My Mag 3.6Rs are about 3 years old and one of the tweeters has
developed slight audible distortion which is noticeable on female
vocals. I've swapped the amp outputs and the crossover boxes but the
distortion remains in the right speaker. Looks like I'm going to have
to replace the tweeter. How long do these tweeters typically last? Has
anyone else been through this? Did you ship the entire speaker back to
the factory to be checked out, or did they send you a new tweeter for
you to install yourself? I'd greatly appreciate advice from anyone who
has been through this before. Thanks,

wrote in message ...
My Mag 3.6Rs are about 3 years old and one of the tweeters has
developed slight audible distortion which is noticeable on female
vocals.

After further testing it appears to be an intermittent problem with
the amplifier. I hooked up different speakers and they had the exact
same audible distortion as my Maggies. Yet through my headphones there
is not a trace of distortion, so the source components and "preamp"
switchbox are clean.

I wired up a cable to run the amp output into my headphones to listen
to the amp (greatly attenuated) on the headphones. I could hear an
easily audible 60 Hz buzz but no other distortion. But my headphones
are 300 ohms and take only milliwatts so they're a much easier load to
drive than my Maggies.

This question is for the guys who have more EE experience than I do:
Is it possible for a bad transistor to be clean driving high impedance
loads at very low power levels, yet cause distortion only when driving
low impedance loads at medium and higher levels?

Thanks,

I have had the same experience with my Tympani iva's. That is, the ribbon is
so revealing of everything in the signal path, you can hear any distortion
clearly (just as you can hear all that detail to the music that makes them
the best) and I had a preamp that once in a while created a distortion that
wasn't noticed right away with a set of ordinary, Rega, 2 way boxes (my
kids said when are you going to put the REAL speakers back).

I jumped the gun and rebuilt the tweeter using the tweeter kit from the
factory (what a great company and product) to find out it wasn't the tweeter
at all. I changed to a Classe preamp and what an improvement. If I had
waited longer after swapping the channels I would have heard it using the
other speaker/side.

Borrow someone elses preamp or power amp to isolate the problems source. I'm
assuming it isn't a source problem ie, CD player etc??

Congratulations the 3.6's are super speakers and your hearing things your
friends don't even know exist. Just isolate the problem and get rid of it,
then set back and enjoy the music.

Tom eh

wrote in message ...
wrote in message ...
My Mag 3.6Rs are about 3 years old and one of the tweeters has
developed slight audible distortion which is noticeable on female
vocals.

After further testing it appears to be an intermittent problem with
the amplifier. I hooked up different speakers and they had the exact
same audible distortion as my Maggies. Yet through my headphones there
is not a trace of distortion, so the source components and "preamp"
switchbox are clean.

I wired up a cable to run the amp output into my headphones to listen
to the amp (greatly attenuated) on the headphones. I could hear an
easily audible 60 Hz buzz but no other distortion. But my headphones
are 300 ohms and take only milliwatts so they're a much easier load to
drive than my Maggies.

This question is for the guys who have more EE experience than I do:
Is it possible for a bad transistor to be clean driving high impedance
loads at very low power levels, yet cause distortion only when driving
low impedance loads at medium and higher levels?

Thanks,

I used to own Maggies with ribbon tweeters, and while I do nelieve the ribbons
do wear out over time, the biggest factor in a ribbon's lifespan is how hard
you drive the speakers. The fuse isn't fast enough. I would up needing a new
ribbon when my speakers were about 6 or 7 years old.

Swapping the ribbon was easy. There are 2 options. You can buy a new ribbon,
and glue it in place. They send you an extra ribbon in case your first attempt
fouls up.

Not wanting to go that route, I bought an entire tweeter assembly. Took about
10 minutes to remove the old ribbon tweeter assembly and put the new one in.
You send back the old tweeter, and they rebuild it for the next customer. I
seem recall there's a core charge for the old tweeter.

It's an easy job, and a ribbon's fragile nature, the fact that they do need
periodic replacement, etc., is the price you pay for a great tweeter.

Oh, and you might want to replace the ribbons as a pair- you know, balanced
sound from both sides-

Your later post indicates that it wasn't the tweeters at all, but let me
answer it anyway. I replaced the tweeters in my MG IIIA's, blown by my
own fault. Magnepan ships you new ones, and you install them yourself.
It's quite simple to do. Follow directions exactly, never just drop the
panels on the floor (the air pressure will destroy the ribbon) and
you're good to go.

