New 'ultimate' SSL buss comp clone ;-)

[Dr nEon]

Ok Jonkan , I see what you mean about the link happening at point F .

I've just been studying all the pics of peoples builds so much ,lately, that it jumped out at me that you didn't have that linked.... but now I think can see why!

Now ,where the instructions on page 10 say:

"If you intend to defeat the make up gain in bypass then link the 2 points where the defeat switch connects to then connect the switch to the master switch pcb in the normal way. "

Then how can this be correct? Of the two pads on the slave switch board,where this link would connect , one pad goes direct to"point E"(which is unconnected ), and the other pad also goes nowhere (it would normally go to m/up pot centre) .

I can't see those links on Kevin or Keiths pics , so it would be nice to get a opinion from someone who has a working ugssl as to whether this part of the mod is necessary...


Cheers all


nEon.

 

[Sleeper]

EDIT: oops

 

[Dr nEon]

Thanks for the info, Sleeper!

Just one thing... when you say that

Since e and f are linked on the ratio boards

, do you mean the "switch pcbs"?

As far as I can tell reading the instructions , E and F should be linked at the master and slave MAIN pcbs .

IF that is the correct way to do it , then that would leave E&Fs traces on the slaves switch board unconnected to anything , which is why I'm having a problem understanding the point of linking the M/up defeat switch , when it leads nowhere..

On the other hand , maybe i'm not up to date here , and if u think the E and F links should occur on the SWITCH boards ( not the Main boards) , then I can see some mileage in trying out the defeat switch links, as the slave switch board is now 'in circuit' , as it were...

I appreciate your input ,Kelly....I'm now just hours away from starting to wire this stuff up , so I'm just looking to start out with as much clarity in my mind , incase I have to go troubleshooting later!


Cheers

nEon.

 

[Jonkan]

Ok, back after some further testing.

I have now tried disconnecting the Slave completely from the Master PCB, including the power to the slave.
Basically connecting the Master as a mono compressor.

Result: Everything measures just fantastic!

I then repeated the same procedure with the slave channel, but connecting the master control board to the slave pcb.

Result: Everything measures just fantastic.

So i can conclude that both Main Pcbs are working correctly, and that the Master control board is working correctly.

I must have messed up the wiring between the boards somehow. I connected it exactly like KHstudios pics.

Have you measured your unit Kevin, to check for problems? My unit is working as is, but measures badly. I cant hear the problems by ear though, so thats why im asking.

Could someone send me a big pic of their units insides so that i can double check my wiring against a fully working unit. Keith??
I remember reading about a big pic of "keiths guts" earlier.

It would really help me!!
/J

 

[Sleeper]

hi Neon.
You are right,
I removed the post.
I worked off of Kevins' diagram on page 11 of this thread.
That should work
Kelly

 

[syn]

Kevin's wireing diagram should be OK as I've used it and I have a working unit that measures quiet cool. Will post data later today...

 

[syn]

 

[Jonkan]

Thats about the same values im getting from whichever pcb is my master channel. Thanks for posting them.

I started redoing the wiring today, soldering the wires directly to the pads instead of using connectors.
I hope that it will solve my problems.

/J

 

[Jonkan]

I found it!!!! :thumb: :thumb: :thumb: :thumb: :green:

YAY!! Beers for everyone :sam: :sam: :sam: :sam:

It was the led for the meter that was the culprit. As soon as i disconnect it, everything becomes a hell of a lot better.

Im currently measuring at -3dbfs from my motu traveler (im not sure what that translates to in actual level since i dont know how the motu is calibrated):
THD:
L ch 0.0042%
R ch 0.0037%

THD+noise:
L ch, 0.0068%
R ch, 0.0065%

Thats pretty damn good, right? (not that i know if i can totally trust RMAA or not...but still)

Finally i can move on to other projects! Well as soon as i can figure out how to connect the led without it messing things up. Im using a Ptownkid behringer meter. Perhaps i should change the diode resistor? Its currently 1k and the power is taken instead of using a powerled from the master controlboard.


/J

 

[Dr nEon]

Hurrah!! Well done Jonkan!

Syn , what is that software u r using to measure thd, noise etc. ?

cheers!

nEon

 

[Jonkan]

Thats RMAA that i was talking about earlier.

Best of all..Its free!
/J

 

[syn]

[quote author="Dr nEon"]Hurrah!! Well done Jonkan!

Syn , what is that software u r using to measure thd, noise etc. ?

cheers!

nEon[/quote]

http://audio.rightmark.org/products/rmaa.shtml

It is free, and it is better than nothing :wink:

Jonkan, great that is working. Are you using on board PSU? I usually use
more than 1k, depends on "how much light" do I want from the leds...

 

[Sleeper]

Hi Jonkan,
Glad to hear you got it worked out.
I found a solder bridge and ???? I hadn't placed a coupling cap on the board that was working strangely. boy that was dumb.
Haven't retested yet.

You could also make a little powersupply if you have any lm317's around.
I do this for all my meter lamps, it's about a 2cm x 4 cm piece of perfboard
and if you build it with a trimpot you can dim it too.

I have my meter LEDs fed from my relay psu, it doesn't have to drop nearly as much voltage that way and once you've got relays happening it's easy to implement true bypass, lighted switches, etc.

Kelly

 

[Dr nEon]

Cheers guys , I checked out RMAA , but unfortunately I'm a mac man!

Mind u , my girlfriend does have an old pc laptop :twisted: ....ah but only with crappy minijack MIC input.. :sad:

..maybe there's something similar out there for mac, hopefully.


Anyway , fired my ugssl up yesterday , and everything seems nice...the dual sidechain (the whole point of the build) is really worth the effort!

Serious thumbs up to Keith for bringing us ssl veterans closer to home, with this part of the design!

