Tonight I'm going to try this out in my Little Dot I+ headphone amp. A member of Headfi shared his experiences doing this with the same amp and claimed it went a long way in improving the performance of the op-amp. Whether or not this is true is something I'd like to hear for myself, but I have a question.

In one picture he shows a .1uf capacitor with one lead soldered to pin 4 (-V) and the other to pin 8 (+V). Seems simple enough and he said it worked. In another picture he tried using two .1uf capacitors, one with a lead to -V, the other with a lead to +V, and the two "free" leads of the capacitors were connected to each other, and a wire was used from there to a ground point.

As I only have one .1uf capacitor on hand that's suitable in quality and (more so) size for this application, I'm going to have to try it by itself first. Will either of these methods work fine for what I'm trying to achieve (stabilization of the op-amp and shunting noise to ground, according to this person), or is there a reason why one should be better than the other?

Thanks for your help.

This seems to work OK with some opamps, and not so well with others. You'll simply have to try it and see.

I've always thought the best was was to have two caps, each from the '+' and '-' supplies to ground.

This is typically called by-passing, or decoupling. Modern people call them DCAPs. Most common is to have a cap per supply leg to ground. But, in the rare instance of no ground reference in the op-amp circuit (such as non-inverting gain of 1), going from rail-to-rail with the bypass has certain advantages. In addition to help remove power supply noise, the by-pass caps can reduce distortion, especially at higher frequencies.

This is typically called by-passing, or decoupling. Modern people call them DCAPs. Most common is to have a cap per supply leg to ground. But, in the rare instance of no ground reference in the op-amp circuit (such as non-inverting gain of 1), going from rail-to-rail with the bypass has certain advantages. In addition to help remove power supply noise, the by-pass caps can reduce distortion, especially at higher frequencies.

Well I suppose I'm a modern person, so perhaps I'll adopt that terminology. :D So what you're both saying is that either way will work, but the best is probably to have two that go to ground? I just don't want to waste my time going from rail-to-rail if it's unlikely to have any effect whatsoever, and at the moment I only have on appropriately sized .1uf cap on hand, like I said.

Thank you both for your help btw.

It is generally an accepted practice to use a .1mf cap on every chip. the tiny .1mf/25v yellow ceramics work well.

Well I suppose I'm a modern person, so perhaps I'll adopt that terminology. :D So what you're both saying is that either way will work, but the best is probably to have two that go to ground? I just don't want to waste my time going from rail-to-rail if it's unlikely to have any effect whatsoever, and at the moment I only have on appropriately sized .1uf cap on hand, like I said.

Thank you both for your help btw.

It would be best to have two decoupling caps, one for the -vcc and one for the +vcc, both to the ground. Sometimes, especially for op-amp only circuits, there may be no ground refrence, so decoupling caps must go between the + and -, these capacitor do nothing except reduce noise. They are also called bypass capacitors.

Here is an example I whipped up for one with a ground refrence (I hope it is correct:yes:):
http://www.majhost.com/gallery/go-go/random/centertapsupply.jpg

And here is an example for an op-amp with no ground refrence:
http://www.majhost.com/gallery/go-go/random/bypass-opamp-example.jpg


As you will note, these both do the same thing. since these are just for noise elimination, it is not imperative to have a ground refrence, since all they do is
"couple" the two different voltages to cancel out stray interferance.

Hope this helped somehow.

I know I learned something from reading this topic, so thanks to the people who already posted the great information. (especially the noise reduction at higher frequencies).

That's helpful too 33&athird, but what exactly do you mean when you say op-amp only circuit? The headphone amp in question is a tube/solid state hybrid, is that what you're referring to? This amp has served as a crash course in op-amps for me, that's why I'm a little in the dark about the whole thing.

So for now, since I'm only putting one cap across the op-amp, that will at least serve to reduce noise? Not sure I'll notice anything as this amp is pretty quiet, but we'll see. I ordered more of the proper value capacitors last night.

