Understanding the 1176 circuit
« on: October 28, 2010, 01:57:23 PM »
So I've been searching for a couple days for a discussion about the 1176 circuit.  I'm building one (on MNATS rev. J boards), and I figure if I'm going to invest this amount of time and effort, it's foolish to not learn more about how these circuits work.  Since I couldn't find anything on this topic, I thought I'd start one myself.  I've attached an image for your perusal in the discussion.  I suppose this is a bit self serving, but it's really difficult to find practical discussion on circuit design and function.  I've read several books, but nothing really seems to prepare you for circuit analysis.

Basically, I'm trying to go through one section at a time and figure out what's going on.  I'll start with the input section;  Please let me know if I'm on the right track.

So I've labeled the 100pf caps as RFI filters, as I believe they will shunt high frequency information to earth.

Following are the 6k8 EL's that should pass AC and block any superimposed DC, right?  I believe the value is larger so it lowers the corner frequency of the HPF to the bottom of the audio spectrum.

Then we've got input caps.  10k on each leg.  (+/- of the input) those resistors couple with the feedback resistors on each half of the IC to make a unity gain inverting amplifier.

I've labeled what I assume is the correct current flow through each half of the IC.  Blue for the + leg of the input and Yellow for the - leg.  I've also labeled polarity with respect to the incoming signal.  (I'm assuming a balanced input will be symmetrical for the purposes of CMRR).  I'll start with the + leg.

The + signal hits pin 6 (inverting input) of the IC, simply gets flipped over, because the IC is providing unity gain.  At that point, I'm not sure of the function of R1-B (10k).  The + input (now having been inverted; though the common mode will be opposite polarity with respect to the - input) now combines with the - input.  The common mode cancels because of the polarity difference, but the input signal combines, both negative with respect to the desired input.  The combined signal hits pin 2 and gets its polarity reversed at the exit of the IC, so now it's right way round again and can be passed through the input pot to the rest of the circuit.  (Via C33; again a 6k8 EL).

I understand that R1-A and C32 in parallel are creating some kind of HPF, though I'm still not sure how that works.  I also think that C30 in parallel with R4 is somehow necessary because of the relationship with R1-A and C32.  Though I'm not sure exactly how that works, either.

Obviously, R3 and R4 are creating a voltage divider to provide +15 volts to the IC.  It's also worth noting that pin 8 (vcc) is connected to the 30 volt rail on the PCB, though it's not noted in the schematic.

So, does it look like I'm on the right track here, or did I totally botch this?

Any comments would be sincerely appreciated.

Thanks!
Mike



dmp

Re: Understanding the 1176 circuit
« Reply #1 on: October 28, 2010, 02:25:57 PM »
Hi Mike,
 Way to go, you'll have much more fun and success with building if you want & try to understand what the circuits are doing. Some times you'll see things that are really confusing and some times really brilliant. It is a lot of fun though.
This section is a differential amplifier with unity gain. It's function is to take a balanced input signal and output a unbalanced signal.
Quote
At that point, I'm not sure of the function of R1-B (10k).
This is part of the basic differential amp, it sums one inverted signal with the other - common mode cancels, differences are amplified at unity gain. I think you had it.
Quote
I understand that R1-A and C32 in parallel are creating some kind of HPF, though I'm still not sure how that works.
Actually, high frequencies see C32 as a short circuit, while low frequencies see it as 10k. So what happens? At high frequencies the negative feedback resistor decreases, which drops the gain of the circuit. Although they form a high pass filter, in this kind of a amp, they rolloff high frequencies. The 3dB point of 10k and 220pf is 72kHz, ie oscillations, etc... things you DON'T want.

In previous revs, this circuit was a single transformer - which also unbalances an input signal and cancels common mode.

Re: Understanding the 1176 circuit
« Reply #2 on: October 28, 2010, 03:14:41 PM »
Thanks for the reply.

I'm actually using the Lundahl in my 1176, but I wanted to understand how the input section worked regardless.

Quote
Although they form a high pass filter, in this kind of a amp, they rolloff high frequencies. The 3dB point of 10k and 220pf is 72kHz, ie oscillations, etc... things you DON'T want.

That makes sense now.  Do you use standard RC circuit calculations to find the 3dB down point? 

So why would you want to use the electrolytics as coupling caps instead of non polar caps?  Is it because of the necessity for DC bias for AC to pass? 