-- Bob T.

wrote:

My Mag 3.6Rs are about 3 years old and one of the tweeters has
developed slight audible distortion which is noticeable on female
vocals. I've swapped the amp outputs and the crossover boxes but the
distortion remains in the right speaker. Looks like I'm going to have
to replace the tweeter. How long do these tweeters typically last? Has
anyone else been through this? Did you ship the entire speaker back to
the factory to be checked out, or did they send you a new tweeter for
you to install yourself? I'd greatly appreciate advice from anyone who
has been through this before. Thanks,

I finally found it -- it's not the amp either!

It's about the very strangest thing I've ever seen.

It happens only when my CD Burner is plugged in to my passive pre's
tape loop.

Each source device driving the passive pre drives the CD Burner in
parallel with the amplifier. Total impedance seen by the source is
about 9 kOhm for both, or 10 kOhm if it's the amp only (CD Burner RCA
jacks unplugged).

Whenever the CD Burner is in the loop, there is distortion on each of
the 3 inputs. Whenever I take the CD Burner out of the loop,
distortion goes away. So something about the CD Burner being in the
loop, causes the distortion.

My first hypothesis was that the CD Burner had a short and the input
impedance was too low, overdriving the source and causing distortion.
But I measure between 50 kOhm and 100 kOhm for its inputs, so that's
not the issue.

So how is this possible? Any ideas? I can think of only 3, all equally
unlikely:

1. The output stage of my CD Player (Rotel RCD 1070) is fried and it
can't drive a 9 kOhm load? But it can drive a 10 kOhm load. Yet its
output impedance is 100 ohms, so it should have no problem.

I had a tape deck with 50 kOhm input impedance for over a year before
the CD Burner, and the problem started about when I added the CD
Burner (a few weeks ago). So (1) is highly unlikely. This leaves (2)
and (3).

2. The CD Burner's analog input is fried somehow and its input
impedance at musical frequencies is very low, overdriving the sources.
Yet it's 90 kOhm for DC (e.g. my multimeter).

3. Some kind of weird grounding issue. All of the RCA inputs & outputs
are grounded together at the passive pre (left & right channel
grounded separately). And the left channel is grounded to the frame.

Thanks in advance for any help with this very strange problem.

I finally figured it out and it's even stranger than I thought.

The CDR-630 causes this distortion whenever it's in the tape loop,
turned off. The moment I turn it on, the distortion goes away.

Turning it on must power up the input stage changing the impedance or
grounding. I've had a lot of different units in my tape loop and this
one is the ONLY one that's ever done this. VERY VERY strange. But the
behavior is consistent and repeatable.

Has anyone encountered this? Can anyone explain this?

wrote:
I finally found it -- it's not the amp either!

It's about the very strangest thing I've ever seen.

It happens only when my CD Burner is plugged in to my passive pre's
tape loop.

Each source device driving the passive pre drives the CD Burner in
parallel with the amplifier. Total impedance seen by the source is
about 9 kOhm for both, or 10 kOhm if it's the amp only (CD Burner RCA
jacks unplugged).

Whenever the CD Burner is in the loop, there is distortion on each of
the 3 inputs. Whenever I take the CD Burner out of the loop,
distortion goes away. So something about the CD Burner being in the
loop, causes the distortion.

My first hypothesis was that the CD Burner had a short and the input
impedance was too low, overdriving the source and causing distortion.
But I measure between 50 kOhm and 100 kOhm for its inputs, so that's
not the issue.

So how is this possible? Any ideas? I can think of only 3, all equally
unlikely:

1. The output stage of my CD Player (Rotel RCD 1070) is fried and it
can't drive a 9 kOhm load? But it can drive a 10 kOhm load. Yet its
output impedance is 100 ohms, so it should have no problem.

I had a tape deck with 50 kOhm input impedance for over a year before
the CD Burner, and the problem started about when I added the CD
Burner (a few weeks ago). So (1) is highly unlikely. This leaves (2)
and (3).

2. The CD Burner's analog input is fried somehow and its input
impedance at musical frequencies is very low, overdriving the sources.
Yet it's 90 kOhm for DC (e.g. my multimeter).