Mr. Lin:

There are some circuits that use both a positive and negative voltage to power an Op-amp but do not bring the ground to the Op-amp. If there is no ground (or resister to ground) on any of the pins of the Op-amp, then the device really does not have any ground reference. In such a case, if you are going to try a bypass cap, one cap must go across the +V/-V rather then 2 caps +V/Ground Ground/-V.

I hope this helps.


Shelly_D

Mr. Lin:

There are some circuits that use both a positive and negative voltage to power an Op-amp but do not bring the ground to the Op-amp. If there is no ground (or resister to ground) on any of the pins of the Op-amp, then the device really does not have any ground reference. In such a case, if you are going to try a bypass cap, one cap must go across the +V/-V rather then 2 caps +V/Ground Ground/-V.

I hope this helps.


Shelly_D

Thank you Shelly. I actually just finished soldering the "orange drop" cap onto one of my op-amps. It was a bit of a pain as this capacitor is rated at 600V, total overkill, and the leads are thick, and copper so they kept pulling heat away from the joint. Anyway, I'm still not sure if this one falls in the category of an op-amp requiring rail-to-rail, here's a link to the pdf of the manufacturer's specs for the 2107 that I got from Digikey:

http://focus.ti.com/lit/ds/symlink/opa2107.pdf

And of course a picture for your viewing pleasure. I'm about to try it.

I just realized I mixed the capacitors up and soldered on the wrong value. The one in the picture is .0033uf, which I assume is not going to do anything here. I had intended to use a .1uf cap. Can't believe I did that.

That's helpful too 33&athird, but what exactly do you mean when you say op-amp only circuit? The headphone amp in question is a tube/solid state hybrid, is that what you're referring to? This amp has served as a crash course in op-amps for me, that's why I'm a little in the dark about the whole thing.

So for now, since I'm only putting one cap across the op-amp, that will at least serve to reduce noise? Not sure I'll notice anything as this amp is pretty quiet, but we'll see. I ordered more of the proper value capacitors last night.

As shelley_d pointed out, sometimes with an op-amp there are no grounds. Now if you had another component around the op-amp that utilized a ground, you may want to put the capacitors to that (I think that will work :scratch2:). The capacitor may reduce noise, but is really there as "insurance" from noise, since it passes stray AC signals that should not be on the DC rails.

I just realized I mixed the capacitors up and soldered on the wrong value. The one in the picture is .0033uf, which I assume is not going to do anything here. I had intended to use a .1uf cap. Can't believe I did that.

yeah, a 3.3nF would do something, but not nearly as much as a 100nF capacitor. If it makes you feel any better, it is worth noting I screw up all the time, especially with capcitors. Heck, that powers supply circuit I posted is incorrect, the -Vcc diode should be reversed.

Out of curiousity, Is this amp using the op-amps as unity gain buffers for impedance matching for a tube, rather than an output transformer?, I was considering a tube & solid state headphone amp using sub-miniture tubes, and a 072 op-amp.

Heard some good things about the little dot 1+, so I hope you get it running well!

As shelley_d pointed out, sometimes with an op-amp there are no grounds. Now if you had another component around the op-amp that utilized a ground, you may want to put the capacitors to that (I think that will work :scratch2:). The capacitor may reduce noise, but is really there as "insurance" from noise, since it passes stray AC signals that should not be on the DC rails.



yeah, a 3.3nF would do something, but not nearly as much as a 100nF capacitor. If it makes you feel any better, it is worth noting I screw up all the time, especially with capcitors. Heck, that powers supply circuit I posted is incorrect, the -Vcc diode should be reversed.

Out of curiousity, Is this amp using the op-amps as unity gain buffers for impedance matching for a tube, rather than an output transformer?, I was considering a tube & solid state headphone amp using sub-miniture tubes, and a 072 op-amp.

Heard some good things about the little dot 1+, so I hope you get it running well!

