Hi everyone. First post here, I've been lurking for a few days reading up on DIY Class AB designs as I'm part way through building a SC480 kit. Unfortunately that's not what triggered my first post. I just received a Quad Vena 2 from the UK (I'm in Australia). I'm aware that the only British part of the amp is the name, but it's still very nice and appears well build.
I had the amp running for ~30 minutes when it stopped... at the same time the oven timer went off. The brown smell was there but all the lights were on. The electronics tech in me couldn't resist taking off the cover for a look. No obvious destruction so I pulled out the multimeter. The pre amp and headphone amp still work, along with all the digital electronics. All the rail voltages are there. I checked the small signal input to the LM3886's and it's getting a signal. Absolutely zilch coming out of the chip though. I checked the mute pin and interestingly it sits at +4V relative to ground. The data sheet for the LM3886 shows a PN junction between the ground pin and the mute pin, so at minimum the mute pin should be -0.6V relative to ground. I know that the mute function needs -0.2mA to work, but I'm not willing to measure this just yet, for warranty reasons. I'd be interested to know what other's have on their mute pin; is it a negative voltage?
Here's the kicker, the amp came from the UK so it's set for 230V mains. There's a wire link that sets this, in my neck of the woods it should be in the position for 240V. That means the secondaries on the transformer will be too high. I think that the oven turning off and putting a jolt on my powerpoint may have been the straw that broke the camel's back. I've measured 75V across the input to the LM3886.
I'm reluctant to send the amp back to where I bought because of the shipping costs, and I haven't heard great things about Quad/IAG's customer service. If I can isolate the fault to the LM3886's then I'd consider swapping them out.
I had the amp running for ~30 minutes when it stopped... at the same time the oven timer went off. The brown smell was there but all the lights were on. The electronics tech in me couldn't resist taking off the cover for a look. No obvious destruction so I pulled out the multimeter. The pre amp and headphone amp still work, along with all the digital electronics. All the rail voltages are there. I checked the small signal input to the LM3886's and it's getting a signal. Absolutely zilch coming out of the chip though. I checked the mute pin and interestingly it sits at +4V relative to ground. The data sheet for the LM3886 shows a PN junction between the ground pin and the mute pin, so at minimum the mute pin should be -0.6V relative to ground. I know that the mute function needs -0.2mA to work, but I'm not willing to measure this just yet, for warranty reasons. I'd be interested to know what other's have on their mute pin; is it a negative voltage?
Here's the kicker, the amp came from the UK so it's set for 230V mains. There's a wire link that sets this, in my neck of the woods it should be in the position for 240V. That means the secondaries on the transformer will be too high. I think that the oven turning off and putting a jolt on my powerpoint may have been the straw that broke the camel's back. I've measured 75V across the input to the LM3886.
I'm reluctant to send the amp back to where I bought because of the shipping costs, and I haven't heard great things about Quad/IAG's customer service. If I can isolate the fault to the LM3886's then I'd consider swapping them out.
The nominal 230 volt AC UK mains specification covers the range 207 to 253 (230 -6% and +10%) and so that should not be an issue.
If the amp has a 230 and a 240v setting then 240 would give a wider safety margin.
If the amp uses a split rail which is the normal configuration then there should be two supplies, a plus and a minus going to the LM3886. The total supply voltage must never exceed 94 volts (so -/+47) and would normally be considerably lower. If your 75 volt is really -/+ 38 ish then that is perfect really.
Both channels faulty?
Could the unit really be in mute for some reason? Chips like the LM3886 typically show no mercy if they fail and they would usually be split open or have a hole blown in the side of them.
The mute voltage you measure suggests it is in mute to me.
If the amp has a 230 and a 240v setting then 240 would give a wider safety margin.
If the amp uses a split rail which is the normal configuration then there should be two supplies, a plus and a minus going to the LM3886. The total supply voltage must never exceed 94 volts (so -/+47) and would normally be considerably lower. If your 75 volt is really -/+ 38 ish then that is perfect really.
Both channels faulty?
