43v SMPS for dual bridge mono LM4780

[juan_caliente]

Hello Everyone,

I recently purchased a SMPS from Ebay: http://cgi.ebay.com/ws/eBayISAPI.dl...MEWN:IT&viewitem=&item=230101751981&rd=1&rd=1

I was hoping to use it to power my two lm4780 boards from Chipamp.com.

Has anyone ever used this smps for the dual mono application. I also want to build a drv134 bridging module so that both boards will have a mono output. This is going to be used to power a bass guitar monitor speaker, probably an Omni 10, 12 or 15, for practice with my pc running guitar rig 2.

Thanks for your assistance.

Sincerely,

John

 

[Minion]

Well First off that PSU will not supply enough Current for a 2 LM4780...Actually it will barely supply enough current for Half of a LM4780.....

Also +/-43v Rails is actually Too high of a Voltage for the LM4870, It states in the Datasheet that the Absolute Maximum it can stand (with a signal applied) is +/-42v and at that Voltage it will run very hot...+/-35V is Probably the Max that I would use but +/-28v what I aim for.....

If you had one of those PSU"s for each Half of the LM4870 (the LM4870 is actually 2 LM3886 in One Package) then it would work but you would probably have to regulate the Voltages down closer to +/-35v for best performance useing maybe an Off Board Linear regulator......


Good Luck

 

[AndrewT]

Hi,
that ebay site is telling lies.

I see they cover themselves by offering a modifcation to get the voltage down to +-30V.

For 8ohm speakers (not 4 to 8ohm) you can run @ +- 35Vdc.
for 4ohm speakers you can use +-28Vdc.
for 6ohm or 4 to 8ohm probably some supply voltage in between.

You may be able to modify the SMPS to allow either 8ohm or 6ohm speakers to be used.
But the problems go on.

I think a 3886 has a peak output current of 7Apk.
A 4870 has two 3886 inside so the total output current is 14Apk.
Fortunately the supply pins for the two internal amps are separated so you can supply each chip separately.

So four of these SMPS can supply your two chipamps with less than half the peak current that they need.

I would suggest you try to sell on the SMPS to some unsuspecting purchaser, but try to be a bit more truthful.
Then go and buy a proper transformer/rectifier/smoothing capacitor PSU and do the job right.
Each 3886 can deliver upto 60W so the 4 chips need about 1.5Times (60+60+60+60) = 360VA @ 25Vac-0 & 25Vac-0 for 8ohm speakers or 360VA @ 20Vac-0 & 20Vac-0 for 4ohm speakers.

 

[justblair]

Can's say much about the SMPS, through ignorance more than anyhting else. However I have used that particular E-bay seller without any problems. In fact just had some lm3875 chips delivered today. Its the second time that I ordered there. Stadholders prices were the best I could find. Got 4 for £9.18 this time, though I did win a pair in auction bought the other two buy it now.

The description does make it clear that the voltages are working at the absolute maximum for the chips.

I looked at this supply, but ended up going for four skynet 8080's.

I was more confident doing this as I could then use Decibel dungeons instructions to build an array i knew would work. I had my teething problems certainly, but i knew that I was not trying to achieve the impossible. I am very pleased with the SMPS amp, I have not built a transformer/rectifier/smoothing capacitor PSU as suggested by Andrew to compare with the SMPS but NUUK has tried both and recommends the SMPS.

Dont know if my experiance helps but I thought I would chip in.

 

[juan_caliente]

Thanks for all of the replys. I have spoken to a couple of members here about modifying this smps to work with a single board set, and getting a second smps for the other channel. I am sure it is going to be some work getting it all done, but I think that it is going to be fun to experiment with the smps, rather than the time tested transformer route.

Does anyone know of a good source for 1u and 2u rack mount cases? I am also building a dual channel tube blackface preamp and a acoustic 360+. I need 2 of the 1u for thoes and probably a 2u for the power amp. Has anyone ever made their own using steel from a hardware store cut to size? Mabey bolted to a frame of extruded aluminum?

Thanks again,

John

 

[justblair]

I have never made a rack, but I have seen a DIY rack made from MDF on the internet somewhere.

It was months ago when I saw it. But a bit of googling might be worth it. It is likely I saw a link to it through hackaday, cause its the sort fo thing I would have come accross through browsing rather than by looking for. If I recall correctly it was used for mounting a server rather than audio equipment.

Sorry I dont have time to search it out again. I have some heavy work on right now. If you have no joy by hte weekend I will see if I can find it for you.

