| theChris |
ok, this is a subject which i think is pretty interesting. i must say that there is an interest on this forum about balenced inputs and outputs.
now so far the best gainclone only balenced implementaion was just 2 amplifers giving a differential in - differential out.
the only single gainclone implementation i've seen was a AC-coupled difference amplifer. the problem is that this circut isn't balenced unless there is no signal or noise. otherwise the noise rejection will depend upon the voltage to the noninverting terminal, so more noise = less rejection of noise!
i made a simple volume control that is balenced, but it is obviously conditionally balenced -- it works on the principle that the source and destination is balenced.
now i was looking over the specs of the lm4780, mind you not the choice of many, but still the CMRR of the chip falls severly, so any noise rejection will be limited more to lower frequencies. fine this works for hum if hum is the problem.
so i wanted to start a hopefully meaningful discussion on how best to implement a balenced input amplifer |
|
|
| GregGC |
| What do you say about that? |
|
|
| Steve Eddy |
| quote: | Originally posted by theChris
so i wanted to start a hopefully meaningful discussion on how best to implement a balenced input amplifer |
Here's my two cents worth. :D
se |
|
|
| theChris |
| quote: | Originally posted by GregGC
What do you say about that? |
the input impedce for input 1 is dependent on the input voltage for input 2, thus the circut isn't really balenced, just a difference amplifer.
as for the transformer, well, no arguments there |
|
|
| Mad_K |
This is one of the pages I read when making my bal. gainclone ;)
I can't understand what all the fuzz is about. min. 45dB CMRR, and 0,00078dB signal imbalance (10 ohm SE output Z from source). Who needs higher performance than this for audio? |
|
|
| GregGC |
Mad_K,
I guess the fuzz is that +in has 110k impedance to GND and -in 9.1k impedance to gnd in your case. |
|
|
| Mad_K |
| my XBZLS couldn't care less ;) |
|
|
| Brett |
| quote: | Originally posted by Steve Eddy
Here's my two cents worth. :D
se |
A much better idea than most of the others presented. |
|
|
| Steve Eddy |
| quote: | Originally posted by Brett
A much better idea than most of the others presented. |
Thanks. It's worked for me for quite a few years. Though it can be considerably more expensive than some of the others presented.
se |
|
|
| theChris |
| quote: | Originally posted by Mad_K
This is one of the pages I read when making my bal. gainclone ;)
I can't understand what all the fuzz is about. min. 45dB CMRR, and 0,00078dB signal imbalance (10 ohm SE output Z from source). Who needs higher performance than this for audio? |
well, the circuit isn't balenced. while it is capable of canceling equal noise on both inputs, it is not capable of picking up the same noise on both inputs. if it were balenced, i would not be able to pick up all that hum on my crossover (which uses the circiut with less gain as an input buffer). i can pick up hum by moving the cables near a transformer. |
|
|
| Brett |
| quote: | Originally posted by Steve Eddy
Thanks. It's worked for me for quite a few years. Though it can be considerably more expensive than some of the others presented.
se |
You're welcome. The initial cost is an issue, but there's all the advantages the Tx's have, and providing you don't toast them, will hold their value well for later resale.
Well John, the 2142's had better have improves a lot since I last used them, because they sounded dire. |
|
|
| Steve Eddy |
| quote: | Originally posted by Brett
You're welcome. The initial cost is an issue, but there's all the advantages the Tx's have, and providing you don't toast them, will hold their value well for later resale. |
Perhaps in the pro audio market. With the exception of transformer-based attenuators and MC step-ups, transformers are almost unheard of in high-end home audio.
se |
|
|
| Brett |
| quote: | Originally posted by Steve Eddy
Perhaps in the pro audio market. |
Absolutely, but if you bought some of the Jensens and decided to sell them later, they would sell readily on ebay.
| quote: | | With the exception of transformer-based attenuators and MC step-ups, transformers are almost unheard of in high-end home audio. |
That's their misfortune then isn't it? |
|
|
| Steve Eddy |
| quote: | Originally posted by Brett
Absolutely, but if you bought some of the Jensens and decided to sell them later, they would sell readily on ebay. |
Let's watch:
http://cgi.ebay.com/ws/eBayISAPI.dl...9&category=4665
:)
| quote: | | That's their misfortune then isn't it? |
Hehehe. Naaaaaaah. Hey, if they're happy, I'm no one to argue. :)
se |
|
|
| Mad_K |
yep, a basic difference amp... very informative datasheet also ;)
edit: was supposed to quote dhaen... |
|
|
| greyhorse |
The only input transformers that are mentioned in these forums seem to be ones specially designed for music (pro-fi, or hi-fi) use, and are thus fairly expensive.
Has anyone thought of using cheaper alternatives like those made by Tamura or Hammond? The following transformers from Tamura look interesting, but what do you guys think of them? Can anyone suggest a good circuit using one of these transformers for making a bridged amplifier?
On page 3:
http://dkc3.digikey.com/pdf/C041/1160-1165.pdf |
|
|
| theChris |
| quote: | Originally posted by Mad_K
yep, a basic difference amp... very informative datasheet also ;)
edit: was supposed to quote dhaen... |
just out of curiosity, how is that IC balenced. i agree that it is a difference amplifer, but i don't agree that it has equal input impedances. as i understand it, the idea is to have equal impeances to pick up equal noise then use a difference amplifer to cancel it out.
no one has been able to show me that this circuit is balenced with any input signal. it is balenced without a signal though, so i guess during silence - when noise could be heard - the noise rejection increases. |
|
|
| millwood |
| quote: | Originally posted by theChris
just out of curiosity, how is that IC balenced |
I think it is not balanced too.
However, if you put two buffers in front of it, then the whole thing should be balanced, :) |
|
|
| usekgb |
What about using the BB INA134 for converting balanced to unbalanced? It looks like a great chip: INA134 .
Cheers,
Zach |
|
|
| theChris |
that IC looks to have the same issue as the previous ones -- it does convert a voltage differance into a voltage with respect to a referance, but it does not have equal input impedances unless the voltage at both terminals is 0v. i would think a true IN-AMP would be required to get both conditions -- same noise on both lines and differential inputs.
The high CMRR of the chip is nice though, as the noise rejection will be affected by the ability to reject common-mode signals, but the input imedance of the inverting terminal is a function of the signal on the output. |
|
|
| djk |
"I guess the fuzz is that +in has 110k impedance to GND and -in 9.1k impedance to gnd in your case."
Close, the inverting input would be 10K. Hardly qualifies as balanced, do you think?
SE is right, transformers are a better solution.
Studio and pro gear frequently had octal sockets for your transformers. You invest in good transformers and keep them when you sell the gear.
Jensen makes some of the best.
Electronically balanced inputs are a waste of time and money, even those with laser trimmed CMRR.
http://www.edcorusa.com/sound/matchers/s2s10k_10k.htm
90% of the Jensen quality for 10% of the price. |
|
|
| usekgb |
I do agree with using the Jensens. I was just trying to find a more cost effective alternative to this problem. You would be surprised how little pro gear is loaded with transformers these days.
Cheers,
Zach |
|
|
|
