Help With 60Hz Hum PLEASE

What happens to the hum if the Marantz is muted, with the mute function set to "full" rather than the other attenuation settings that are provided for that function?

Regards,
-- Al
 

Bgoeller, note the most recent post above by Jea48. I agree with him that the findings stated so far are not consistent with the problem being caused by a ground loop.

Good suggestion by Imhififan to see if the 8801 produces a hum when connected to the sub.

One note about muting. When I do use mute there is a "pop" whenever I mute or un-mute the system. Is something I should be concerned about?

Not sure that any of us can answer that, and of course it would depend on how severe it is. I wouldn’t be totally surprised, though, if the underlying cause turns out to be the same as whatever is causing the hum problem.

But on the other hand take a look at this thread, starting with post no. 245. That poster also had a popping problem with an 8801, which occurred at power on/off, mute on/off, and Audyssey on/off. Some posts later he incorrectly concluded that it was due to an interchange in the 8801’s design of the signals on XLR pins 2 and 3 relative to the USA convention of pin 2 hot and pin 3 cold, which of course would not be responsible for that kind of problem. (At most it would result in a polarity inversion, which on most recordings would be audibly insignificant, and which in any event can be compensated for by interchanging + and - at the speakers or the amp outputs. And if an XLR input and an XLR output are being used it wouldn’t result in a polarity inversion at all, since the two inversions would cancel out). Later in the thread, though, it appears that the popping problem remained, and no resolution was indicated.

Finally, FWIW, in post no. 245 that poster linked to a post by someone else who had a popping problem with an 8801 at power on/off and mute on/off, which arose only after a firmware update had been incorporated in the unit. Apparently no resolution was indicated there either. But that would seem to suggest that if perchance any firmware updates were released subsequent to whatever version is in your unit, updating to the latest firmware MIGHT cure the popping.

Regards,
-- Al

Al can tell you the whys you will need balanced cables over single ended for the lowest noise floor possible.

A properly designed balanced interface will reject nearly all noise that is present equally on its two signal lines, since a balanced receiver circuit responds essentially just to the difference between the voltages on those two lines. An unbalanced interface has only one signal line, of course, so it will not do that. Also, a **properly designed** balanced interface will be less susceptible to ground loop-related hum and noise than an unbalanced interface, since a properly designed balanced interface will not put signal current through the ground connection in the cable. Although as was said earlier it appears in this case that a ground loop is not responsible for the problem.

A suggestion that occurs to me at this point, although it is just a hunch: In addition to obtaining longer XLR cables, purchase some RCA shorting plugs and insert them into the amp’s RCA input connectors, while connecting the amp to the preamp with the XLRs. Just a hunch, as I say, taking into account that both the specs for the amp and the Stereophile measurements at the link Jim provided make it clear that the amp’s RCA and XLR inputs are received by separate and independent input stages. (If that were not the case shorting plugs on the RCA inputs would probably short out one of the two signals in the balanced signal pair).

You can find a number of sellers at eBay offering inexpensive RCA shorting plugs.

Regards,
-- Al

JO3533, yes by all means try to somehow post or link to the schematic you have for the amp. I’ve tried to find one online and have not been able to.

Jim, a couple of your recent statements have touched upon what I’ve been thinking that led me to suggest putting shorting plugs on the RCA inputs. Without a schematic I’m shooting in the dark to some extent, but here is my speculation:

First, even if there is some commonality between the input circuitry for the RCA inputs and the XLR inputs, given that the input impedance of the RCA inputs is vastly higher than it is for the XLR inputs, and given that per JA’s measurements the gain of the RCA inputs is even significantly higher than for the XLR inputs, I can’t envision a scenario in which putting shorting plugs on the RCA inputs would adversely affect the signals on the XLR inputs. (In contrast to the design approach that is used by many amps providing XLR and RCA inputs, in which the center pin of the RCA connector is connected directly to pin 2 of the XLR connector, with the balanced/unbalanced switch simply grounding pin 3 of the XLR connector when the switch is set to unbalanced).