3. Some kind of weird grounding issue. All of the RCA inputs & outputs
are grounded together at the passive pre (left & right channel
grounded separately). And the left channel is grounded to the frame.

Thanks in advance for any help with this very strange problem.

Does the problem go away when the CD burner is powered up?

wrote in :

I finally found it -- it's not the amp either!

It's about the very strangest thing I've ever seen.

It happens only when my CD Burner is plugged in to my passive pre's
tape loop.

Each source device driving the passive pre drives the CD Burner in
parallel with the amplifier. Total impedance seen by the source is
about 9 kOhm for both, or 10 kOhm if it's the amp only (CD Burner RCA
jacks unplugged).

Whenever the CD Burner is in the loop, there is distortion on each of
the 3 inputs. Whenever I take the CD Burner out of the loop,
distortion goes away. So something about the CD Burner being in the
loop, causes the distortion.

My first hypothesis was that the CD Burner had a short and the input
impedance was too low, overdriving the source and causing distortion.
But I measure between 50 kOhm and 100 kOhm for its inputs, so that's
not the issue.

So how is this possible? Any ideas? I can think of only 3, all equally
unlikely:

1. The output stage of my CD Player (Rotel RCD 1070) is fried and it
can't drive a 9 kOhm load? But it can drive a 10 kOhm load. Yet its
output impedance is 100 ohms, so it should have no problem.

I had a tape deck with 50 kOhm input impedance for over a year before
the CD Burner, and the problem started about when I added the CD
Burner (a few weeks ago). So (1) is highly unlikely. This leaves (2)
and (3).

2. The CD Burner's analog input is fried somehow and its input
impedance at musical frequencies is very low, overdriving the sources.
Yet it's 90 kOhm for DC (e.g. my multimeter).

3. Some kind of weird grounding issue. All of the RCA inputs & outputs
are grounded together at the passive pre (left & right channel
grounded separately). And the left channel is grounded to the frame.

Thanks in advance for any help with this very strange problem.

Does the problem occur when (a) the CD burner is powered off or (b) when
it's powered on or (c) in both cases? If it only occurs when it's off,
then it's likely that there are some protection diodes or other active
devices at the input to an op amp that are getting enough voltage from
the signal for them to conduct. If (b) or (c), then the CD burner may have
a problem. Since you haven't said otherwise, I assume that CDs you've
burned sound OK so I'm guessing it's case (a).

In any case, it seems like an isolation amplifier or a pre-amp with
multiple, separately driven outputs would be better than paralleling the
loads.

-- JS

wrote:
I finally figured it out and it's even stranger than I thought.

The CDR-630 causes this distortion whenever it's in the tape loop,
turned off. The moment I turn it on, the distortion goes away.

Turning it on must power up the input stage changing the impedance or
grounding. I've had a lot of different units in my tape loop and this
one is the ONLY one that's ever done this. VERY VERY strange. But the
behavior is consistent and repeatable.

Has anyone encountered this? Can anyone explain this?


I posted a reply to your earlier message, and it had not shown up yet. I
was asking whether the problem went away when the CD recorder is turned on.

The reason why you are seeing this "problem" is because of clamping
diodes from the input terminals to the supply voltages. These may be
discrete diodes on the PC board, or diodes inside integrated ciruits
(e.g. analog switches or op-amps). When the recorder is turned on, the
supply voltages are at the nominal levels, and the diodes will not
conduct unless the input voltage exceeds the supply voltages in
magnitude. However, when the CD recorder is turned off, the supply
voltages drop to ground level, and effectively the inputs are clamped to
ground by diodes. Whenever the input voltage gets higher than a few
tenths of a volt, you will see appreciable clipping due to the diodes
starting to conduct, since the source impedance is not zero (in other
words, distortion current flowing throught the source resistance and
appearing as a distortion voltage).

The solution is either to always turn on the CD recorder, or disconnect
the CD recorder when it is not used. You can also isolate the CD
recorder by putting a switch in series with its inputs.

wrote:
I finally figured it out and it's even stranger than I thought.

The CDR-630 causes this distortion whenever it's in the tape loop,
turned off. The moment I turn it on, the distortion goes away.

