Script

Re: GSSL HELP THREAD!!!
« Reply #8200 on: January 17, 2020, 07:41:03 PM »
Clip or bend SC VCA pin 4 when using  THAT2181 there. Personally, I prefer bending at a 90 degree angle. Just make sure it doesn't connect.

I haven't checked in detail but difference in distortion should be minimal, as in measurable but not really audible -- above all cos this is the sidechain control signal and not the audio path.


AmosMoses

Re: GSSL HELP THREAD!!!
« Reply #8201 on: January 20, 2020, 04:58:31 AM »
Sidechain is supplied by 12V, audio path VCAs by 15V, hence the difference.

Gustav

On the latest gyraf layout REV11 5k1? Can someone explain what does that do?

AmosMoses

Re: GSSL HELP THREAD!!!
« Reply #8202 on: January 20, 2020, 06:32:40 AM »
Ok so it seems that the compressor works as should, compresses and the audio signal is evenish -0.3 dB difference.

But now I have the problem when I connect XLR to the output mixer or daw i get a piersing signal. it's constant if sidechain is on, and if i turn it off i only get it when i play something.

If i measure with audio probe i get that have that same signal even on the input XLR. but when I disconnect the output XLR from the daw or mixer everything is ok....

Update:

When no playback. treshold turned down and sidechain on turning on the makeup gain makes the meter increase in db from 0 to +20. When the makeup gain is turned down no sound and compression is showing 0 and all the way up, the opposite, compression +20 and a loud sound.

the wave starts at 600 hz

With output XLR disconected i get this signal. it doesn't change besides the output one when the output xlr are connected.

when I send i 1 kHz signal measuring AC between inputs I get 5.9v
when measuring output xrl pins i get 4.0 and 4.2v

and even on the lowest treshold setting i still get 6 dB reduction.

I think my signal is comming in too strong?

UPDATE 2:

I got my input signal to 0.573v at 1000 hz sine wave. i get output 0.638v at output pins when when makupis off and when fully engaged makuep i get 4.2v.

When I listen to the sound without output xlrs inserted i get normal good compressed and working sound, when I insert and output xlr and connect it to any any device audio interface or mixer I get that problem i mentined above...

UPDATE 3: it might have been somethinh with the audio interface... currently testing at it seems it's working :D tomorrow will know more.
« Last Edit: January 20, 2020, 09:26:01 AM by AmosMoses »

AmosMoses

Re: GSSL HELP THREAD!!!
« Reply #8203 on: January 20, 2020, 09:43:51 AM »
Ok, so whole new post as it seems it's working, it was settings on interface that was messing me up.

I even get even channels :D

Now my only problem it seems is that the makeup doesn't boost anything, turned all the way up it makes the signal unity with the input.

And when measuring the AC I get a stronger input signal then the output. But in the DAW I see the input and output is the same signal in dB.

Script

Re: GSSL HELP THREAD!!!
« Reply #8204 on: January 20, 2020, 07:40:13 PM »
So was it the 100 pf caps then?
-----

For makeup gain, first check the makeup gain range resistor on the main board. It's the 620K(?) one. Correct value? And check connectivity between that resistor and 50K makeup gain pot.

Then check for correct values the 470R and 47R voltage divider resistors sitting on the input to the SC VCA.
« Last Edit: January 20, 2020, 07:46:44 PM by Script »

AmosMoses

Re: GSSL HELP THREAD!!!
« Reply #8205 on: January 21, 2020, 09:36:19 AM »
So was it the 100 pf caps then?
-----

For makeup gain, first check the makeup gain range resistor on the main board. It's the 620K(?) one. Correct value? And check connectivity between that resistor and 50K makeup gain pot.

Then check for correct values the 470R and 47R voltage divider resistors sitting on the input to the SC VCA.

It all seems to work.

And in the end all the problems was down to bad component values, bad direction of components, and good cleanup of the soldering and the board. Easy simple fixes in the end.

Even when doing it step by step you you can't be confident enough and not triple check for errors thinking that you were correct in the first place. Triple check the values, orentation and soldering.

A bit frustrated with myself but happy that it was just that.

Now on to add the Turbo and then gonna make me a second one :D

gyraf

Re: GSSL HELP THREAD!!!
« Reply #8206 on: January 22, 2020, 02:20:06 AM »
Quote
A bit frustrated with myself but happy that it was just that.

