[jacobmj]

Hi.

I'm experiencing much trouble while trying to connect my Hagerman
Chime directly to a Parasound A23 amp, without a preamp.  In
particular, I experience a clearly audible (at 10-15ft) hiss from the
speakers any time the Chime is powered on.  This hiss doesn't occur if
I connect the Chime to my receiver, or just use the receiver as a
buffer between the Chime and the A23.

This is my first try with this amplifier.  I have previously tested
with a Lexicon 212 (Bryston 3B-ST) amplifier, and had no trouble.  I'd
appreciate any suggestions, and would like to know if anyone else has
had success/failure with the Chime and any amplifier.

I don't believe this is a ground loop, but have gone through the usual
steps to alleviate one.  So far, I've had no luck.  To be clear,
here's the sort of connectivity I have tried:

A: Chime -> A23

B: Denon receiver -> A23  (nothing but power connected to the
receiver.  Don't worry about where I'll get sound.)

C: Chime -> Denon -> A23

D: A23 (no input)

The A23 is connected to a pair of Sound Dynamics 300ti speakers. All
connections (->) are line level RCA unbalanced.  Only power is
connected in addition, and I've tried various cases of ground lifting,
different circuits, and line conditioners with no apparent change in
anything below.

Case A, my planned connectivity suffers from a rather loud hiss, akin
to white noise. It is very obvious from across the room, and entirely
obvious when playing music.  This isn't really the noise floor in the
Chime -- I've been testing with the volume at nil.

Case B doesn't have hiss, but may have a small amount of hum, audible
within a foot of the speaker.  The level would not concern me right
now.

Case C doesn't have hiss either, and any hum is as with B.

Case D has a very slight hiss audible only with an ear to the tweeter.
Shorting the input with a spare length of wire made it quieter, and I
expect correct input or shorting eliminates this entirely.

I tested the Lexicon amp in case A without any trouble.  It sounded
improved over C, and very nice.

I've spoken with the Parasound technical people, but they say (1) that
this isn't an issue of ground loops and (2) they don't really have any
suggestions.  They offered to test the amp, but believe they will find
nothing.  The fellow (Todd) spoke that it may be an incompatibility
between the Chime and the A23, possibly to do with it being directly
coupled or the Chime needing more loading.  One shot in the dark might
be to try using the balanced inputs on the A23 with a balun.  They
appear to be pretty pricey though.

Might anyone have any suggestions?  What can be said about the Chime
"needing more load"?

Thanks much.
~Jacob

[poty]

Tough puzzle... Too many to check.
I'd not reject the testing of the A23. Somehow the situation is closely connected with it (I don't think the hiss in case D should be audible at all with the input short circuited). It may be a good starting point.
Then... You have not said anything about interconnects. I can barely think about what can cause this behaviour, but you shouldn't eliminate this part of your system.
Next thing to try - tubes. Here in the forum are a lot of notes about noisy tubes. Maybe your receiver has less amplification for certain high frequencies which blocks the hissing?
Then you can try the balanced input of the A23 without any balun. Just some simple wire converter between RCA and XLR will do. It may certainly not sound good, but at this point we are more concerned about your problem, not quality of sound.
Then I think of two opposite problems - very high impedance input of the A23 and very low one. It is more speculations than well thought analysis, but:
In case of very high impedance input it should be very sensitive for RFI and other interferences (then - interconnects comes to play too). As soon as Chime has digital circuits with very wideband signals it may be the matter. You can try to temporaly disable power for any digital circuits in Chime and test.
In case of very low impedance input - the output stage of the Chime may not be capable of driving it and the C409 may work as Hi-Pass filter, which singles out the high frequencies and makes them nore notable.
Sorry if I confuse things more.

[jacobmj]

poty,

Thanks for your reply. I have done a bit more testing, and thought I'd
report back.

To clarify, case D with the input shorted is very, very quiet.  It
takes an ear <3" from the tweeter, and the other ear plugged to hear.
I assumed any residual noise was my crude shorting with an aligator
clip on male end of a short interconnect.  About interconnects, I've
tried a few.  My "best" is a ~5' length of Belden 1505F with Canare
RCA connectors, from blue jeans cable.  I've also tested with a
variety of generic cables--I can't notice any difference.  I happen to
have other cables arriving today, and could test more, but I can't
imagine any change.

My tubes are a Telefunken 12au7 and a Brimar 12ax7.  I did try a set
of Sovteks in their place, with no difference in measurements (below)
or sound.  The rectifier is a Sovtek, and I don't have another to try.

