Control room amp with stereo to mono design trial

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

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

mik

Well-known member
GDIY Supporter
Joined
Jun 4, 2004
Messages
461
Location
Milano
hi gentlemen, here is a drawing for a control room section for my summing amp. this design works quite well with out the Stereo to Mono option, My goal is to fit this option in it, would some one check it to recognize errors before starting protos ?
thank you.
P.S this drawing works with single ended PSU ( 24VDC )
M.
clik to enlarge:
controllroomresize5rp.jpg
 
[quote author="mik"]stange thing, mine is perfect visible I will try to resize it
m. try this:
click on to enlarg image
  • http://img233.imageshack.us/img233/6059/controllroomresize0um.jpg
[/quote]

There's something odd about the file. It looks like you created it with Photoshop CS, but I don't have that here. 'xv' complains about an unsupported color conversion request, and The Gimp tells me that it can't open a JPEG with four color channels.

Steve.
 
[quote author="mik"]try this
  • http://img314.imageshack.us/img314/4292/controllroom36qa.jpg
[/quote]

Sorry, that's no better.

I wonder if your original image has transparency or layers that are causing problems. Can you use "Save For Web" to create your JPEG? Alternatively, you could try saving as a GIF.

Steve.
 
[quote author="mik"]... My goal is to fit this option in it, would some one check it to recognize errors before starting protos ?
P.S this drawing works with single ended PSU ( 24VDC ) [/quote]

Hi Mik,

The R1 and R2 must be also DC decoupled from the input pots.
The R9 and R10 are too small (also too complex circuitry around K3).
You don't need R4,R12 network.
The IC 2 pin 8 is +Vc.
Rotate verticaly the IC2D.

Regards,
Milan
 
ok Milan, I ' ve to coupled R1, R2 with 1µF cap, then flip verticaly ic 2 ( :oops: ) R9 , R10 100 or 120 ohm ( I guess ) arrangement around K3 is too complex..... ?im looking for some thing for avoid pop 'n tik wen switch on it..
R4,R12 network : are you mean pin 3 on ic to go diretly to the voltage divider R49, R50 ?
2nd version:
controllroom6bis7mi.jpg
 
[quote author="mik"]ok Milan, I ' ve to coupled R1, R2 with 1µF cap, then flip verticaly ic 2 ( :oops: ) R9 , R10 100 or 120 ohm ( I guess ) arrangement around K3 is too complex..... ?im looking for some thing for avoid pop 'n tik wen switch on it..
R4,R12 network : are you mean pin 3 on ic to go diretly to the voltage divider R49, R50 ?
[/quote]

Now that you've added C12 and C13 you can remove C1 and C2.

Can you also remove R12 and make the bias network the same as you have on IC1A and IC1B?

You show 1M variable resistors in the feedback loops of the op-amps. Why so big? If these are just for trimming the gain then I'd suggest, for example, 43K fixed plus 10K variable, to give a range of possible values above and below 47K. I'd also consider missing out the variable resistors altogether. They're expensive and not as reliable as fixed values. Use 1% resistors and I bet your gains will be close enough.

I'm a bit confused about what you're trying to achieve with the arrangement around K3. You say you want to avoid clicks and pops. Can you explain how you think this works? You're going to have trouble with R9 and R10. These are still far too small - almost a short circuit connecting your L and R channels together. Why not just replace this whole arrangement with a DPDT switch selecting between the stereo channels and the mono mix?

Do you expect a clicking problem here? What causes it? Different DC levels on the outputs of the three op amps?
Would it help if you used one DC bias network for the whole system, instead of three? That way you would know that everything was using the same DC reference.

I'm also wondering why you need both K1 and K3 switches. If you're selecting the outputs with K3, does it matter that the mono mix amp is always supplied with signal? You could save a whole relay.

Steve.
 
Smorphet, pleas be patient to me, I confused you cause I'm beginner and this the first task.

ok,
sistem simplified , R 31, 51, 15 ( 1 M Pot ) are not important for now. the walue for R9,10 is the most important, does 47K works there?

is DC reference right, how to do to have only one bias network ?

3rd try
controllroom7resize7ru.jpg
 
[quote author="mik"]
sistem simplified , R 31, 51, 15 ( 1 M Pot ) are not important for now. the walue for R9,10 is the most important, does 47K works there?
[/quote]

R9 and R10 could still be a problem. If they're too big then the output stages will see a high impedance input, and that might affect their gain.

I think it would be easier if you replaced the K3 switch, which would let you avoid any resistors in this area. Then you could consider leaving out K1.

However, I have an idea that might be even better, and would allow you to save a relay and an op amp. I will need to sketch this to explain. Can you wait until I get home this evening?

is DC reference right, how to do to have only one bias network ?

Better, but now you have connected R45/R46 to R49/50. This will work, but the parallel 2K resistors are just like having a single network with 1K resistors. Again, I think it is easier if I show you some of the possibilities on a diagram.

Steve.
 
OK, here's what I was thinking about. Your two existing op amps (IC1A and IC1B) are inverting configurations, so it's simple to adapt them to sum the channels:

mono_stereo_01.gif


With the switch in the mono position, both inputs are connected to the summing junction of each op amp through 67K resistors. With the 47K feedback resistor that gives the same -3dB gain that you used in your original circuit.

With the switch in the stereo position an extra 90K resistor is switched in. 90K + 67K in parallel with 67K is 47K, so each channel has unity gain.

The connection between IC1A, IC1B and the output drivers can remain unchanged from your existing stereo circuit. Another advantage is that if you keep the same number of op amps then you can continue to use the same DC bias arrangement that obviously works well for you at the moment.

A potential disadvantage of this idea is that the input impedances change slightly when you switch between stereo (47K) and mono (33K). I think that's very unlikely to cause any problem.

I've sketched some of the options for DC bias (if you choose not to use my stereo/mono switch idea). To use a single bias network, you would simply use one resistor divider, and connect it to every point in the circuit that needs a reference voltage. You're already sharing the R45/R46 network between the IC1A and the output driver circuit IC2A and IC2B. The same concept can be extended to additional devices.

However, as you connect more things to it, the reference voltage is more likely to be disturbed by the load on it. You could reduce the resistor values to make the reference voltage "stiffer", but this would consume more power. Eventually you would need to consider an active method, i.e. a voltage regulator, to produce the reference voltage.

dc_bias_01.gif


I hope this helps.

Steve.
 
Steve............ :sam:..............yes................... :sam:................this really help me........ :thumb:
:sam: :sam: :sam:
I'll let you know , now I've to study it.
Many many thanx :sam: :sam: :sam: :sam: :sam:
M.
 
[quote author="mik"]here's the corrected version with the Steve Stereo Mono Switch, and one bias network for all Ic:
:sam: :sam:[/quote]

The mono/stereo switch looks good. I think you should turn your capacitors, C12 and C13 around, so that the + side is towards the opamps. The whole resistor network will be at the DC bias reference voltage, which is more positive than the voltages on the volume pots.

You will find that you can't buy resistors of exactly 67K or 90K. I think if you do the calculations you'll find that E24 standard values of 68K and 91K will be close enough.

In the bias system you have four points that require the reference voltage: IC1A pin 3, IC1B pin 3, the R39/R37 junction, and the R21/R25 junction. Each of these nodes needs to be connected back to the reference point at the junction of R47/R48. You're missing just one link, from R47/R48 to IC1A pin 3.

Steve.
 
Back
Top