The Opamp thread - Page 267

Quote:

Originally Posted by kiteki 

Thanks for the swift answers Amarok!!  The loaded OPA627 doesn't sound very good.  I will experiment with voltage supply in OPA627 and OPA827 to see if I can hear the difference!

 

Edit:  My amplifier won't turn on when 4x 9V are connected.

how exactly have you connected them? have you taken the center tap (ground reference) from the middle, or just connected the 4 in series and connected across it?

Like that.

I could smell something burning, lucky it was an empty saucepan in the kitchen ^^

Kiteki,

Quote:

Originally Posted by kiteki 

Is there any reason why a class-A biased OPA627 with some silly resistors would sound worse than a regular OPA627?  I don't like it.

 

When the PDF says absolute max +-18V that's 4x 9V batteries right?  What happens if I overclock the OPA627 to +-19.2V?

I'm not the thread guru, by a long shot, but I'm pretty sure that the +-18V absolute supply voltage spec is reached with only two 9V DC batteries in series, NOT four.  And, except for so-called rail-to-rail op-amps, where output voltage is nearly as high as supply voltage, most datasheets will show performance curves that are limited to a supply voltage no higher than 15V - because the op-amp sounds better when operating at something LESS than the absolute maximum allowable supply voltage.

The Meier Audio Stepdance, for example, uses OPA1611s, rated with an absolute maximum supply voltage of +-18V, but its designer documents that external power supplies should be limited to 15V DC. 

The iBasso PB2 can be used with various op-amps, but its designer documents that external power supplies should be limited to 16V DC.  The iBasso's internal battery is only 12.6V when fully charged - it's a 3-cell Lithium-ion pack (4.2V per cell, when fully charged).

The RSA SR71B's internal battery is only 16.8V when fully charged - it's a 4-cell Lithium-ion pack.

I really don't think you should use the 36V serial battery (your first picture, above).   I realize that the 54V pack was meant to be a joke.  

Comments from anyone else?

Thanks,

Mike

stupid question, Yulong Sabre?

@kiteki

sounds like something a few engineering buddies did, but it involved charging a rather large cap with every battery in the class, and then smacking it onto a switchboard. Ever heard of a University laboratory tripping? Yeah me neither until then.

Quote:

Originally Posted by zilch0md 

Kiteki,

 

I'm not the thread guru, by a long shot, but I'm pretty sure that the +-18V absolute supply voltage spec is reached with only two 9V DC batteries in series, NOT four.  And, except for so-called rail-to-rail op-amps, where output voltage is nearly as high as supply voltage, most datasheets will show performance curves that are limited to a supply voltage no higher than 15V - because the op-amp sounds better when operating at something LESS than the absolute maximum allowable supply voltage.

 

The Meier Audio Stepdance, for example, uses OPA1611s, rated with an absolute maximum supply voltage of +-18V, but its designer documents that external power supplies should be limited to 15V DC. 

 

The iBasso PB2 can be used with various op-amps, but its designer documents that external power supplies should be limited to 16V DC.  The iBasso's internal battery is only 12.6V when fully charged - it's a 3-cell Lithium-ion pack (4.2V per cell, when fully charged).

 

The RSA SR71B's internal battery is only 16.8V when fully charged - it's a 4-cell Lithium-ion pack.

 

I really don't think you should use the 36V serial battery (your first picture, above).   I realize that the 54V pack was meant to be a joke.  

 

Comments from anyone else?

 

Thanks,

 

Mike

no, to get +/-18 you need 4 or 8 ~9v batteries (8 for a 2P4S, 4 for 4S) but the point between 2->3 is your ground reference so you have  +18v[+9v-] -> [+9v-] -0/G- [+9v-] -> [+9v-] -18v   , since that ground connection is missing youve just connected 36v across the opamp but with no ground its got nowhere to go, so wouldnt have damaged anything. sorry zilch0md hes done it exactly right, just missing ground.if it has to be batteries then he hasnt got much choice but +/-9v or +/-18v unless he wants to do a 3 channel (I think this would be silly) or rail splitter (also silly given hes got the cells to do a real return.

