which points even more strongly towards a power supply issue.
buildmesomething: your hum/noise will be because being a battery powered device with virtual ground (the O2) when its off the inputs of 'the wire' are effectively floating because there is no stable 'ground' reference voltage, thus free to pick up any stray noise.
FR frugal horns with the wire and dps600? seems overkill no?
on the other hand I realize there is merit to the 'build once and build well' mantra. you are much more likely to build another set of speakers (or have more than one set) than build another 4 channels of poweramp (you have 4 channels dont you?)
my main concern with the frugalhorn and this amp is the gain structure, then again you are using the amp with some seriously efficient speakers arent you Owen?
buildmesomething: your hum/noise will be because being a battery powered device with virtual ground (the O2) when its off the inputs of 'the wire' are effectively floating because there is no stable 'ground' reference voltage, thus free to pick up any stray noise.
FR frugal horns with the wire and dps600? seems overkill no?
on the other hand I realize there is merit to the 'build once and build well' mantra. you are much more likely to build another set of speakers (or have more than one set) than build another 4 channels of poweramp (you have 4 channels dont you?)
my main concern with the frugalhorn and this amp is the gain structure, then again you are using the amp with some seriously efficient speakers arent you Owen?
Last edited:
Yes I did try that and I did find an unbalanced Ground current between the planes. That current developed, as far as I can remember 8mV across the 10r resistor.
I also reported that I would and did put back in the 0r0 between the Ground Planes.
So that reverts my implementation back to one of the 4 given in the build sheets.
I think I also reported which of the 4 options I had built: AC coupled Single ended.
There is no other grounding that can be implemented on the as supplied PCB.
The screechy power down version is assembled using the official build instructions, except for adopting the NFB/gain components of the AC balanced in the AC single ended, i.e.30k/1k2 instead of 5k49/220r and reducing input cap from 330uF to 2u2F C78 and adding RF attenuation of 100pF across the 30k of R46. This is the version I have been listening to for the past few months.
Can anyone see any technical reason why any of those should leave a screechy output on power down?
However, I have modified the second and third PCBs to separate the Signal Ground from the Power Ground. This I have not reported since my experiments to remove the oscillation were abandoned some months ago, to devote time to some symasym style amplifiers.
I have also looked at, but not implemented a different version of the mute circuit, but due to a lack of suitable smd components that too is in limbo.
Last edited:
Thanks qusp, I can now remember that my O2 does generate some noise with the Pioneer cdp I'm using, but is quiet once both are on. Plus the O2 has it's own on/off transitions.
Just doing two channels, though did say I'd have liked to do more [5.1 amp - future project]. Just the idea of building the speakers is tempting, but I'll never be a chippy so cabinate work would be difficult in the extreme. The XTZ speakers appear to be good speakers and offer some tuning abilities as standard. Their UK distributer has closed so I'd have to go direct, & they have good representation on the UK based AVForum.
So with the dps-600's, the best way forward is to use the stanby circuit. This is something I was planing to do as part of the final build, a mains supply switch [rear mounted] & seperate standy switch [front mounted].
I too am using the AC Coupled SE build & everthing is as spec'd in the work sheet. With mine, I would expect all or most of the start-up noise to disappear once it's in a chassis & have moved away from using the O2.
Just doing two channels, though did say I'd have liked to do more [5.1 amp - future project]. Just the idea of building the speakers is tempting, but I'll never be a chippy so cabinate work would be difficult in the extreme. The XTZ speakers appear to be good speakers and offer some tuning abilities as standard. Their UK distributer has closed so I'd have to go direct, & they have good representation on the UK based AVForum.
So with the dps-600's, the best way forward is to use the stanby circuit. This is something I was planing to do as part of the final build, a mains supply switch [rear mounted] & seperate standy switch [front mounted].
I too am using the AC Coupled SE build & everthing is as spec'd in the work sheet. With mine, I would expect all or most of the start-up noise to disappear once it's in a chassis & have moved away from using the O2.
I am using regulated linear supplies for testing till the parts I need to implement higher LME supply voltages arrive. With no input signal, the output has a soft low frequency approximately 1 volt thump at turn off. With an AC signal applied while the amplifier is shutting down, there is a bit of buzzing when the supplies reach around +/-10volts. This I believe is due to the sine waves distorting while the supply voltages decrease way below the permitted operating range +/- 20volts. I don't see any high frequency oscilations. Bias is around 200 ma
Zener orientation
Hi BuildMeSomething,
I think that you are the second builder to get this wrong, Is it not obvious which way to mount the diode? ie line on device to line on PCB?
