Hi,
are your speakers 6ohm or 4 to 8ohm?
You need this information to allow a transformer to be chosen to power the amplifiers.
A two way 6ohm speaker will have a 6ohm Bass/Mid driver and a 6ohm treble driver.
A two way 4 to 8ohm speaker will have a 4ohm Bass/Mid driver and an 8ohm treble driver.
You can measure Re of the drivers to determine the nominal driver impedance. Re ~ 70% to 80% of Rnominal
are your speakers 6ohm or 4 to 8ohm?
You need this information to allow a transformer to be chosen to power the amplifiers.
A two way 6ohm speaker will have a 6ohm Bass/Mid driver and a 6ohm treble driver.
A two way 4 to 8ohm speaker will have a 4ohm Bass/Mid driver and an 8ohm treble driver.
You can measure Re of the drivers to determine the nominal driver impedance. Re ~ 70% to 80% of Rnominal
I have single LM4780 driving Energy eXL-16 bookshelf.
50 - 20 KHz +/-3dB
Tweeter(19mm)
Aluminum Dome with Cloth Suspension
Mid Driver(16.5cm)
Mica/polypropylene
XO cut 2.2KHz
91dB
8-ohms
15 - 110 watts
Not really satisfied with the sound, may be cause of the eXL-16's XO(full of electrolytic caps). Shall I mod it?
So getting a new speaker, either Wharfedale Diamond 10.0(6ohm nominal) or Dali Concept 1(8ohm nominal)/Dali Ikon 1(6ohm nominal). The only speakers I can get locally.
Here is the Wharfedale 9.1 measurements. I guess the drivers should be the same as newer Diamond 10.1.
http://www.stereophile.com/content/wharfedale-diamond-91-loudspeaker-measurements
50 - 20 KHz +/-3dB
Tweeter(19mm)
Aluminum Dome with Cloth Suspension
Mid Driver(16.5cm)
Mica/polypropylene
XO cut 2.2KHz
91dB
8-ohms
15 - 110 watts
Not really satisfied with the sound, may be cause of the eXL-16's XO(full of electrolytic caps). Shall I mod it?
So getting a new speaker, either Wharfedale Diamond 10.0(6ohm nominal) or Dali Concept 1(8ohm nominal)/Dali Ikon 1(6ohm nominal). The only speakers I can get locally.
Here is the Wharfedale 9.1 measurements. I guess the drivers should be the same as newer Diamond 10.1.
http://www.stereophile.com/content/wharfedale-diamond-91-loudspeaker-measurements
Last edited:
Yes.
What does that mean? Can this be summed up as 15-100 W? How did you arrive at those figures? Most Gainclone owners have 20-25 W of output power per channel. Most of them believe they have 68 W and hardly any of them notices the difference or needs more power than he has. The few that do need more, won't even be satisfied with 100 W per channel.
I doubt that you will be able to fit even a single channel into that. My 2 x 3,5 W LM1876 hardly fits into a 100x100x63 mm case with power supply and it only has a single pair of input connectors. Maybe you can achieve such a small thing, if you opt for a class D amp with external power supply.
do you know Rth s-a of this sink?
How many 3886 chipamps will be fitted to this heatsink?
Because most of them use 2x18 V transformers and 8 Ohm speakers. 30 W is what you get from such a configuration with a stable power supply. Gainclones usually have low capacitance PSUs, which does not create a stable power supply and limits the output power accordingly.
Seems you are listening at very low power levels. Even the over-optimistic Overture Design Guide calculates 1 K/W for 150 °C junction with 36 VDC and that is with 8 Ohm speakers.
Short answer to OP's question:
Yes, your 36V LM4780 can drive your 6ohm, 86dB speakers.
It should be able to peak at 40W/ch. Your heatsink can't handle 40W continuous but no problem, you (and nearly everyone else) don't listen at high power levels anyways. The Wharfedale's should peak at 102dB with your amp, plenty loud for a small room.
The bigger question is will you like the sound? Can't help you with that. It's a matter of your entire audio chain. Just audition the amp+speaker in your normal listening environment and see.
Yes, your 36V LM4780 can drive your 6ohm, 86dB speakers.
It should be able to peak at 40W/ch. Your heatsink can't handle 40W continuous but no problem, you (and nearly everyone else) don't listen at high power levels anyways. The Wharfedale's should peak at 102dB with your amp, plenty loud for a small room.
The bigger question is will you like the sound? Can't help you with that. It's a matter of your entire audio chain. Just audition the amp+speaker in your normal listening environment and see.
If the power supply is up to it, it should peak at more than 87 W.
Right. And with ±36 V and 6 Ohm you get something like 50 W of heat dissipation per channel. Two channels like that on a single heatsink will limit the average listening level severely.
That is true. Except when we use our equipment for parties or as PA system we hardly even use a single Watt at average. And with the above configuration you can achieve healthy average levels of around 2 x 6 W (LM3886T), 2 x 3 W (LM3886TF) or 2 x 1 W (LM4780). If it takes so little power to be happy, why not reduce the power supply to ±25 V, get 2 x 40 W into 6 Ohm out of it and be safe with that heatsink even when the weather is hot? As positive side effects you save electricity, rule out amp destruction by operating error and heat up your room less which is good during summer.
have a look at the graph: chip dissipation Vs output power into 4 r0 and 8r0 loads.
