Hi All,
I'm interested to know how reliably and gracefully the 3875's thermal overload protection works. I'm having a party this weekend and suspect it may get a sustained workout so I'd like to know the risks of doing so and what to look for when it happens. Does it simply cut out until it cools? I've never overheated mine so I have no idea. Is there any chance of speaker damage? Your advice is appreciated.
Regards,
Greg.
I'm interested to know how reliably and gracefully the 3875's thermal overload protection works. I'm having a party this weekend and suspect it may get a sustained workout so I'd like to know the risks of doing so and what to look for when it happens. Does it simply cut out until it cools? I've never overheated mine so I have no idea. Is there any chance of speaker damage? Your advice is appreciated.
Regards,
Greg.
Well, I guess you should be far enough away not to annoy me then
Not graceful at all I'm afraid. You'll get gross distortion when the "SPIKE" protection starts to operate.
If you have adequate heat-sinking, then this isn't going to be a problem.
Frank
Thanks Frank. My heatsink is fine for normal listening levels (I have one 200mm length of 40mm x 40mm x 3mm Aluminium angle for each chip - see attached), but it's probably marginal for extended listening at high volume levels. What do you think? The chips barely get warm at normal levels. How reliable do you think the thermal overload protection is? Is there any chance of speaker damage if I cut the volume at the first sign of problems. Kinda nervous as they are worth $4k!
Attachments
Last edited:
What I would do is: Get a heftier power supply at least 10000uf per channel and get a real heat sink.That heat sink is way to weak for the IC. What is the SPL of your speakers? What is the impedance of those speakers?.
I know many people like the sound of small power supply but if you like strong bass go for bigger power supply you will not sacrifice treble or mid-range as many people say. I have try with many brands and also I started from 4.7uf, 10uf, 1000uf, 1500uf, 10000uf, 20000uf. But I Like 10000uf per channel better. Just give a try.This is all about diy buddy!
Those heatsinks do look a insufficient. If you don't want the thermal protection kicking in, and it does sound NASTY, you really should consider beefing them up a bit.
That being said, my LM3886LF has heatsinks that some people would really bash me about, but it's proven itself trough the "drunken-party-continious-high-volume" test very well. After 6 hours of continuous use, the entire enclosure got pretty warm, but not to a level of concern. My higher-power NAD amplifier has actually been hotter at similar events, with the same speakers.
Taking a look through the photo gallery thread will give you some ideas about size, if you don't have heatsink specs and power dissipation calculations.
Do note that while the chips can run really cool at "normal listening" levels, an extended high-power issue is a completely different story. You can not expect a miniature amp to be able to handle this type of usage, without proper cooling. The thermal protection is there to protect you from fatal failures and is not supposed to trigger in any intended use of the amplifier.
Trust us - you do not have "party mode" cooling for those chips.
When it overheats it has a nasty cut-off, which does not turn off until the chips have went certain amount of degrees below their critical value. It's all in the datasheet. I have not managed to trigger it myself to know how it is in practice and I have been trying
Generally it's a very good idea to have a DC-protection circuit, if only for making sure that any mistakes you make will not blow your speakers. The thermal overload is there to keep you safe, in general, but there are other things that might go wrong. You've got the space in the chassis and the protection circuits are not pricey. Add one, be safe, always.
That being said, my LM3886LF has heatsinks that some people would really bash me about, but it's proven itself trough the "drunken-party-continious-high-volume" test very well. After 6 hours of continuous use, the entire enclosure got pretty warm, but not to a level of concern. My higher-power NAD amplifier has actually been hotter at similar events, with the same speakers.
Taking a look through the photo gallery thread will give you some ideas about size, if you don't have heatsink specs and power dissipation calculations.
Do note that while the chips can run really cool at "normal listening" levels, an extended high-power issue is a completely different story. You can not expect a miniature amp to be able to handle this type of usage, without proper cooling. The thermal protection is there to protect you from fatal failures and is not supposed to trigger in any intended use of the amplifier.
Trust us - you do not have "party mode" cooling for those chips.
When it overheats it has a nasty cut-off, which does not turn off until the chips have went certain amount of degrees below their critical value. It's all in the datasheet. I have not managed to trigger it myself to know how it is in practice and I have been trying
Generally it's a very good idea to have a DC-protection circuit, if only for making sure that any mistakes you make will not blow your speakers. The thermal overload is there to keep you safe, in general, but there are other things that might go wrong. You've got the space in the chassis and the protection circuits are not pricey. Add one, be safe, always.
