, original signal gets interrupted as it should. With no (external) signal at the external-in, U1-pin7/gnd is 0V (inverted 0V is still 0V). Compared to your 1.4V input signal (only of academical use, as signal is interrupted by the switch) is your 1.4V readout. All fine.

Voltage at pin6 of your controlvoltage attenuator buffer at the DBX202_to_THAT2180 substitution circuits is close to but not 0V when compressing. (will be -6.1mV=-0.0061V per dB attenuation).

Screw your SCF boards into your box, close the lid and compress the s**t out of it.

Your meter seems a little insensitive. Nothing obviously wrong with your numbers. When switched to External mode, original signal gets interrupted as it should. With no (external) signal at the external-in, U1-pin7/gnd is 0V (inverted 0V is still 0V). Compared to your 1.4V input signal (only of academical use, as signal is interrupted by the switch) is your 1.4V readout. All fine.

Click to expand...

So technically it works and I think it works from the beginning (since I have rewired with the modifications you've made on my drawing). I just did another test with a 50Hz sine signal and I see my db meter move depending on the SSC switch position....what I could not see with a 500Hz sine signal (out of range for the SSC filter). However, of a musical perspective I still hear no difference with or whitout SSC whereas with or whitout turbo it's night and day....I'll have to go check my ears!

Voltage at pin6 of your control voltage attenuator buffer at the DBX202_to_THAT2180 substitution circuits is close to but not 0V when compressing. (will be -6.1mV=-0.0061V per dB attenuation).

Click to expand...

I've checked this and You're right. DC voltage on Pin6 of the two NE5534 near the That2181 increase with the amount of compression.

One last thing, I have to push up the main mix volume quite high to have more than 0v on inputs. For instance, the 1.4V I measured previously were with a so high level that I had to cut off my speakers (despite the fact it was a noisy sine signal !!). And I suppose that the more the input level is, the "best, earliest" it compress. For a reasonable volume of work in my room, it seems that the input level is quite low and I have to turn the treshold to half way in order that it becomes to compress. Hope it's clear enough to understand my feeling 
My audio chain looks like this: cubase (Master 0db) => Mackie Onyx internal firewire audio card => Mackie Main out with switch level to +4db => GSSL => Dynaudio BM5A with level switch to -10db.

Almost done !