Second, I would discount as irrelevant all of the findings in which hum occurred when an unterminated XLR cable was connected, and when an XLR cable was connected to the AVR while the AVR was unplugged from AC power (which probably also represents an essentially unterminated condition).

Third, what I’m envisioning relates to Jim’s comment that:

If jo3533 disconnected all ICs between the Marantz and Adcom, powered up both units, and then checked with an AC volt meter for an AC voltage from one chassis to the other chassis he should measure zero AC volts, Correct?

I’m envisioning that the Marantz AVR has significant AC leakage to its chassis and its signal ground (which Jim had indicated earlier are common with each other), the leakage occurring either in its power transformer or perhaps as a result of a leaky line filter capacitor. As long as the leakage is within reason, it won’t cause a problem under **most** circumstances.

Now imagine hypothetically that the amp has an infinite input impedance. Since the AVR has a two-wire power cord, the amp would then see the input signals it receives as "floating" relative to its own safety ground and signal ground. Meaning that the two signals in each balanced signal pair would both be seen as riding up and down over a significant voltage range at a 60 Hz rate. In other words, a significant amount of 60 Hz "common mode" noise would be present.  The finite (and relatively low) input impedance of the amp’s XLR inputs will then greatly reduce that range, but I’m envisioning that there may still be enough 60 Hz common mode noise on the signals, relative to safety ground and to the amp’s signal ground (which is probably connected to safety ground through a low impedance) to cause the issue. And conceivably it could be causing the issue by coupling into the very high impedance of the RCA input circuitry (although other pathways are possible as well). Which is why I am suggesting the shorting plugs, assuming I am correct in thinking that they will not affect the signals received via the XLR connector.

To respond to some other questions that have been raised: The polarities of pins 2 and 3 on the XLR connectors of the components have no relevance to the issue. I wouldn’t expect a DC offset in the AVR’s outputs to be relevant either. For relatively low-priced but high quality XLR cables I suggest Mogami Gold Studio, in either a 6 foot length or a 3 foot length. Mogami Gold Studio is the de facto cable of choice in most recording studios and other pro audio applications, and has been used in home audio applications and recommended here by me and many others in the past. Although as is usually the case among audiophiles opinions about it are not unanimous.

Good luck. Regards,
-- Al

Hi Jim,

Please consider the paragraph in my first post of today in which I used the word "floating."  And in doing so, re your comment that:

... the Signal ground of the Adcom was extended through the IC to the chassis/signal ground of the Marantz.

consider that the Adcom amp may connect pin 1 of its XLR connectors to chassis, rather than to its signal ground (the former, in fact, generally being best practice).  And its signal ground and chassis may be connected together through a significant impedance (that also generally being best practice).

My point being that the amp may not "see" the balanced pair of signals it is receiving as being tightly referenced to its own signal ground.  Resulting in the possibility that the amp might see those signals as having some degree of 60 Hz common mode noise riding on them, that may in turn somehow couple into its signal path to a degree that results in hum.

Best regards,
-- Al
 

Please explain your reasoning for when connected to a preamp that is powered off being the same as the cable not connected to anything, open ended.

Hi Jim,

While the output impedance of an unpowered solid state component is probably unpredictable in most cases, especially without detailed knowledge of its design, I would expect that in general it would be very high. Very high output impedance = essentially no termination at that end of the cable.

I have had instances when a power amp was left powered up and the preamp was turned off. There was not any hum/buzz coming from the speakers. Of course I have never owned a 300 watt per channel amp either.

Not relevant; different components.  And chances are those preamps had 3-wire power cords, as opposed to the OP's AVR (see the comments in my first post of today about "floating").

I still can’t wrap my head around that the Adcom amp is quiet when nothing is connected to any of it’s inputs, but when an XLR balanced cable is connected to the amp it hums/buzzes. I have been assuming the hum/buzz is only from the speaker connected to that channel of the amp.

If the unterminated cable is not picking up radiation from some external source of EMI, perhaps it is picking up such radiation from some part of the amp itself. And if so, that pickup would probably not be an issue when the input end of the cable is terminated by the very low output impedance of the powered up AVR. I would expect that the very low output impedance of the powered up AVR would load down and "kill" any such pickup.

