Luxman R117 Relay Issue

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I have an R117 with burned/melted speaker relays on channel A and I need help brainstorming how it happened.
Background on the unit is that I got it in non-working condition, fixed the right and left channel amp sections, a selector board selector IC issue, and tone board opamp and tone defeat switch issues to get it up and running. It worked fine for months so I did a total recap/upgrade which made the unit sound spectacular. It worked fine after recap for about a year of daily use with no issues. Recently I put it aside to listen to something else for a couple of weeks, but yesterday I put it back into the mix and it played fine for hours at a level that was not particularly stressful to the amp. This morning switched it on and it played fine for several minutes then went into a loud distressed static tone for a couple of seconds. By the time I got to it to turn it off the sound was normal again. I powered it down and got the pop that occurs when the relays do not do their job. I smelled burning at the speaker terminals so I put it on my bench. On the bench I hook up a pair of expendable speakers to A channel and powered the receiver up. I got the turn on pop but it played relatively fine. I checked the DC level and L was 40mV, R was 24mv. I then hooked the speakers to Channel B and turned it on and there was no pop and it sounded fine on B channel. DC on B channel L was .1mV and on R channel it was .5mV I disassembled it down to the relays and found A channel L relay melted and fused and the other R channel not looking much better, see the pics.
I have the parts on order but before they get here and I install them I want to try and find how this happened in the first place. The only thing going on in the house circuit at the time the receiver was on was an old washer kicked in with its huge motor drawing down on the house current, but I don’t think this a likely cause. I would appreciate any help or insight that anyone would like to offer.
 

Attachments

  • REL 007.JPG
    REL 007.JPG
    323.2 KB · Views: 242
  • REL 009.JPG
    REL 009.JPG
    337.5 KB · Views: 244
It sounds like there was an intermittent failure that caused a high DC offset, and as a result, a high current into the speakers. The relays may have gotten fried while trying to open under these conditions. Normally you would expect speaker relays to be up to the job of disconnecting a speaker load while there was a DC offset at or close to one of the power rails. If they were not, the contacts might get welded shut if the protection circuit tried to open them.

To find out if DC offset is the cause, I would probaly run the amp without speakers for an extended period of time while monitoring the DC voltage on the output with a DMM in VDC setting. A DMM with MIN/MAX hold function comes in handy in case of an intermittent fault as it doesn't need constant monitoring but will display the highest or lowest occurrence.
If you don't have such a DMM, run it with your expendable test speakers and start measuring when the noise occurs...
 
It sounds like there was an intermittent failure that caused a high DC offset, and as a result, a high current into the speakers. The relays may have gotten fried while trying to open under these conditions. Normally you would expect speaker relays to be up to the job of disconnecting a speaker load while there was a DC offset at or close to one of the power rails. If they were not, the contacts might get welded shut if the protection circuit tried to open them.

To find out if DC offset is the cause, I would probaly run the amp without speakers for an extended period of time while monitoring the DC voltage on the output with a DMM in VDC setting. A DMM with MIN/MAX hold function comes in handy in case of an intermittent fault as it doesn't need constant monitoring but will display the highest or lowest occurrence.
If you don't have such a DMM, run it with your expendable test speakers and start measuring when the noise occurs...

Thanks for the input jitter, once I get the parts in I will do as you suggest. My DMM has a hold feature, I will give it a shot.

I am a a loss as to how I can have a higher DC offset showing up on the A channel than on the B channel at the same exact time. This suggests to me that the fault does not stem from the amp section per se but maybe coming in from the ground loop. Do you think this is a possibility?
 
Did you have speakers connected to both channels A and B or only to A when the fault occurred?
If the latter is the case, it would explain why channel B relay is OK. DC offset cannot damage a relay if there's no current going though it (i.e., if there are no speakers connected or if the channel is not switched on).

You can run the test without relays if you want. Just connect the DMM to the output of the power amp, in front of the relay.

A note on the hold function of a DMM: it must be a MIN/MAX-hold function. A button with only "hold" on it will probably freeze the reading at the moment that button is pressed whereas (e.g.) a MAX-hold updates the reading whenever the previous maximum is surpassed.
If it is a MIN/MAX-hold, run the amp a while on MAX-hold first for positive offset. If that doesn't show anything strange, repeat on MIN-hold for negative offset.
 
Did you have speakers connected to both channels A and B or only to A when the fault occurred?
If the latter is the case, it would explain why channel B relay is OK. DC offset cannot damage a relay if there's no current going though it (i.e., if there are no speakers connected or if the channel is not switched on).

You can run the test without relays if you want. Just connect the DMM to the output of the power amp, in front of the relay.

A note on the hold function of a DMM: it must be a MIN/MAX-hold function. A button with only "hold" on it will probably freeze the reading at the moment that button is pressed whereas (e.g.) a MAX-hold updates the reading whenever the previous maximum is surpassed.
If it is a MIN/MAX-hold, run the amp a while on MAX-hold first for positive offset. If that doesn't show anything strange, repeat on MIN-hold for negative offset.