After this session of testing, I think I should add that Case A does
have hum and some buzz in the larger drivers.  At any distance, it's
overwhelmed by the hiss though.

I haven't yet tried anything with the balanced connection.  I did take
a closter look at the Chime, and took a few oscilloscope measurements
to try and get a handle on things.  Not surprisingly, I'm left with
more questions than answers.  Note that I didn't build this Chime.  I
bought it used.  I don't know whether it was bought pre-built from
Hagtech or not.

To be sure, ought the chassis of the Chime have continuity with the
RCA shields?  This persists when I remove the "earth" lead from the
chime board.  The RCAs are clearly insulated away, but the USB
connector likely is not.

I spent a long time scoping outputs of all devices.  With nothing
connected to the A23 (Case D), the voltage difference across +/-
speaker terminals is *extremely* quiet, and zero. Any noise I did see
was in excess of 20Mhz. A speaker was connected.

The line-level output of the Chime was a bit more noisy (<1-3mV peak
to peak--I can't accurately measure very low values), and exhibited a
strange DC drift of +/- 60mV (120mV p/p) at slow (1-2 second)
timescales, with intermittent peaks off the 200mV scale I happened to
use for this.  The drift is such that it will move to a voltage, hold
close to it for a second or two, and then slew on to another.  By my
estimation, positive and negative values were equally likely, and more
extreme voltages were more rarely attained.  Both channels of the
Chime were absolutely identical.

I don't have a solid enough understanding of the Chime to reason much
about this.  It does surprise me that C409 could pass so much DC.
There seemed to be some correlation of slew with other equipment in
the house (fridge, furnace, etc.), but I tried power from an APC line
conditioner, an APC Smart UPS, the wall, and from another outlet from
my neighbors, who are on a different mains service transformer.
Finally, I ran on UPS power (it is a sine-wave UPS, and produces a
sine wave better than my sharp-peaked utility power wave), with the
mains entirely shut off to my house.  None of this changed the
behavior, or the hiss/hum when connected through to the speakers.

Most of my measurements were made using two probes in a differential
setup, to avoid inducing a ground with the oscilloscope.  I tested
line-level signals with a single probe and found no difference.  I was
afraid of grounding a speaker terminal.

The next thing I considered is case A connectivity, with a probe on
one speaker terminal, and another on the other channel of the Chime.
The DC offset is passed through to the speaker, at approximately 1.5x
the line-level voltage.  Curiously, the speaker offset seems to
*preceed* the line level change.  I think what is really going on is
that the A23 is amplifying the offset, making it move very quickly for
small incident offsets, and then slowing down up to some limit.

I know DC is bad for speakers, but how much is bad?

One thing that did occur during my initial testing, is that I
apparently destroyed the tweeter on the speaker I was testing with.
The tweeter no longer conducts electricity, and clearly has a burned,
disconnected coil upon dissassembly.  I did accidentally try to connect
a cable while the amp was on before noticing the tweeter was dead,
producing a momentary static.  Sadly, I thought I'd fixed the hiss,
only to find the speaker broken.  I can't say for sure that I was the
cause, but will take the blame absent some other explanation.  Your
note about high frequency did get me thinking.

(Incidentally, given that this tweeter unbolts, exposing the voice
coil which is wrapped around the diaphragm, are there ways to patch up
the split wire so I might limp along until new speakers or luck in
finding a new tweeter?)

Thanks again.
~Jacob

[poty]

case D with the input shorted is very, very quiet.... I've
tried a few.

Then at this time we eliminate obvious problems with the amplifier and interconnects.

My tubes are a Telefunken 12au7 and a Brimar 12ax7.  I did try a set
of Sovteks in their place, with no difference in measurements (below)
or sound.  The rectifier is a Sovtek, and I don't have another to try.

If it is possible - try to check with another set of tubes. Just in case...

... ought the chassis of the Chime have continuity with the
RCA shields?  This persists when I remove the "earth" lead from the
chime board.  The RCAs are clearly insulated away, but the USB
connector likely is not.

Interesting observation. Normally the RCAs should be connected to chassis only through one way - the ground wire. But it is possible it have some way of grounding through the digital ground plane too (with some resistance though).

The line-level output of the Chime was a bit more noisy (<1-3mV peak
to peak--I can't accurately measure very low values), and exhibited a
strange DC drift of +/- 60mV (120mV p/p) at slow (1-2 second)
timescales, with intermittent peaks off the 200mV scale I happened to
use for this.