Blair

 

[space]

With some hard work you probably can make the SMPS deliver lower voltages but you can't (easily and economically) make it deliver more current. I would drop that SMPS and look for something with the right spec (~+/-30V ~10A).

space

 

[kscharf]

I don't know that I'd want to use a SWITCHING power supply in an audio amplifier application anyway. Asking for noise problems IMHO.

 

[justblair]

I don't know that I'd want to use a SWITCHING power supply in an audio amplifier application anyway. Asking for noise problems IMHO.

No noise issues with mine, silent when its supposed to be. Very detailed, controlled sound.

I have 4 skynet 8080's in series and parrallell to give +/-24v 6a. I folllowed NUUK's advice to make it and I am very happy.

 

[FastEddy]

I use two +48 VDC switchers and my own "secret sauce" humbucking torrid line filters to get a really nice quite source. Result is +/- 48 VDC rails driving two 500 watt modular MOSFET amps (into 4 ohms) with reduced power of ~ 220 watts per channel = a very cold running amp for 24/7/365 usage without ever needing a fan ... This one is a work in process ...

 

[AndrewT]

Hi Fasteddy,
500W into 4 ohms from +-48Vdc supply rails.
Are these amps bridged?

 

[FastEddy]

Make that ~ 250 watts from each amp module / each channel = ~ 500 Watts in total output. Yes, these can be combined or bridged, but for the purpose, not.

(The alternate configuration, +/- 60 VDC from dual 60 VDC supplies is closer to the total "One Horse" configuration = ~360 watts per channel = ~ 725 Watts if bridged = ~one horse power.
From the module maker: http://aussieamplifiers.com/nx500ln.htm =
Specifications
500 Watts RMS into 4 Ohms RMS Min Continuos
300 Watts RMS into 8 Ohms RMS Min Continuos
Frequency Response: 5hz to 200khz
SIgnal to Noise Ratio -125db) ... )

 

[!]

No noise issues with mine, silent when its supposed to be. Very detailed, controlled sound.

I have 4 skynet 8080's in series and parrallell to give +/-24v 6a. I folllowed NUUK's advice to make it and I am very happy.

It's quite possible to like the sound when it's powered by a SMPS. There's another factor, that the PSRR of these chipamps drops quite a lot when dealing with high frequency noise like any switcher will generate. For example the SMPS might switch at about 50KHz and have a quite broader band but to focus on 50KHz alone, PSRR (V-) is already down from ~ 120dB to 45dB.

Given enough time and a scope, appropriate filters could be made to mitigate this, but in general the noise is approaching something audible from the amp.

I'm not saying a SMPS can't work, only that unless you are actively seeking to alter accurate amp output, it is technically best to stick with the traditional low freqency supply then if you want to tweak that, put more filtration like a linear or cap multiplier stage in it.

I suppose what I'm really getting at is if someone had a beefy switcher and wanted to save some $ or just play around, it makes more sense to try it than for someone who has yet to buy a supply and isn't looking to alter the expected sound of a gainclone.

 

[justblair]

It's quite possible to like the sound when it's powered by a SMPS. There's another factor, that the PSRR of these chipamps drops quite a lot when dealing with high frequency noise like any switcher will generate. For example the SMPS might switch at about 50KHz and have a quite broader band but to focus on 50KHz alone, PSRR (V-) is already down from ~ 120dB to 45dB.

Given enough time and a scope, appropriate filters could be made to mitigate this, but in general the noise is approaching something audible from the amp.

I'm not saying a SMPS can't work, only that unless you are actively seeking to alter accurate amp output, it is technically best to stick with the traditional low freqency supply then if you want to tweak that, put more filtration like a linear or cap multiplier stage in it.

I suppose what I'm really getting at is if someone had a beefy switcher and wanted to save some $ or just play around, it makes more sense to try it than for someone who has yet to buy a supply and isn't looking to alter the expected sound of a gainclone.

I didn't understand much of that !. Suppose I were to go the low frequecy supply route... Would my sound get better? Its already very good, but I am open to suggestion. NUUK on his site reckoned to the SMPS being the better sounding option. I'm a mere sheep on this one. I was building my first, I just copied the easiest to understand site I could.

I can understand the PSRR is reduced (I'll take your word for it...) but is it not reduced enough not to be an issue anymore? I'm trying my best to learn but as you can probably tell I have a bit to go...

Regards

Blair

 

[!]

Above "down from ~ 120dB to 45dB" is wrong, should have been roughly ~ 100dB to 45dB, drop.