Also, why would you want the non inverting inputs of the op amp to see 15 volts? 


dmp

Re: Understanding the 1176 circuit
« Reply #3 on: October 28, 2010, 03:42:43 PM »
Quote
standard RC circuit calculations
I used 1/(2piRC)
Quote
why would you want to use the electrolytics as coupling caps instead of non polar caps
You need high values to bring the 3dB point below 20 Hz. Again, 1/(2piRC). 6.8uf and 10k gives you 2.34 Hz
Quote
Also, why would you want the non inverting inputs of the op amp to see 15 volts?
Biasing. The V+/V- of the opamps is 30v / 0v, right? You want your amplifier biased in the middle so it can swing both ways without clipping against the power rails.
By the way, a great reference to learn all the building blocks of electronics is "The Art of Electronics" by Horowitz and Hill.

Re: Understanding the 1176 circuit
« Reply #4 on: October 28, 2010, 03:50:30 PM »
Sold.  I've bought enough other books trying to figure it out.  Getting a recommendation from someone that actually knows what's up is nice.

Thanks again!

elektroivo

Re: Understanding the 1176 circuit
« Reply #5 on: October 28, 2010, 06:22:02 PM »

Also, why would you want the non inverting inputs of the op amp to see 15 volts? 




It would be better to say that this voltage divider ensures virtual ground (half supply), beacouse OPAMPS usually work with bipolar PSU, rather rare with unipolar, like in this circuit. In bipolar circuit, those non inverting inputs would be tied to ground. Normally You would have bipolar PSU of typicaly +/-15V, and all signal would be referenced to ground. Anyway, every bipolar PSU circuit could me reworked to unipolar, and viceversa (if we're talking about OPAMP designs). IMO bipolar designs are better, under premise that You can deal with zero-crossing distortion and some other issues. main advantage in bipolar design is that you can omit caps between stages if circuit is well made, so there is no DC offset between them (if sircuit is rather complex, caps make "fogy" sound).

regards, Ivica

Re: Understanding the 1176 circuit
« Reply #6 on: October 29, 2010, 09:13:50 AM »
Thanks, Ivica, that makes a lot of sense.  No wonder every circuit example I could find was referenced to ground instead of a voltage in this way.

If I understand you correctly, DC offset is a bigger danger in this arrangement, because the rail voltage, and as a consequence, the voltage in the middle of the divider, could change potential?

Thanks,
Mike

isophase

Re: Understanding the 1176 circuit
« Reply #7 on: October 29, 2010, 09:40:26 AM »
great thread !
i'm about to finish my 1176 project (revJ) and i had a few questions concerning the output section:
i can't see any negative feedback tapped off the secondary of the output transformer, and from reading an article on the different 1176 rev models, the original 1176 seems to have some negative feedback. how come ? i know the output gain would change with feedback, but why does the clone have no feedback and how critical is this ?
also, i'm totally broke at the moment and so i can't afford to buy an Lundahl output transformer right now but i have a few neve transformers laying around that i hope i can use in the meantime. was thinking of using one Carnhill orange torodial VT24499 1:1 (600:600). any thoughts on using a 600:600 output transformer on the 1176 output.
i thought since the boards comes with the option of using the lundahls either in a 2:1 or reversed 1:2 configuration, i'm thinking a 1:1 output won't do any harm ???
sorry for hijacking this thread, but i'd love it if someone could give some details on the output section
thank you DI Wires !
"we'll fix it in the shrinkwrap" fz

elektroivo

Re: Understanding the 1176 circuit
« Reply #8 on: October 29, 2010, 11:01:53 AM »
No, Mike, there is no danger of change in rail voltage, beacouse, voltage divider always gives one half of supply, so virtual ground is where it should be, or, rail voltages are simetrical acording to "ground", a as they would be in bipolar design.


 

elektroivo

Re: Understanding the 1176 circuit
« Reply #9 on: October 29, 2010, 11:14:25 AM »
@ isophase

you should not have any problem implementing that transformer, only, You'll have 50% lower signal on output. transformer driver have's low impendance, so, diferent transformer should work fine, just, check your's transformer's datasheet for that load resistor on output, it might be diferent than used with lundahl transformer. You could even make some experiment with it's value. bigger value makes sound lighter, and, smaller value makes darker sound.


you've mentioned feedback on transformer.... look at schematic with little better.. there is big difference in output amplifier, it is not just transformer involved, that is completly different design of amplifier

regards, Ivica


isophase

Re: Understanding the 1176 circuit
« Reply #10 on: October 29, 2010, 11:26:41 AM »
thank you Elektroivo !
"we'll fix it in the shrinkwrap" fz


 

Related Topics

  Subject / Started by Replies Last post
6 Replies
3275 Views
Last post December 01, 2004, 02:17:29 AM
by idylldon
1 Replies
1284 Views
Last post December 22, 2004, 04:42:25 PM
by NewYorkDave
36 Replies
12579 Views
Last post May 16, 2005, 02:55:37 PM
by matthias
6 Replies
1701 Views
Last post August 20, 2006, 12:54:37 PM
by Eric H