Ok, I reacently picked up two LM4780 chips from ebay and thaught that I would make a amp with a bit more power then my lm3875. Is there a good idea to use one bridged chip per channel with a 4 ohm load? Or do I need to use them in parallel config instead?
And one other thing i might add. Do you think the regulated powersupply by CarlosFM using lm338 could be used for this settup aswell, though with a 300VA trafo?
And one other thing i might add. Do you think the regulated powersupply by CarlosFM using lm338 could be used for this settup aswell, though with a 300VA trafo?
you can bridge or parallel into a low value load -- just have to be mindful of the maths when it comes to PDMAX.
i would prefer that you go parallel.
there really is no need to regulate, you just turn decibels into heat -- there are studies which indicate that louder is better 🙂
i would prefer that you go parallel.
there really is no need to regulate, you just turn decibels into heat -- there are studies which indicate that louder is better 🙂
Ok, but since I already got that regulated powersupply for my lm3875 I was hopping I cpould use it for my 4780 instead. But is the 5 amperes that the lm338 is good for enough to drive the two chips in parallel config?
(and btw, whats PDMAX?)
(and btw, whats PDMAX?)
use the design tool on Nat Semi's website:
http://www.national.com/appinfo/audio/files/Overture_Design_Guide15.xls
http://www.national.com/appinfo/audio/files/Overture_Design_Guide15.xls
thank so far for all the help!
I cant really understand the designtool, I cant see how much current it draws when paralleled into a 4ohm load with a suuply voltage of about 25-30 volts?
I cant really understand the designtool, I cant see how much current it draws when paralleled into a 4ohm load with a suuply voltage of about 25-30 volts?
Download the applications note AN1192 (BPA-200) from National, This will outline almost everything you need to know regarding engineering for bridged/parallel of the LM3886 (LM4780)
Hi,
a 4ohm load will draw about 6Apk from an AC supply of 24Vpk. That's equivalent to 72W from the pair of chipamps so they should be able to cope.
BUT
a 4ohm speaker probably has a minimum impedance of nearer 3ohms and that would draw 8Apk.
Some commentators say that the maximum current draw can be a few times more than plain ohm's law predicts.
If this becomes more than 1.3times 6Apk then the problem gets even worse.
Regulated is going to become quite a development exercise.
Chipamps into low impedance loads is not an easy bolt it together option.
Take care.
Bridged, forget it!
a 4ohm load will draw about 6Apk from an AC supply of 24Vpk. That's equivalent to 72W from the pair of chipamps so they should be able to cope.
BUT
a 4ohm speaker probably has a minimum impedance of nearer 3ohms and that would draw 8Apk.
Some commentators say that the maximum current draw can be a few times more than plain ohm's law predicts.
If this becomes more than 1.3times 6Apk then the problem gets even worse.
Regulated is going to become quite a development exercise.
Chipamps into low impedance loads is not an easy bolt it together option.
Take care.
Bridged, forget it!
If you can lay your hands on 2x 50w zener diodes cheep, you could try the regulated power supply I used for my paralleled LM4780 amplifier. (I recently posted 2x photos in the chip amp photo gallery)
It is very simple (extra 3 parts) but you would also need 2 separate windings on your transformer or 2x transformers (as there are no equivalent negative regulators)
It uses the 5A (7A-limit) LMS 1585-1.5v regulator with a 5k trimmer in the ground lead to set the output to 35v.
An 8.2v/50w zener diode across the LMS 1585 limits its maximum differential to the zener voltage, (Under the regs 13v limit).
The LMS 1585 is then protected from capacitor charging inrush, and placing a short on the PS and turning on the power only blows my 5A DC protection fuses (No Damage to the PS)
With full power both channels into a 4ohm load it can lower the PS input voltage to below the output voltage (I have 43v input) so the PS becomes unregulated at this point.
But as the LMS 1585 is a low dropout regulator, only 1.1v is lost so almost the full PS voltage is still available although now unregulated.
It is essential to have large capacitors (I used 10,000uf) on the input and output of the regulator to handle the peak currents.
Remember the output current to the speaker is shared from both the positive and negative supplies (3A @ 2x35v = 210W)
I find using the amplifier to clipping point listening to music (very loud) and monitoring the PS and output on a dual trace scope the PS never looses regulation.
I salvaged the 50w zeners from some old power supplies given to me.
It is very simple (extra 3 parts) but you would also need 2 separate windings on your transformer or 2x transformers (as there are no equivalent negative regulators)
It uses the 5A (7A-limit) LMS 1585-1.5v regulator with a 5k trimmer in the ground lead to set the output to 35v.
An 8.2v/50w zener diode across the LMS 1585 limits its maximum differential to the zener voltage, (Under the regs 13v limit).
The LMS 1585 is then protected from capacitor charging inrush, and placing a short on the PS and turning on the power only blows my 5A DC protection fuses (No Damage to the PS)
With full power both channels into a 4ohm load it can lower the PS input voltage to below the output voltage (I have 43v input) so the PS becomes unregulated at this point.
But as the LMS 1585 is a low dropout regulator, only 1.1v is lost so almost the full PS voltage is still available although now unregulated.
It is essential to have large capacitors (I used 10,000uf) on the input and output of the regulator to handle the peak currents.
Remember the output current to the speaker is shared from both the positive and negative supplies (3A @ 2x35v = 210W)
I find using the amplifier to clipping point listening to music (very loud) and monitoring the PS and output on a dual trace scope the PS never looses regulation.
I salvaged the 50w zeners from some old power supplies given to me.
Ok, thank you all very much for the info. But if
"a 4ohm load will draw about 6Apk from an AC supply of 24Vpk. That's equivalent to 72W from the pair of chipamps so they should be able to cope." then shouldnt it draw less with my closer to 30 VAC supply? It´s not that important to me to have a regulated supply but it would be nice to compare the two options
"a 4ohm load will draw about 6Apk from an AC supply of 24Vpk. That's equivalent to 72W from the pair of chipamps so they should be able to cope." then shouldnt it draw less with my closer to 30 VAC supply? It´s not that important to me to have a regulated supply but it would be nice to compare the two options
Hi J,
what voltage are you running your chip from?
I assumed that you could achieve about 24Vpk from +-30Vdc.
what voltage are you running your chip from?
I assumed that you could achieve about 24Vpk from +-30Vdc.
I got a transformer with dual 30 volts secondaries, after rectifying that should be around 40 volts right? That may be a bit to much for the 4780 chip so then I regulate it down to whatever but atleast 34 volts since the lm338 requires atleast 6 volts regulation (i think). So maybe 30-34 volt supply to the chips. And again lm338 max output 5 amps, and I want to run one chip in paralell config per channel and there will be two channels, so two chips to drive. I hope you understand, cause i lost myself somewhere...
A 4ohm load is going to demand a lot more current than an 8ohm load.
A 5A limit from the regulator is not, in my opinion, going to deliver the goods.
A 5A limit from the regulator is not, in my opinion, going to deliver the goods.
Also a 2x40 VDc supply is a bit to much for a 4 Ohm load. I suggest looking for a transformer with a 2x22 Vac secundairy. Given the very good PSRR of the Overture chips, I do not see the need of a regulated supply.
I've bridged a LM3875 chip amps (Audiosector kits - Dual Mono) running 6 ohm (nominal) speakers without any sonic or heat problems. Admittedly, listening levels are always kept at low to medium levels and the chips are well heat sinked.
- Status
- Not open for further replies.