This is from the description of the I+:

The vacuum tube stage is run in Class-A, the MC33078 operational amplifier (socketed) provides voltage amplification, and BD139/BD140 transistors serve as output buffers. The Little Dot I+ is perfect for driving low impedance headphones that require gobs of current to perform to their full potential.

It uses one dual op-amp. I've acquired quite a few so far, and been through approximately five of them, but haven't had the time with each one to decide which I like best. The 2107 I'm using here ranks among the better ones. The most important thing I've found with this amp is that the op-amp it comes with does choke performance, so IMO replacing it with something a little better is necessary to really hear what it can do. Then there's that whole tube rolling thing... :banana:

Most recently I've been putting single and/or SOIC op-amps on Brown Dog adapters, which has been entertaining. You can read more about my experience with the I+ in this thread from the headphone forum:

http://www.audiokarma.org/forums/showthread.php?t=259926

Hmmm. In a previous life, one of our products was a mixed signal (analog and digital) distribution amplifier that was pretty cool and had a response from DC to blue light (just kidding). It had to drive 50-ohm loads and had very wide bandwidth.

The digital stuff was bypassed with a 0.1 uF part to ground. For the analog op amps, we used 4.7uF tantalums at the supply pins to ground. There were measurable differences as we decreased that capacitance, and we found no improvement beyond that value.

Just $0.02...

Hmmm. In a previous life, one of our products was a mixed signal (analog and digital) distribution amplifier that was pretty cool and had a response from DC to blue light (just kidding). It had to drive 50-ohm loads and had very wide bandwidth.

The digital stuff was bypassed with a 0.1 uF part to ground. For the analog op amps, we used 4.7uF tantalums at the supply pins to ground. There were measurable differences as we decreased that capacitance, and we found no improvement beyond that value.

Just $0.02...

Very intresting, that the maximum value was .1uF. However, from what I have heard tantalum capacitors are not great for audio use, though as coupling capacitors they should be fine. They are harder to find and cost more (I think), if you want the best of all worlds, get a small film capacitor, they are cheap and generally high-quality. Standard ceramics would work, But as with the tantalums, some people have issues with there use in audio paths.

That being said, the biggest difference you will get in that amp (very nice, by the way) is with tubes, or the op-amp...so use whatever caps you have.

Is that adaptor you are using making it use one-half of an op-amp for each channel, from different op-amps? what would be the benefit of that?:scratch2:

Is that adaptor you are using making it use one-half of an op-amp for each channel, from different op-amps? what would be the benefit of that?:scratch2:

I don't know, you tell me? :D

Seriously, this is all experimentation, meant to squeeze a maximum amount of performance out of this ridiculously low-priced amp, and so far it's working. Did you take a look further into my thread in the headphone forum? I have pictures of some of the adapters I've tried. One allows you to use a dual SOIC op-amp in the DIP 8 socket, another allows you to use two single SOIC op-amps in the socket (not pictured in the thread, just made it last night), yet another allows you to use two single DIP 8 op-amps in the circuit that requires a dual DIP 8 op-amp. Get it? :thmbsp:

My first thought with this is that at least I'll be able to try reputable op-amps that only come in single and/or SOIC packages. I do also wonder if there's a benefit to using two separate single op-amps as opposed to one dual op-amp, where everything is in one package. Sort of along the lines of using discrete transistors (Burson) vs. ICs, but obviously not going nearly as far.

In the end I'm just having a blast playing around with this stuff, and I'm still enjoying the music of course.

EDIT: Regarding the type of capacitor to use, I think I found something on Mouser last night that will solve the space problem. It's a Mallory polypropylene film cap, .1uf 230V. Very small, in fact, the smallest of this type and value I've been able to find. Note Brown Dog also offers the adapters with "test pins" that stick out the top-side of the adapter circuit, which will make is very easy to attach one of two capacitors to one of those. I've also read that it's not advisable to use ceramic caps here as these may have a negative impact on the sound.