Best regards,
-- Al

Thanks for providing the documents, Joe.  After looking at the schematic I withdraw my suggestion of RCA shorting plugs.  Consistent with your earlier comment it shows the center pin of the RCA connector being wired directly to pin 2 of the XLR connector.  Some unusual things that are done with various resistors account for the unusual differences in input impedance and gain between balanced and unbalanced input modes.

I'm assuming, btw, that the schematic you provided, which is for a GFA-7800, represents the circuitry for each of the five channels of your GFA-7805.

I don't know if the fact that the cables you are using connect pin 1 to the shells of the XLR connectors might factor into the problem.  But in any event I would suggest ordering the Mogami cables I referenced earlier, as I suspect they will sound at least a little bit better than what you are using even if changing to them doesn't help the problem. 

With the XLR cables disconnected from pre-amp. (NO NOISE)

This is different than before.  Perhaps when you tried this previously, with noise resulting, most of the length of the cable was closer to the amp than in the present arrangement that is shown in the photos, resulting in the unterminated cable picking up EMI from the amp itself (as I speculated earlier).

Not sure what else to suggest at this point.

Regards,
-- Al
 

Hi Jim,

When I first looked at that photo last night I interpreted it the same way as you did, that two pins (probably 1 and 3) were jumpered together. But after noting Joe’s statement that:

I took the XLR cable apart on both ends and both end have pin 1 connected to the shield ground of the plug itself. You can see it in the photos.

... and after looking at the photo a little further, I concluded that pin 1 is most likely not connected to pin 3, but rather to a lug which in turn connects to the shell of the XLR connector. Which was the reason I said in my last post that:

I don’t know if the fact that the cables you are using connect pin 1 to the shells of the XLR connectors might factor into the problem. But in any event I would suggest ordering the Mogami cables I referenced earlier....

But I agree that it can’t hurt, and could conceivably help, to clip that jumper at both ends of the cable and see what happens.

Best regards,
-- Al

Wow! Congratulations! Can’t wait to hear what the solution was.

Jim, sorry for not responding to your post of yesterday morning sooner. It’s a busy weekend here. Your points were well taken, IMO, and as you’ve seen Joe subsequently confirmed that continuity existed per your suggestion.

I would highlight, though, that R49 and R47 are jumpers, not resistors, as indicated by the zero ohm value noted on the schematic ("0R00"), and as stated explicitly in the parts list in the service manual. So analog signal ground and chassis/safety ground are connected directly together in this design, rather than through a low value resistor (which would lessen the chances of ground loop issues, as Atmasphere has stated in a number of past threads here).

Also, as indicated in the Rane paper you linked to earlier, ideally XLR pin 1 should be connected to chassis ground and NOT to signal ground, although in this case that is probably a moot point since both grounds are jumpered together.

So although the design is not ideal in a couple of respects, in ways that could conceivably contribute to hum problems depending in part on how grounds are handled in the component the amp is connected to, and perhaps also depending on whether XLR or RCA interconnections are used, my understanding (consistent with what is said in the Rane paper) is that lots of other designs are similarly flawed.

But we’ll soon see what the upshot is in this case.

Best regards,
-- Al

Good find, Jim. Consistent with your comment, very close to that clip are a couple of locations that appear likely to be designed and designated for resistors, but appear to have jumpers installed. (I expanded the photo to 200% to be able to see that better). My guess (which appears to be confirmed by the diagram you referred to in the service manual) is that one of them is R49, and that there is a connection within the board from one of the R49 pads to the pad the clip is attached to. With the other end of the clip being physically connected to chassis.

Although as you indicated it appears nearby that there is an actual resistor installed at what the service manual diagram depicts as R47 (the pin 1 to analog ground connection). So the parts list and the photo are inconsistent.