@AmarokCZ: I didnt look at the 627 datasheet at all and havent used one in ages, I was talking more generally. I thought it would be deeper than that though, since so many people use it basically as a buffer. the dedicated buffer chips are much better in this regard, for instance that paralleled buffer amp I posted is in class A for a decent chunk of its 15w output into expected loads. I cant remember how much exactly from memory.

not that modern class AB is bad mind you, as long as the PSU is up to it (Class AB is actually much harder on PSUs than Class A for others reading, due to the constant power draw of class A; regardless of load currents)

Thanks for the responses AmarokCZ, qusp, WiR3D, zilch0md, viveksaikia212.  I've had a LOT of fun tonight, first with really successful results with my first DIY amplifier, and then really successful results with an IEM modification too! http://www.head-fi.org/t/614513/jvc-fx500-yellow-filter-black-filter-and-cotton-removal-review

I couldn't actually hear much difference in this experiment from 2 batteries to 6, and I only used OPA552 with 6 batteries since it can safely handle +-27V.  AD797 and OPA627 at +-18V seem to sound pretty nice though.  My favorite line tonight was with AK4396 -> OPA827 (in the DAC) -> buffer -> OPA627 (+-18V) -> IEM's.

Oh, I overlooked my new MUSES!  Time to put that one in, continue drinking alcohol and just enjoy the music! =)

Qusp,

Quote:

Originally Posted by qusp 

no, to get +/-18 you need 4 or 8 ~9v batteries (8 for a 2P4S, 4 for 4S) but the point between 2->3 is your ground reference so you have  +18v[+9v-] -> [+9v-] -0/G- [+9v-] -> [+9v-] -18v   , since that ground connection is missing youve just connected 36v across the opamp but with no ground its got nowhere to go, so wouldnt have damaged anything. sorry zilch0md hes done it exactly right, just missing ground.if it has to be batteries then he hasnt got much choice but +/-9v or +/-18v unless he wants to do a 3 channel (I think this would be silly) or rail splitter (also silly given hes got the cells to do a real return.

 

[snip]

More confused than ever...

So, does this mean that if my PB2 is equipped with op-amps rated for an absolute supply voltage of +- 18V, I can feel free to connect a 36VDC external power supply?  

I suspect you will answer, "No," but I don't understand why you will say, "No" other than to advise me to stick with the manufacturer's recommendation to not exceed 16V DC at the external supply jack.

Mike

(the villiage idiot)

Mike, there are other components in there that may limit what you can safely pump in.

Quote:

Originally Posted by SpudHarris 

Mike, there are other components in there that may limit what you can safely pump in.

Thanks Nigel,

I can readily imagine that's true, and would not want to exceed the manufacturer's documented maximum permissible external supply voltage, but I was hoping to gain an understanding of how it is that kiteki can use 36V DC battery packs to supply power to op-amps that have the same +-18V absolute maximum Vs rating as many of the op-amps that I use (including the AD797 - which we both use).  

I have no problem understanding that a swing from -18V to +18v is 36V (doh!), but does this mean that if kiteki were to use only two 9V batteries in series (for a total of 18V) instead of four 9V batteries (for a total of 36V), kiteki's op-amps would only be getting HALF their maximum permissible supply voltage?  If the answer to that question is, "Yes," then does this mean that when I use a 16V external DC battery pack with my PB2, I'm only getting a swing of -8V to +8V, which is less than HALF of my op-amps' maximum permissible +-18V supply voltage?   I don't think so.  Here's why:

Jan Meier once told me that the Meier Stepdance uses a "voltage doubler" that will take an internal 9V battery up to 18V or an external 15V supply voltage up to 30V.

Maybe that's what's going on with the PB2, also - when I connect a 16V external power supply, the op-amps see 32V because of some sort of voltage doubling circuit - and maybe kiteki's amp doesn't have this voltage-doubling feature...

Mike

I'm not aware of any volting doublage, though I am interested in one since I can't fit 4x 9V batteries in a portable amplifier lol.

Do not put a class-A biased OPA627 in the wrong way, I turned on the DAC and it looked like it was going to catch on fire.

Edit:  Threw away the chip all fixed.

Two 9V batteries means only +/-8V, so approx. half of maximum voltage.

It seems that IB2 has an voltage inverter to make neagative supply (-16V) from positive external supply (e.g. on PCB there is one compoment that looks like MOSFET - this would be inverters switching transistor).

Quote:

Originally Posted by zilch0md 

Qusp,

 

 

More confused than ever...

 

So, does this mean that if my PB2 is equipped with op-amps rated for an absolute supply voltage of +- 18V, I can feel free to connect a 36VDC external power supply?  

