BEHRINGER MODS / UPGRADES - MY FINDINGS (T1953-Centric)

[falcon]

I know this is an old topic, I have completed my phase 1 modification on this 1953 preamp. I have replaced all 10 opamps with OPA2228. (it only took me 1 year to get them, they must be rare or super popular)
Now i need to do some upgrades on the signal path capacitors and shielding on the cabling. What kind of and brand of capacitors can i use to replace the signal path capacitors that would work 20hz to 20 khz.

 

[falcon]

I have ordered ELNA RFS silmic capacitors (audio electrolytic caps) for the signal path, once completed i will start working on the power supply section.

I have also shielded the cables that run between the pcb's

 

[falcon]

Elna signal capacitors are in and are sounding great. This unit now has a very low noise floor. I had to solder some of the capacitor on the back side of the PCB because of size. I also have changed the tubes to Mullards. Now to work on the power supply and determine what to do.

 

[Majestic12]

One short question. Which type of OPA2228 should I use ? Digikey has a bunch of different types (UA, A 2K5 2K5AE...) and I have no idea which is the right one to use.

 

[Harpo]

One short question. Which type of OPA2228 should I use ? Digikey has a bunch of different types (UA, A 2K5 2K5AE...) and I have no idea which is the right one to use.

For what application? A OPA2228 contrary to a 4580 is not unity gain stable (voltage gain 5 or greater required), and looking at the parts datasheet, you'd know if you want a 2500 parts reel, a soic or pdip package, a variant with better input offset voltage specs if so required, or this part at all.

 

[falcon]

One short question. Which type of OPA2228 should I use ? Digikey has a bunch of different types (UA, A 2K5 2K5AE...) and I have no idea which is the right one to use.

The opamp that i used is OPA2228UA-ND

 

[abechap024]

I've used 228 quite a bit and when working correctly I think they sound great. But you have to be careful because they are sometimes more difficult to get stable...And when they aren't fully happy they can sound very "grungy" trust me -- i know....consider the 227 or just make sure you have a scope...if not just make sure everything is properly bypassed and hope for the best!

 

[gswan]

The reason I use the square wave for analysis is pretty simple.  The square wave contains very high frequencies and very low frequencies.   They're all in that waveform.  The way a particular circuit "distorts" that waveform tells me a great deal about the frequency response of the circuit.   The leading edge of the square wave is comprised of high-frequency content.   The trailing edge is comprised of low frequency content.

Well not quite.
A square wave consists of the summation of an infinite number of odd harmonics of the fundamental frequency, although in practice a reasonable approximation of a square wave can be had using just the 1, 3, 5, 7, 9, 11, and 13th harmonic. You don't get frequencies lower than the fundamental. So if you use a 10kHz square wave, it will contain sine waves (of reducing amplitudes) at 10kHz, 30kHz, 50kHz, 70kHz, 90kHz, 110kHz, 130kHz ... So your circuit will need to have a flat frequency response of at least 130kHz for the waveform to appear square at the output.

An audio circuit may only have a flat response to 50kHz, however if you feed a square wave greater than about 4kHz into it, the output will not be square since the high order harmonics will be reduced in amplitude. Similarly, distortion in the circuit will manifest itself as either changes in the relative amplitudes of the odd harmonics and even introduce some even order harmonics to upset the waveform. A 1kHz square wave is a good test here since you would expect your audio circuit to be able to pass 13kHz without a problem.

 

[Majestic12]

I've used 228 quite a bit and when working correctly I think they sound great. But you have to be careful because they are sometimes more difficult to get stable...And when they aren't fully happy they can sound very "grungy" trust me -- i know....consider the 227 or just make sure you have a scope...if not just make sure everything is properly bypassed and hope for the best!

So there's more to it than simply change the IC's and bypassing the caps? Do I need to adjust something in the circuit to get them "stable" ?

Or should I better use the opa2227 to be on the save side ?

 

[falcon]

the OPA2227 i would not use because of the really slow slew rate, buy the opa2228 and put them in, i have no regrets and excellent sound.

 

[abbey road d enfer]

The 27 and 28 are identical, except that one is compensated for unity gain, hence slower.

 

[abbey road d enfer]

 NJM4580 OPERATION AMPLIFIER ISSUES  
The op amps have a HUGE THD + Noise specification.

0.0005% cannot be reasonably qualified as HUGE. OTOH, there is no THD spec for OPA2227/28. It could be better, because the open-loop gain is higher. That doesn't make the 4580 a pig.

 
The op-amps have a low slew rate

NJM4580: 5v/us; OPA 2227: 2.3; OPA2228: 10 Again, the difference is not that significant; with 5V/us, it passes a 350kHz sinewave at +4dBu.

The op-amps have a low CMRR (Common Mode Rejection Ratio) - the ability to reject (cancel) a signal that's the same on both inputs.

4580: 110dB; OPA: 138dB. But does it have any practical consequence? NO, because the tolerances on passive components will restrict this to about 60dB in the best case.

Noise Figure - the NJM4580's are very high noise, to say the least.

4580: 3nV/sqrtHz; OPA: 3nV/sqrtHz. Serious difference!

Bandwidth limitations - the NJM4580's have some pretty severe bandwidth limitations, as well.

"Power" response: 4580: 100kHz; 2228: 140kHz; 2227: 25kHz.

When the leading edge of the square wave is lower than the trailing edge this equals poor high frequency response.   When the trailing edge is lower than the leading edge it signifies poor low frequency response.

