Intro and PSU opinions

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
I was thinking from the aspect of noise, THD, etc.  If the EQ is going to be just a noise machine, I'll look elsewhere for a different design.  On the other hand, if the noise floor only slightly goes up when the EQ in ON (Yea, I had already planned to put in a bypass) then I would call that "equal".  I guess I can answer that on my own just by looking at the datasheets for each.

The unbalanced phase circuit from ESP is posted below.  From what I have read so far, it's better to flip it after the mic signal has been amplified.

Would it be possible to do something like this with the OPA606 in the MicPre circuit?

14612915671_7afb22150c_b.jpg
 
Giggity said:
If the EQ is going to be just a noise machine, I'll look elsewhere for a different design.  On the other hand, if the noise floor only slightly goes up when the EQ in ON (Yea, I had already planned to put in a bypass) then I would call that "equal".
ALL EQs will increase the noise floor.  The 'cure' is to do EQ at high level.  Which is why all those 'passive EQs' are very poor.

The quietest EQs are those based on State Variable Filters like Gyraf's Calrec EQ .. but you may not gain any practical advantage over your circuit.  And the advantages can be lost by poor design / choice of values.  I'd suggest you breadboard the circuit and hear for yourself.

The unbalanced phase circuit from ESP is posted below.  From what I have read so far, it's better to flip it after the mic signal has been amplified.

Would it be possible to do something like this with the OPA606 in the MicPre circuit?

14612915671_7afb22150c_b.jpg
This is as good as any and better than most.  You are unlikely to get any practical advantage from OPA606 over TL072 in this circuit.  Noise is dominated by the 22k resistors.

Yes.  Flip stuff after the mike signal is amplified.
 
OK.......Add the 220Ω 220µF on the P48v (even though I still don't understand what they are for), add the phase circuit after the preamp and drop the EQ.  Probably better off doing the Black Lion mod on the old Alesis 3630 and get some use out of it anyway.

Given the components I am working with and my familiarity with the program, I think DIYLC will work just fine for laying out the PCB traces.  Will have to work up a BOM so I can get the dimensional specs for all the parts before I start the PCB trace layout.
 
Giggity said:
OK.......Add the 220Ω 220µF on the P48v (even though I still don't understand what they are for), add the phase circuit after the preamp and drop the EQ.  Probably better off doing the Black Lion mod on the old Alesis 3630 and get some use out of it anyway.

Given the components I am working with and my familiarity with the program, I think DIYLC will work just fine for laying out the PCB traces.  Will have to work up a BOM so I can get the dimensional specs for all the parts before I start the PCB trace layout.

That RL has to do with pin 1 problem, as you need the front end to be referenced to chassis, pin 1 is connected as directly as it can, but the 48V phantom power is referenced to analog ground, for practical reasons, because it's referenced to all the PS ground which goes to an analog ground, they should be the same, but they aren't since they have some inductance and resistance. The thing is that's ok to have some noise between one and the other, in fact, that's why shields and chassis are, to pick up the noises from external world and keep them out of the signal and signal reference, so as phantom power is referenced to chassis, looking from the mic side which is what we are interested, and your 48V PS is referenced to ground, you need to decouple it somehow, and the easier way to do so is like this, a series resistor from the 48V PS to a cap, the bigger both the better, more decoupling, too big resistor has too much voltage across and 48V are no longer 48V, but at 10mA 220Ω is just 2.2V, which is fine, with 220µF which is a relatively small cap, you have about 25dB of decoupling at 50/60Hz which is something, bigger cap may not make a lot of sense, expensive, big, etc. and anyway the noise inserted at this point will came as common mode and rejected by the preamp. If you have another transformer or winding for the 48V PS you reference it to chassis and avoid all those caps, but for a 2 channel mic pre is cheaper and easier to use 2 more caps than another transformer. Also you can take this to make a slow turn on/off switching before the resistor, I know 300ms is not a lot to charge it but is better than a spark and for discharge you can go from there to another resistor to ground which will make it as slow as you want, you have to be sure that the resistor is able to charge the cap, will limit the power to 10W but that's just the peak, in 1/3s will be around 3W and when a second and a half later pretty much nothing, I think 1W resistor can take that without problems as long as you don't start to play with it on and of all the afternoon.

JS
 
What is the accepted way to add a LED to the P48v to show that is powering the mic?  All the calcs I found online show 48v as to high.  Just go with it like one normally would with a lower voltage, only make sure the resistor is of a high enough wattage to handle the reduction?

I believe I have the MicPre sorted out.  Below is the latest version.  If there are any glaring mistakes or obvious improvements, please let me know.  If it looks good, I'll start on the PCB layout.  Also, if you don't mind, give the PSU schematic a look for the same thing.  I'll probably work on the PCB layout for it first.

And THANKS AGAIN everyone.

14437534850_11a59a2a5f_b.jpg


14426740057_674ed1fc04_b.jpg
 
> calcs I found online

You GOTTA get some sense of basic electric math.

There isn't an online calc for EVERYthing.

Many of the online calcs make dubious assumptions (like "nobody would be foolish enuff to feed a 2V LED with a 48V supply, whatta waste!").

Some are just bogus. I like pushing them to display "NaN". Sometimes it is too easy.

> add a LED to the P48v to show that is powering the mic?

LEDs can take 20mA, but indoors you rarely need more than 5mA.

48V is practically 50V.

We should allow for the LED drop (1.5V-5V) but compared with 48V, let's skip it.

50V/5mA is a 10K resistor.

50V dropped for 5mA passed is 250 milli-Watts or 1/4 Watt (use a 1/2W part!).

