THAT1512 based pre

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

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

aortizjr

Well-known member
Joined
Sep 9, 2005
Messages
63
Location
St. Paul, MN
Hi...

I was poking around and couldn't find much info about the THAT1512. It has a bit different gain characteristics making not quite the same as an INA163 or something like that. The 1510 is better for that.

So I started working on the schematics:

Here is what I have so far:

THAT1512-pre-schem-orig.gif


So a few questions:

Is there any optimizations or improvements that can be done to this?
I did this following THAT's data sheets and articles. I also stole some that I saw with some INA163 based schematics.

I would like this to be standalone that feeds a recorder line-in. I know that the THAT doesn't have much as far as ouptut. So it needs a balanced line driver. Would it be best to use say a DRV134 or SSM2142? Or should I build it "by-hand" with other Op-Amps.

If I do go with the DRV134/SSM2142, should I add a buffer like on Kev's DMP? Or can I go without?
http://www.celestial.com.au/~rosswood/diy/dualmicpre/dualmicpre.htm

Also regarding the output DC servo, are there better OpAmps that I can use? Do they really make a difference?

In regards to price, I think it would be cool to have two versions, the budget/bang-for-the-buck version and the best components available version.

Anyway, I will keep chugging away on this slowly, but any help at all would be excellent. Hopefully someday I can post up the full project with all the details, assuming it sounds good of course :)

EDIT:
------------------------------------------------
Here are the schematics after suggestions:

THAT1512-pre-schem.gif
 
Some thoughts:
* C3 470 pF instead of 47 pF (for DIY, simply match C1 and C2 and ground them directly without C3)
* add a little series resistance for the 48 V supply (R1 on the Kevin Ross pre)
* I do not see much point in using R3/R4 - where do you have that from?
* I don't think you need a true floating balanced out as long as you don't drive half a mile of cable around a television tower; simply add a 100 ohm series resistor at the output (gives "hot") and a 100 ohm to ground (for "cold"). This thing is called a "impedance balanced output" and works very well.

Samuel
 
[quote author="Samuel Groner"]
* I do not see much point in using R3/R4 - where do you have that from?
[/quote]

Current limiting for the diodes?
 
[quote author="Samuel Groner"]Some thoughts:
* C3 470 pF instead of 47 pF (for DIY, simply match C1 and C2 and ground them directly without C3)
* add a little series resistance for the 48 V supply (R1 on the Kevin Ross pre)
* I do not see much point in using R3/R4 - where do you have that from?
* I don't think you need a true floating balanced out as long as you don't drive half a mile of cable around a television tower; simply add a 100 ohm series resistor at the output (gives "hot") and a 100 ohm to ground (for "cold"). This thing is called a "impedance balanced output" and works very well.

Samuel[/quote]

The C1,C2,C3 series was taken from the THAT Datasheet. Could you maybe explain why to match C1 and C2 for DIY? Why possibly 470pF for C3?

For the 100Ohm series resistance on the +48V, what does that buy you? Isn't that already supplied by R5 and R6?

R3/R4 was also from the THAT datasheet. I am not even remotely an EE or something. I can read schematics, know some basics (modded pedals, fixed amps, etc.), and can follow directions really well. This is the first that I am building from the ground up.

So the guess from cuelist is probably the best. I can't really say why.

I could probalby go unbalanced and may have an unbalanced out. For that I will definitely need a buffer. But I think for completeness, it should have a balanced out to work well in my studio as well as others.
 
Current limiting for the diodes?
That's what I've thought as well - but a 4r7 is not very helpful (1 A instead of 2 A, perhaps), is it?

For the 100 Ohm series resistance on the +48V, what does that buy you?
Together with C6 it forms a low-pass filter which attenuates noise and hum from your PSU as well as any crosstalk between different channels. Not a necessity with a good PSU and grounding, but a simple add-on to avoid nasty surprises.

Could you maybe explain why to match C1 and C2 for DIY? Why possibly 470pF for C3?
Not easy to explain - let me try: interference (hum from mains transformer, RFI from cell phones etc.) typically arrives "common mode" - it has equal amplitude and phase on both hot and cold wire. Your THAT1512 suppresses these common mode signals by design (the advantage of having balanced lines). If C1 and C2 are grounded and mismatch to some degree, part of the interference is converted to "differential mode" (your mic signal) and does not benefit from the THAT1512-built-in attenuation. If you add C3, this conversion is much less sensitive to C1/C2 mismatch - a good thing if you want to use cheap caps.

Wait - before we go on, we need to now why we have these caps: they are simply here to direct high-freq. stuff to ground instead of letting it pass into our amp where it might do nasty things.

By adding C3, we make this filter much weaker (47 pF in series with 470 pF is about 47 pF - ten times less than the 470 pF alone). The value of C3 is a compromise: larger values do more filtering, but make things more susceptible to C1/C2 imbalance and vice versas.

