mic preamp modification

[rex]

I recently bought a Marantz PMD670 -- a digital audio recorder. It seems
pretty good except the mic preamps have a pretty high noise level for a
recorder that is good otherwise. I'd like to see if I can make it
better. (It isn't broken, its a design issue.)

As a general description, the mic inputs are balanced XLR with
switchable phantom power. It only seems to work decently with pretty
high-level condenser mics. A lower dynamic seems to require cranking the
gain way up and the crappy noise floor becomes apparent.

One guy offers a mod for the preamps that improves things considerably.
link:
http://www.oade.com/digital_recorders/hard_disc_recorders/PMD-670.html
He only does it on stuff he sells and I already have mine. I asked him
if he would share anything about the nature of his change, but he
declined.

I don't have any schematic, but I opened the thing up and had a look at
what is in the preamp area. Assuming the signal flows sensibly across
the board, there are (per channel) two largish electrolytics. Then a
couple sot-23 devices (transistors I assume) marked 62Z. Then an HC4052A
analog mux. Then a JRC 2068 op amp.

I was surprized to see that 4052 so close to the input. Not sure what
exactly it is doing.

If I was just to replace the 2068 with something else, do you think it
would help? If so, what might be good choices for replacements?

Anyone know what the 62Z devices are? I did some searching but didn't
find anything that matched. Do you think that bears looking at?

What are the chances that the caps are contributing? I'm not sure what
these are. They have a symbol that is an S in a circle. They are marked
CE105 C above 0402. Is 402 the value? What does it mean?

Any thoughts about the HC4052A?

I know this isn't much to go on without a schematic. I don't have much
experience with quality audio circuits. I thought I would start by
asking opinions here about whether swagging out any of these parts is
worth the effort.

Thanks for any opinions.

 

[Joerg]

Hello Rex,

FWIW, the 4052 is a dual 4:1 multiplexer. Sometimes also used to switch
resistors in and out for gain control. I don't know about this
particular opamp but check it's data sheet for noise performance. If it
is considerably worse than others a swap might help.

Anyways, you might want to print all the data sheets of the parts and
draw a schematic to understand how it works.

Phantom power sources can be a noise contributor but if so that could be
filtered.

Regards, Joerg

http://www.analogconsultants.com

 

[Pooh Bear]

rex wrote:

I recently bought a Marantz PMD670 -- a digital audio recorder. It seems
pretty good except the mic preamps have a pretty high noise level for a
recorder that is good otherwise. I'd like to see if I can make it
better. (It isn't broken, its a design issue.)

I'll guess it's a design issue. ;-)

As a general description, the mic inputs are balanced XLR with
switchable phantom power. It only seems to work decently with pretty
high-level condenser mics. A lower dynamic seems to require cranking the
gain way up and the crappy noise floor becomes apparent.

One guy offers a mod for the preamps that improves things considerably.
link:
http://www.oade.com/digital_recorders/hard_disc_recorders/PMD-670.html
He only does it on stuff he sells and I already have mine. I asked him
if he would share anything about the nature of his change, but he
declined.

Tuh ! Silly wanker.

I don't have any schematic, but I opened the thing up and had a look at
what is in the preamp area. Assuming the signal flows sensibly across
the board, there are (per channel) two largish electrolytics. Then a
couple sot-23 devices (transistors I assume) marked 62Z. Then an HC4052A
analog mux. Then a JRC 2068 op amp.

I was surprized to see that 4052 so close to the input. Not sure what
exactly it is doing.

If I was just to replace the 2068 with something else, do you think it
would help?

NO. The 'front end' is what matters.

If so, what might be good choices for replacements?

Anyone know what the 62Z devices are?

That's a typical SMD code. There isn't enough space to fit the full part
number. Sorry !

I did some searching but didn't
find anything that matched. Do you think that bears looking at?

NO.

What are the chances that the caps are contributing?

So close as next to nil that you might as well forget it.

I'm not sure what
these are. They have a symbol that is an S in a circle. They are marked
CE105 C above 0402. Is 402 the value? What does it mean?