Think I found the root problem, got the Zener Diode mounted the wrong way... Then probably shorted something trying to figure out why I wasn't getting any sound...
R39/46 are shot & will replace C71/81, one or both of these went poof... and reorientate the zener. Hopefully everything else is ok...
Hi BuildMeSomething,
I think that you are the second builder to get this wrong, Is it not obvious which way to mount the diode? ie line on device to line on PCB?
I got my first build right, I got my second build wrong.
It took me a while to find the low voltage across the Zener. A very long while !
Now I state an observation.
The Zener is not needed.
We need a mute that has a delay on un-mute at start up and a near instant mute at power down. The Zener prevents near instant mute at power down.
I have words that describe this design decision, but I am not allowed to use them here.
It took me a while to find the low voltage across the Zener. A very long while !
Now I state an observation.
The Zener is not needed.
We need a mute that has a delay on un-mute at start up and a near instant mute at power down. The Zener prevents near instant mute at power down.
I have words that describe this design decision, but I am not allowed to use them here.
Last edited:
With giving it more than a few seconds thought and going though a flow diagram or looking at the circuit, etc, I suggest this possibility is investigated:
You need to fit a micro relay such that a pair of normally closed contacts would short the Zener or cap or whatever. The relay coil is joined to a suitable power rail and earth. Instant enough power up, the de-thump works with the cap and diode as normal, then the relay goes off as the power rail drops, and mutes the chip. You'd have to do it such that the relay's coil is de-energised before the rail drops to the point where noise is output.
You need to fit a micro relay such that a pair of normally closed contacts would short the Zener or cap or whatever. The relay coil is joined to a suitable power rail and earth. Instant enough power up, the de-thump works with the cap and diode as normal, then the relay goes off as the power rail drops, and mutes the chip. You'd have to do it such that the relay's coil is de-energised before the rail drops to the point where noise is output.
simply use a resistor feed to the mute pin instead of a Zener stabilised feed. It's the Zener that holds up the mute pin voltage on power down.
I would have experimented with a resistor only feed on my second PCB if I had the more correct resistor values, but my stock of smd resistors is very small in comparison to my through hole stock.
I would have experimented with a resistor only feed on my second PCB if I had the more correct resistor values, but my stock of smd resistors is very small in comparison to my through hole stock.
Andrew:
The zener is the only reasonable way to ensure the correct current into the mute pin given that people will be running supply voltages ranging from +/-20V up to +/-90V.
If just a resistor were being used, then I would have to supply several different values and people would have to populate the correct value based on what supply voltage they are using. There is a lot more room for error with that compared to getting a diode in the right way around.
Besides all that, if you don't want the Zener, then take it out! The resistor to Vcc is still present, you just need to change the value to suite your supply. You not having the correct SMD resistor values at hand is hardly a good reason to gripe about my design decisions.
Cheers,
Owen
Wire Amp Muting?
Hello folks,
Been following this thread, as i am working on my own version of this design. Thanks for the feedback, helps me figure out some design issues I have in my implementation.
So based on this Mute issue, I take it that you need to have a AC present & absent detector.
A simple transistor ckt to ground the mute pin quickly in the absence of an input AC voltage, sounds like what you are requiring. That and a mute delay at power up. This is why we prototype, test and then improve, all part of the design process, know one is at fault here, so strong words are not required.
I am undecided how I will do my speaker protection part of the design. Contemplating a full speaker protection ckt, AC line detect, o/p DC detect, over current detect, turn-on delay, relay for the speakers? I understand the reasons to have these features and why not. All part of the design process. If anyone has input for me please express it in the thread that I have created, rather than de-rail this thread
http://www.diyaudio.com/forums/solid-state/220962-lme49830-latfet-yet-another-design-review.html
AndrewT, not sure why you are saying that the zener "Z1", 12V is not needed. It has to be there to limit the input voltage to the LME49830 mute pin. Looking at the LM49830 spec, I can not determine the maximum V the mute pin can tolerate, they only say 120uA min to put the part into play mode. So I assume that Owen (OPC) just followed the guidelines from the data sheet and app note info.
Hi Owen, thanks for your design contribution, helps me alot in my efforts. If you are open to me asking you some questions about my design, I would appreciate it very much. We could do this off line. Who knows one day I might be able to buy you a beer or two/lunch/dinner, I realize that you are in Ottawa, so I am not that far from you or if you are in the GTA at some time in the future.