I don't know, you tell me? :D

Seriously, this is all experimentation, meant to squeeze a maximum amount of performance out of this ridiculously low-priced amp, and so far it's working. Did you take a look further into my thread in the headphone forum? I have pictures of some of the adapters I've tried. One allows you to use a dual SOIC op-amp in the DIP 8 socket, another allows you to use two single SOIC op-amps in the socket (not pictured in the thread, just made it last night), yet another allows you to use two single DIP 8 op-amps in the circuit that requires a dual DIP 8 op-amp. Get it? :thmbsp:

yup, very impressive thread, makes me want to get one!...maybe for christmas :scratch2: I may check out all the mods again...its a very neat amplifier, and somewhat unique because it seems built around allowing the user to try things out.

My first thought with this is that at least I'll be able to try reputable op-amps that only come in single and/or SOIC packages. I do also wonder if there's a benefit to using two separate single op-amps as opposed to one dual op-amp, where everything is in one package. Sort of along the lines of using discrete transistors (Burson) vs. ICs, but obviously not going nearly as far.

Perhaps the idea is that when in the same package, the op-amps heat up more and would cause more distortion to each other?...Either way, if that mod gets even a little more performance out of just using another op-amp, go with it!

In the end I'm just having a blast playing around with this stuff, and I'm still enjoying the music of course.

Thats really all that matters, sometimes its more fun to tweak and troubleshoot than to sit idle...and when your moding it to sound how you want it to, its very exciting.

EDIT: Regarding the type of capacitor to use, I think I found something on Mouser last night that will solve the space problem. It's a Mallory polypropylene film cap, .1uf 230V. Very small, in fact, the smallest of this type and value I've been able to find. Note Brown Dog also offers the adapters with "test pins" that stick out the top-side of the adapter circuit, which will make is very easy to attach one of two capacitors to one of those. I've also read that it's not advisable to use ceramic caps here as these may have a negative impact on the sound.

That should work very well, and the size is always a bonus. I hope adding a coupling capacitor will have a notable diffrence...since the op-amp already has a fantastically low noise ratio, but even if you can't hear a difference, it wouldn't hurt anything being in there.

I'll keep checking the headphone topic for updates, this is a very neat project :yes:.

The capacitors came, here's a picture of one next to a 408 tube for size reference. Haven't had a chance to try it out yet, but these appear to be the perfect size, and the leads are much thinner so soldering them to the op-amp pins should be a lot easier than it was with the orange drop cap.

yup, very impressive thread, makes me want to get one!...maybe for christmas :scratch2: I may check out all the mods again...its a very neat amplifier, and somewhat unique because it seems built around allowing the user to try things out.





Because these are so inexpensive I'm getting one for my friend as a Christmas/birthday present. Then I can take my Pro-Ject Headbox SE back from him. :D

The capacitors came, here's a picture of one next to a 408 tube for size reference. Haven't had a chance to try it out yet, but these appear to be the perfect size, and the leads are much thinner so soldering them to the op-amp pins should be a lot easier than it was with the orange drop cap.

Because these are so inexpensive I'm getting one for my friend as a Christmas/birthday present. Then I can take my Pro-Ject Headbox SE back from him. :D

Heh heh heh:D...sounds good, that pro-ject is a neat looking amp, and i'm sure your friend will appreciate the little dot. On a seperate note, that capacitor will fill your needs perfectly, and will be much easier to solder...probably will look better too.:yes: the orange drops are nice, but almost certianly tube-only for most applications (especially at 600v ratings). I like the yellow-film caps as well, just because they look nice.

Heck, with all this work maybe your next project should be a clear box for it all :D

Heh heh heh:D...sounds good, that pro-ject is a neat looking amp, and i'm sure your friend will appreciate the little dot. On a seperate note, that capacitor will fill your needs perfectly, and will be much easier to solder...probably will look better too.:yes: the orange drops are nice, but almost certianly tube-only for most applications (especially at 600v ratings). I like the yellow-film caps as well, just because they look nice.

