UA 2192 Converter - PROBLEMS!!!

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

khstudio

Well-known member
Joined
Dec 27, 2005
Messages
2,116
Location
New Jersey, USA
I just bought one of the UA 2192 Converters for my studio & it's causing problems with most of my gear that I plug INTO it! (not as much, if at all out of it)

I first noticed the problems with my mixing board (SOUNDTRACS TOPAZ Project 8) Going from the MAIN L/R out INTO the UA 2192:

I noticed - Switches were POPPING in the master section & the audio sounded dirty & constricted!
When I unplug the UA the popping switches SLOWLY dissapear & the sound of the board opens up :shock:

Sounds like caps charging & DC to me... so I got my scope out & tested the XLR input of the UA 2192:

2 VOLTs of DC - measured from GND to pin 2 or 3
50mv of DC from 2 to 3 (+ to -)

WTF... My board don't like that :mad:

How can ANY device without a transformer OUTPUT be happy with 2 VOLTs of DC coming back into it... & is this normal???
What can/should I do about it?

I found this article
http://www.kellyindustries.com/computer/universal_audio_2192.html

(READ THE PART BELOW IN BOLD)
It starts with the use of a newly design analog portion that doesn’t use capacitors or DV-servo circuits as decoupling devices. They found these where causing phase problems that became very noticeable at higher samplerates. The sound would appear to shift up and down the audio spectrum first giving clear highs with a muddy bass and then shifting again to giving a tight bass with a loss in the high end. Capacitors are normally used to block any DC offset that might be present in the signal. The problem is they are basically operating as high-pass filters with a very noticeable phase distortion at low frequencies. The sound produced is missing the presence and detail that should normally be there. It was solved by removing the capacitors and servos and running the preamp in Class A. The preamp section’s op amp was actually designed by UA and biased (a DC current is applied to the input terminal) to eliminate the cross-over distortion that would occur if the circuit was run in bipolar mode (using a Positive and Negative Voltages). The op amp they designed used minimal component stages and only enough ‘negative feedback’ to insure the op amp remained stable. Because there are no DC blocking capacitors or DC servo components used in the 2192 the output of the op amp is biased using a “digital offset calibration scheme”. This allowed the system to maintain its maximum headroom without the caps or the use of digital high pass filters.

Could they have done this on purpose :?:

So much for "UNIVERSAL" audio :?
 
That's really a pretty stupid idea from UA. Looks you'll need a pair of AC coupling capacitors. Or just send it back...

I think Rupert Neve once startet this (if you ask me rather doubtful) idea to superimpose a DC in order to avoid crossover distortion at low levels. But he does AC couple at the input and output.

Samuel
 
[quote author="Samuel Groner"]That's really a pretty stupid idea from UA. Looks you'll need a pair of AC coupling capacitors. Or just send it back...

I think Rupert Neve once startet this (if you ask me rather doubtful) idea to superimpose a DC in order to avoid crossover distortion at low levels. But he does AC couple at the input and output.

Samuel[/quote]

I agree.

I talked to tech support today & they just wanted to blame my equipment or hookup to be the problem.

After I disconnected everything & measured the DC coming out of the input I called back & FINALLY, he said he'd let me talk to the designer (a REAL TECH... what I asked for in the first place) but he left for the day.
I have to wait til Monday... but I have the feeling this is how it was designed... almost $3000 for this :cry:

Can someone PLEASE tell me how ANY piece of gear can handle 2 volts of DC being feed to their output? I don't get it.

I wonder how my Distressors & FATSO will like it?
 
indeed the 2192 does have DC offsets on the input and output. The rational about leaving out the coupling caps is somewhat flawed when most transformerless gear's inputs and outputs are capacitively coupled. So when the full signal chain is considered it really doesn't matter wether the capacitors live in the output stage of unit A or the input stage of unit B.
With transformer coupled devices they won't care if there is DC present on the device into which they are connected.

One important thing we've come across in our studio running an SSL4000 where the output of the mix buss is connected to the inputs of a number of two track devices ( Studer A810, RME converters, and 2192 ) simultaneously is that if the 2192 is NOT powered up it will grossly distort the output of the SSL. This is due i imagine to the input amplifers of the 2192 being not powered up and therefore looking to the outside world like a bunch of diodes.

