ECM8000 Mods

Marik,

It may be similar to the pics you posted on page 2, but the variations required enough modifications to the PCB to regard this as "version III".

What I have just noticed is that either (or both) the larger 22uf electrolytic near the XLR connector and the 152 ceramic next to it have been generating enough heat to start to fuse the two together as they were virtually in contact. On my other ECM, even though there is a gap between the two caps, there are signs that heat is causing changes to the cap surfaces.

Gus,

The two small ceramics at the connector end are 222. The other two are both 152. The larger electro is 22uf 35V. The small electro in that group I can't identify other than it is 50V. The larger blue electro is 47uf and the small one next to it appears to be 22uf and the last electro at the capsule end is possibly 10uf. The polyester at the end is 332J.



The transistor is marked .... S9014 C 331, and the IC is A1349. :cool:
 
<It may be similar to the pics you posted on page 2, but the variations required enough modifications to the PCB to regard this as "version III".>

Ausrock,

I meant the schematics is the same, just different PCB layout.

<The two small ceramics at the connector end are 222. The other two are both 152. The larger electro is 22uf 35V. The small electro in that group I can't identify other than it is 50V. The larger blue electro is 47uf and the small one next to it appears to be 22uf and the last electro at the capsule end is possibly 10uf. The polyester at the end is 332J.
The transistor is marked .... S9014 C 331, and the IC is A1349>

All the values are the same as on my version, except of coupling small electrolitics, which are changed to SMs on mine. On one of my boards the transistor is BC118, on another--BC331--same as yours. The A1349 is a 2SA1349 monolitic dual transistor.
 
ausrock said:
What I have just noticed is that either (or both) the larger 22uf electrolytic near the XLR connector and the 152 ceramic next to it have been generating enough heat to start to fuse the two together as they were virtually in contact. On my other ECM, even though there is a gap between the two caps, there are signs that heat is causing changes to the cap surfaces...
Click to expand...
This is very strange. I've never heard of caps heating up like that - unless you happen to be holding a soldering iorn to them. :p

Are you sure they weren't just glued together?
 
Tim,

That was my first thought, so I opened up my second ECM and found that the two caps in question were totally separated but two others were almost touching and showing similar signs. It is always possible that a small quantity of a solvent based liquid has splashed on them and caused some minor damage. I can't imagine flux retaining sufficient heat to do this although the residue around the "melted" surface could be a flux deposit........maybe. Remember, what I am seeing is only a couple of mm square if that and only where the caps are VERY close together.

:cool:
 
Here is complete schematics of the stock one:
 

Attachments

  • behringerecm8000schema.jpg
    behringerecm8000schema.jpg
    35.8 KB · Views: 1,585
ausrock said:
...Remember, what I am seeing is only a couple of mm square if that and only where the caps are VERY close together...
Click to expand...
Ha! I know what happened - they were touching each other when they were being soldered to the board. The heat from the soldering iorn (or flow machine) was just enougn to cause some melting.
 
OK folks,

Hope I did not bore you to death with this one and you are still interested in this project...
The mic is ready to go.
As I wrote before, I tryed a lot of different circuits and made at least five prototypes. I gave up WM60 capsule for two reasons. First, it is more noisy, and second, it is out of production.
The easiest mod would be to change it for WM61 with Linkwitz mod, and to throw away transistor in the circuit with associated parts. Apart of some noise and SPL handling improvement, the sonics are also better. The output will be the smae as a stock one.
Scott Dorsey's mod, as somebody advised, did not work well with this capsule--bass was noticably rolled off.
Finally, I made a circuit based on JFET source follower as a buffer, with constant current source. If you are interested I can draw schematics (right now it is in my head only). I am wondering if I use this circuit for Oktava 012...
I am getting very busy, as I am preparing for major recording of mine. Meanwhile, for this recording I made a new tube preamp, based on Kevin Carter's phono pre, and now working on my new ribbon mic coupled with omni capsule.

Harvey kindly agreed to check the modded ECM8000, so around mid March I will send it to him for review.
 
Thanks, Marik. I think we are still very interested in your work on this mod and others. Good luck on your recording project.
 
Marik said:
Hope I did not bore you to death with this one and you are still interested in this project...
Click to expand...
Not a chance! ;)
The mic is ready to go... The easiest mod would be to change it for WM61 with Linkwitz mod, and to throw away transistor in the circuit with associated parts. Apart of some noise and SPL handling improvement, the sonics are also better. The output will be the same as a stock one... Finally, I made a circuit based on JFET source follower as a buffer, with constant current source. If you are interested I can draw schematics (right now it is in my head only)...
Click to expand...
Yes, a schematic would be great! BTW, earlier at TT you said the output was low with the Linkwitz mod. Did you figure out a way to get around that?
...I made a new tube preamp, based on Kevin Carter's phono pre...
Click to expand...
How did it turn out?
 
Here is the schematics. Please note (it was idea of PRR) that the capsule is floating, and not connected to the ground for even greater than Linkwitz SPL handling. The noise is considerably less than stock one. But even bigger improvement was in clarity, transparensy, details, and bass definition.
 

Attachments

  • behringermodedscema.jpg
    behringermodedscema.jpg
    48.3 KB · Views: 1,764
Thats a nice welcome here, I just joined and the schematic appears. thanks !

