GSSL HELP THREAD!!!

[Harpo]

1 -hrmm.. which one would be the 1300 ms attack? 1m * .47u?..
2- how did you calculate the attack for each stage? id like to do that to be more precise!!
tbh i wanted actually one with attack VERY sharp (around 0.05ms one with mild attack (around from .2 - 1ms) and a long one (around 60 - 150 ms (to have a longer attack time for kick drums etc...)

sorry, misread your obviously part of the part of the release section. About 1.3sec attack would have been 270k series R from the attack switch with 6.8uF cap to gnd. [270k * 6.8 * 10^-6 / SQR(2)]. No exact numbers, as these are depending on level and current drive ability of the sending stage.
Your wanted VERY sharp attack around 0.05ms, sure, what driving stage? The TL074 won't do this.

4- the variable ones are rheostats ??

3 normal pots, wired as rheostats, substituting the 2 lorlin type rotary switches and the 3M3.

5 -and the 820R would that be the one on the attack switch pos1?

yes, min.value, couldn't get much lower with a TL07x in front.

6- the (auto) that you written there, does that mean the 6.8u loop is the second release time? (kinda the slower one of the 2)???

my sketch is more or less the same as the complete GSSL attack/release timing section, but variable without steps. For release, dialing the lower pot across the 6.8uF full ccw (shorting out the cap) giving a timing in range of previous steps 1 to 4 from the release switch, controlled by the upper pot across the 0.47uF cap. Dialing the upper pot full ccw and the lower pot cw giving a timing of previous step 5 for auto. Mix it to taste.

 

[ptron]

wow thanks so much for the detailed explanation!! ill look into this but im not sure i want to change the original design so much hehe ..

ok i realize now im more of a noob than i gave myself credit for.. i got to read more.

sorry, misread your obviously part of the part of the release section. About 1.3sec attack would have been 270k series R from the attack switch with 6.8uF cap to gnd. [270k * 6.8 * 10^-6 / SQR(2)]. No exact numbers, as these are depending on level and current drive ability of the sending stage.
Your wanted VERY sharp attack around 0.05ms, sure, what driving stage? The TL074 won't do this.

so you are saying it is indeed going to be 1.3 sec of attack or that you misread and it will keep its original attack?
i thought if i kept the values of .47 and 6.8u the attack would remain like it was?

i guess my question is.. for that formula is the attack time? or release time? or both? i dont understand

sorry for all the questions and thanks for your patience! good things will come to you if i end up understanding this! (from karma tho..  i cant make anything happen) ;D

this helps? http://www.northcountryradio.com/PDFs/column008.pdf

thanks!!

 

[Harpo]

so you are saying it is indeed going to be 1.3 sec of attack or that you misread and it will keep its original attack?
i thought if i kept the values of .47 and 6.8u the attack would remain like it was?

1.3sec attack for constant level pure sine for my misread part of your partly schematic, which I already appologiesed for. Your program might/will be not and other more likeley varying waveshapes would rise the divisor /SQR(2) to a higher number. As already said, no exact numbers.

 

[ptron]

thanks a lot harpo i know my schematic drawing skills arent the best hehe apologies for that ;D

would this be more or less somewhat aproximate to the values we'd be dealing with? i found some sort of calculator online..

[ 1 ] 6.8 microfarad * 15000 ohm
    = 0.102 s
   
[ 2 ] 6.8 microfarad * 10000 ohm
    = 0.068 s
   
[ 3 ] 6.8 microfarad * 7500 ohm
    = 0.051 s
   
[ 4 ] 6.8 microfarad * 4700 ohm
    = 0.03196 s
   
[ 5 ] 6.8 microfarad * 3300 ohm
    = 0.02244 s
   
[ 6 ] 6.8 microfarad * 2700 ohm
    = 0.01836 s
   
[ 7 ] .47 microfarad * 1500000 ohm
    = 0.705 s
   
[ 8 ] .47 microfarad * 15000 ohm
    = 0.00705 s
   
[ 9 ] .47 microfarad * 1500 ohm
    = 0.000705 s

thanks again!

 

[Harpo]

i found some sort of calculator online..
[ 1 ] 6.8 microfarad * 15000 ohm
     = 0.102 s
....

??? Your OS or your mobilephone would have a calculator on board for such heavy math.
[R*C=t] with R in ohm, C in farad and t in sec. 15000*0.0000068=0.102.
You could multiply this 0.102s with 1000 to get a value in ms if you like it better.
Most often it is easier to calculate with cap values written exponentially (imagine a cap value in the pF range, thats 6 additional zeros after the decimal dot), so 6.8*10^-6 is the same 6.8uF as 0.0000068F.
Multiplier for nF would be *10^-9 and for pF would be *10^-12 to get a F reading.
Time would be 102ms for attack if the source was DC, but audio is AC, and time would be 102ms for release if the cap was fully charged, so this aproximation will be somewhat off, especially in this auto release timing circuit with another R/C in series.

