New 'ultimate' SSL buss comp clone ;-)

[Jonkan]

So im stuffing these pcbs according to the images posted by khstudio and have a few questions.

6. On the slave board omit the 3M3 & 2K resistors next to where the board connects to the switch pcb.
Im guessing I can ditch the 2k resistor to the meter on the slave. But what about the 3m3?


I am also wondering about the differential conneted VCAs to get less distorsion. I have at the moment only stuffed left side on each pcb like KHstudio did it, but i actually have all the parts to stuff both channels on the pcbs except for two more 2180 VCAs.

Would it be worth it to get two more VCAs, or probably not that much difference in the end?

Thanks!
/J

 

[khstudio]

My Pics are complete & my unit is fully functional. :wink:

I wasn't so sure how much I liked the 202x VCA's I used & have been wanting to swap all the VCA's out for the newer 2180's.
I finally did it a few days ago & like it very much... I think it's a lot smoother & more responsive. I have not completely checked or refined my meter & ratio's but it does sound great... still working on it.

It is still loaded up exactly like the pictures I posted except I'm using the 2180's & loaded the "DBX202 emulation circuit"

USING THE 218X VCA's

These links are VERY helpful to see & understand what has to be done when using the 2180 & 2181 VCA's.
Take note: The 3.9k should stay for the sidechain (because of the 12v) BUT the 3.9 on the emulation circuit (main signal path) should change to a 5.1k (because of the 15v)

Also note the need to remove the 10k between pins 3 & 5.

Matt's DIY
http://diy.fischerworks.com/gssl_vca.shtml

GSSL Calibration
http://homepage.mac.com/marten.thielges/gssl/calibration.html

 

[Jonkan]

Ok, thanks alot!
Then you have your unit set up exactly like mine. Cool!

So the 3m3 resistor stays then? Im not sure what it does, so i am a bit confused about that part (i couldnt tell from the picture of your build since the connectors were in the way.)

Im still toying with the idea though to get two more 2180 Vca:s to get distorsion figures down a bit...its not very much more expensive since i already have all the other parts at hand. But perhaps that would be overkill, i dont know.

/J

 

[Jonkan]

Hey! I just realised that i have some 2150A vcas that i could use for the sidechain in the compressor.

That way i could use my four 2180 in differential mode for the added benefits that brings.

Seems like a reasonable idea, right? As long as the 2150A:s are good enough to use in the sidechain (which im guessing they are..)

What do you think?



/J

 

[khstudio]

[quote author="Jonkan"]Hey! I just realised that i have some 2150A vcas that i could use for the sidechain in the compressor.

That way i could use my four 2180 in differential mode for the added benefits that brings.

Seems like a reasonable idea, right? As long as the 2150A:s are good enough to use in the sidechain (which im guessing they are..)

What do you think?
/J[/quote]

They are fine for the sidechain.

I thought about doing the exact same thing since I have 4x 2180's as well.

My only question is, what if I'm running it UN-BALANCED??? :?
(Keith's Version)

 

[Jonkan]

ok, great!

Im not sure how it would work unbalanced...probably fine.

How about the 127k resistor that controls the ratios (the one next to the 620k resistor that originally was 100k). Will there be issues if im using a different vca for the sidechain than for the actual gain reduction vcas?

I guess my question is: do i still use the 127k resistor here? Im having trouble reading the schematic, i guess i must read up on my circuits to have a better grasp on how this thing actually works. I only understand the very basic topology of a vca compressor at the moment.

/J

 

[khstudio]

With a 2150 or 2151 in the side chain a 127k worked great for me - using 218x for the mains L/R.

I did a lot of testing on my first 2 units... not sure about the 127k with the 218x in the sidechain... still need to test to be sure.

Take your time & read a LOT & you'll build up some knowledge... it takes time. Trust me... I knew nothing 4 years ago.

Good Luck.

 

[Jonkan]

Thanks for the advice. I really appreciate it!
/J

 

[Jonkan]

Ok, i made the ugliest hookup diagram i have ever seen. Can someone take a look at this and see if it is correct.

I want to make one switch that has a higpassfilter for the sidechain, and one switch that changes the compressor between gyraf and oxford modes.

Yes, I made this with windows paint. Nice huh!?

Does that look ok?

Im also wondering about the 3k9 resistor in the sidechain vca. I have read everything here that i could find, but it still seems inconclusive. Do i change it to a 5k1 resistor or not? The main VCAs have a 5k1 in my build just as the THAT application notes. But im not sure about the sidechain VCA since its running on 12v instead of 15v. Im using 2180LA for main VCAs btw, and 2150A for the sidechain.

Im guessing it doesnt matter too much, since there hasnt been more discussion about it.

Thanks!!
/J

 

[Dr nEon]

Hi!

A quick question re the power transformer for this ultimate version.

Everyone seems to be using 18-0-18v , like Keith did .... but as the original gyraf version called for a 15v tx , I'm wondering about this.

Would a 15v tx still be acceptable , with a decent va rating , or is it a necessary part of the nu design that the power tx be 18v?

cheers and thanx in advance!


nEon.

 

[radiance]

A 15v with a decent VA rating will do....18v will run the regulators hotter, that's all.