I think, you should add some load to the line output of the Chime (say - 100k) while do measurements. The 1-3mV is rather low voltage to be shure it is no oscilloscope's wires induced current.
The DC drift is actually (IMHO) very low frequency oscillations (0.5-1Hz). I'd try to measure the test points in the Chime itself (caution - High voltage). It may be just PS filter oscillations, which are reflected by CCS of the last stage. As soon as you add load, the low frequencies should be supressed by the C409 - it is what I pointed in the previous message as high impedance load.
What we have left? Digital circuits, another tube rolling trial and balanced input I think.
If you don't know who built the device it may be some inappropriate chosen parts inside too.

(Incidentally, given that this tweeter unbolts, exposing the voice
coil which is wrapped around the diaphragm, are there ways to patch up
the split wire so I might limp along until new speakers or luck in
finding a new tweeter?)

If you can get to the single broken wire - it is possible to solder the ends together and restore the tweeter. But usually the wires are so thin that it is almost mission impossible.

[jacobmj]

poty,

Thanks again.

I forgot to mention that had tried termination of the RCA into a 100k
resistor.  There was no apparent difference in measurements between
terminated or not.  I mean to go back and remove components ahead
of the capacitor to start to get a closer idea of what is going on.  I'm also
close to giving up, and believing parasound that it is "incompatible."

I've also since given balanced input a try, connecting the RCA to pins
2 and 3, and leaving 1 disconnected.  I tried with the amp ground
lifted with a cheater plug, and not, as well as combinations of the
ground lift switch on the amp (which shouldn't matter here).  In all
cases, there's a rather substantial hum (some cases more than others),
and the white noise remains clearly audible.

I'll write back once I get a chance to test more points in the Chime.

Thanks.
~Jacob

[dnewman]

So what is the input to the Chime?  S/PDIF over USB or RCA?  I ask as I once had a similar issue where I could hear hiss when going from my Chime to a Bottlehead Paramount amp (2A3) but not when going to an Transcendant SE OTL amp.  The problem actually turned out to be my source into the Chime.  It had a faulty transformer which was causing the source's ground to be fluctuating considerably.  This got carried over the RCA cable's ground into the Chime.  For whatever reason, it was audible & very annoying with the Paramounts but not with the (inherently quieter in and of itself) SE OTL amp.  I eliminated it by introducing an isolation transformer into the path between the source and the Chime.  Not much different than the isolation transformer used in the USB circuit on the Chime.

If, OTOH, you are using USB to feed your Chime, I wouldn't expect this to be the problem.  Jim's design is quite careful in isolating the Chime from the USB source.  No worries about fried computers there.

Dan

P.S. And, of course, I was using very sensitive speakers.  A must with a 3W/channel 2A3 or 1.5W/channel SE OTL.

[jacobmj]

Dan,

For all of the cases I noted, the Chime had no input at all--just power. 

Out of curiosity, I did at one point try  connecting *only* the center pin and not the RCA ground to the amplifier.  The hiss persisted with no apparent change.  Of course, the ground in the Chime could still be noisy, perhaps if it were not tied well to the signal ground.  Any noise relative to the powerline ground would probably also occur on the signal line.

Thanks. ~Jacob

[poty]

I ... had tried termination of the RCA into a 100k
resistor.  There was no apparent difference in measurements between
terminated or not.

IMHO it means that the source of the noise is outside the Chime. Or there is something wrong with measurements.

I mean to go back and remove components ahead of the capacitor to start to get a closer idea of what is going on.

I'd started with disconnecting the digital part of the Chime from power. If such big hiss persists - it means that something wrong with your amplifier input. By the way - can the hiss be heard with normal listening level or you try to eliminate the hiss at full volume?
And you haven't done the internal Chime measurements.

I've also since given balanced input a try, connecting the RCA to pins 2 and 3, and leaving 1 disconnected.  I tried with the amp ground lifted with a cheater plug, and not, as well as combinations of the ground lift switch on the amp (which shouldn't matter here).  In all cases, there's a rather substantial hum (some cases more than others), and the white noise remains clearly audible.

Hum is a different problem in this area. I remain sure that something wrong with the A23. Lifting the ground in the amp should be clearly audible.

I did at one point try  connecting *only* the center pin and not the RCA ground to the amplifier.  The hiss persisted with no apparent change.  Of course, the ground in the Chime could still be noisy, perhaps if it were not tied well to the signal ground.  Any noise relative to the powerline ground would probably also occur on the signal line.

The ground in that case was achieved externally, through the mains common ground. How the hiss in that case could remain unchanged? It's unbelievable!