 

[ptron]

this is what doesnt add up to me...  (this are the good ol RC circuits from the release) up to the 4th step of the filter...

1- how did jakob ended up with .1, .3, .6 and 1.2ms from 84, 126, 263 and 564 ms ??
2- should i expect this kind of difference on my calculations or is the 3m3 resistor doing something there that im not calculating?? ;D


     T= R*C                                                 GSSL

[ 1 ] .47 microfarads * 1200000 ohm
    = 0.564 s                                             0.0012 s
   
[ 2 ] .47 microfarads * 560000 ohm
    = 0.2632 s                                            0.0006 s            
   
[ 3 ] .47 microfarads * 270000 ohm
    = 0.1269 s                                            0.0003 s
   
[ 4 ] .47 microfarads * 180000 ohm
    = 0.0846 s                                            0.0001 s
   
EDIT: ---- K i think i cracked it.. ill post some more in the near future to double check ---

3- i know im going a bit berserk with this but when im mastering i normally use more than 30ms attack time so most transients go thru the compressor without being squashed instantly... is there a way to calculate attack time for this circuit in order to step it up to 8 steps for attack and add 2 longer attack times?    

 (EDIT)... if the attack keeps going in that proportion... 510k = 60 750k = 90 ms....iwould this be a bald assumption??
eg..
      RES   |    SCHEM  |in my numbers                    
              |               | using .1ms as reference
     820R   |   .1ms     |                                    
      2k7    |   .3ms     |  .329                              
      8k2    |    1 ms    |   .99                            
      27k    |    3ms     |   3.29                          
      82k    |     10ms   |   9.9                                  
      270k   |    30ms   |  32.9                                  
     
       
      510k *|    60ms * |    62.1                                
      750k * |    90ms  *|   91.3                                

with this calculation:  2k7 = time i counted 820R =.1 as TRUE)


thanks a lot!

 

[ptron]

ok numbers...

would these charts look more like it??


ATTACK

      RES | "T" in pannel  | "T"  in my numbers                    
            |                   | using .1ms as reference
     820R   |   .1ms       |                                    
     2k7    |   .3ms       |  .329                              
     8k2    |    1 ms     |   .99                            
     27k    |    3ms     |   3.29                          
     82k    |     10ms  |   9.9                                  
     270k   |    30ms   |  32.9                                  
     ---------------------
    - 470k*  |    55ms * |   57.2  
      510k *|    60ms  *|    62.1                                
    - 620k * |    75ms * |   75.4
     670k * |    80ms  *|   81.4
    - 750k * |    90ms  *|   91.3
*not in original
-+-+-+-+-+-+-+-+-+-+-+-+-+-+

RELEASE

RC loop 1                        RC loop 2

RES   CAP     T      |      cap     T

22k           0.02            6.8     0.58
27k           0.028          6.8     0.71
47k           0.049          6.8     1.2
62k           0.066          6.8     1.5
82k           0.08           6.8     2.1
91k                             6.8   2.4(*)
100k   .47    0.107ms      6.8    2.6
180k   .47    0.19     |    6.8    4.7
270k   .47    0.289    |   6.8    7.1
360k   .47    0.385    |   6.8    9.4ms
470k   .47    0.49     |    6.8    12.3
560k   .47    0.59 (0.6*)  "       14.7
620k      "    0.66     |      "    16.3
750k     "     0.8      |        "  19.7
1m       "     1.05     |       "   26.3
1m2      "     1.2      |      "    31.6
1m8     "      1.89     |          47.5ms
2m3           2.4      |          60.6
3m3           3.47     |          87  
3m9           4.1ms    |          102.8

(*) used as reference for the rest

    if this is true then i can start etching the pcbs for this .. !! also SSC pcb just arrived on the mail today!! started soldering but since this is my first project where it involves more than 10 resistors and a couple of opamps.. well im a bit excited/nervous/anxious/ ;D havin the time of my life crunchin this numbers  

btw how fast can the tl074 perform? u said i need other driving stage to reach faster attack 0.05ms kinda times (i assume release too?)

thanks!

 

[ruckus328]

Just my 2 cents, build first, then modify  Since this is your first build and you are already incorporating Turbo and SSC, there's a very high likelihood of issues just trying to get a default unit up and running.  Some people get lucky, but many do not (myself took a week of troublshooting to get my first GSSL working with just turbo mod).  By all means, continue your efforts for adding your own additions, but I would leave them out of the equation until you have it working first.