Last edited:
Hi,
if the amps are clipping severely due to being run overloaded then the extra treble sent through the crossover can damage your treble driver and sometimes the crossover. There are detailed articles on the web that say this excess treble energy hypothesis is all balderdash.
If your speakers are of low sensitivity then this problem of gross clipping will be much worse.
The chipamps, with the tiny onboard smoothing caps, look like Peter Daniel's.
He designed his circuit to sound best with his 95dB/W @ 1m speakers.
BTW,
your internal heatsinks are far too low in dissipation capability.
if the amps are clipping severely due to being run overloaded then the extra treble sent through the crossover can damage your treble driver and sometimes the crossover. There are detailed articles on the web that say this excess treble energy hypothesis is all balderdash.
If your speakers are of low sensitivity then this problem of gross clipping will be much worse.
The chipamps, with the tiny onboard smoothing caps, look like Peter Daniel's.
He designed his circuit to sound best with his 95dB/W @ 1m speakers.
BTW,
your internal heatsinks are far too low in dissipation capability.
I've had my LM3875's cut off often when driving a big subwoofer. It cuts off like an on off switch, without any noise or anything. But I won't give you any guarantees.
If a small amount of dc is passed each time it cuts off. That should not damage your tweeters. Clipping will. I once burned out a tweeter in my Wharfedale 505.2's during a party. Because the NAD 3020 was not powerful enough in the big room with all the people.
If a small amount of dc is passed each time it cuts off. That should not damage your tweeters. Clipping will. I once burned out a tweeter in my Wharfedale 505.2's during a party. Because the NAD 3020 was not powerful enough in the big room with all the people.
Hi Andrew & Bas,
Firstly, yes well picked. They are indeed Peter's PCBs. Regarding the heatsinks, I made the amp in a hurry on the cheap as a "proof of concept". Unfortunately I never expected it to sound so good, and never thought I would even consider replacing my mega$$ commercial amp with it, so this proved to be short sighted. I am also pretty stuck with the configuration as I used heatsink glue (plus screws of course) to attach the 3875s and don't expect it will be easily disassembled.
My speakers are Soavo 1s, and can be seen here:
Soavo 1
While not ideal for the amp impedance or sensitivity wise, the combination is the best I have heard. The amp drives the speakers to astonishingly high levels with good control. Lots of detail yet surprisingly warm. The bass was a pleasant surprise and I would never want any more. Even my girlfriend has been asking me to play music more so she can hear it. It just gives the music more life and makes the listening experience more pleasurable. We both love it. It has really softened up the "dissective" sound of my benchmark dac-1 too.
The amp runs cool at normal levels so I will swap it out with my Yamaha amp for the party and use it for personal listening only. On the side I have ordered another set of PCBs, a tranny and heatsinks to make a better version over the coming months.
Bas, thanks for your reply. It's nice to know that someone has pushed the chips without ill effect, but I'll stick with my old amp for the party to be safe.
Regards,
Greg.
Firstly, yes well picked. They are indeed Peter's PCBs. Regarding the heatsinks, I made the amp in a hurry on the cheap as a "proof of concept". Unfortunately I never expected it to sound so good, and never thought I would even consider replacing my mega$$ commercial amp with it, so this proved to be short sighted. I am also pretty stuck with the configuration as I used heatsink glue (plus screws of course) to attach the 3875s and don't expect it will be easily disassembled.
My speakers are Soavo 1s, and can be seen here:
Soavo 1
While not ideal for the amp impedance or sensitivity wise, the combination is the best I have heard. The amp drives the speakers to astonishingly high levels with good control. Lots of detail yet surprisingly warm. The bass was a pleasant surprise and I would never want any more. Even my girlfriend has been asking me to play music more so she can hear it. It just gives the music more life and makes the listening experience more pleasurable. We both love it. It has really softened up the "dissective" sound of my benchmark dac-1 too.
The amp runs cool at normal levels so I will swap it out with my Yamaha amp for the party and use it for personal listening only. On the side I have ordered another set of PCBs, a tranny and heatsinks to make a better version over the coming months.
Bas, thanks for your reply. It's nice to know that someone has pushed the chips without ill effect, but I'll stick with my old amp for the party to be safe.
Regards,
Greg.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.