Regards
Rick
Hello folks,
Been following this thread, as i am working on my own version of this design. Thanks for the feedback, helps me figure out some design issues I have in my implementation.
So based on this Mute issue, I take it that you need to have a AC present & absent detector.
A simple transistor ckt to ground the mute pin quickly in the absence of an input AC voltage, sounds like what you are requiring. That and a mute delay at power up. This is why we prototype, test and then improve, all part of the design process, know one is at fault here, so strong words are not required.
I am undecided how I will do my speaker protection part of the design. Contemplating a full speaker protection ckt, AC line detect, o/p DC detect, over current detect, turn-on delay, relay for the speakers? I understand the reasons to have these features and why not. All part of the design process. If anyone has input for me please express it in the thread that I have created, rather than de-rail this thread
http://www.diyaudio.com/forums/solid-state/220962-lme49830-latfet-yet-another-design-review.html
AndrewT, not sure why you are saying that the zener "Z1", 12V is not needed. It has to be there to limit the input voltage to the LME49830 mute pin. Looking at the LM49830 spec, I can not determine the maximum V the mute pin can tolerate, they only say 120uA min to put the part into play mode. So I assume that Owen (OPC) just followed the guidelines from the data sheet and app note info.
Hi Owen, thanks for your design contribution, helps me alot in my efforts. If you are open to me asking you some questions about my design, I would appreciate it very much. We could do this off line. Who knows one day I might be able to buy you a beer or two/lunch/dinner, I realize that you are in Ottawa, so I am not that far from you or if you are in the GTA at some time in the future.
Regards
Rick
Last edited:
You don't need the zener if you calculate the resistor value properly:
IMUTE = (VMUTE – VBE) / (1kΩ +RM) where VBE ≅ 0.7V
The problem lies in the fact that the calculated resistance only really works with one supply voltage (over a reasonable range) and needs to be adjusted as per the above formula for different supply voltage.
Using the zener simplifies this as you can use any rail voltage you want and the current into the mute pin is always the same.
If you were designing this into a product with a set rail voltage, it would be cheaper and easier to simply calculate the right resistance and omit the zener.
Cheers,
Owen
IMUTE = (VMUTE – VBE) / (1kΩ +RM) where VBE ≅ 0.7V
The problem lies in the fact that the calculated resistance only really works with one supply voltage (over a reasonable range) and needs to be adjusted as per the above formula for different supply voltage.
Using the zener simplifies this as you can use any rail voltage you want and the current into the mute pin is always the same.
If you were designing this into a product with a set rail voltage, it would be cheaper and easier to simply calculate the right resistance and omit the zener.
Cheers,
Owen
So...
Another resistor can be inserted between the capacitor and the mute pin. The relay joined to the mute pin. This way, when the relay earths the mute pin, the capacitor discharges slowly and doesn't weld the relay contacts.
During the period of earthing there will be some small current flowing through the resistors, via the relay, to earth.
Would that do it?
Same relay can also be operated remotely from, say, a pre-amp. (by interrupting the voltage to the relay)
Another resistor can be inserted between the capacitor and the mute pin. The relay joined to the mute pin. This way, when the relay earths the mute pin, the capacitor discharges slowly and doesn't weld the relay contacts.
During the period of earthing there will be some small current flowing through the resistors, via the relay, to earth.
Would that do it?
Same relay can also be operated remotely from, say, a pre-amp. (by interrupting the voltage to the relay)
Last edited:
LME49830 mute circuit
I guess they could have spec'd the maximum current into the mute pin node, but I am sure is not to be over say ~100mA.
See why I am asking for help!!
Ya, it would help if I looked at the LME49830 block diagram to see the 1K res in series with the diode to gnd. I am a real Homer Simpson at times
I guess they could have spec'd the maximum current into the mute pin node, but I am sure is not to be over say ~100mA.
See why I am asking for help!!
National specifically states that exceeding 2mA into the mute pin could damage the IC, so that's your limit. If I were designing it, I would target about 900uA at nominal Vcc since that's in the middle of the specified allowable range. The last thing you want is the IC muting during a large transient that loads the supply rails, or getting destroyed under high line conditions.
See the appeal of the zener?
Cheers,
Owen
See the appeal of the zener?
Cheers,
Owen
I wish
I'll take one of these... definitely the nicest triple at the LCBO right now, if you can find one.
Product Information
Cheers,
Owen