Heck, with all this work maybe your next project should be a clear box for it all :D

I like the looks of the Mallory caps too, but I also like the orange drops. Oh well, I'm in the midst of putting together a Bottlehead Foreplay III preamp, and soon thereafter I'll be making the Paramour mono blocks. I plan to upgrade the capacitors in both models, but slowly, so maybe the first small step will be the orange drops.

That cap would make an excellent coupling or filter cap. However, it is much too large to use as a +/- supply bypass on an opamp. The whole point of +/-bypassing is to try and stabilize a high-bandwidth chip by lowering the local supply impedance, in the multi-MHz range. The cap you show is designed for excellent performance at lower frequencies, and has too much inductance to work +/-.

The standard for this kind of bypassing is a ceramic. Or, if you are not comfortable with a ceramic even on the supply, a silver mica or small styrene.

http://www.intersil.com/data/an/an1325.pdf

-k

Can this bypassing technique help to tame oscillations in a dual op amp? I've got a THS4032CD on an adapter board that sounds great in the application but runs pretty toasty. No scope available to be positive, but this chip is apparently prone to oscillation.
I'm trying it as a direct sub for an LM833.

Thanks!

Dan

That cap would make an excellent coupling or filter cap. However, it is much too large to use as a +/- supply bypass on an opamp. The whole point of +/-bypassing is to try and stabilize a high-bandwidth chip by lowering the local supply impedance, in the multi-MHz range. The cap you show is designed for excellent performance at lower frequencies, and has too much inductance to work +/-.

The standard for this kind of bypassing is a ceramic. Or, if you are not comfortable with a ceramic even on the supply, a silver mica or small styrene.

http://www.intersil.com/data/an/an1325.pdf

-k

Wish I had seen this earlier- very interesting, and a good point:thmbsp:

Can this bypassing technique help to tame oscillations in a dual op amp? I've got a THS4032CD on an adapter board that sounds great in the application but runs pretty toasty. No scope available to be positive, but this chip is apparently prone to oscillation.
I'm trying it as a direct sub for an LM833.

Thanks!

Dan

Maybe... First I would check the datasheets for the originals voltage rating and compare it to the new one. You may find they aren't quite the same, and that can lead to heat. I think coupling caps can tame oscillilation, but it can't tame everything. The potential across the V+ and V- are still the same. In a dual op-amp (what this topic is about, by the way) has only one positive and one negative supply pin per IC, since both channels use the same pin for power. In all honesty, if your comfortable with it, adding a small non-polarized cap can't really hurt anything.

But first, check the voltages, you may find the original chip ran at a higher operating voltage, and the new one has a lower maximum operating temp. Could be oscillation, though:scratch2:.

If you have any other questions, go ahead and ask:thmbsp:

Edit; Darn, missed this the first time around, Welcome to AudioKarma!

Very intresting threat, To bad i did not find it sooner.

Anyway, im going to upgrade my DAC (a denon DA-500) Im replacing the IV and buffer with dual Bursons, and 2 other double opamps that are used in the analog selection with OPA2604.

Im a bit new with this stuff, but i did a lot of reading on forums and other places. I understand that You sould pretty much always use 100n caps between + and gound and - and ground. If my understandings are correct its a bad thing to use a higher rating because they crap that was filtered out of the psu will get into the ground, and have a effect on the output of the opamp. I was also told that it can have a positive effect to place a 2.2-4.7u cap between both psu pins of the opamp.

Any thoughts on this? Im also using a low noise PSU just for the analog selection. I got it of ebay. Its claimed to be less than 0.1mV noise and im planning to run at 20-25V.

Im a bit new with this stuff, but i did a lot of reading on forums and other places. I understand that You sould pretty much always use 100n caps between + and gound and - and ground. If my understandings are correct its a bad thing to use a higher rating because they crap that was filtered out of the psu will get into the ground, and have a effect on the output of the opamp. I was also told that it can have a positive effect to place a 2.2-4.7u cap between both psu pins of the opamp.