 

[Dr nEon]

Ok , Thanks radiance , I'll stick with the 15v tx ! :thumb:

HOWEVER , I do have one major issue with this version which is troubling me , which I don't think anyone has questioned so far...

We have two fully populated gssl boards..effectively two gssls...paralleled from one transformer , so two rectifiers etc from one source.

Elsewhere on this forum, advice is being given to avoid this sort of arrangement , i.e.:
http://www.groupdiy.com/index.php?topic=6530&start=0&postdays=0&postorder=asc&highlight=ssl+transformer+voltage

Indeed , a year or so back , when I was looking to build a dual mono g1176 ,I asked if I could power both units from one tx , and jakobs advice was that this could cause unpredictable results.So I used two separate txs,following his good advice.

Then again ,in this case, SSl tech ,and some of the big guns here , have got into some serious measurements looking into distortion figures ,s/n etc ,and shown some grand results. So there would appear to be no problem in arranging the power distribution this way..

So in the end , to a (still) relative newcomer like myself , I would love an explanation of why it may be ok, in this instance, to parallel from one tx. I have boards stuffed , and just want to pursue with confidence with he last part of the build.

Cheers!

 

[nielsk]

Consider that entire consoles are powered off one transformer, as long as it can supply enough current and voltage it really does not matter. It can help in very high transient current demand situations like power amps to have seperate power transformers, but to run a sidechain in a compressor it will not matter

 

[Dr nEon]

Thanks for the reply nielsk!

I understand your point re power consumption , but the issue I am worried about here is the fact that one transformer is going to two separate sets of onboard psus ,so two rectifiers etc... something that I was advised against doing , in another project.

BTW I'm not set out to criticize what is being done..I'm not qualified to do that...just merely trying to understand why it is ok in this situation.

cheers, nEon

 

[SSLtech]

There are several things that are bad to combine... two outputs into one input is a bad idea, and two transformers into a single rectifier CAN be a recipe for trouble...

But I can't easily think of a situation where two rectifiers connected to a single transforemr secondary would cause trouble, other than simply being a case of the two loads adding up to a total load size which causes an issue.

Classic case in point: the SSLtech ±18V + 48V PSU... one transformer, two rectifiers... hundreds built, no problems with the two reccy's, EVER.

Keith

 

[Dr nEon]

Hi Keith!

Great stuff, man , thanks for explaining that ...I can go ahead with confidence now :grin:

..and big respect for coming up with this version of the gssl , I can't wait to get this one finished!

Cheers

nEon.

 

[Jonkan]

I am having a very strange problem with my UGSSL clone. My psu voltages on the positive side come and goes. I get 20VAC into both cards. 28VDC after the rectifiers and into the 15v regulators. But occasionally i get 0.7v out of the positive 15v regulators on both boards, and then suddenly when i reset the power without doing anything, it works again and i get +-15.7V instead. I still get 28Vdc into the regulator though.

This only happens on the positive side of the psu.

It also seems to happen more often when i connect both cards.

I blew one of the 5534 opamps one one of the cards (i put it in the wrong way...duh, im stupid, its replaced now), but im not sure if that could have been the cause, as i recall the problem happening before that also)

Any ideas? Should i just replace the regulator and see what happens?
/J

 

[mhelin]

So how well a mid/side encoder + summing (adjustable) circuit in GSSL could emulate this UGSSL?

Like a THAT1240 M/S encoder/decoder:
http://www.groupdiy.com/index.php?topic=21795&highlight=mid+side+encoder

 

[SSLtech]

[quote author="mhelin"]So how well a mid/side encoder + summing (adjustable) circuit in GSSL could emulate this UGSSL?

Like a THAT1240 M/S encoder/decoder:
http://www.groupdiy.com/index.php?topic=21795&highlight=mid+side+encoder[/quote]
Not at all well.

It would be usable, and occasionally useful, but for M/S compression, you'd oftan want to use SEPARATE sidechains and detectors, and in addition, there's a LOT more to it. A GSSL sums, so you ALREADY have a sidechain doing the 'M' part, and if you add the 'S' in the other channel, you'd end up with a sidechain reacting to the left channel ONLY...

M = L+R.
S = L-R.

The GSSL sidechain sees the two added together, therefore since:

M+S = (L+R) + (L-R),

and since L+R+L-R = 2L

you're left with 2L (left channel, plus 6dB) in the sidechain. Signals on the right channel would NOT do anything to compress the signal.

The ultimate "greater-of-two" sidechain would process both signals without preference or summing, and only the single-pole DC rectification is added, so it WOULD be able to do M/S.

The GSSL really wouldn't be usable for that purpose.

Keith

 

[Jonkan]

this is driving me nuts. Half of the time im switching on the unit without touching anything im gettin 0.7v on the 7805 regs. Half of the time i get +17V

I mean, wtf?

(I get ´-17v on the 7915 regs all the time)

And after the 10r (now in my build =270r, cause i didnt have 300r at home) going to the one of the 78L12 regs i am getting 13v into it, where on the other regs i am getting 22v. The resistors all measure fine, and everything is dandy before them (im getting about 26v before all the resistors going to the 12v regulators).

This is really strange to me, what could be the cause here? Bad/broken regulators? There are no shorts, or obvious bad solderjoints, and everything measures fine continuitywise.

????
/J