105 means 105 degrees C rating - that's fine. 0402 is normally an SMD
footprint size.

Any thoughts about the HC4052A?

Nope.

I know this isn't much to go on without a schematic. I don't have much
experience with quality audio circuits.

I do !

I thought I would start by
asking opinions here about whether swagging out any of these parts is
worth the effort.

Unlikely. Once the circuit topology is fixed there's usually bugger all that
you can do to improve it.

The Eastern guys don't have a clue how to make a good mic amp - apart from
Yamaha maybe ?


Graham

 

[Pooh Bear]

Joerg wrote:

Hello Rex,

FWIW, the 4052 is a dual 4:1 multiplexer. Sometimes also used to switch
resistors in and out for gain control.

Erk ! Thinks *non linear resistance with applied voltage*.

I don't know about this
particular opamp but check it's data sheet for noise performance.

It's actually quite good.

If it is considerably worse than others a swap might help.

It won't.

Anyways, you might want to print all the data sheets of the parts and
draw a schematic to understand how it works.

Phantom power sources can be a noise contributor but if so that could be
filtered.

Damn well shouldn't be a noise source unless the designer was mentally
defective !

Graham

 

[Joerg]

Hello Graham,

FWIW, the 4052 is a dual 4:1 multiplexer. Sometimes also used to switch
resistors in and out for gain control.

Erk ! Thinks *non linear resistance with applied voltage*.

Yes. The true audio freak would frown and at least use an SD5400 here.

Phantom power sources can be a noise contributor but if so that could be
filtered.

Damn well shouldn't be a noise source unless the designer was mentally
defective !

Then I guess I have diagnosed mental defects in two designers
life-to-date...

Both systems looked very expensive and "pro", BTW. Needed ferrites and
better electrolytics.

Regards, Joerg

http://www.analogconsultants.com

 

[Larry Brasfield]

"Joerg" <notthisjoergsch@removethispacbell.net> wrote in message
news:yMy5e.877$RQ7.31@newssvr14.news.prodigy.com...
Hi, Joerg.

Hello Graham,

FWIW, the 4052 is a dual 4:1 multiplexer. Sometimes also used to switch
resistors in and out for gain control.

Erk ! Thinks *non linear resistance with applied voltage*.

Yes. The true audio freak would frown and at least use an SD5400 here.

In one project I worked on, one of the most competent and
conscientious engineers I've known used the 8:1 mux from
that family for a wide, RF (2 to 13 MHz), barrel switch
where noise and distortion were both critical parameters.
(The signal at that point was boosted enough that thermal
noise from the switch was not too harmful, and judicious
use of carefully biased followers eliminate most distortion.)

Phantom power sources can be a noise contributor but if so that could be
filtered.

Damn well shouldn't be a noise source unless the designer was mentally
defective !

Then I guess I have diagnosed mental defects in two designers life-to-date...

I expect you were more charitable than to do that.

Both systems looked very expensive and "pro", BTW. Needed ferrites and better electrolytics.

Often, I think, appearance is the primary design
goal in expensive audio gear sold to consumers.

Best regards,
--
--Larry Brasfield
email: donotspam_larry_brasfield@hotmail.com
Above views may belong only to me.

 

[Joerg]

Hello Larry,

Yes. The true audio freak would frown and at least use an SD5400 here.

In one project I worked on, one of the most competent and
conscientious engineers I've known used the 8:1 mux from
that family for a wide, RF (2 to 13 MHz), barrel switch
where noise and distortion were both critical parameters.
(The signal at that point was boosted enough that thermal
noise from the switch was not too harmful, and judicious
use of carefully biased followers eliminate most distortion.)

Being a cheapskate when designing stuff I often use PIN diodes with a
long enough carrier lifetime for RF muxing. The SD5400 is usually
reserved for more valuable tasks such as servoed gain or phase control.

Damn well shouldn't be a noise source unless the designer was mentally
defective !

Then I guess I have diagnosed mental defects in two designers life-to-date...

I expect you were more charitable than to do that.