 

[ptron]

thanks a lot for that ruckus i know u are right and have thought of just building the unit as it is for now .. still i feel the urge to keep workin on this thingy i think its just i want to learn very much and am getting whatever tangent i can to try and understand this compressor and electronics in general.. heh
btw its only ssc for now that im modding it up.. ssc and crc

thanks again ! i wont mod my compressor yet but ill keep trying to make this circuit on ARES.. i already have it in isis but i wanted just one layer board and its making 2 layers :/ going to have to figure that out eventually. but its my first time etchin a pcb.. im going to start with a pc osciloscope safety device so i dont blow up my soundcard. and will go from there.

anyways do u see those numbers in those charts "right??" or within reason?

thanks

 

[stereokillah]

hi there,

i built a gssl with turbo and crc for a friend who i s in another region, before sent it I ve decided to put 5534 signetics instead of texas but i did a mistake and and I put them in place of the 5532

everything worked great since a postman play with it like football

now my firend sent me back after changed a gain pot and some parts damaged i have a noise with a frequency of 800hz and 400hz

i checked the signal of my ground it's ok
tried another psu toroid
changed 7815 and 7915
changed the 1000uf
changed all the dip
still have this sound

it's not from the turbo,not from the crc not the control board


when the unit is powered off still have this sound since i disconnect the toroid

got 15v and -15 even 12v and -12

every thing work nice just have this bad noise


i tried another xlr wire in case of but nothing



if someone could give me a direction

thanks

 

[tomcat]

Hmmm... I assume a 7915 should output -15v even when there is -26 volts on the input? Datasheet says up to -35V is ok. On all my 7915 they stay at -15v up to around -22volts input, then they go up with the input voltage. So my -26volts (idle) on input create -22volts on output.

7815 is fine. 7812/7912 measure also fine.

Did i get bad 7915 or is this behavior normal in idle (gssl board without semis)??? Trafo is 2x18volts.

 

[stereokillah]

Hmmm... I assume a 7915 should output -15v even when there is -26 volts on the input? Datasheet says up to -35V is ok. On all my 7915 they stay at -15v up to around -22volts input, then they go up with the input voltage. So my -26volts (idle) on input create -22volts on output.

7815 is fine. 7812/7912 measure also fine.

Did i get bad 7915 or is this behavior normal in idle (gssl board without semis)??? Trafo is 2x18volts.

every time i use 7815 and 7915 CV each time i used another i had problem with the voltage, check the datasheet of yourmodels the input voltage seems to hicg for your 7915

 

[tomcat]

Thats the problem, datasheet (its an ST L7915CV chip) says for -5 to -18 Vo up to 35 Vi is ok. Like i said, it starts to go up beyond 15V with 22V on input, this margin is imho way too low. Maybe i got some chinese copies

Last time i ordered spares for all 78 models and i forgot to order 79 spares. Arghhh  :-\

 

[Harpo]

..Did i get bad 7915 or is this behavior normal in idle (gssl board without semis)??? Trafo is 2x18volts.

Min.load current is 5mA for 78xx or 79xx. Sometimes they might throw out the correct voltage without load, but don't count on it. If you wanna be sure without chips in, just temporary put a load resistor across +15V/0 and 0/-15V -maybe 1k5 for a 10mA load- and measure again.

 

[stereokillah]

bump

hi there,

i built a gssl with turbo and crc for a friend who i s in another region, before sent it I ve decided to put 5534 signetics instead of texas but i did a mistake and and I put them in place of the 5532

everything worked great since a postman play with it like football

now my firend sent me back after changed a gain pot and some parts damaged i have a noise with a frequency of 800hz and 400hz

i checked the signal of my ground it's ok
tried another psu toroid
changed 7815 and 7915
changed the 1000uf
changed all the dip
still have this sound

it's not from the turbo,not from the crc not the control board


when the unit is powered off still have this sound since i disconnect the toroid

got 15v and -15 even 12v and -12

every thing work nice just have this bad noise


i tried another xlr wire in case of but nothing



if someone could give me a direction

thanks

 

[tomcat]

..Did i get bad 7915 or is this behavior normal in idle (gssl board without semis)??? Trafo is 2x18volts.

Min.load current is 5mA for 78xx or 79xx. Sometimes they might throw out the correct voltage without load, but don't count on it. If you wanna be sure without chips in, just temporary put a load resistor across +15V/0 and 0/-15V -maybe 1k5 for a 10mA load- and measure again.

Yeah a friend pointed me to that fact already. Will check again.

 

[Harpo]

bump

If you don't have a scope, you could try to tracelisten to locate the fault.

 

[tomcat]

..Did i get bad 7915 or is this behavior normal in idle (gssl board without semis)??? Trafo is 2x18volts.

Min.load current is 5mA for 78xx or 79xx. Sometimes they might throw out the correct voltage without load, but don't count on it. If you wanna be sure without chips in, just temporary put a load resistor across +15V/0 and 0/-15V -maybe 1k5 for a 10mA load- and measure again.

Oh yes, reading the complete datasheet helps sometimes

Output voltage
IO = -5 mA to -1 A, PO ≤15 W VI = -18.5 to -30 V

EDIT: Thanks Harpo, the Load resistor helped

 

[stereokillah]

thank you harpo, you are always ready to help.

at time i think i found the solution , two 100uf on the output section seems to be die, i would changed it and see if it's case

 

[stereokillah]

it's not case still have the problem what should give a noise with 400 and 800hz

it's not psu not regulator or 1000uf not dips

i will test trace listen but before i'll checked all the point with signal to see if the pcb have a cut somewhere