You will see that maximum dissipation occurs roughly at 2/3 maximum output power. You will also see that dissipation falls only slightly from that peak to ~80% of maximum dissipation over the power output range of 30% to 100% of maximum output power. This tells you that the chip demands lot's of heatsink capability if the output is anywhere between 20W and 60W for a 60W into 4r0 amp and for a 60W into 6r0 amp and for a 60W into 8r0 amp.
Now look at dissipation when output power is 10% & 20% of maximum output power. The dissipation when outputting just 6W from a 60W amplifier is much higher than most commentators seem to want to refer to.
Even at 2W output (getting difficult to read) there is still a significant heatsink requirement.
Use National's table of heatsink vs operating conditions and double it.
If you have 2 chipamps on the sink then quadruple it.
If you use a dual channel chip then doubling may not be sufficient.
Hi Andrew,
Good points. The OP and other interested chipamp builders should read it. However, the OP's point is that his chipamp is running just fine (< 70C) at his typical listening levels. Therefore, the answer to his question is still a yes, his chipamp can drive his desired 86dB 6R Wharfedale speakers. If he dislikes the sound (as he seems to do so with his current speakers), then it is not likely a question of the amp's drive capability but something else in his signal chain.
Hi Pacificblue,
Reducing VCC to one's chipamp is an option when the power is not required. I like to run mine with 22V transformers, yielding 30Vdc so I can use 35V caps with a decent margin. Most builders here seem to want that magical 60W-70W capability so they run higher voltages. Some even claim the sound is better (maybe due to extra headroom?). Let's just give the OP the right numbers and let him figure out if his chipamp is suitable for his desired speakers. Those of us who can work the numbers usually have no problems building reliable chipamps.
Good points. The OP and other interested chipamp builders should read it. However, the OP's point is that his chipamp is running just fine (< 70C) at his typical listening levels. Therefore, the answer to his question is still a yes, his chipamp can drive his desired 86dB 6R Wharfedale speakers. If he dislikes the sound (as he seems to do so with his current speakers), then it is not likely a question of the amp's drive capability but something else in his signal chain.
Hi Pacificblue,
Reducing VCC to one's chipamp is an option when the power is not required. I like to run mine with 22V transformers, yielding 30Vdc so I can use 35V caps with a decent margin. Most builders here seem to want that magical 60W-70W capability so they run higher voltages. Some even claim the sound is better (maybe due to extra headroom?). Let's just give the OP the right numbers and let him figure out if his chipamp is suitable for his desired speakers. Those of us who can work the numbers usually have no problems building reliable chipamps.
that is hot for heatsink.
Does he have any idea of Tc and Tj during operating conditions?
Can you estimate what the peak current limit is adjusted to, when Tc~70degC? I can guess it is no longer 7Apk. Is it 6Apk or 5Apk or ? What does National say about spike limits and protection limits for elevated temperature operation? Very little !
Hi,
the Rth s-a (thermal resistance between sink and ambient air) is how well the sink transfers heat from the whole back face of the sink to the air surrounding the whole sink when the difference in temperature between the back face and ambient air is equal to the specification. The spec usually calls for delta T s-a = 70 to 75 Centigrade.
Most 1.3C/W heatsink that are specified in the conventional way will have similar volume and similar fin area. Fan assisted heatsinks are completely different.
As an aside, an external heatsink will have a very different Ta from one enclosed inside an improperly ventilated box.
the Rth s-a (thermal resistance between sink and ambient air) is how well the sink transfers heat from the whole back face of the sink to the air surrounding the whole sink when the difference in temperature between the back face and ambient air is equal to the specification. The spec usually calls for delta T s-a = 70 to 75 Centigrade.
Most 1.3C/W heatsink that are specified in the conventional way will have similar volume and similar fin area. Fan assisted heatsinks are completely different.
As an aside, an external heatsink will have a very different Ta from one enclosed inside an improperly ventilated box.
The trouble with heatsink specs is that dealers and some manufacturers have an interest in giving good looking numbers. So they only tell you a thermal resistance, but they do not tell you to which power dissipation it belongs.
To give an example a heatsink 160 x 40 x 50 mm is sold by a dealer here as 1,2 K/W. If you look that same heatsink up in one manufacturer's catalog, you will find that it has 1,2 K/W, when the power to dissipate is 60 W. If you only need to dissipate 40 W, that same heatsink is only specified for 1,32 K/W. Another manufacturer shows a hard to read graph that seems to indicate 1,4 K/W without any further information except that "all heatsinks are measured at practical temperatures and power dissipations".
Such methods explain why you can find heatsinks with the same thermal resistance in K/W that are not comparable at all.
To give an example a heatsink 160 x 40 x 50 mm is sold by a dealer here as 1,2 K/W. If you look that same heatsink up in one manufacturer's catalog, you will find that it has 1,2 K/W, when the power to dissipate is 60 W. If you only need to dissipate 40 W, that same heatsink is only specified for 1,32 K/W. Another manufacturer shows a hard to read graph that seems to indicate 1,4 K/W without any further information except that "all heatsinks are measured at practical temperatures and power dissipations".
Such methods explain why you can find heatsinks with the same thermal resistance in K/W that are not comparable at all.
there's the delta T s-a = 72Cdegrees. You have to read the manufacturer's spec sheet.
...if the manufacturer gives those specs. That's what my former post was about. Not all dealers and manufacturers reveal the required data.
The figures I gave were for an identical heatsink from Alutronic (1,2K/W@60W and 1,32K/W@40W) and Fischer (~1,4K/W@unknown power). The dealer sells that heatsink as 1,2K/W@unknown power. And it is not always easy to find out where a dealer buys his stuff which makes it hard to get by a manufacturer's spec sheet.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.