Could the unit really be in mute for some reason? Chips like the LM3886 typically show no mercy if they fail and they would usually be split open or have a hole blown in the side of them.
The mute voltage you measure suggests it is in mute to me.
Fully agree with mooly. You should check mute resistor to ensure that it is supplying enough current to unmute the chip, 10-33Kohm is fine.
Tell him to measure voltage, what is the range for that power supply for muting.
And what else to check?
A dim bulb tester?
Any other thoughts?
And what else to check?
A dim bulb tester?
Any other thoughts?
Thanks for the replies everyone. It’s the same for both channels. I think it would be good if it was an issue with the mute input, it looks like LM3886 is pretty hard to get at the moment. By the way the mute function on the remote stops the pre amp, so I don’t think they’d bother using the mute function of the chip unless it’s a protective feature.
I can’t check the mute resistor without removing the double-sided PCB. It’s on one big board so it’s a bit off a mission to remove and I’d probably void the warranty. My current theory is that the power glitch didn’t kill the whole chip but just the mute transistor.
What’s odd is that the amp has two voltage ratings - 100/230 and 120/240. These are the two options for the jumper so how does the amp get enough voltage at 100V or 120V?
I can’t check the mute resistor without removing the double-sided PCB. It’s on one big board so it’s a bit off a mission to remove and I’d probably void the warranty. My current theory is that the power glitch didn’t kill the whole chip but just the mute transistor.
What’s odd is that the amp has two voltage ratings - 100/230 and 120/240. These are the two options for the jumper so how does the amp get enough voltage at 100V or 120V?
Maybe, in which case it´s possible no current is being fed into pin 8 by (presumed) microprocessor
For testing, I woukd tack solder a 33k resistor, as suggested by nirupam from pin 8 (mute) to -V (pin 4).
You are removing nothing and when needed you just remove it and nobody´s the wiser.
In any case I would avoid sending an amp both ways around the Globe unless strictly necessary.
For testing, I woukd tack solder a 33k resistor, as suggested by nirupam from pin 8 (mute) to -V (pin 4).
You are removing nothing and when needed you just remove it and nobody´s the wiser.
In any case I would avoid sending an amp both ways around the Globe unless strictly necessary.
The mute pin usually sits at about -3 V relative to ground when the amp is running. That voltage does depend some on temperature (three base-emitter junctions), but it should definitely be a negative voltage if you want the chip to enter active mode.
Is there a protection circuit that activates the mute pin on the LM3886?
Tom
Is there a protection circuit that activates the mute pin on the LM3886?
Tom
Tomchr, I’ve spent some time on your site, so I know you know your stuff regarding the LM3886! The mute pin isn’t sourcing 0.5mA if it’s got +4V on it, but whether that caused by a fault in the chip or a fault in whatever drives the pin I can’t confirm. Considering that both chips are showing the fault, they may be driven in parallel in which case the fault is in a transistor somewhere that should be pulling the mute pin down.
I got a response from the shop I bought the amp from and they want me to send it back to the UK.
I got a response from the shop I bought the amp from and they want me to send it back to the UK.
Put a 20mm 275 volt MOV on your house mains, phase to neutral, permanently connected. Each phase to neutral if you have many phases.
For 110 V nominal, 130-145 V should be OK.
It will short out anything above 275 Volts, and below that it is an excellent arc quencher.
So all the equipment in your house will benefit. And the jolt from the oven or whatever will be absorbed.
Water heaters, air conditioners, others using make and break switches will have extended switch life.
For 110 V nominal, 130-145 V should be OK.
It will short out anything above 275 Volts, and below that it is an excellent arc quencher.
So all the equipment in your house will benefit. And the jolt from the oven or whatever will be absorbed.
Water heaters, air conditioners, others using make and break switches will have extended switch life.
That seems quite likely.
Just to be clear: No current is sourced by the mute pin. It's up to the amp designer to ensure that at least 0.5 mA is drawn from the mute pin to make the chip active. That's probably what you meant.
I find that 500 uA is a bit on the low side if you also want low THD. I generally design for 1 mA at 80% of the nominal supply voltage.