Don't be hard on yourself - the whole aim of this project is to show you just that, and motivate you to go beyond. Read back in this thread, and you'll see numerous tales of just this.

Once you realize that electronics WILL work as it is supposed to once you have done your part, you're free to go out and do just about anything.

/Jakob E.
..note to self: don't let Harman run your company..

AmosMoses

Re: GSSL HELP THREAD!!!
« Reply #8207 on: January 22, 2020, 11:11:20 AM »
Don't be hard on yourself - the whole aim of this project is to show you just that, and motivate you to go beyond. Read back in this thread, and you'll see numerous tales of just this.

Once you realize that electronics WILL work as it is supposed to once you have done your part, you're free to go out and do just about anything.

/Jakob E.

Thanks :)

Now after instaling the turbo board i get more compression on the same settings with lower ratios then with higher, is this normal? Anyone?

And one thing changed now with turbo added and SC bypassed, when sending 0 dB i get back at daw -6 dB. I think i used to get 0 to 0 if i remember correctly, cant be sure.

And one more thing only the first positions of threshold makes the unit compress, maybe removing the 47k resistor would change that if I want the unit to start compressing earlier?


Nevermind, it was again software related. This digital to analog thing is more tricky then it seems.
« Last Edit: January 22, 2020, 03:19:57 PM by AmosMoses »

AmosMoses

Re: GSSL HELP THREAD!!!
« Reply #8208 on: February 20, 2020, 01:55:59 PM »
Hey, building new gssl and on this one i have a problem that there is no DC on the board. Just ~67v ac almost everywhere. I got this https://www.don-audio.com/Audio-Toroidal-Transformer-Prim-2-x-115-Sec-2-x-15v-Center-tap transformer, and connected it as should. When measured secondaries i get 36 or 17 and 17.

Where should I start looking? I haven't put any IC on.

Potato Cakes

Re: GSSL HELP THREAD!!!
« Reply #8209 on: February 20, 2020, 09:19:13 PM »
Did you install the rectifier right next to the AC input on the board? If you're getting AC in the circuit you are either missing a part, or have a bad solder joint. Look at the schematic. It goes AC to rectifier to the rest of the board. If AC is getting to the board then problem has to be right where power comes into the circuit.

Thanks!

Paul


AmosMoses

Re: GSSL HELP THREAD!!!
« Reply #8210 on: February 21, 2020, 04:37:03 AM »
I know, I just installed a new rectifier... I will try to replace it with new one, maybe i damaged it while soldering... then will see

Replaced with a new bridge, still the same. When not connected to the board the transformer secondaries show 17 and 17 across 36.

When connected to the board the one to the corner shows 52v, the middle one 66v, and the one on the right side 80v AC...
« Last Edit: February 21, 2020, 06:32:01 AM by AmosMoses »

gyraf

Re: GSSL HELP THREAD!!!
« Reply #8211 on: February 22, 2020, 09:36:45 AM »
AmosMos: You are probably measuring wrong. Your reference to measure against IS the center tap secondary of the mains transformer - also our 0V for the circuit.

Lars: that belongs in the  "GSSL add-on help thread"

/Jakob E.
..note to self: don't let Harman run your company..

AmosMoses

Re: GSSL HELP THREAD!!!
« Reply #8212 on: February 22, 2020, 10:45:51 AM »
AmosMos: You are probably measuring wrong. Your reference to measure against IS the center tap secondary of the mains transformer - also our 0V for the circuit.

Lars: that belongs in the  "GSSL add-on help thread"

/Jakob E.

But when measuring voltage on the board with voltmeter I connect one to the ground (chassis) and the other I connect to the point I wanna measure.

gyraf

Re: GSSL HELP THREAD!!!
« Reply #8213 on: February 23, 2020, 02:13:24 AM »
Chassis ground ONLY counts when safely connected to circuit ground, i.e. 0V.

/Jakob E.
..note to self: don't let Harman run your company..

Re: GSSL HELP THREAD!!!
« Reply #8214 on: February 25, 2020, 06:46:42 AM »
Anybody found a loss of 5db compression when switching between Turbo (Oxford) and GSSL (Aarhus) modes  (no SSC board)  ?

Not found why but when I plugged 1 SSC board the compression is at same level.