Back to your problem; the DC offset should not be a problem if your equipment connected to the 2192 is connected up balanced and has suitable coupling caps ( and hopefully someone has flipped a coin somewhere and put the caps in with the +ve leg facing the output socket )

Rob
 
Thanks Rob :thumb:

All I can say is... WOW :sad:

The outputs of my console are UN-BALANCED ONLY... not sure if there's a special way I could or should wire up some cables?

It's funny you mention the distortion of your console... besides the popping switches, my board gets distorted sounding (the output & control room sections at least) when it's ON... never listened to it off... YET.
- UA 2192 MANUAL -
"For Un-Balanced operation, Pin 3 can be Grounded"

Is it possible you guys (ROB) could measure the DC of your units & post the results
:?:

Man.... I'm NOT diggin this at ALL.
I don't just use this converter for one thing... there's always something different plugged into it - Board, Pre's, Comps, Keyboards (sometimes), EFX units... how are they going to like it?

The ONLY thing besides returning it (which may not be possible) is getting some nice IRON (Transformers) in front of it... any suggestions if I HAVE to go that route?
 
The ONLY thing besides returning it (which may not be possible) is getting some nice IRON (Transformers) in front of it... Any suggestions if I HAVE to go that route?
As I said in my first post and think might be rather obvious, AC coupling will solve the problem.

Can someone PLEASE tell me how ANY piece of gear can handle 2 volts of DC being feed to their output? I don't get it.
Once again, if it's AC-coupled there won't be any problem. And usually there should be no problem otherwise, as long as the DC does not reach a pot.

Samuel
 
[quote author="gyraf"]This is a known problem. Imo a design error.

Jakob E.[/quote]

Unbelievable... I researched this unit for weeks before buying it & NOBODY mentioned it.

Thanks for posting Jacob :thumb:
 
[quote author="Samuel Groner"]
The ONLY thing besides returning it (which may not be possible) is getting some nice IRON (Transformers) in front of it... Any suggestions if I HAVE to go that route?
As I said in my first post and think might be rather obvious, AC coupling will solve the problem.

Can someone PLEASE tell me how ANY piece of gear can handle 2 volts of DC being feed to their output? I don't get it.
Once again, if it's AC-coupled there won't be any problem. And usually there should be no problem otherwise, as long as the DC does not reach a pot.

Samuel[/quote]

This is semi good news... but I know my board already has coupling caps AFTER the fader & fader booster amp... so DC MUST be getting on the master card somehow & at least effecting the FET control room switching.

#1
Could it be getting on the GROUND of my console?
I mean... what happens to the 2v when pin 1 & 3 are connected, like in my console? When I put ONE leg of my meter on the tip or sleve (+,-) it oscillates... I wonder if it does that to my boards ground?

#2
What about the DIRECTION of the Coupling Caps (+, -)
Aren't electrolytic caps directional?
or doesn't it matter.


Thanks for posting Samuel (& everyone)
:thumb:

I like this converter & think it sounds pretty good... just trying to understand what's happening & what I should do about it.

I only know so much about electronics. :oops:
 
"2 VOLTs of DC - measured from GND to pin 2 or 3
50mv of DC from 2 to 3 (+ to -)"

Well you could connect a 1:1 transformer across pins 2 & 3
 
To know why exactely the problem appears as it does it would be necessary to see the schematic of both console and converter. I'd be surprised though if connecting 220 uF caps in series with both pin 2 and 3 for input and output does not work. Positive terminal faces the UA gear, voltage rating is uncritical (I'd suggest 16 V, but 6.3 V is enough). I might consider it to be good design practice to include a 100k resistor for each capacitor from negative terminal to ground.

A transformer would work as well, though I think it is a unnecessary costly and distortive solution.

Samuel
 
Samuel,

good idea with the resistors, especially if he unplugs his converter from time to time for some reason. Many times the main A/D stays connected to the mixer outputs.

The input impedance of the 2192 is as low as 1.5k so the caps could be even larger, suggested 470 uF + a 10 uF polyprop in parallell (for better performance at higher frequencies) on gearslutz. Would keep freq and phase mostly unaffected in the audio region...

Finally, since the console output is not balanced anyway, and the A/D pin 2 can be grounded according to the manual, it should be enough with one capacitor, not two, or what do you think?

