Hello to all,
I have develop very recently this PCB which use 5 LM3886 they can be installed in the PCB all in gainclone/parallel,fully independent channel's or mixed
configuration, and I want to develop a simetric regulator which can power this, so in terms of amps my target is 20A.
About the circuit for the regulator I am thinking of using power op amps as current source biasing some zeners which will set
the voltage on the output, after the zeners I will use a driver and then two devices in parallel for the output section, I am thinking
of using TIP35/36 but would like to ear some suggestions.
Regarding output overcurrent I am thinking of standart VI limiter but I would like to add also a protection, for creating a safe operation area, this by trip the
circuit if the differential voltage between in/out be greater than 7 volts this for temperature don't rise to fast, but this might bring some problems, for
now I would like to if operating an LM3886 and then remove one side of the supply does it burn/short or no problem of removing one side of the supply?
I hope this resonate with more people as this regulator would also work for any amplifier, after I develop this regulator,I will make it available alongside
with my other modules like this one, you can check it here:
https://www.ebay.de/usr/rocksandsound
Regards
Congratulations and I hope the regulator you develop, turns out to be rugged and high performance and exceptionally robust!!
You are the developer of this regulator -- so I will not presume to offer design assistance or circuit schematics. I'm not the developer, you are. Therefore I will confine myself to typing out a few mild suggestions and ideas to study (or reject) on your own. Such as:
20 Amperes for the +V rail and another 20 Amperes for the -V rail, may require quite a few discrete transistors operated in (ballasted) parallel, for each rail, to achieve a large margin of safety.
Considering that the LM3886's metal leadframe interface to the heatsink on the "back side" of the LM3886 package, has a contact area of about 3x the contact area of a discrete transistor TO-247, and also considering that the regulator's (Vout - Vin) is approximately (1/3) of the Vce of the internal transistors inside the LM3886 . . . . . then perhaps you need approximately as many TO-247's as you need LM3886's. If N=6 LM3886 amplifier ICs, you might decide that you need N=6 PNP discrete power transistors in the negative power supply rail of the voltage regulator, and you might also decide you need N=6 NPN discrete power transistors in the positive power supply rail of the voltage regulator.
But, I repeat myself, you are the developer . Not me. You make the decisions and you live with the consequences. Best of luck and congratulations on choosing a challenging project!
Well, for sure would like that more people would be interested on this and actually I will make it available like all my other diy modules.
Since very young I really enjoy music but couldn't afford it so I started exploring, now that I have some years on this, I want to use my
experience and knowledge to try do the best as possible in áudio modules, I want that People can experience really great product without
staying broke but it needs to put their hands on it to build "diy" if I were developping a complete comercial product for a brand I wouldn't do it
for free "which I can do it also" but here like all my modules they are diy kits so I hope that I am adding solutions/helping to the diy community.
About designing, I am looking for suggestions, tips or shared knowledge but I will develop it, I have been given suggestions to another
diy members and of course I am not charging anything for that.
Regards
Five LM3886 would be able to draw 55 A total, assuming you get the 'typical' data sheet performance. 35 A if you get the worst case performance. Good luck designing a regulator that can deliver that.
You then want to add a current limiter? I hope for you that it'll be of the fold-back type or at least have some clever circuitry in there to limit the power dissipation in the regulator during the over-current condition, because if not you're looking at dissipating around 2 kW in the regulator. Those transistors will go POOF in a hurry.
As Mark has pointed out above, designing such a regulator is no small task.
I too am surprised that this thread hasn't been moved to the Vendors' Bazaar.
Tom
You then want to add a current limiter? I hope for you that it'll be of the fold-back type or at least have some clever circuitry in there to limit the power dissipation in the regulator during the over-current condition, because if not you're looking at dissipating around 2 kW in the regulator. Those transistors will go POOF in a hurry.
As Mark has pointed out above, designing such a regulator is no small task.
I too am surprised that this thread hasn't been moved to the Vendors' Bazaar.
Tom
Hi Mark thanks for energy, I will do my best, for now I would like to share some rough numbers about this and listen some opinions on those from the Community.
Considering that if for example we have some big speakers with low impedance of 2 Ohm and the voltage after the regulator being 40V I conclude that the peak current would be 20A AC leadind to DC current of 14,18 amps on the regulator,in meantime I would like to mention that this would represent 397 watt RMS so power levels that probably no one would listen in Home.
By having an input voltage of 45V on the input of the regulator I conclude a total wasted energy per rail of 70watts, TIP35/36 do have a collector power dissipation of 125 watt on max that is the reason I thought on 2 devices, "what do you think Mark?" with this data.
I do pretend to install a circuit that activate a protection if the differential voltage between IN/OUT be greater than 7 volts does someone can advise something to implement this? Does anyone know if one of voltage rails being removed from an LM3886 in operation, burn or get damaged or in other kind of class AB amplifier?
Next week I will finish skematic and share it here for being more easy to view all this.
Regards
Don't know about that. I use Sound Lab electrostatic panels. They are rather inefficient, and difficult loads. Present amplifiers are Marantz MA9-S2, which are 300W/channel at 8R, and 600W/channel at 4R. Minimum speaker impedance drops down to around 2R at some frequencies. If I turn them up loud I have to watch the power amps to avoid clipping. And of course, the power amps have speaker protection circuitry, more or less similar to protection circuitry in other amplifiers of similar power ratings.
I would question the need, or percieved need, for a regulated supply in a power amp in the first place.
Most amps do not use them and for good reasons.
Additionally, the LM3886 has 120 dB of PSRR - Power Supply Rejection Ratio, which basically means its output is not affected by rail voltage fluctuations, until clipping.
Attached is a power supply design application note from Texas Instruments, maker of the LM3886, which contains a lot of info on the subject.
This will help you design a good unregulated supply if you decide to go that route.
Also suggest reading chapter 9 of The Audio Amplifier Design Handbook by Douglas Self, which has great practical info on the advantages / disadvantages of both.
Most amps do not use them and for good reasons.
Additionally, the LM3886 has 120 dB of PSRR - Power Supply Rejection Ratio, which basically means its output is not affected by rail voltage fluctuations, until clipping.
Attached is a power supply design application note from Texas Instruments, maker of the LM3886, which contains a lot of info on the subject.
This will help you design a good unregulated supply if you decide to go that route.
Also suggest reading chapter 9 of The Audio Amplifier Design Handbook by Douglas Self, which has great practical info on the advantages / disadvantages of both.