Now I know why. ;D
« Last Edit: February 25, 2020, 08:28:48 AM by lars on »

Re: GSSL HELP THREAD!!!
« Reply #8215 on: February 25, 2020, 08:13:39 AM »
Thank's again Script.
 I have to go back to my books  ;D

Re: GSSL HELP THREAD!!!
« Reply #8216 on: March 29, 2020, 07:16:11 AM »
Hi gents,
first post here after days of lurking and studying this infinite thread.
In this quarantine time i had the "great idea" to buy a kit to spend some time for fun, but i think contracting the virus would be more amazing than that!  ;D

A little background:
That's my first kit, soldered alot before but mainly cables and some repairs (including playstations) but without a deep knowledge of electronics.
The full kit came from a known chap here, but the guy sent me wrong ICS, all 5532 instead of 4 5534 and 2 5532. ICS were labelled in bags so i installed them as per manual and during first switch, obviously, had not working unit at all.

I realized in few seconds that ICS, supposed 5534s, at the input were getting hot, so started immediately a long trouble-shooting.
After 30 mails with the vendor and 150kg of frustration i realized with a 10x glass that i had wrong chips...  >:(

Long story short i had the missing 4 5534s, installed them and the unit started to work properly like magic BUT for 1 hour or so.
Probably for a temperature related matter the gain reduction started jumping all around with proper effect on audio. 
At least i discovered that audio path was right!
Read about various cases about shorts, cap/chip/vca orientation, 704 chip, 702, 10 pin molex, jumpers and so on. Reflowed several times critical areas but troubles troubles troubles.

I'm pretty sure that's my fault somewhere, tried to follow paths on GSSL Roadmap, looked with 10x glass a number of times without being able to spot a short. Changed the bypass switch. Nothing. Deep inside there's always the idea that i had something other wrong in the kit since the experience with ICs but is out of question to desolder and measure all.

Status quo:
rails seems to be ok both +-12 and +-15
IC sockets seems to be ok
bypass is fine, sound passes also with comp engaged but no GR.
SC HP switch (rev 11 board) causes the gr to slam at max both on meter and audio).

After last 10p molex and company reflow audio is still fine in bypass but seems the makeup is now less effective :D
At least the meter and GR are not triggering randomly...

I have on order both 704s and 702s but i'm getting a bit itchy since major sellers are in lockdown.

Any idea about what to do while waiting for supplies ?
(I have here both a Fluke 187 and a 200 mHz scope i use for testing audio paths but without a proper knowledge of how to use them to test parts of a board).

Thanks from Italy.

DB



 

Script

Re: GSSL HELP THREAD!!!
« Reply #8217 on: March 29, 2020, 09:19:36 AM »
Long post, sounds like an Odyssey indeed...

What I have learned through bitter lessons over the years:
ALWAYS double check all components before finally installing one at a time :)

Anyway, so the unit now works all fine as long as you don't activate the SC HP ?

Re: GSSL HELP THREAD!!!
« Reply #8218 on: March 29, 2020, 09:30:14 AM »
no gain reduction, just makeup gain if comp engaged. Unresponsive to threshold at any rate, attack or release.
Is a gain stage at the moment :D

Imagine that no part's sellers are available at the moment so i cannot change a 22u cap if blows... ;-)
To have back the right ICs quickly i spent quite the amount of a new board... :D
Now hoping sometimes in the next week i'll be able to have 704s and 702s, major suspects at the moment.

gyraf

Re: GSSL HELP THREAD!!!
« Reply #8219 on: March 29, 2020, 09:41:59 AM »
..you mean (TL)072 and 074, just so you get the right parts.

For no GR, but makeup-gain working, check integrity of your bypass switch wireing? There are two parts, both circuits open in bypass.

Check for varying DC voltages (correlating to input level and/or threshold) after the rectifier. And check for it again after the bypass switch.

If it worked correctly for ~1hour, there are not really any hardware component burnouts that are possible - both the power regulation and the opamps are short-circuit and SOA safe.
..note to self: don't let Harman run your company..


 

Related Topics

  Subject / Started by Replies Last post
422 Replies
107434 Views
Last post August 25, 2018, 10:11:21 AM
by musika
5 Replies
2781 Views
Last post September 13, 2005, 06:21:43 PM
by clintrubber
5 Replies
2111 Views
Last post November 09, 2005, 07:52:12 AM
by Soeren_DK
0 Replies
1292 Views
Last post May 03, 2006, 10:57:52 AM
by jensenmann