 

I suspect you will answer, "No," but I don't understand why you will say, "No" other than to advise me to stick with the manufacturer's recommendation to not exceed 16V DC at the external supply jack.

 

Mike

(the villiage idiot)

that depends on how the power supply is handled, or more specifically, how the ground reference is generated and; as covered, whether there is any type of DC switching/upconverting performed to boost voltage, or generate the negative rail voltage. of course if the manufacturer says not to exceed 16vdc, do not exceed 16vdc, they will have reason for saying this.

the use of a voltage doubling circuit is not all that common in high end headphone audio, even portable. 1. because it is a process that generates noise, mostly switching noise (out of band noise, but noise all the same) 2. it means a virtual ground is created and this is higher impedance than a real split supply 3. most headphones are perfectly happy with the swing available from +/-8-9vdc rails, with only 600ohms needing more IMO. for instance in my portable balanced dac-.amp i'm 'only' running a 4 series connected A123 LiFePO4 for a total of +/-6v6->7vdc, this is regulated down to +/-5.9vdc with a dual mono/bipolar supply and I challenge anyone to go past 60% with HD600. these batteries are capable of 140A bursts, thus I dont use them directly lol even though their noise and output impedance is lower than the regulators (2 x LTC3032)

Jan has some clever tricks in the stepdance overall, but i'd rather balanced than pseudo balanced and i'd rather bipolar rather than pseudo bipolar, both mechanisms introduce noise and impedance

Quote:

Originally Posted by SpudHarris 

Mike, there are other components in there that may limit what you can safely pump in.

what he said

Quote:

Originally Posted by zilch0md 

 

Thanks Nigel,

 

I can readily imagine that's true, and would not want to exceed the manufacturer's documented maximum permissible external supply voltage, but I was hoping to gain an understanding of how it is that kiteki can use 36V DC battery packs to supply power to op-amps that have the same +-18V absolute maximum Vs rating as many of the op-amps that I use (including the AD797 - which we both use).  

 

I have no problem understanding that a swing from -18V to +18v is 36V (doh!), but does this mean that if kiteki were to use only two 9V batteries in series (for a total of 18V) instead of four 9V batteries (for a total of 36V), kiteki's op-amps would only be getting HALF their maximum permissible supply voltage?  If the answer to that question is, "Yes," then does this mean that when I use a 16V external DC battery pack with my PB2, I'm only getting a swing of -8V to +8V, which is less than HALF of my op-amps' maximum permissible +-18V supply voltage?   I don't think so.  Here's why:

 

Jan Meier once told me that the Meier Stepdance uses a "voltage doubler" that will take an internal 9V battery up to 18V or an external 15V supply voltage up to 30V.

 

Maybe that's what's going on with the PB2, also - when I connect a 16V external power supply, the op-amps see 32V because of some sort of voltage doubling circuit - and maybe kiteki's amp doesn't have this voltage-doubling feature...

 

Mike

I think kiteki is playing with the output stage of his dac, arent you? but yeah everything else is covered above. i've not looked at the pelican PSU, I really dont see the need for +/-16v, but hey...

Quote:

Originally Posted by kiteki 

I'm not aware of any volting doublage, though I am interested in one since I can't fit 4x 9V batteries in a portable amplifier lol.

 

 

Do not put a class-A biased OPA627 in the wrong way, I turned on the DAC and it looked like it was going to catch on fire.

 

Edit:  Threw away the chip all fixed.

haha yeah batteries will do that, no current limiting

Quote:

Originally Posted by WiR3D 

stupid question, Yulong Sabre?

 

nope, fully DIY 2 x Ackodac AKD12P, with various additions and it leaves the Yulong in the dust. well some is kit stuff, some i've started with blank PCBs and some parts i've built from scratch like the pre-regs and soon to be MCU. with the i2s fifo and titan USB-i2s convertor, its up there with 5 figure dacs and so it should be, by the time i'm done there'll be about 5k into the project and a lot of time. this is for a 4 channel and soon to be 6 channel balanced sabre (tri differential with a sabre for each Bass, Mid, Tweeter and matching amps, not using the 8 channels in the 9018) ive got about the same into the amps and speakers.

i'm not trying to save money, just build something that cannot be bought, built to taste, as well as having some fun and learning along the way.