Yes, one of my rules is not to place blame on a person. I just corrected
the problem and went on with life.

Both systems looked very expensive and "pro", BTW. Needed ferrites and better electrolytics.

Often, I think, appearance is the primary design
goal in expensive audio gear sold to consumers.

Oh yeah. I have seen gear in polished boxes with gold plated connectors
that cost hundreds yet contained less than 50 Cents in parts. The best
was a "compressor". It contained two 1N4148 diodes.

You can also see that effect in some lab gear. I remember the first time
I opened a "power splitter" that retailed in the four-digit range. It
contained a plain old hybrid with three resistors. Ok, they were laser
trimmed. But I have designed a lot of laser trimmed circuitry and it
doesn't cost an arm and a leg.

Regards, Joerg

http://www.analogconsultants.com

 

[Joel Kolstad]

"Joerg" <notthisjoergsch@removethispacbell.net> wrote in message
news:VLz5e.21285$zl.6000@newssvr13.news.prodigy.com...

Oh yeah. I have seen gear in polished boxes with gold plated connectors
that cost hundreds yet contained less than 50 Cents in parts. The best was
a "compressor". It contained two 1N4148 diodes.

This is too easy it's not even fair, but I love Monster Cables and how they
can even manage to take a regular old phone cord for dial-up Internet
connections and turn it into "Ultra-High Speed Internet Phone Cable"
(http://www.monstercable.com/computer/productPageComputer.asp?pin=448&LastPage=Internet%20Cables)...
Oh yeah, 3kHz and 30dB SNR clearly needs $2/ft cables!

You can also see that effect in some lab gear. I remember the first time I
opened a "power splitter" that retailed in the four-digit range. It
contained a plain old hybrid with three resistors.

I've seen designs where cable TV power splitters were actually disassembled
and their parts then removed and soldered down to a different board because
the commercial alternatives (e.g., those from MiniCircuits) were actually
more expensive overall!

---Joel Kolstad

 

[Joerg]

Hello Joel,

I've seen designs where cable TV power splitters were actually disassembled
and their parts then removed and soldered down to a different board because
the commercial alternatives (e.g., those from MiniCircuits) were actually
more expensive overall!

Done that, for lab experiments. It's just so tough to crack these
splitters open if you don't want the dreaded F-connectors.

Regards, Joerg

http://www.analogconsultants.com

 

[rex]

On Fri, 08 Apr 2005 09:23:30 +0100, Pooh Bear
<rabbitsfriendsandrelations@hotmail.com> wrote:

rex wrote:

I recently bought a Marantz PMD670 -- a digital audio recorder. It seems
pretty good except the mic preamps have a pretty high noise level for a
recorder that is good otherwise. I'd like to see if I can make it
better. (It isn't broken, its a design issue.)

I'll guess it's a design issue. ;-)

As a general description, the mic inputs are balanced XLR with
switchable phantom power. It only seems to work decently with pretty
high-level condenser mics. A lower dynamic seems to require cranking the
gain way up and the crappy noise floor becomes apparent.

One guy offers a mod for the preamps that improves things considerably.
link:
http://www.oade.com/digital_recorders/hard_disc_recorders/PMD-670.html
He only does it on stuff he sells and I already have mine. I asked him
if he would share anything about the nature of his change, but he
declined.

Tuh ! Silly wanker.

Well, he won't work on mine since I didn't buy it from him, so I though
he might be willing to give me a general idea.

I don't have any schematic, but I opened the thing up and had a look at
what is in the preamp area. Assuming the signal flows sensibly across
the board, there are (per channel) two largish electrolytics. Then a
couple sot-23 devices (transistors I assume) marked 62Z. Then an HC4052A
analog mux. Then a JRC 2068 op amp.

I was surprized to see that 4052 so close to the input. Not sure what
exactly it is doing.

If I was just to replace the 2068 with something else, do you think it
would help?

NO. The 'front end' is what matters.

If so, what might be good choices for replacements?

Anyone know what the 62Z devices are?

That's a typical SMD code. There isn't enough space to fit the full part
number. Sorry !