Turning it on must power up the input stage changing the impedance or
grounding. I've had a lot of different units in my tape loop and this
one is the ONLY one that's ever done this. VERY VERY strange. But the
behavior is consistent and repeatable.

Has anyone encountered this? Can anyone explain this?

Almost any opamp has its input protection consisting of 2 diodes against the
supply voltages. if the device is powered on the diodes are reversed biased
and have just a small capacitance, but when the supply is zero, any input
voltage above 600mV or 300mV for a Schottky-diode will be shortened to gnd.
Usually in a good design there is a 1k resistor to the input and this
isolates the diode clippers, but when your input impedance is high, as it is
in a passive preamp with a volume control, a distortion is there when the
level is higher than the threshold voltage of those diodes. Get an active
preamp with buffered outputs.
--
ciao Ban
Bordighera, Italy

wrote in news:_d6Hc.43875$IQ4.19071@attbi_s02:

I finally figured it out and it's even stranger than I thought.

The CDR-630 causes this distortion whenever it's in the tape loop,
turned off. The moment I turn it on, the distortion goes away.

Turning it on must power up the input stage changing the impedance or
grounding. I've had a lot of different units in my tape loop and this
one is the ONLY one that's ever done this. VERY VERY strange. But the
behavior is consistent and repeatable.

Has anyone encountered this? Can anyone explain this?



Not strange at all. The CDR-360 probably uses op-amps for its input stage.
In the absence of bias provided by the power supply (i.e., with power
turned off), protection diodes (or other active devices) at the inputs to
the op-amps will conduct when an input signal that exceeds their threshold
is applied. When they conduct, the apparent input impedence changes
abruptly, causing the distortion that you hear. Even if the signal went
directly to the A/D converter or a CMOS switch (i.e., no op-amps), an
effect like this is possible.

I guess this problem qualifies as a design flaw on the part of the CDR-360,
but it's probably something that the designers didn't even consider. I've
been bit by it more than once, most notably in a power-supply that was
designed to operate in parallel with other supplies. The internal
monitoring circuits were being kept alive by leakage through an op-amp's
input, resulting in a supply that insisted it was on and working OK even
when its plug had been pulled.

You have several choices. Keeping the recorder powered on when you're
listening is probably the simplest. If easily accessible, you could also
disconnect it when not in use, but that will cause wear and tear on the
interconects. Better choices would be to use an active preamp with
multiple separately driven outputs or an isolation amplifier.

-- JS

wrote in message
news:_d6Hc.43875$IQ4.19071@attbi_s02...
I finally figured it out and it's even stranger than I thought.

The CDR-630 causes this distortion whenever it's in the tape loop,
turned off. The moment I turn it on, the distortion goes away.

Turning it on must power up the input stage changing the impedance or
grounding. I've had a lot of different units in my tape loop and this
one is the ONLY one that's ever done this. VERY VERY strange. But the
behavior is consistent and repeatable.

Has anyone encountered this? Can anyone explain this?


I'm somewhat puzzled as to what you mean. In the tape loop, but turned off?
Are you saying that whenever the 630's analog outputs go into one of your hi
level preamp's inputs having in/out monitoring capability? If so, I'm either
happy or sorry (depending upon how you look at it) to say that I cannot
confirm any such distortion. I listened to Olivia Newton-John's "Jolene" and
"Come on Over" tracks for my tests. I use 2 different "vintage" pre-amps;
Audio Research SP3-A-1 and Adcom GFP-565. The pre-amps drive a pair of
Bryston 7B-STs which feed a Tympani IVa. I am so satisfied (actually
overwhelmed) by the results I obtain, that I will never change anything.
Some of my components have been with me over 30 years and if I could never
hear any sound improvements way back then, my Medicare ears certainly can
hear any today. I had my Tympani's ribbons replaced only once, about ten
years ago and it probably was unnecessary. My dealer suggested that I look
at the ribbons with a flashlight to detect disruptions, bending or physical
distortion. I saw none then and see none today. Although I have soldering
equipment, I'm not a practiced veteran in soldering and don't excel at
placing those tweeter panels in a horizontal position, and therefore let the
dealer order the ribbons and do the replacement for me, total cost being
$150. To add to the story my dealer's soldering *failed* after several years
losing contact in one ribbon, and I got them back to re-do the job. The
dealer is a "hot-shot" Maggie merchant, or so he considers himself as being.

chung wrote in message ...
Does the problem go away when the CD burner is powered up?