And I like the twist with the ground of the capsule at a mid-tap. Sounds like this mic just begs to live inside a kick-drum :-)

But I'm struggling a bit what the mid-tap accomplishes - techs usually start talking they need sleep when they see something they want to understand but just don't grasp at some moment.

OK, let's see, the mid-tap in fact decreases the polarizing potential ? This won't be it.

It also increases the biasing of the FET - that's closer to the explanation perhaps. Please tell (or forward PRRs description - likely he has mailed you a nice little story about this twist ?)

Bye,

Peter
 
If I'm correct the 2SK389GR you specified is a dual JFET. Are you using this mainly because it fits the existing layout well ?

Or do you consider the matching essential for the FET-stage ? (it won't hurt of course)

Thanks,

Peter
 
Oops, I just payed attention I missed 47uf/50V electrotolytic before 15K resistor....

Peter welcome!

<And I like the twist with the ground of the capsule at a mid-tap.>

Normally, to make the capsule's FET to swing a lot of voltage, you would need to use +/- voltage (+ to drain, and - to source through bias resistor) and to connect the capsule's body (internally connected to diaphragm) to ground. Since we have only + polarity of the phantom supply, we fake it. This way diaphragm in reference to the FET is still at the same relationship.

<If I'm correct the 2SK389GR you specified is a dual JFET. Are you using this mainly because it fits the existing layout well ?
Or do you consider the matching essential for the FET-stage ? (it won't hurt of course)>

Yes it is a dual matched FET. It got to the place of 2A1349 (same package). You could use matched 2SK170 here, as well.
I soldered everything on the stock board, however, I needed to cut some traces and put many parts on the other side of the board.
It would be much more convinient to make a dedicated PCB for this circuit.
 
Hi,

I dared to write:

If I'm correct the 2SK389GR you specified is a dual JFET. Are you using this mainly because it fits the existing layout well ?

Or do you consider the matching essential for the FET-stage ? (it won't hurt of course)
Click to expand...

and remembered having seen nice stuff about JFETs. Marik, you probably are familiar with these articles:

http://www.borbelyaudio.com/index15.htm

JFET buffer is in fig.15 of http://www.borbelyaudio.com/ae699bor.pdf

and it also says:

If the two JFETs are matched and the
two source resistors are equal, then the
DC offset will be very small.

OK, that explains it.

Let me end with two small questions that are just based on the mentioned article, and definitely not on the hands-on experience you have with this circuit. (so please tell if they're perhaps true from the technical point of view but still can better be done otherwise):

#1 could it be that the 1M biasing resistor should also be halfway the supply i.s.o. ground ?

#2 the output to the 47 uF cap, should it be attached to the bottom terminal of the upper 10 Ohm resistor ?
(if so, the 47 & 82 output resistors would of course also change)

FWIW...

Bye,

Peter
 
Since I'm pretty dumb and all these schematics mean nothing to me, can you answer in laymans terms, what is the difference between the mic before and after? :) Lowered the self noise? Ta
 
<Marik, you probably are familiar with these articles:
http://www.borbelyaudio.com/index15.htm
JFET buffer is in fig.15 of http://www.borbelyaudio.com/ae699bor.pdf
and it also says:
If the two JFETs are matched and the
two source resistors are equal, then the
DC offset will be very small.>

Yes, I am familiar with this articles, however in the Borbley circuit there is +/- supply, so carefull matching is essential for DC offset
reason and thus eliminating coupling capacitor. Since we have to use this cap anyway carefull matching is not crucial. At first I played without input (1uf) cap as well, using 5V DC from the capsule for biasing output buffer. However, it requared dropping the voltage on the buffer to 10V, which reduced its overload capability.

<#1 could it be that the 1M biasing resistor should also be halfway the supply i.s.o. ground?>

No, I use this resistor as a gate leak one, so the buffer is self-biased. If you connect it as a bias resistor the stage will draw too much current and the supply voltage will drop considerably. (know that, done that)

<#2 the output to the 47 uF cap, should it be attached to the bottom terminal of the upper 10 Ohm resistor ?
(if so, the 47 & 82 output resistors would of course also change)>

Yes, it can, however the difference would be small.
 
<Since I'm pretty dumb and all these schematics mean nothing to me, can you answer in laymans terms, what is the difference between the mic before and after? Lowered the self noise?>

Canopus, good and probably most important question. :)
The improvement was in three areas:

1) Lower noise.

2) Much better SPL handling.

3) Dramatical improvement in sound quality--clarity, transparency, details, bass definition, and transient response.
In fact, I would like to check it against Earthworks.
 
Since, CanopuS and I are both intrested in this, and have NO idea how to go about doing it...;-)

How much would someone charge to perform that MOD that Marik just posted? Would you be willing to do it Marik?
I'd love to hear how they now compare to something like MXL 603s (what I was originally going to buy, but maybe I can save some $$$$), or even Rode NT5's. Or like you said Marik, compare them to Earthworks.

Anyways, you can get one ECM-8000 at 8thstreet for 40 bucks, how much would someone such as Marik charge to perform that MOD?
 
You must log in or register to reply here.
Back
Top