Yes, I know. There are some lists online and I have been successful with
matching some in the past, but not this time.

I did some searching but didn't
find anything that matched. Do you think that bears looking at?

NO.

What are the chances that the caps are contributing?

So close as next to nil that you might as well forget it.

I'm not sure what
these are. They have a symbol that is an S in a circle. They are marked
CE105 C above 0402. Is 402 the value? What does it mean?

105 means 105 degrees C rating - that's fine. 0402 is normally an SMD
footprint size.

These are not smd parts. They are radial lead, about 5mm dia x 10 mm
long. It was hard to read them in the circuit but I didn't see anything
other than the 0402 that could be a value, so I assume it might mean
4000 somethings. 4 uF maybe? Didn't see any voltage rating though.

Any thoughts about the HC4052A?

Nope.


I know this isn't much to go on without a schematic. I don't have much
experience with quality audio circuits.

I do !

I thought I would start by
asking opinions here about whether swagging out any of these parts is
worth the effort.

Unlikely. Once the circuit topology is fixed there's usually bugger all that
you can do to improve it.

But the Oade guy I mentioned above seems to have found a way to get S/N
from 65 to 80 dB and distortion from 0.08 to 0.06. Getting the unit
apart is a pain, I did it to look at the board, so I think he can't be
doing too much to the circuit for what he is charging.

The Eastern guys don't have a clue how to make a good mic amp - apart from
Yamaha maybe ?

Is Marantz design done in Asia? I though they were based in Europe.

 

[Pooh Bear]

Joerg wrote:

Hello Graham,

FWIW, the 4052 is a dual 4:1 multiplexer. Sometimes also used to switch
resistors in and out for gain control.

Erk ! Thinks *non linear resistance with applied voltage*.

Yes. The true audio freak would frown and at least use an SD5400 here.

Phantom power sources can be a noise contributor but if so that could be
filtered.

Damn well shouldn't be a noise source unless the designer was mentally
defective !

Then I guess I have diagnosed mental defects in two designers
life-to-date...

LOL !

Both systems looked very expensive and "pro", BTW. Needed ferrites and
better electrolytics.

Noisy 48 volts ? I like to keep the noise on my phantom power supplies down to
a few hundred microvolts ( audio band ).

Graham

 

[rex]

On Fri, 08 Apr 2005 09:27:05 +0100, Pooh Bear
<rabbitsfriendsandrelations@hotmail.com> wrote:

Phantom power sources can be a noise contributor but if so that could be
filtered.

Damn well shouldn't be a noise source unless the designer was mentally
defective !

Phantom power is not the problem, or at least not the main problem. I
connected a condenser mic, turned phantom off to disable it and cranked
up the volume. That's where I realized that there is more hiss from this
thing than I would have liked.

 

[Pooh Bear]

rex wrote:

What are the chances that the caps are contributing? I'm not sure what
these are. They have a symbol that is an S in a circle. They are marked
CE105 C above 0402. Is 402 the value? What does it mean?

Correction. CE is short for capacitor electrolytic.

0402 doubtless means week 2 of 2004.

Chances are that some where there'll be a value like 10/16 meaning 10uF 16V
working.

Graham

 

[Joerg]

Hello Graham,

Noisy 48 volts ? I like to keep the noise on my phantom power supplies down to
a few hundred microvolts ( audio band ).

In these cases it was proximity to switchers and the phantom supply
picked up noise. The electrolytics were, well, bottom-of-the-line garden
varieties. After all, had they used the Philips caps which I put in or
another good brand the profit margin would probably have plummeted from
98.5% to 98.4%.

When I talked to the (stunned) audio people about what else had been
done and I mentioned ferrites all I got were those deer in the
headlights looks.

Regards, Joerg

http://www.analogconsultants.com

 

[John Woodgate]

I read in sci.electronics.design that rex <notat@hotmail.invalid> wrote
(in <i5qd51dstese4da78r80l933v66id016vg@4ax.com&gt about 'mic preamp
modification', on Fri, 8 Apr 2005:

These are not smd parts. They are radial lead, about 5mm dia x 10 mm
long. It was hard to read them in the circuit but I didn't see anything
other than the 0402 that could be a value, so I assume it might mean
4000 somethings.