Martin
 
Thanks for the help :thumb:

I just got back from a VERY short vacation & have to address this problem soon.

I can post the sections of my boards schematic but I bet it'll be a cold day in hell before UA will let theirs out.

I'm starting to get a grip on this a little. My FATSO & Distressors are balanced out & most of my pres are Tranny or Differentially balanced out so they "should" be OK.

My board is a different story. What I THINK is happening is the DC is getting onto the master card & OR into the FET CONTROL ROOM SWITCHING & making it distort... but it may be more serious.

I know for a fact that there's a cap after my output Op-Amp (Fader booster) then after that I think the signal splits - one to the Control room switching & one to the main out L/R. So this could be it... no cap between the output of the console & the FET switching.

I'm not home but will be soon & I'll check the schematic.

I will try a cap & tranny to see (Just wish I didn't have to)

QUESTIONS :idea:

#1
What about my other equipment CAPS like FATSO & Distressors, etc... - what if they're NOT facing the "right" direction? ... am I supposed to change all of my gear to match this thing?

I would bet my studio that none of my other gear has over 2 volts DC coming out & that they're OUTPUT CAPS have the + facing the Op-Amp which would be the wrong direction to match the UA, NO? :?

I must mention that I think the UA2192 sound good & I like it.
I also feel my mixes are turning out as good if not better that when I was using an APOGEE ROSETTA.
I'm just shocked that no one else has any problems because of this & that they don't warn you or give solutions. It shouldn't be my job to fix it but thank GOD I can adjust.

Also, my console NORMALLY goes thru the FATSO, which STOPS the Popping! this is why I couldn't hear or track down the problem... it was intermittent. When the FATSO is OFF it has a true bypass relay that kicks on... so when it was on = no popping & distortion... still in bypass but the output caps absorbed the DC & the popping faded away.

#2
So I don't need to worry about a cap on the GND of my board output?
Can my board really take 2 volts of DC on the GND :?:
EDIT:
Well the more I think about it it shouldn't develop there.
 
[quote author="Martin B. Kantola"]Samuel,

Finally, since the console output is not balanced anyway, and the A/D pin 2 can be grounded according to the manual, it should be enough with one capacitor, not two, or what do you think?

Martin[/quote]

Correction!
Pin 3 can be grounded :wink:

Good question... is one cap enough?
 
I know for a fact that there's a cap after my output Op-Amp (Fader booster) then after that I think the signal splits - one to the Control room switching & one to the main out L/R. So this could be it... no cap between the output of the console & the FET switching.

I just looked over the schematic & this definitely looks like a problem :idea:

The LAST coupling cap (before leaving the console) is at the Fader booster amp.

The problem with this is the signal splits AFTER that cap to:

MAIN L & R OUT JACKS
&
CONTROL ROOM SWITCHING (FET)

So the DC can't get back to the amp or fader BUT CAN reach the control room... this is probably the bad sound I'm hearing coming out of the speakers. :sad: :? :green:

Could this have caused ANY DAMAGE to the FET Switching circuit :?:
 
[quote author="squib"]

Back to your problem; the DC offset should not be a problem if your equipment connected to the 2192 is connected up balanced and has suitable coupling caps ( and hopefully someone has flipped a coin somewhere and put the caps in with the +ve leg facing the output socket )

Rob[/quote]

So my (hopefully) last concern with the 2192 & connecting to other devices is the DIRECTION of the Coupling Caps. :?:

Can someone explain how caps are affected by this:
Higher voltage to the (-) side???
:oops:

Can it cause them to sound bad or even damage them?


Thanks everyone for your help :thumb: :sam: :guinness:
 
[quote author="gyraf"]
Can it cause them to sound bad or even damage them?

An output electrolytic, polarized with 2-4VDC at wrong polarity?

Yes, most definitely!

Jakob E.[/quote]

Thanks Jacob - I REALLY appreciate & respect your input on this!!!

WTF were these people thinking :sad:

I'm going to be calling the main tech/designer tomorrow (They finally said I could talk to him :? )



Tomorrow I will be running some tests... like with 600:600 trannys in front of the 2192 & after the FATSO... I'd like to hear if it's making the (Balanced) FATSO sound worse because of the reverse CAP issue.
 

Latest posts

Back
Top