DOA AMEK 2500 preamp

UPDATE: I'm picking up where i left last year. Read here: http://groupdiy.com/index.php?topic=55946.msg768092#msg768092
---
I was looking around for a small-ish circuit to make a compact multi-channel preamp box, and stumbled upon this over at gyrafs site: http://www.gyraf.dk/schematics/Amek_2500_micamp.gif
Not too complex, fairly low parts count, no obscurities yada-yada..

I've redrawn it in Eagle: https://docs.google.com/file/d/0BxbSxLy_O-iOWFZxbVF3RVUwY2c
A second set of eyes would be appreciated - just in case

I'll be adding connections to add phantom power to the schematic, otherwise leaving it as is. No bells and whistles needed here...

If everything goes well, I'll make a shared project on OSHpark. My preliminary design sketch implies that the boards should fit vertically inside a shallow 1U rack case at 40x60mm, and the price would be about $18 for three boards shipped directly from OSHpark, in case anyone else would be interested

Does anyone have experience with this circuit - good or bad? Cool mods? Whatever?

Peace out // grrrunge

Schematic with added Phantom power:
EDIT: R26 and R27 should obviously be 6K8 - It's 2AM here, I really should've hit the sack a couple of hours ago

what transistors for Q1 - 10?

It is kind of interesting, I cannot say I understand some aspects of it (even before your conversion).

I disagree with the "no obscurities..." if they aren't obscurities they are at least befuddlements.

The two series polarized capacitors at the front.  Was that a size/voltage issue (they couldn't find a 48V cap small enough)?  The capacitance of the two in series is what 5uf?  Typical coupling caps in mic pre's in that spot are 47uf ish or a little less.

It is not transformer coupled and it uses a gang of PNP's (is the gang to reduce shot noise or something... someone far geekier than me would need to chime in) as input.

I am pretty sure there is no benefit to your KEEPING the 2 series CAPs, both polarized the same way.

The schematic has a differential design for most of it.  Meaning it amplifies the two signal lines separately, which has the advantage of allowing common mode noise to remain so and stay out of the signal... but then wham, they make it single ended at the end.  I don't get that.  Amek's 9098i mic input schematic remained differential through the whole thing, balanced in balanced out, and was the inspiration for the green pre (the green pre is a copy).

Anyway I have not looked at the circuit in detail, but these are rough impressions.  Take a look at the green pre, or the Amek 9098i input section.

I am interested in what folks chime in about the 4pnp's though (is the idea to make the noise incoherent so it doesn't get amplified as much?).. and the series caps on the input.

Also, board seems incomplete, is there power to the IC's?  IC's should have decoupling caps on the power rails (often not shown on schematic if power is not shown)

Amek 9098i is easy to find

Green pre is here, you have to look around.

Schematic
http://www.mhumhirecords.org/DIYpics/Green/gren_pre_cct4.gif

Boards
http://www.mhumhirecords.org/DIYpages/Green/_V14-All.gif

ricardo said:

what transistors for Q1 - 10?

Click to expand...

Schematic says ZTX214P, I found ztx214.. http://hardware.speccy.org/datasheet/ztx213.pdf

thanks bruce0.  ztx214 is the Zetex version of TI bc214; an ancient and very popular 'LN' pnp.  It's not very LN by modern standards having rbb' (equivalent series resistance on the base) of about 150R ... except for source resistances of more than 5k.

Paralleling them also parallels rbb'.  Five of them gives an effective rbb' of 30R .. nearly as good as the 25R of 2sa970.

This is an old, complex & naive circuit striving for clean & neutral.  Simple THAT1510/12 circuit will likely be better for 'clean'

There's no point in more complexity unless you want less than 1nV/rtHz.  If you do, the best for that is Graeme Cohen's circuit that is used by Millenia Media and probably Earthworks .. probably the quietest practical mike preamp in the known universe.
http://www.leonaudio.com.au/double.balanced.mic.amp.notes.pdf

None of the many Cohen copies, including the Greens and various kits come near its potential performance.