Listening to the two modes on decent cans (I used sennheiser 600s) , the panoramic shape of the way the compressor is working is so evident .

I was only playing commercial cd's yesterday , not a raw mix , but it becomes obvious why the gyraf mode has been sounding so much better than a standard ssl when squashing ambiences and drums and the like...

...and why, as stated before, oxford is way to go on the mix buss...oxford really DOES give u the glue , and the stereo image is awesome.


Sleeper , could u point me to that psu circuit u mentiioned please?

I have my relay card fed DC from just before the 7815 reg , on my standard one-board gssl .

I was just about to do the same with this new ugssl , but that would mean that one board has that extra load and one board doesn't , and maybe this might cause problems..noise, distortion, etc?

Then again , I am tapping from BEFORE the regulator..

:?

Cheers!

nEon

 

[Sleeper]

hi dr neon, I've got a standard LM317 supply straight from the trafo with it's own bridge rectifier. as drawn in the LM317 datasheet with just slightly more capaciticance as seen in the JLM power module. I think I have 100uf/10uf
(only because i have a lot of these)

if you have a 7812 or 9 or 5 or whatever suits your relays and led dropping resistors you could use that too. also straight from the datasheet.
or the ssltech psu!!!

there's mac the scope, but I'm not totally sure it will work (in demo mode anyhow) to do the calibration. search for blue peter to calibrate/vca distortion adjust without a meter. just been reading that myself and getting ready to give it a try. search any of the bold stuff and you should get it
Kelly

 

[Dr nEon]

Sleeper , thanks for all that info!

I'll go looking for those datasheets and for the ssl tech psu stuff..half of the nice stuff with diy is the research!

I'll check out mac scope... i've built my ugssl with the pre trimmed that2180's , so i won't be calibrating as such , but it would be great to have some kind of analyser anyway , to check stuff in more detail.

i.e. I have only one small concern with this ugssl, which is a very slight rounding off at the very top end ; Similar to the way in that a small phase shift can dampen decays,and apparently dull or dry up a sound .
I am talking way up in the 'AIR' frequency zone,here... it doesn't sound like an actual problem , more of a character thing. But I'm interested to explore why this is..

I used Elna Starget's for the first time .Only for the 22uF sig path caps.(35volt) at the input stage. The rest are FC's.

Could I just be hearing the Starget's character?

My first Gyraf gssl is FC's throughout , and has a clearer top end, in this area.

I know that the general view is that cap brand isn't critically important in the Gssl..

So maybe it's down to using a different batch of VCAs and opamps??

Maybe i simply need to leave it powered up for a while , before jumping into assessing it , within an hour of powering up for the first time!!

Any thoughts?

BTW I am a/b ing between hard bypass and compressor in , but with threshold wound clockwise , away from onset of compression.

What I'm saying here sounds a bit vague , which is why i could do with an analyser to confirm my ears' hunch!


cheers

nEon
:razz:

 

[Sleeper]

Hi Dr. Neon
I've read many threads regarding caps and their effect on sound, and I would hazard a guess that a circuit as clean as the Ussl would be one of the few places where cap selection has a noticable difference...
This is a guess and I'll leave making an argument for that to somone else :roll:

I went exotic on the coupling caps at the output and used nichicon muse 100uf bipolars there... I had a bunch of them and I thought I'd put them to use, couldn't think of a better place to do that.

I'm sure someone will say it doesn't matter but I'm giving it a try anyways.
I opened up an alansmartcompressor and that one uses exotic black gates in the coupling positions. I'll bet someone will say it doesn't matter, but I doubt anyone would say you would be worse off.(although you'll be a bit poorer)

I read a post years ago regarding the 5532s and 5534s and it seems that on some ssl consoles the clients were not pleased until signetics 5532's were installed... It was said that you could hear the difference between chip manufacturers... don't really know about that, but you might read up on that one a bit.
Cheers
Kelly

 

[Dr nEon]

Cheers , Sleeper , more good stuff for me to think about !

One question..

I went exotic on the coupling caps at the output

I've also noticed somewhere on one of the BOMs for the gssl , fc's suggested for input , and PW's for output.

This leads me to wonder : if there is any mileage at all in esoteric caps , is it more likely to be at the output stage, than the input?

In my case , I've already built lots with fc's ,and wanted to try something different , so I got a load of stargets , but the 100uFs weren't in stock at the time.

So in the end ,I've only put the stargets at the input, but they are a lot larger than the tiny fc's in diameter and height . I'd thought this would probably be a good thing, but maybe they are a bit sluggish compared to the fc's ..i dunno yet?

I'll definitely read up on the op amp brand issue...I've seen some discussion on that , and of course its easy to try alternatives with everything being in ic sockets!

BIG thanks again!

:thumb: :guinness:

nEon

 

[Sleeper]

Neon,
I can't really respond to your question except to say that a whole lot has been written, I've read it, and I still don't really know when and where the conditions are most favorable for specific brands of caps.

I did build a 5532 output circuit very similar to the gssl for a buffer amp on my monitor controller. It wasn't scientific and it was running on 18volts, but I liked the sound of the nichicon muses. the other upside is that a bipolar on the output means that if I accidentally patch the output into a channel with phantom power turned on I should be protected. This is a possibility because I have a number of electrodyne modules in my studio and I love their EQ. It's very unlikely but possible that phantom could show up at the outputs of my Ussl.

On top of that, this stuff all takes me far too long to build to get hung up on an extra couple of dollars here or there. I just always use the best quality I can get and even better when I have something on hand.

Kelly

 

[Dr nEon]

Ok, thanks Sleeper !

I'm gonna rack this one up now and live with it for a while , get to know it , and then probably play around with different o/p caps & op amps at a later date.

Cheers

nEon.