Best regards,
-- Al

Hi Jim,

Although we can’t be completely certain, I’m thinking that R49 and R47, which are explicitly noted in the parts list in the service manual as being "jumpers," are most likely just short lengths of wire. And I’m envisioning that during the design process the 1/4 watt 0R00 designations were just used by the electrical designers to specify to the mechanical designers that the corresponding holes and pads in the printed circuit board should be spaced and sized such that they would be suitable for insertion of 1/4 watt resistors. The intent being to provide the flexibility of being able to install 1/4 watt low value resistors later in the development process, if it proved to be necessary.

Best regards,
-- Al

Good call! Yes, it appears that it could be a thermistor. Ironically, though, I believe that the symbol for R55 shown nearby on the schematic (connecting digital ground to chassis) is one of the symbols that are used to represent thermistors, yet R55 appears in the photo to be just a jumper. Also, I don’t see any listing at all for R55 in the parts list.

So if the photo is considered to be gospel, as opposed to the schematic and the parts list, it appears that XLR pin 1 is connected directly to chassis (via the R49 jumper); digital ground is also connected to chassis; and analog ground is connected to those points through a thermistor.

Given the flexibility that is provided by the board design for how these circuit points are utilized I doubt that there would have been a reason for them to modify the related circuit traces under or within the board. But given the inconsistencies between the various documents it’s all hard to figure.

Best regards,
-- Al

Great news indeed, Joe!

We provided some thoughts and leads, and you did an exemplary job of following up on them in a thorough manner, and taking them further. Truly a collaborative effort, on a difficult problem (as grounding and hum problems often are).

Regarding the popping issue, all I can suggest is what I said in my post dated 2-7-16, which raised the possibility that it might be resolved by a firmware update, if one is available. As I indicated in that post, someone had reported elsewhere a few years ago that what sounds like the same problem was **introduced** by a firmware update. Which suggests the possibility that the issue may have been addressed in a subsequent update. Aside from that possibility, I would suspect that the popping is being caused by a hardware problem, perhaps a faulty capacitor or relay, or a diode associated with a relay.

Enjoy! Best regards,
-- Al

Looking at the Adcom Amp schematic wiring diagram again, trying to figure out why when all the outer contacts of the RCA jacks of the 5 amps of the Adcom are connected together and then connected to the Marantz chassis by an extended wire, the hum/buzz is eliminated.

Your guess is as good as mine, Jim.  Although I'll add that if the modules in the Adcom conform to the photo you supplied (as opposed to the schematic), as we said earlier it would appear that the Adcom's analog ground (which is common with the ground sleeves of the RCA connectors) is connected to XLR pin 1 through a thermistor.  And if that is the case then the solution Joe arrived at resulted in a reduction of the impedance between the Adcom's analog ground and the chassis of the Marantz Pre/Pro (which you had indicated is common with the signal ground of the Marantz), from whatever impedance the thermistor has during normal operation to essentially zero.  The introduction of that more direct connection between the signal grounds of the two units would seem consistent with a reduction or elimination of hum, that might have been caused by small amounts of AC leakage to chassis in either component.

In any event, considered purely from an intuitive standpoint it seems to me to be very plausible that what Joe did would result in a solution.

Best regards,
-- Al
 

I suppose a Faraday Cage effect is a possible explanation, but a couple of factors seem to suggest otherwise.  First, the problem didn't seem to be sensitive to the positions of the components, which lessens the likelihood that EMI pickup (that would be prevented or minimized by a Faraday Cage) was responsible.  Second, a similar shielding effect was apparently present all along, since the two chassis had always been interconnected via the shield of the XLR cables.  The shields of the cables having apparently been connected to the chassis of each component both via pin 1 and via the jumpers in the cable which connected pin 1 to the shells of the connectors.

Putting my previous speculation in more general terms, I suspect basically that the grounding approaches in the two components, possibly in combination with some condition-related issue (such as a leaky line filter capacitor), somehow caused an interaction between the two components resulting in small (and perhaps expectable) amounts of AC leakage to chassis in one of the components (perhaps occurring via stray capacitances in a power transformer or via a leaky line filter capacitor) to be seen by the amp as signal.

Best regards,
-- Al