It has unbalanced out cos it is probably NOT a standalone preamp but feeds other unbalanced stuff in a mixer strip.  You can make it 'balanced' out by having another differential stage in parallel with your output OPA but connected in reverse.  I would just make it Impedance Balanced

This was a project years ago, I built a pair of 2500 pres and two greens, i use more often thegreen but not for an especific reason, it is possible to have an balanced impedance output just adding a resisitor.

That project is not public anymore

12afael said:

it is possible to have an balanced impedance output just adding a resisitor.

Click to expand...

The issue I was raising was keeping it differential.  Like the AmeK 9098i, the Lola, the green pre, and for that matter the ThatCorp 1570 and 1583. 

I understand that you can make a balanced out, but keeping the signal balanced throughout  provides an opportunity to be pretty immune to the "G-Word"  (ground)

bruce0 said:

I understand that you can make a balanced out, but keeping the signal balanced throughout  provides an opportunity to be pretty immune to the "G-Word"  (ground)

Click to expand...

Impedance Balanced is as immune to ground as any more complex design.  Resistors are pretty good compared to extra OPAs for THD, noise, sound bla bla ;D

But do whatever floats your boat.

Ricardo:

Not sure I understand what you mean by Impedance Balanced (do you mean both signal legs have the same output impedance?)

My understanding was that if the two signal paths were entirely separate, and never became a ground referenced unbalanced signal (even internally in the amplifier) that the advantage was that any ground noise would remain common mode.  I realize that with PCB layout and good design all of this is angels dancing on the head of a pin (single ended can have low noise), but I though that is why differential signal paths were used.  Am I wrong here?

I thought that's why green, amek, Thatcorp and for that matter Graeme Cohen's circuit that you mention were fully differential.


Thanks for the thoughts, I alway appreciate them... and thanks for reminding me of my boat, it makes me smile (My boat is still out of the water, but summer is coming!)..
b

ricardo said:

bruce0 said:

I understand that you can make a balanced out, but keeping the signal balanced throughout  provides an opportunity to be pretty immune to the "G-Word"  (ground)

Click to expand...

Impedance Balanced is as immune to ground as any more complex design.  Resistors are pretty good compared to extra OPAs for THD, noise, sound bla bla ;D

But do whatever floats your boat.

Click to expand...

bruce0 said:

My understanding was that if the two signal paths were entirely separate, and never became a ground referenced unbalanced signal (even internally in the amplifier) that the advantage was that any ground noise would remain common mode.  I realize that with PCB layout and good design all of this is angels dancing on the head of a pin (single ended can have low noise), but I though that is why differential signal paths were used.  Am I wrong here?

I thought that's why green, amek, Thatcorp and for that matter Graeme Cohen's circuit that you mention were fully differential.

Thanks for the thoughts, I alway appreciate them... and thanks for reminding me of my boat, it makes me smile (My boat is still out of the water, but summer is coming!)..

Click to expand...

The definitive treatment on balanced lines is by Bill Whitlock
http://sound.westhost.com/articles/balanced-2.htm

This lists ALL the REAL advantages of properly implemented output 'balancing'.  Note most of the electronic solutions DON'T have these advantages .. especially the common ones like Green etc. but introduce yucky behaviour.

Also worth careful study are
http://sound.westhost.com/balance.htm
http://sound.westhost.com/project87.htm

References to all the important original papers are in these articles.

In the next thread, I am crucified for my heretical beliefs for IMHO, marketing reasons  :'(
http://www.proaudiodesignforum.com/forum/php/viewtopic.php?f=6&t=464

Loadsa good measurements except the ones I really wanted to see for THAT 1600 .. RF immunity.  Immunity to rubbish is the ONLY reason we use balanced lines.

The Cohen (of which the Greens etc are poor copies) has output balancing for marketing reasons.  There's a Cohen version, which I can't find again, using a pseudo floating output instead of the crude double OPAs that recovers some (but not all) of the advantages.
_____________

My boat wasn't very happy with Cyclone Ita and is unlikely to float again  :'(
https://www.facebook.com/photo.php?fbid=753526097999217

ricardo said:

The definitive treatment on balanced lines is by Bill Whitlock
http://sound.westhost.com/articles/balanced-2.htm

This lists ALL the REAL advantages of properly implemented output 'balancing'.  Note most of the electronic solutions DON'T have these advantages .. especially the common ones like Green etc. but introduce yucky behaviour.