I did not frame my question clearly. I understood why channel A relays got fried due to that channel being in use at the time of the incident. My question is to the current state of the amplifier section. It is simultaneously showing high DC offset on channel A while showing minimal DC on B channel. This is suggesting to me that the amp section is ok, due to both channels sharing the same amp, and that I need to look to the ground of A channel as the source of the increased DC offset reading. Does this make sense?
You are right on the hold function on my DMM, it just freezes the last reading shown. I am currently pondering whether it is better to finally get a good Fluke DMM or a new Rigol DS to replace my old broken Tektronix 465M.
I got in the new relays today. I will take a look at the unit on Sunday to see if I can find the source of the high DC offset on channel A.
 
Given the fact that channels A and B are driven by the same amps (L and R), and that all black speaker terminals are tied together, I don't see how the fault would be showing on channel A and not on channel B unless channel B relays aren't working.

I regard 25-40 mV offset as highish, but nowhere near dangerous. The kind of damage the speaker relays suffered would indicate offset close to a power rail.
Since this fault could be intermittent, it may have occurred a few times before you got the amp and the last time the channel A relays were finally damaged too much.

Since both L and R channels suffer from the problem, I would start looking in a circuit common to both L and R channels, like e.g. an intermittent connection in a power supply line.

Do you have the service manual? If not, you can find it here.
 
often use of exotic parts and ""upgrades"that have never been verified by instruments and only by the ears might end up like that

excessive high at listening , high frequency oscillations might fry the contacts capacitive or wrong loads also DC in the output and finally an already faulty relay that got much worst by starting to operate again after the "upgrade ""

Often people understand high frequency oscillation as ...""plenty of high "" which is a condition that except the "plenty of high " might cause some overheating and some other issues to the amp that totally will not be fatal for the amp unless pushed hard ..

Replace the relay, verify offset , scope the amp for any other issues

Kind regards
sakis
 
I regard 25-40 mV offset as highish, but nowhere near dangerous. The kind of damage the speaker relays suffered would indicate offset close to a power rail.
Since this fault could be intermittent, it may have occurred a few times before you got the amp and the last time the channel A relays were finally damaged too much.

I see your point about the offset not being all that bad. I guess I am just overly sensitive about it due to I have Adcom GFA 565s and I am always checking to see that they are in the 10mV or less range. I'm sure everyone is aware of the leaky cap issues with these units and the resultant damage that ensues.

I was thinking the same about the relays being damaged before their final demise. When I got the unit it was in very bad shape, damaged power transistors, loose disconnected wiring etc., so it is very possible this last power flux was the last straw for the relays of A channel. B channel relays looked excellent and I just gave the contacts a quick deoxit cleaning and reinstalled them.

Testing out the unit currently.
 

Attachments

  • Relay2 001.JPG
    Relay2 001.JPG
    327.4 KB · Views: 148
Last edited:
often use of exotic parts and ""upgrades"that have never been verified by instruments and only by the ears might end up like that

excessive high at listening , high frequency oscillations might fry the contacts capacitive or wrong loads also DC in the output and finally an already faulty relay that got much worst by starting to operate again after the "upgrade ""

Often people understand high frequency oscillation as ...""plenty of high "" which is a condition that except the "plenty of high " might cause some overheating and some other issues to the amp that totally will not be fatal for the amp unless pushed hard ..

Replace the relay, verify offset , scope the amp for any other issues

Kind regards
sakis

Hi sakis,

Thanks for the interest and input.

I was very careful to use OEM spec replacement parts and tested/recorded the values of all the replacement caps, even the ones I removed. I wanted to have a record of everything done in case there were issues and to use the information on my next R117 recap. I have a couple more that I would like to upgrade, maybe with a different brand name cap. Just to see if there is a difference in sound.
 
That can be a very interesting thing to listen ..please let us know

kind regards
sakis

I will post when I do. I am still researching which brand to use as the primary replacements. I used the Elna SilmicIIs this time because I read a thread where N. Pass casually noted that these were interesting.

Some pics of the replacements on my rehab project R117 and the new relays. I had to shoehorn a couple of caps into the confined spaces, but it worked out ok. The unit isn't pretty but it sounds very good.

I adjusted idle on the two power boards to 5mv, tested on the disposable speakers for a day and then on my shop Magnepan MGI Improved for a few hours with no issues. It is now back in my normal setup. Thanks jitter and sakis for the input.
 

Attachments

  • Relay2 002.JPG
    Relay2 002.JPG
    338.6 KB · Views: 75
  • Relay2 007.JPG
    Relay2 007.JPG
    344.8 KB · Views: 111
  • Relay2 006.JPG
    Relay2 006.JPG
    338.3 KB · Views: 90
  • Relay2 005.JPG
    Relay2 005.JPG
    347.5 KB · Views: 96
  • Relay2 004.JPG
    Relay2 004.JPG
    341.9 KB · Views: 97
  • Relay2 003.JPG
    Relay2 003.JPG
    331.2 KB · Views: 104
  • Relay2 008.JPG
    Relay2 008.JPG
    339.8 KB · Views: 78
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.