Well, depends on the supply. Op-amps don't necessarily have a ground, so sometimes the coupling is directly between + and -, no problem with that, and sometimes it's easier that way(keep in mind the voltage across the cap will be the positive voltage+ the negative voltage, or their potential). If a ground is handy, then what you describe is perfect, + to ground, ground to -, keep in mind these caps won't be polarized. Coupling caps are weird, you can't really determine affect unless simulated, or tested. That being said, 100nF is a great value, and seems like a general value for coupling, i've used in in a PS build before. Anything higher won't do much better, and may be detramental anyhow.

I don't get the 2.2-4.7uF cap values. These would almost certainly be polarized (for that value), and putting them between the pins would be bad news, I think. You could add the 2.2-4.7uF's to the op-amp if you did + to ground, and ground to - (observing polarity), however, this would only increase the filtering, and not help couple to any large degree; this is because coupling caps have to be non-polarized to work as such.

Any thoughts on this? Im also using a low noise PSU just for the analog selection. I got it of ebay. Its claimed to be less than 0.1mV noise and im planning to run at 20-25V.

That's a good, low noise supply...it probably (read; should) have coupling already in it, but I don't know how its constructed. Adding coupling to the op-amp may already be there as well, but maybe not.

I think all my information is correct, if I find something off, i'll edit it in or out. Good luck on the project, sounds like you already have a good grasp on things:thmbsp:.

like i said, it is what i have been reading on forums and other places, so my information is probaly not perfect. They way i see it there is always a ground used with a opamp. Maybe its not directly connected, but its always one of the pins on the signal output. (maybe its diffrent with ballanced cirquits, dont know) and ofter there is a pin connected to ground by a resistor or cap. So here always is a ground. While a psu can be made very low noise, there are many reasons why there still could be noise at the opamp i gues. The 100n caps filter this away into the earth, so less noise will reach the opamp. The idea of the bigger cap (yes, a polarized one) would be to deliver maximum current to the opamp i gues. I gues it would basicly work like the caps in the PSU, but much closer the the actual opamp.

But again, those i just the ideas that i got from readign forums and stuff. I cant be sure the info is right, and i cant be sure i understand it the right way. But please discuss this issue. I would love more opinions on this.

like i said, it is what i have been reading on forums and other places, so my information is probaly not perfect. They way i see it there is always a ground used with a opamp. Maybe its not directly connected, but its always one of the pins on the signal output. (maybe its diffrent with ballanced cirquits, dont know) and ofter there is a pin connected to ground by a resistor or cap. So here always is a ground. While a psu can be made very low noise, there are many reasons why there still could be noise at the opamp i gues. The 100n caps filter this away into the earth, so less noise will reach the opamp. The idea of the bigger cap (yes, a polarized one) would be to deliver maximum current to the opamp i gues. I gues it would basicly work like the caps in the PSU, but much closer the the actual opamp.

Very true...there is normally a ground somewhere. I say normally, because there is an off chance that there isn't, as there are several different op-amp circuit designs. For the most part, you are 100% correct (actually, ona ll accounts), in step-by step;

1)use a good, small value coupling cap connected to ground, straight to ground, not to another component that goes to ground. The coupling cap basically "shorts out" the AC noise to the ground. Keep in mind a cap of any nature blocks DC and passes AC, even polarized caps do this. This is why amplifier stages use a cap in series between stages, to pass the AC signal, but not the DC voltage. In this case, by putting the cap in parallel your not affecting the DC voltage, but eliminating the AC voltage to ground.

2), very correct on the polarized cap near the op-amp, although don't go crazy on the value (I think 2.2 or 4.7uF would be fine). These caps would be decreasing the latency (slowing down of) of the current supply, effectivly increasing the current, or more truthfully, insuring the current will be more redily available for the IC. This will provide a smoother response time, and in theory, better performance.

But again, those i just the ideas that i got from readign forums and stuff. I cant be sure the info is right, and i cant be sure i understand it the right way. But please discuss this issue. I would love more opinions on this.