Click to expand...

I was aware of that, and had read it.  I even built a discrete version of his "bootstrapped" receiver for "fun", (the circuit he licensed to ThatCorp for the InGenius receiver).

ricardo said:

Also worth careful study are
http://sound.westhost.com/balance.htm
http://sound.westhost.com/project87.htm

References to all the important original papers are in these articles.

In the next thread, I am crucified for my heretical beliefs for IMHO, marketing reasons  :'(
http://www.proaudiodesignforum.com/forum/php/viewtopic.php?f=6&t=464

Loadsa good measurements except the ones I really wanted to see for THAT 1600 .. RF immunity.  Immunity to rubbish is the ONLY reason we use balanced lines.

Click to expand...

I had read the ESP articles too (but I admit I had not seen your crucifixion on Pro Audio design (interesting bit of trivia... turns out on that site 21% of your posts are related to Balanced Lines... so maybe I am walking into the meat grinder here)).  I plan to read through that thread, it looks interesting.

ricardo said:

My boat wasn't very happy with Cyclone Ita and is unlikely to float again  :'(
https://www.facebook.com/photo.php?fbid=753526097999217

Click to expand...

Your question on the ThatCorp 1600 series is interesting. ThatCorp has a very innovative receiver chip (120x inGenius series based on Whitlock's bootstrapping thing) which I think does offer some real interfacing advantages when dealing with "whatever gets plugged in".  I think ThatCorp is trying to piggyback (from a marketing point of view) on that innovation with the 1600 series "OutSmarts", but I can't say I see the advantages there.  I haven't measured them though, but I am using a couple.

All of the articles you mention seem to deal (thoroughly) with interfacing using balanced lines.  I think I have a handle on the issues there pretty well (having read everything I could from Whitlock, he writes so clearly).

My comments were related to the benefits (or lack thereof) of maintaining the balanced signal internally in the amplifier.  Which is something that some designs do (including ones I mentioned before).  Any thoughts on that?

And I am genuinely sorry about your boat... the link says "temporarily unavailable"... or is that the joke? ;D

bruce0 said:

ricardo said:

My boat wasn't very happy with Cyclone Ita and is unlikely to float again  :'(
https://www.facebook.com/photo.php?fbid=753526097999217

Click to expand...

Your question on the ThatCorp 1600 series is interesting. ThatCorp has a very innovative receiver chip (120x inGenius series based on Whitlock's bootstrapping thing) which I think does offer some real interfacing advantages when dealing with "whatever gets plugged in".  I think ThatCorp is trying to piggyback (from a marketing point of view) on that innovation with the 1600 series "OutSmarts", but I can't say I see the advantages there.  I haven't measured them though, but I am using a couple.

All of the articles you mention seem to deal (thoroughly) with interfacing using balanced lines.  I think I have a handle on the issues there pretty well (having read everything I could from Whitlock, he writes so clearly).

My comments were related to the benefits (or lack thereof) of maintaining the balanced signal internally in the amplifier.  Which is something that some designs do (including ones I mentioned before).  Any thoughts on that?

And I am genuinely sorry about your boat... the link says "temporarily unavailable"... or is that the joke? ;D

Click to expand...

The link was to a Cyclone Tracking Map.  It shows Cat 4, with us in the Very Damaging Winds area when it hit.  Cooktown was lucky.  Only 1 roof gone (local pub).  Some verandahs & fences.  Only a few windows.  Loadsa big trees down but they all seem to miss houses.  No one injured.

The 'balanced signal internally' advantages don't extend to the output.  The Amek is no better or worse than Cohen etc in this respect.  I'm not knocking it.  But it's slightly worse than a good THAT1510/12 implemenation while being more complex than Cohen.

If I did something as complex, I'd want Cohen performance .. while the better Greens etc only give THAT performance at best.

In case it isn't obvious, I really like the THAT stuff.

ricardo said:

The 'balanced signal internally' advantages don't extend to the output.

Click to expand...