 

[justblair]

Above "down from ~ 120dB to 45dB" is wrong, should have been roughly ~ 100dB to 45dB, drop.

What effects would this have... I know I'm asking dumb questions... But I dont know why they are dumb

Is there something in the sound I should be looking for?

Thanks for your patience

Blair

 

[!]

I didn't understand much of that !. Suppose I were to go the low frequecy supply route... Would my sound get better? Its already very good, but I am open to suggestion. NUUK on his site reckoned to the SMPS being the better sounding option. I'm a mere sheep on this one. I was building my first, I just copied the easiest to understand site I could.

Obviously audio is a subjective thing. I can't tell you what "sounds better" to you. I can instead tell you that SMPS introduces distortion and I think most people will tend to want to avoid that, while[/I} striving for the sound they like. In my opinion the first gainclone one makes should not use a SMPS, and then if one wants to swap in a SMPS to compare the sound, with the same or an identical amp, they will understand more how it effects the sound. I also encourage such tests, including linear regulation of an unreglated supply instead of or in addition to the SMPS trial, but I think it only fair to start out with the known universal an technically correct way to get the best agreed upon, and measurable, sound, which is not using a SMPS. After you achieve that, you have a frame of reference to what others are referring to from similar gainclones and can better contrast changes from SMPS trial.

I can understand the PSRR is reduced (I'll take your word for it...) but is it not reduced enough not to be an issue anymore? I'm trying my best to learn but as you can probably tell I have a bit to go...

Regards

Blair

I don't know what you're trying to say with your double negative statement.

You don't want PSRR reduced at all, ideally it is as high as possible on any amp. It is bad for it to be 45dB @50KHz "IF" the supply is at 50KHz (which it may not be, but in roughly this 20-100KHz range is likely and the PSRR drops quite a bit more as frequency climbs). It is an issue. Clean power is part of the reason why some amps sound different, sometimes better, than others.

All I'm really saying is, you owe it to yourself to hear what the amp was meant to sound like before adding or changing things that have a strong potential to interfere with fidelity. This is not a new topic, SMPS are known to be a bad thing to use for an amp for many years by the overwhelming majority of audiophiles. Its only a certain niche thing, if you like what it sounds like. Most people apparently don't.

 

[!]

What effects would this have... I know I'm asking dumb questions... But I dont know why they are dumb

Is there something in the sound I should be looking for?

Thanks for your patience

Blair

You should be looking for a non-SMPS so you can hear the amp without it being distorted by the power supply.

Then you know what you have for an amp. If it doesn't sound right, adding distortion is usually not the answer unless it's getting a signal from an electric guitar, or it's a certain type of distortion like from a tube preamp.

The answer is do not use a SMPS. Build it correctly first then listen to it. Now after you have listened to it in it's technically correct form with as little distortion as possible, then add the distorting SMPS and listen again.

There's no words that will do as much as just listening. I advise NEVER EVER EVER to build something with a SMPS the first time. Only switch to a SMPS temporarily, do not do the wrong thing then ask about it, do the right thing the first time so you know what you were supposed to hear, what the amp sounds like when it's not being distorted by the PSU. I can't subtractively tell you what noise to ignore or how much of a factor the amp design was in what you hear because these are variables.

 

[justblair]

Hmmm

I will perhaps try a normall one in the future, fortunately I like the results I have so far, So I can try out other stuff later.

NUUK though does seem to prefer smps, so there is one or two different opinions there. (isn't there always!)

Thanks for the explanation... Even if I only half understand it probably.

Blair

 

[!]

If it's opinion, one opinion is held by the vast majority in addition to being technically correct. SMPS is considered a very bad thing to power an amp with by the entire world of audiophiles except for a very few who either sell SMPS, finished products that were cheaper because they used a SMPS, or those who have never heard optimal versions of both PSU types.

Cost or space and weight constraints are the only real reasons to use one. Deliberately causing the power rails to be dirty instead of using legitimate audio-related changes is like rejecting everything you were trying to learn and achive and just spinning the wheel of chance to see if you are lucky.

Reasons to promote use of SMPS is another matter, anyone who is selling SMPS might promote them for uses even if they aren't good for the purpose. There are only a handful of people who really seem to think SMPS seem ok. Nuuk is one of those very few and just happens to hang out here, so you are getting a very lopsided perspective. Nuuk may honestly prefer the sound of SMPS, but most people do not and Nuuk should know well enough to make that clear to anyone before having them only hear a distorted amp without a way to hear what it was supposed to sound like.