Yes it does. No problem any more.

I knew I needed a device with a high input impedance for my passive
setup, which is part of the reason I picked this burner with its 50
kOhm input impedance. I just didn't know I'd have to have it turned on
all the time. No other device I've had in the tape loop has required
that.

Thanks everyone for all the info.

chung wrote:

(CLIPPED: description of how the behavior of clamping diodes on the
input of the CD recorder causes a nonlinear and low-impedance load when
the machine is not powered on)

The solution is either to always turn on the CD recorder, or disconnect
the CD recorder when it is not used. You can also isolate the CD
recorder by putting a switch in series with its inputs.

I ran across this on a friend's cassette recorder, some years ago. The
preamp did not have a separate Record Output switch (which would solve
the problem handily) and the user did not want to reach behind to
disconnect cables. The recorder had a motor which ran all the time the
unit was powered up, so leaving it on all the time was a bad choice,
too. The solution we came up with was to plug a pair of 1/4 inch plugs
into the microphone jacks, which had built into them the switching
needed to disconnect the line inputs when microphones were used.

"Norman Schwartz" wrote in message news:yboHc.50341
I'm somewhat puzzled as to what you mean. In the tape loop, but turned off?

YES

Are you saying that whenever the 630's analog outputs go into one of your hi
level preamp's inputs having in/out monitoring capability?

NO, the opposite. Whenever the switchbox (preamp) tape out goes into
the CDR-630's analog input, it causes distortion in the selected line
level source device (if the CDR-630 is turned off). Turn on the
CDR-630 and the distortion goes away.

Having a passive attenuator makes all the difference. It's wired so
source devices drive the power amp & tape loop directly in parallel
with no intervening circuitry. There is no active preamp to isolate
the devices. The signal path is about as pure as it can get, but it
does make the source devices sensitive to the input impedances of the
power amp & the CD burner. They have input impedances 30 kOhm so
it's all matched up well. Given the ultimate clarity and resolution my
passive setup provides, I don't intend to change anything. I'll just
leave the burner on all the time now.

wrote:
"Norman Schwartz" wrote in message news:yboHc.50341
I'm somewhat puzzled as to what you mean. In the tape loop, but turned off?

YES

Are you saying that whenever the 630's analog outputs go into one of your hi
level preamp's inputs having in/out monitoring capability?

NO, the opposite. Whenever the switchbox (preamp) tape out goes into
the CDR-630's analog input, it causes distortion in the selected line
level source device (if the CDR-630 is turned off). Turn on the
CDR-630 and the distortion goes away.

Having a passive attenuator makes all the difference. It's wired so
source devices drive the power amp & tape loop directly in parallel
with no intervening circuitry. There is no active preamp to isolate
the devices. The signal path is about as pure as it can get, but it
does make the source devices sensitive to the input impedances of the
power amp & the CD burner. They have input impedances 30 kOhm so
it's all matched up well. Given the ultimate clarity and resolution my
passive setup provides, I don't intend to change anything. I'll just
leave the burner on all the time now.

You only need to turn it on when you are listening to your system .

Seriously, even if you do not have a passive attenuator, there could
still be a problem with your CD recorder. Most active preamps have a
source resistance somewhere in the range 50 ohms to several hundred
ohms. That will cause noticeable distortion driving a load that clamps
to ground, even through a K or so of series resistance inside the load.
You need a separate buffer just for that tape recording loop, or a
series switch that opens when you are not recording, to eliminate the
problem. Many preamps and receivers do not have separate buffers for the
tape loop, and the loads on that loop are in parallel with the main load
which is the power amp input.

chung wrote in message news:Ob4Ic.31985$WX.5763@attbi_s51...
Many preamps and receivers do not have separate buffers for the
tape loop, and the loads on that loop are in parallel with the main load
which is the power amp input.

Yes you are right. I've seen at least a couple different preamps that
could allow a device in the tape loop to record a source, without the
preamp being turned on. This would mean the source is directly exposed
to the recording device's inputs just like they are in my passive
attenuator, right?