I get your point.... but it depends upon implementation of course. In many designs (like ThatCorp's) it does create equal output impedance when driving a balanced line (because both halves are the same) which gets better performance out some (lesser) receivers.

bruce0 said:

I get your point.... but it depends upon implementation of course. In many designs (like ThatCorp's) it does create equal output impedance when driving a balanced line (because both halves are the same) which gets better performance out some (lesser) receivers.

Click to expand...

Equal output impedance is ESSENTIAL and in all the good solutions.

Whitlock deals with the advantages of the various drivers when using 'lesser receivers' ... as does the proaudiodesign thread.

Standing back as far back as possible, only THAT16xx (not the SSM & DRV versions) and what I call "Self's Fig 5" have significant advantages over El Cheapo in the proaudiodesign thread.

I don't have any or know of any 'real life' experience with Self's Fig5 except some snide remarks by Self and Uneeda Audio.
_____________

But these balanced output issues have nothing to do with AMEK 2500 or the Cohen variants.

I searched with no luck, for the "el cheapo.gif" schematic mentioned.  Any link left?

ricardo said:

Equal output impedance is ESSENTIAL and in all the good solutions.

Standing back as far back as possible, only THAT16xx (not the SSM & DRV versions) and what I call "Self's Fig 5" have significant advantages over El Cheapo in the proaudiodesign thread.

Click to expand...

Respectfully:
Equal output impedance is not essential, in fact it doesn't help a bit, if you use a transformer.

It is also not essential if you do the bootstrapped impedance to ground thing (InGenius) but it does at least help ( at higher frequencies) in that case.

As far as standing back, well the thread you referred me to pretty much exposes everyones ability in the "standing back" department ( ) 

But I am only on page 16... So maybe by page 21 everyone is calm and collected.
But it is an interesting thread...  (FYI... I do know the definition of BEST driver and/or receiver for ME. What is best for me is that I can screw up a connection and it still works well).

bruce0 said:

Equal output impedance is not essential, in fact it doesn't help a bit, if you use a transformer.

It is also not essential if you do the bootstrapped impedance to ground thing (InGenius) but it does at least help ( at higher frequencies) in that case.

Click to expand...

Why don't you study Bill Whitlock's stuff carefully.  Also Rod Elliot etc

In proaudiodesign, I show an important practical case where matched output Z isn't strictly essential.  But Whitlock's 'balanced bridge' concept is the important one to determine how matched things need to be.

Yes.  Transformers are the most resistant to mismatching .. hence InGenius.

My prejudices are from investigating 'real life' practical behaviour of Balanced Transmission systems at Calrec for the Soundfield Mk4.  HF matching is essential if microphone lines are not to pick up clicks, pops & thyristor theatre lighting.  Mobile phones are probably the biggest evil today.

For that, a symmetrical PCB layout is essential for the input stage.  Far more important than 0.1% matched resistors.  We just used 1% resistors from the same batch.

(FYI... I do know the definition of BEST driver and/or receiver for ME. What is best for me is that I can screw up a connection and it still works well).

Click to expand...

I started the thread to find out what people thought were the REQUIREMENTS for BEST.  Somehow that got lost along the way ... though the info is still there if you look beyond the crucifixion.

Your requirement is an important one I share.  Thanks.

ElCheapo.gif is something I've used with great success in commercial gear and I show it with PROS & CONS in proaudiodesign.

I got only a non tech comment about this guy.

This is the roughly same micpre circuit as my Blue Angela's A10 module. 
These guys sound DAMN GOOD !!  I bread boarded one
up a few years ago using BC560Cs and it worked great.

Best of luck with this.. I will be interest in PCBs
when you get all the bugs out.

GARY

BTW.. use .1% guys for the 6.8ks

Sorry to hijack your thread ggrrunge.

Let us know how it works out.  Also I would be interested in what the series caps are about.

gar381 said:

This is the roughly same micpre circuit as my Blue Angela's A10 module. 
These guys sound DAMN GOOD !!  I bread boarded one
up a few years ago using BC560Cs and it worked great.

Click to expand...

Good to hear

bruce0 said:

Sorry to hijack your thread ggrrunge.

Let us know how it works out.  Also I would be interested in what the series caps are about.

Click to expand...

I've been googling a bit, but i can't find any info on improved characteristics on caps in series ??? I'm thinking about replacing them with 4µ7 film caps instead.