Official 33609 builder's tread. See 1st page for updates.

[Igor]

Yes, all this is up to specs. PC-based analyzers have tendency
to take their own "zero db" as reference. I prefer my cheap but
hardly modyfied Boonton 1120.

 

[maxheadroom]

any idea to reduce the noise?

 

[Igor]

Where is the ROFL smile???
Neither 33609 nor 2254 cannot be count in as world's quiet compressor's
It can be bit hissy. Not hum or 50/60 hz related noise, but pure hiss.
To reduce it...
There's one hardcore option....
Replacing diode bridge VCA to something conventional and changing whole gain structure
If it will have less noise, it won't be 33609.
Just see schematic... There's about 26db of amplification at unity gain after 5k source
even if we don't count in noise figure of stuff before interstage trunny...
Measured -82@unity gain is close to theoretical lowest possible noise floor from _this_ circuit.
I use 33609 on hot levels... This is simplest way to shift noise floor down.

@ maxheadroom:
Hiss is OK, but 50/60 hz related stuff is not. Please refer to test procedure
and check everything again. Maybe, there's kinda ground loop or
PS trafo placed close to input trunnies...
Pics of your unit can be helpfull.

 

[keithcamilleri]

hey guys do u have a printable vu scale for the sifam al19w for the 33609?

thanks

 

[maxheadroom]

ok here is a top view picture: http://gasman.lima-city.de/diy/33609/33609topview.jpg
test proc was OK. i had no problem with the values.
its not a 50Hz or 100Hz hum. its hiss.
thats because i saw it on the noise chart: http://gasman.lima-city.de/diy/33609/N33609_RMAA_01/noise.png
on the headphones i hear also no hum but hiss.
the strange thing is that the hiss is on both channels.
i played around with the star ground, but no changes.
i disconnected limiter board cable and compressor board cable. no big changes (~0.5dB better)

 

[maxheadroom]

is anybody here with a finished 33609 and running RMAA, to compare the noise level?

 

[keithcamilleri]

maxheadroom :

i havent checked mine properly yet,  but from the top view picture i am noticing that you did not shield your compression and limiting cables ... maybe one of the noise level problem could be that. as soon as i test mine i will let you know....

aswell i noticed that your marinair output is not grounded..  i am using carnhill but i don't know if the marinair needs to be grounded or not.... just trying to point out things that might work for u

keith

 

[maxheadroom]

hi keith,

thank you for your answer!

i solder the comp and lim cables new with shielded cables, but with no changes.
also i tryed to ground the marinair output, but also with no changes.
???

a comparison in RMAA would be nice 

 

[rob_gould]

[quote author=keithcamilleri]
i don't know if the marinair needs to be grounded or not.
[/quote]

Can anyone confirm whether this is necessary or not?

Cheers,

Rob

 

[rob_gould]

Another question: 

What do the red numbers on this section of the schematic refer to?  They also appear on output connections and around output transformer too...

link to schemo

 

[keithcamilleri]

hello rob those numbers corresponds to the numbers on the pcb so no 1 on the pcb is + ( on the xlr) no 2 on the pcb is the  shield ( on the xlr ) and no3 on the pcb is the the negative ( on the xlr )

hope this helps.

cheers
keith

 

[rob_gould]

hello rob those numbers corresponds to the numbers on the pcb so no 1 on the pcb is + ( on the xlr) no 2 on the pcb is the  shield ( on the xlr ) and no3 on the pcb is the the negative ( on the xlr )

hope this helps.

cheers
keith

Thanks Keith - it's obvious now you mention it :

Rob

 

[maxheadroom]

[quote author=keithcamilleri]
i don't know if the marinair needs to be grounded or not.

Can anyone confirm whether this is necessary or not?

Cheers,

Rob
[/quote]

i noticed that i did not measure any changes if the marinair grounded or not

 

[JdJ]

It can be bit hissy. Not hum or 50/60 hz related noise, but pure hiss.

I use 33609 on hot levels... This is simplest way to shift noise floor down.

Hot levels is my approach to reducing hiss as well.  That and I pretend I'm mixing to tape instead of the Rosetta... 
What a great sounding unit!  After using it for 6 months or so, it lives on the mix bus.  BTW it KILLS on hugely smashed room mics.

Happy New Year everyone!

 

[Igor]

Hi JDJ, you meant clone or original unit?

 

[JdJ]

Hi JDJ, you meant clone or original unit?

Clone.  I have only had the pleasure of hearing an original in other people's studios.  That's what got me hooked on this project in the first place!

 

[maxheadroom]

is it possible to detect the noise source by shorting some wires step by step from the output side to input side (or otherwise)?

or can my noise source occured by the interstage tranny i use. i used the 31267 wired as a mic input?

 

[Igor]

No, 31267 cannot add noise. "Noise-wise" it is allmost same impedance as micinput.
It has different inductance, transient response, DCR, freqresp etc when connected
as micinput than 10468...but can't affect noise figure by more than tiny fractions of dB..
Maybe j109's are noisy (_never_ had this exp though) or you refer to very low level as 0dB.
Just take dB chart and compare to your card's reference level
From other side, best is take original 33609 and compare one by one,
as I did when built first clone.
Yes, it is a bit hissy in original as well. Never cared about this.
Cannot upload image (seems site is dead) but IIRC you did not shielded sidechain cables.
This is not reason for hiss, but for additional compression artefacts maybe-very small leakage
from CV signal on fast peaks can produce hearable harshness.

 

[maxheadroom]

thank you igor for your message!

i fixed the compressor and limiter cable wiring with shielded cable as you described in the wiring image.
i took the dB chart from the schematic and measured following values:

- Input Voltage: 0dBv (=1V)
- After IP Tranny R47: -6,8dBv
- After IP PAD R11: -31,8dBV
- Interstage PRIM: -32,1dBv
- Interstage SEC R33: -27dBv
- After BA340 R23: -9dBv
- Output Voltage: 0dBv

I think the values are in range.

i am a little bit dissatisfied about the "out of range noise floor". i don`t want change ALL the parts, that is absurd.
my next step is to order new J109 from farnell. the used J109 are from RS. but i think there is not a difference.
Also i ordered 2 pcs of Carnhill VTB9045 for comparing interstage (also for audiowise, and pound rate in euro are good at the moment)

I would be grateful for any Help!!

 

[Igor]

@maxheadroom : yes, signal flow looks 100% fine.
Good luck with finding noise source then!