This is a simplistic view. That would be almost true if the circuit had 1st-order response, like a simple RC filter.

I'm not saying the 4580 is the ultimate audio opamp (not one single opamp can make this claim), I just say it's not the piece of junk you consider it to be.

There are a bunch of 10V caps on the board, too.  These are SIGNAL PATH caps.   Ahhh... that's how they do that.

Do you imply there's something wrong there? Considering the voltage here is the offset voltage, probably much less than 100mV, I see nothing wrong in 10V caps. 

About non-polarised electrolytics: Truth be told, that would provide some performance improvement,

Truth, or myth?

The "problem" with the electrolytic capacitor is that it's really only a capacitor until about 5-6kHz, or so.  Then it turns into an inductor and causes all sorts of problems.

What problems?

 To rectify this, I got me a bunch more of the 1uF military grade surface mount caps and shunted all of the electrolytics in the signal path.  This, again, significantly improved my high frequency performance.

Just another audiophool myth. You would be surprised, looking at very professional audio products, they just don't do that, because they just put the right value, because they calculated the right value, instead of guessing and praying.

I am happy for you that you find your mods to be an improvement (although you haven't a single measurement figure to prove your claim), but please, don't try to justify that with concepts that you don't fully understand.

 

[falcon]

Big difference

4580 = 0.0005 THD+N

opa2228 = 0.00005 THD+N

 

[Lurch]

Does anyone know the extent of the Black Lion ADA8000 AD DA Mods Which op amps would be best and how much can or what can you do to the Power supply to make it better? I have not seen this project officially from Black Lion, Hope they do it on mass, but for now I am probably stuck making it happen DIY .I have been told to use Texas instruments Burr Brown and or Analog Devices chips, Not sure what way to go and I am looking for some solid advice on Mods To the ADA 8000 . I read a bunch on it recently but did not  reach a conclusive decisions based on the forums on the ADA  8000 AD DA .I know Black Lion did a couple, where the power supply the op amps in and out as well as some Digital chips that help clocking or counters. That might be over my head right now to take on so I was going to start with the Op amps and see what advice would lead me to the best sound for those converters
Thanks

 

[hobiesound]

bartender. give me what abbey road's having and pour him another one....

 

[abbey road d enfer]

Big difference

4580 = 0.0005 THD+N

opa2228 = 0.00005 THD+N

OK, it's a factor 10, but who can hear the difference between something that's buried 100dB down and something also buried 120dB down?

 

[Rossi]

Exactly! And what's really important is not the specs of the individual parts but the performance of the whole box. I doubt THD of the whole preamp will be lower than say 0,05%. Plus the whole faux-toob thing is about *adding* distortion because depending on application and personal preference, some distortion may sound pleasant.

Personally I'm not a fan of the whole "clean + dirty in one box" approach. The really cool boxes - and all that ever approached classic status - are the ones that have their own "natural" sound. Not stuff that tries to be something else. A real tube mic pre has an input transformer, an all-tube circuit and an output transformer. Don't think you can emulate that with a faux-toob distortion circuit added to a basic clean pre. And when it comes to clean, a box like this is not going to perform in the same league as top classs solid state preamps, either.

So what you really have is a neither-nor design. That's an okay approach for a budget box, but no matter how many parts you upgrade (or sidegrade/downgrade), it is never going to sound like a top class box. The whole design, not just its parts, spells out BUDGET.

In this particular case, if you look at the actual preamp stage, you'll find that it looks like a typical budget mixer input. It is no different than a Behringer mixer. Is it bad? No. Is it great? Hell no! It's good *for the price*. And it's never going to be more: It's a dyed-in-the-wool budget design. Yes, you can improve a few things, but no, you can't make it sound great. Rarely worth the effort, IMO.

This comes from someone who has tried upgrading Behringer boxes. I have whole stack of them that I haven't used in at least 5 years.

 

[gemini86]

The one bad thing I'll say about B'ringer's "designs" is that they ALWAYS skimp on the power supply. They get hot, and are always running at the top of their design limits. If an amplifier is a modulated power supply, that seems like the worst place to cut corners.

....okay, the OTHER bad thing about b'ringer is their marketing team. Some of the lit just cracks me up, but then again, that's not exclusive to them either.

 

[Jidis]

Dag. Looks like this guy's been gone a few years, but maybe one of you others knows one of these anyhow:

1. I notice a few mentions in Behringer "T-series" reviews online of meters/backlights failing. This 1953 I grabbed appears to have all four lights dead. Has anyone ever repaired that, and assuming there's an LED somewhere, is it safe to figure the failure is back at the board somewhere rather than the actual meter? If anyone's dealt with those round VU's they do and knows the pinning or light voltage, that would help.

2. As this post is now a bit old, if I'm looking to try the chip swaps, are the OP's 2228's still the safe choice or is there something better now in that price range? I think Mouser had a price break at 10, but it still put them at $4.71 or something each. FWIW, I did a Xenyx 502 a few years ago, so I've already met Behringer's red "glue blobs".

Much thanks!

 

[Waza]

Elna signal capacitors are in and are sounding great. This unit now has a very low noise floor. I had to solder some of the capacitor on the back side of the PCB because of size. I also have changed the tubes to Mullards. Now to work on the power supply and determine what to do.

hello,

what's the reference of the Elna signal capacitors  you used please?
thanks