Hey, nothing wrong with much of the info on forums, most is true, since I think alot of it is from engineers who honestly enjoy the hobby...i'm currently working on a electronics degree, so I still learn alot from reading around the internet.

Sounds like you already have a good handle on all this:yes:.

"...i'm currently working on a electronics degree, so I still learn alot from reading around the internet."

The internets is a wonderful thing! Plus, once you get your degree and design equipment professionally, no matter how skillful, experienced and diligent you become, the internets will find an easy way to improve your work with a different opamp or an extra cap....

-k

Wish I had seen this earlier- very interesting, and a good point:thmbsp:



Maybe... First I would check the datasheets for the originals voltage rating and compare it to the new one. You may find they aren't quite the same, and that can lead to heat. I think coupling caps can tame oscillilation, but it can't tame everything. The potential across the V+ and V- are still the same. In a dual op-amp (what this topic is about, by the way) has only one positive and one negative supply pin per IC, since both channels use the same pin for power. In all honesty, if your comfortable with it, adding a small non-polarized cap can't really hurt anything.

But first, check the voltages, you may find the original chip ran at a higher operating voltage, and the new one has a lower maximum operating temp. Could be oscillation, though:scratch2:.

If you have any other questions, go ahead and ask:thmbsp:

Edit; Darn, missed this the first time around, Welcome to AudioKarma!

Thanks for the answer, and the welcome!

The original was rated at 15-30V. The THS is rated at 5-15V. The actual measured voltage is 14V, so there's still a bit of margin. I even tried "potting" the chip to help distribute the heat, but still a bit too toasty.
I've got no problem soldering a cap in place (just finished mounting 8 OPA1641s to 4 dual-single SOIC/8PDIP adapters). I think I can dig up a couple of .01uf PP or PS caps.
My greater concern was damaging the circuit. I've been doing this stuff long enough to know that, once you let the smoke out, the components are never quite the same again.:nono:

Thanks Again!

Dan

"...i'm currently working on a electronics degree, so I still learn alot from reading around the internet."

The internets is a wonderful thing! Plus, once you get your degree and design equipment professionally, no matter how skillful, experienced and diligent you become, the internets will find an easy way to improve your work with a different opamp or an extra cap....

-k

Yep, gotta love the internets, always something new:yes:

Thanks for the answer, and the welcome!

The original was rated at 15-30V. The THS is rated at 5-15V. The actual measured voltage is 14V, so there's still a bit of margin. I even tried "potting" the chip to help distribute the heat, but still a bit too toasty.
I've got no problem soldering a cap in place (just finished mounting 8 OPA1641s to 4 dual-single SOIC/8PDIP adapters). I think I can dig up a couple of .01uf PP or PS caps.
My greater concern was damaging the circuit. I've been doing this stuff long enough to know that, once you let the smoke out, the components are never quite the same again.:nono:

Thanks Again!

Dan

Okay, voltage is about right, op-amps, even when run near top-end of their scale, should be pretty stable. 14v is pretty common for op-amps, as normally, 15v is used on the + and - supplies (odd standard, just is...much like digital circuits being 5v). I belive the .o1uF (10nF) caps would work fine for this application, although as mentioned above (in superkris's post) 100nF is a more standard value (still couples, though), it's worth a shot. I don't think theres any way to damage a circuit through a coupling cap unless installed incorrectly or under-valued for the voltage. Your right about that smoke thing, tthat one power filter cap of mine that blew was never quite right afterwords:scratch2:...

A coupling cap may solve your problems, if the heat is caused by oscillation. I belive the effect is sometimes called motorboating in an amp setting, especially when discussing transistors. Typically, this is caused by replacing a transistor with another one that has too high a Hz rating, and messes up how the circuit functions. In op-amps it may be called speed or Hertz, or more commonly, bandwidth. Another term is slew-rate, but it's complex, and I don't think quite the root of it all. Try a small cap, a 10nF may have noticable impact, but a 100nF should do it, if that would indeed solve the problem. Just do it from + to ground, and - to ground, and you'll get it.