I did read the bit about +4 V on the mute pin. That's broken! It should be about -3 V if the LM3886 is to be unmuted. As you mentioned, it is likely a transistor somewhere that's gotten cranky.
That sounds pricey.
Tom
There is some discussion going on here about supplies and authenticity of the LM3886 IC.
Could you, if not too much effort is involved, post pictures of the ICs?
And a general one of the entire innards?
Some help to us all, as to the current state of British design and outsourced manufacturing as well.
Could you, if not too much effort is involved, post pictures of the ICs?
And a general one of the entire innards?
Some help to us all, as to the current state of British design and outsourced manufacturing as well.
I had a look.
Heat sink next to transformer, compact design.
Find the space to put a small 12 v fan, or if you feel like, an AC operated fan on the top cover.
Did not check the price, here I can buy the bits for 40 to 50 GBP.
You would have spent much more.
Heat sink next to transformer, compact design.
Find the space to put a small 12 v fan, or if you feel like, an AC operated fan on the top cover.
Did not check the price, here I can buy the bits for 40 to 50 GBP.
You would have spent much more.
I've attached a photo of one of the LM3886's.. It's in a tight spot so sorry about the poor image quality. The label is really hard to make out in the best conditions.
I called Quad today and after getting a bit of an earful for not buying it locally (I couldn't get the steel chassis here) I was given the email address of one of their engineers. Funny, the lady on the phone said that he was busy in a meeting, but after I emailed I got an automatic response saying that he was on leave for the next week. I'm not holding my breath that they'll help me repair it, the "International Audio Group" is headquartered in China, so chances are that's where all the real engineers are and the UK office is just a front.
I did a bit more probing, the mute pins are shorted together. I haven't been able to trace where they are pulled down to V-, but I've measured ~6k8 from Mute to V-. Who reckons I've fried the mutes in both chips? In the meantime I've ordered some replacement LM3886T's (couldn't find a TF from a reputable supplier) just in case I have to replace them. Those things are like hens teeth!
I called Quad today and after getting a bit of an earful for not buying it locally (I couldn't get the steel chassis here) I was given the email address of one of their engineers. Funny, the lady on the phone said that he was busy in a meeting, but after I emailed I got an automatic response saying that he was on leave for the next week. I'm not holding my breath that they'll help me repair it, the "International Audio Group" is headquartered in China, so chances are that's where all the real engineers are and the UK office is just a front.
I did a bit more probing, the mute pins are shorted together. I haven't been able to trace where they are pulled down to V-, but I've measured ~6k8 from Mute to V-. Who reckons I've fried the mutes in both chips? In the meantime I've ordered some replacement LM3886T's (couldn't find a TF from a reputable supplier) just in case I have to replace them. Those things are like hens teeth!
Attachments
I think the T version is metal tabbed and the TF fully insulated and so beware, the T will not work without insulating washers and bushes.
This fault doesn't sound like classic chip failure to me, it sounds more like a system control issue.
Have you tried the remote control?
Is there any kind of back up cap in the main unit?
Any factory reset procedure?
This fault doesn't sound like classic chip failure to me, it sounds more like a system control issue.
Have you tried the remote control?
Is there any kind of back up cap in the main unit?
Any factory reset procedure?
The LM3886 looks legit as far as I can tell. It would be exceptionally unlikely for a reputable outfit like Quad to use fakes.
It's unlikely that both LM3886es fried. It's much more likely that the protection circuit that controls both LM3886es fried or that it's doing its job and holding the LM3886es in mute due to a fault.
I'd trace out that protection circuit.
Have you measured the DC voltage on pin 3 (output)? If the LM3886 is muted and working correctly it should provide 0 V out.
Tom
It's unlikely that both LM3886es fried. It's much more likely that the protection circuit that controls both LM3886es fried or that it's doing its job and holding the LM3886es in mute due to a fault.
I'd trace out that protection circuit.
Have you measured the DC voltage on pin 3 (output)? If the LM3886 is muted and working correctly it should provide 0 V out.
Tom
That won't work if something pulls the mute pin high with a lower impedance than ~33 kΩ.
Tom