So if you don't mind matching a bit, you can omit C3. If you by 5% caps, C3=470 pF is probably fine. For 10% and 20% caps, use the 47 pF as shown.

Hope that helps!

Samuel
 
Another suggestion...

Most "real world" designs of a higher calibre I have seen also has some inductive filtering on the input. Typically a 50mH common mode choke. It should have negliable effect on audio performance.

Such a choke would go before the 470pF caps on the input.
 
Samuel Groner - thank you so much for the explanations.

I think I am going to edit the original post and put the edits for the final outcome.

cuelist - how do I go about putting inductive filtering? Sorry, yet again my lack of real electronics understanding is showing :) I will do some searching around though.

EDIT:

Found this:

http://www.digikey.com/scripts/dksearch/dksus.dll?PName?Name=M9851-ND&Site=US

Is this what you are talking about cuelist? Does there need to be some additional circuitry? I guess I really don't know how to even hook these up. How much does it really affect the audio?

How about just using ferrite cores? Does that have similar effect?
 
[quote author="aortizjr"]Found this:

http://www.digikey.com/scripts/dksearch/dksus.dll?PName?Name=M9851-ND&Site=US

Is this what you are talking about cuelist? Does there need to be some additional circuitry? I guess I really don't know how to even hook these up. How much does it really affect the audio?

How about just using ferrite cores? Does that have similar effect?[/quote]

Yes, this is the sort of thing you need. Though this one looks VERY large!

A common mode choke only affects common mode signals, i e signals that are roughly equally present on both + and - signal lines, typically induced RF garbage!
 
Yeah that one is pretty big....

Any other suggestions for the inductor?

I have been looking around and mostly I see the use of ferrite cores and like 15uH inductors.

But also, would I need one or two inductors? I am thinking two, one for the regular signal and the other for the inverted signal.

But some smaller ones would be nice :)
 
So I checked out the 9k SSL pre and noticed that it uses 15uH inductor.

So did you mean 50uH (micro) or really 50mH (milli)?

If it is micro, that would explain the size difference :)

Some others like the green pre and Kev's DMP just ferrite beads... would these work as well? What values should those be so not to interfere with the audio?
 
R11 is at the wrong place right now - should go between "+48V" and C6, not after C6.

For a common mode choke, 50 mH is OK. For two single inductors/ferrite beads use about 50 uH.

Samuel
 
[quote author="aortizjr"]But also, would I need one or two inductors? I am thinking two, one for the regular signal and the other for the inverted signal.

But some smaller ones would be nice :)[/quote]

There are certainly smaller 50mH inductors available. The one you found was rated for 2.3A current!!

The common mode choke has two winding around one core. Signals that are of opposite polarity (normal signals) are passed through unaffected while common mode signals are very effectively attenuated.

You can't accomplish this with two separate inductors.

Here's a pretty good tutorial: http://www.murata.com/emc/knowhow/pdfs/te04ea-1/26to28e.pdf
 
[quote author="Samuel Groner"]
Current limiting for the diodes?
That's what I've thought as well - but a 4r7 is not very helpful (1 A instead of 2 A, perhaps), is it?



Samuel[/quote]

THAT specifies the 4.7 ohm R's so that the distortion introduced by the schottky capacitances is small and the noise contribution of the resistors is negligible. Those are hefty schottkies. It is essential in this arrangement that they be schottkies to keep the voltage drop in forward conduction well below where the chip gets into trouble.

They work, but I'd prefer something a little more complex and less brutal---like maybe 1N4148's with a slightly lower-than-rail-magnitude bias.
 
Cool.. thanks for the info. I will look around some more and see what I can find for the inductors.

I updated the schematic to have R11 between +48 and C6. However, doesn't R5 and R9 still need to be after R11? If so I will have to rearrange things some. Or is the effect the same where it is now?

bcarso - can you post a basic schem of what you are talking about? I will do some homework and see if I can come up with something.
 
[quote author="Samuel Groner"]
:grin: Did not thought that there were so many possibilities to arrange a resistor and a cap!

Look here (Dual_Mic_Pre_BATT.PDF), a picture say's more than thousand words or however this proverb translates from German.

Samuel[/quote]

Yeah I was looking at that and wondering, which is why I asked the question.

I think I have it right now :)
 
why the C9 should be a low ESR cap?
I see on a couple of designs that put a few caps on parallel instead a big one. the ESR is in serie with the gain pot so a couple of ohms don`t affect too much, or i`m wrong?
 
[quote author="Carl_Huff"]Who makes a suitable common mode choke for EMI suppression in mic lines?

Anybody have a vendor and part number handy?[/quote]

This one looks excellent (and small);
 

Latest posts

Back
Top