I called Marantz tech and they said they never heard of anyone using a
unit in a passive setup. He said something about CMOS switches being
in the analog input stage and hypothesized that a big input voltage
might trigger them when there is no power to the input stage. I didn't
understand what he was saying. But overall he said as long as it works
fine when turned on why worry about it? I agree.

wrote:

chung wrote in message news:Ob4Ic.31985$WX.5763@attbi_s51...
Many preamps and receivers do not have separate buffers for the
tape loop, and the loads on that loop are in parallel with the main load
which is the power amp input.

Actually that's not true exactly, since the volume control is usually
not in the tape-recorder output path. But the same buffer (if there is
one) is driving that path and the volume control pot most likely. That
buffer (or the output of the upstream device) will still clip because of
the diodes to ground at the input of your recorder when it's off.


Yes you are right. I've seen at least a couple different preamps that
could allow a device in the tape loop to record a source, without the
preamp being turned on. This would mean the source is directly exposed
to the recording device's inputs just like they are in my passive
attenuator, right?

That probably means that there are mechanical switches or relays in the
signal path that allows a connection from source to record-out when the
preamp is turned off.

The difference is that there may be little or no resistance in that
path, whereas in your case there is the attenuator in the path.


I called Marantz tech and they said they never heard of anyone using a
unit in a passive setup. He said something about CMOS switches being
in the analog input stage and hypothesized that a big input voltage
might trigger them when there is no power to the input stage. I didn't
understand what he was saying. But overall he said as long as it works
fine when turned on why worry about it? I agree.

As I said earlier, the likely culprit is the analog switch or op-amps,
or protection diodes on the PCB, in your recorder. The analog switch is
typically a pair of CMOS transistors. When no power is applied to the
unit, these switches act like diodes to the supply voltages, which are
at ground level. Therefore you get the clamping action, meaning gross
distortion starting at a few tenths of a volt. That was why when you
were listening via headphones, you did not hear the distortion, since
the level is much lower.

The fact that you have a passive attenuator makes the distortion fairly
obvious, but even if you have an active preamp, you will still have the
problem, since those diodes clamp the voltage going to the power amp.
You need to leave the recorder powered up, or switch it out so that the
preamp does not see it, using a mechanical switch when the recorder is
powered off. RS and Sony both sell simple mechanical switches for very
little money. Or you can build your own, using a pair of mechanical
switches from RS.

chung wrote in message ...
wrote:
The difference is that there may be little or no resistance in that
path, whereas in your case there is the attenuator in the path.

In my box, the attenuators are not in the tape loop. The input signal
from each source splits. One branch goes to the tape loop, the other
goes through the attenuators to the power amp. The tape loop device is
in parallel with the attenuators and power amp. But even though
they're separate branches of the signal, they're not isolated from
each other so clamping in one branch could affect the signal in the
other branch.

As I said earlier, the likely culprit is the analog switch or op-amps,
or protection diodes on the PCB, in your recorder. The analog switch is
typically a pair of CMOS transistors. When no power is applied to the
unit, these switches act like diodes to the supply voltages, which are
at ground level. Therefore you get the clamping action, meaning gross
distortion starting at a few tenths of a volt. That was why when you
were listening via headphones, you did not hear the distortion, since
the level is much lower.

Actually I did hear the distortion on headphones. I was using HD-580s
which are relative inefficient 300 ohm phones so the voltage needed to
drive them is a little higher than most other headphones. On the
speakers, the distortion wasn't "gross" or obvious. Several different
people listening to the system never heard it. It was audible only on
certain well recorded acoustic music passages, during big dynamics.

But it's true that low level signals did not produce the distortion.
When I used my portable CD player as a source device, the distortion
wasn't there. The line out of the CD player is only 0.7 volts instead
of the 2.05 volts of my Rotel.

The fact that you have a passive attenuator makes the distortion fairly
obvious, but even if you have an active preamp, you will still have the
problem, since those diodes clamp the voltage going to the power amp.
You need to leave the recorder powered up, or switch it out so that the
preamp does not see it, using a mechanical switch when the recorder is
powered off. RS and Sony both sell simple mechanical switches for very
little money. Or you can build your own, using a pair of mechanical
switches from RS.

Yes and active circuits can veil subtle distortions. When I ran the
signal through a high quality headphone amp the distortion was still
audible, but not quite as easy to hear as it was through the passive
preamp. Even a preamp that measures well and sounds clean and neutral
still veils the sound slightly compared to a straight wire.