Check: actual voltage might be 48V-3V= 45V. 45V in 10K is 4.5mA. If 5mA is brite enuff, 4.5mA is not gonna suck darkly. It will work.

In practice, LEDs vary in brightness and also distribution. If you know where your eye is you can use a narrow pattern at low current. If you dance all over the room, you may need a wide pattern and more current to light the places you aren't at the moment. So you tack-in that 10K, but have a 5K (4.7K 1W) and a 20K (22K 1/2W) in-hand for when you decide the 10K is not the best brightness.
 
Oh, I can do the math.  Ohms Law.  Works every time.  But with the calcs coming up "you can't do that", I was just wondering if there was something in particular going on with LEDs, or a better way to do it.

Other than that, everything else look good?

 
I don't know what's R16 and R20 doing there, but shouldn't change much...

R14 goes between the switch and the R1/R2 tap, and the C10 between R1/R2 tap and chassis (where it is now, without the resistor). C10 should go close to the connector or at least take the chassis reference from it, I usually do it PTP on the connector in a nasty view with R1 and R2 and R14 may go there if you don't find any other place.

I usually put to VR1, VR2 and VR3 from the PS a series resistor to limit the range, gives two things, better calibration point, when you only have 14V to 20V range for example and you are looking 15V to 18V use, than 0V to 25V in the same trim resistor, It also makes never goes to 0V and dissipate a lot of heat if loaded hard and some PS I did to go to 0V as ref I blow a couple of big caps. R1, R2 and R5 should have a diode in parallel, reversed to protect the IC. D1, D3 and D4 aren't doing much, just protect a reverse voltage in case the bridge fails reversing it self  ??? For C5, C6 and C13 a non electrolytic cap in parallel may be good for better HF rejection, not really needed though. I always make my boards to have parallel caps in case I find it's noisy and that's the solution so I don't have to solder on the back side a cap holding from the bigger one. C3, C4 and C12 shouldn't be so big, you want some non electrolytic there too.

JS
 
I did the phantom power switch that way as per the THAT design note 140. http://www.thatcorp.com/datashts/dn140.pdf  The bottom of page 1, page 2 and the top of page 3 deal with phantom power as related to their 1510 IC.  With my needs, the "type A variant" is what I went with.

I redrew the schematics once more.  Hoping they are good to go with now.  Below is the latest versions of the PSU and the PreAmp.

R23 for the P48v LED is unlabeled as I have yet to determine what LED will be used and will probably vary anyway.  I will also incorporate a "Power ON" led which is not shown either.

14649905193_5d6d0be951_b.jpg


14627867674_451ab12d1a_b.jpg
 
No bother at all.  That's what THAT's Design Notes, and all the others docs, are for.  To reap info from various manufacturer's products.
 
Changed a couple things on the PSU.  D5 & D6 need to be 1N4002.  D1 thru D4 can change as well but shouldn't need it.  Also, C9,C10,C11,C12 could come down to 75v from the 100v spec.  A little cost saving and a bit small footprint on the PCB.

Making up a BOM so I know the cost as well as to get the physical dimensions and pin spacing for laying out the PCBs.  Time consuming.............
 
Can I assume that the last schematics posted look good?  Better to correct now than after I have the PCB layout done.
 
So......considering the lack of any response, shall I assume that either I or this cheap little micpre are not worthy and beneath contempt?  If so, I shall cease and desist.
 
oh, they do worth I guess, I have no experience with those but not because I don't want but I can't get the ICs easily, there are some projects around with those, look at the META of mic preamps and you'll find some with everything already solved for you.

JS
 
I have looked through the META stuff and found that the majority of the links have a "404 - Page not Found" error.  That is the external links.  The ones that stay internal are another story.  But a good number of those lead to a "access denied" page.  Personally, I have found the META pages to be only partly helpful.
 
Giggity said:
So......considering the lack of any response, shall I assume that either I or this cheap little micpre are not worthy and beneath contempt?  If so, I shall cease and desist.
I've 'liked' your post #29.

It's very close to the THAT datasheets & Design Notes which will give good results .. even if you use SSM or TI equivalents.

Some of your later mods are not required but they don't harm stuff so I haven't said anything.
 
You are right about meta, most of them are like that... Anyway there are some projects out there that may help you, I think it's fine as it is but maybe I;m missing something, also you could find already worked boards, at least the archives to self etch directly and no need to redraw what's already done. Use the search tool, you'll find something.

JS
 
ricardo said:
I've 'liked' your post #29.

It's very close to the THAT datasheets & Design Notes which will give good results .. even if you use SSM or TI equivalents.

Some of your later mods are not required but they don't harm stuff so I haven't said anything.


I appreciate that.

You mention that some of the later mods are not required.  Could you elaborate on that?  This design thing is new to me.  I can follow a schematic no problem and someone can say "built this" and I can.  But, I have almost NO idea what is going on in the circuit.



joaquins said:
You are right about meta, most of them are like that... Anyway there are some projects out there that may help you, I think it's fine as it is but maybe I;m missing something, also you could find already worked boards, at least the archives to self etch directly and no need to redraw what's already done. Use the search tool, you'll find something.

JS

The best link I found in the META was one to a PDF about etching PCBs.  I did find one other (that I can not find again) about PSUs that was very similar to mine.  If I could have found it again in the META, I'm sure it would have saved some questions.  I have spent hours with "search" as well as the META and my results have been somewhat iffy.

One thing I have been wondering about after looking at the THAT1510 datasheet as well as the design notes is about the OPA606.  It was in another schematic I previously posted but nothing like it was mentioned from THAT.  What is it's function, if anything worthwhile?
 
Back
Top