0402 is almost certainly a date code for April 2002.
--
Regards, John Woodgate, OOO - Own Opinions Only.
There are two sides to every question, except
'What is a Moebius strip?'
http://www.jmwa.demon.co.uk Also see http://www.isce.org.uk

 

[Joerg]

Hello Rex,

Phantom power is not the problem, or at least not the main problem. I
connected a condenser mic, turned phantom off to disable it and cranked
up the volume. That's where I realized that there is more hiss from this
thing than I would have liked.

If you really want to figure out where that hiss comes from there is a
trick: Take a cap and ground one side with a short braid. Hold the cap
at the inputs of the various amps in there. But turn down the speakers
since this can cause a loud pop if there is DC on a node. Discharge the
cap everytime before probing a new spot or you could fry things.

When, for example, you capacitively short the input and nothing changes
but you short the next input and the hiss stops then you know the first
stage is the culprit.

Regards, Joerg

http://www.analogconsultants.com

 

[rex]

On Fri, 8 Apr 2005 22:29:21 +0100, John Woodgate
<jmw@jmwa.demon.contraspam.yuk> wrote:

I read in sci.electronics.design that rex <notat@hotmail.invalid> wrote
(in <i5qd51dstese4da78r80l933v66id016vg@4ax.com&gt about 'mic preamp
modification', on Fri, 8 Apr 2005:
These are not smd parts. They are radial lead, about 5mm dia x 10 mm
long. It was hard to read them in the circuit but I didn't see anything
other than the 0402 that could be a value, so I assume it might mean
4000 somethings.

0402 is almost certainly a date code for April 2002.

Ah, yes, never thought of that. The important stuff must have been
somewhere else on the thing that I couldn't see.

 

[rex]

On Fri, 08 Apr 2005 21:33:29 GMT, Joerg
<notthisjoergsch@removethispacbell.net> wrote:

Hello Rex,

Phantom power is not the problem, or at least not the main problem. I
connected a condenser mic, turned phantom off to disable it and cranked
up the volume. That's where I realized that there is more hiss from this
thing than I would have liked.

If you really want to figure out where that hiss comes from there is a
trick: Take a cap and ground one side with a short braid. Hold the cap
at the inputs of the various amps in there. But turn down the speakers
since this can cause a loud pop if there is DC on a node. Discharge the
cap everytime before probing a new spot or you could fry things.

When, for example, you capacitively short the input and nothing changes
but you short the next input and the hiss stops then you know the first
stage is the culprit.

Thanks, that's a good tip.

 

[Phil Hobbs]

Joerg wrote:

Hello Graham,

Noisy 48 volts ? I like to keep the noise on my phantom power supplies
down to
a few hundred microvolts ( audio band ).


In these cases it was proximity to switchers and the phantom supply
picked up noise. The electrolytics were, well, bottom-of-the-line garden
varieties. After all, had they used the Philips caps which I put in or
another good brand the profit margin would probably have plummeted from
98.5% to 98.4%.

When I talked to the (stunned) audio people about what else had been
done and I mentioned ferrites all I got were those deer in the
headlights looks.

Regards, Joerg

http://www.analogconsultants.com

Filtering phantom power is a great application for a capacitance
multiplier, because you don't care much about losing 0.6 volt. One
resistor, one capacitor, one 2n3904--sayonara to pickup. Much cheaper
than LCs, too.


Cheers,

Phil Hobbs

 

[Joerg]

Hello Phil,

Filtering phantom power is a great application for a capacitance
multiplier, because you don't care much about losing 0.6 volt. One
resistor, one capacitor, one 2n3904--sayonara to pickup. Much cheaper
than LCs, too.

Just be careful when there are transmitters around, especially AM.
Transistors need more than a lone cap in that situation or they might
act as demodulators. 80dB of rejection ain't always easy.

Regards, Joerg

http://www.analogconsultants.com