On a seperate note, if this is for a higher-end application, ceramics should still work just fine, although as a personal opinion I perfer film caps...but use what you have. Most people have issue with the ceramics (small tan ones) when they are in the audio path, however I think in a power supply setting, they should work.

I hope that a small fix like this helps your heat problem...IC's are fairly reliable, even when pretty warm, but best not let it get too hot:yes:.

loving this thread. Next question is a bit offtopic, but since were talking about opamps and caps its still somewhat related. Here is the schematic of the oritional analog part of the DAC

http://www.audiokarma.org/forums/attachment.php?attachmentid=193550&stc=1&d=1265376277 I am already replacing all the resistors from the oritional extremly small carbonfirm ones to 0.6W 1% metalfirm ones. Not sure if i should replace the 100n caps. (they are already used in the schematic) but i want the best components in there. Im going to place 150$ opamps (burson) connected to a very high end PSU. All the caps around the DAC chips are going to be replaced with Nichicon MUSE BP caps, and c514, 515 are going to be Nichicon MUSE BP-S.

Now im just wondering if i should replace C503,505,509,511 and c524 and what kind of cap would be best here.

Okay, voltage is about right, op-amps, even when run near top-end of their scale, should be pretty stable. 14v is pretty common for op-amps, as normally, 15v is used on the + and - supplies (odd standard, just is...much like digital circuits being 5v). I belive the .o1uF (10nF) caps would work fine for this application, although as mentioned above (in superkris's post) 100nF is a more standard value (still couples, though), it's worth a shot. I don't think theres any way to damage a circuit through a coupling cap unless installed incorrectly or under-valued for the voltage. Your right about that smoke thing, tthat one power filter cap of mine that blew was never quite right afterwords:scratch2:...

A coupling cap may solve your problems, if the heat is caused by oscillation. I belive the effect is sometimes called motorboating in an amp setting, especially when discussing transistors. Typically, this is caused by replacing a transistor with another one that has too high a Hz rating, and messes up how the circuit functions. In op-amps it may be called speed or Hertz, or more commonly, bandwidth. Another term is slew-rate, but it's complex, and I don't think quite the root of it all. Try a small cap, a 10nF may have noticable impact, but a 100nF should do it, if that would indeed solve the problem. Just do it from + to ground, and - to ground, and you'll get it.

On a seperate note, if this is for a higher-end application, ceramics should still work just fine, although as a personal opinion I perfer film caps...but use what you have. Most people have issue with the ceramics (small tan ones) when they are in the audio path, however I think in a power supply setting, they should work.

I hope that a small fix like this helps your heat problem...IC's are fairly reliable, even when pretty warm, but best not let it get too hot:yes:.

Thanks again!

I can't easily do the + to GND and - to GND as there's no true ground in the immediate area (and I'd like to keep this a "plug-in" for now, but I'll try bridging the + and - with a .1uf cap)
We'll see......

BTW I see that you're using EPI T/E 70s- I've got a pair of a100s and a a pair of a110s--LOVE 'EM! I've made a bit of a hobby of repairing/upgrading EPIs. They're all a tad bit bright sounding with solid state stuff, so the associated electronics are always a challenge; finding a balance between good detail and irritation....:music:

loving this thread. Next question is a bit offtopic, but since were talking about opamps and caps its still somewhat related. Here is the schematic of the oritional analog part of the DAC

I am already replacing all the resistors from the oritional extremly small carbonfirm ones to 0.6W 1% metalfirm ones. Not sure if i should replace the 100n caps. (they are already used in the schematic) but i want the best components in there. Im going to place 150$ opamps (burson) connected to a very high end PSU. All the caps around the DAC chips are going to be replaced with Nichicon MUSE BP caps, and c514, 515 are going to be Nichicon MUSE BP-S.

Now im just wondering if i should replace C503,505,509,511 and c524 and what kind of cap would be best here.

I'd say that still in topic:thmbsp:, if your going all-out like that, go ahead and put film caps in, they last forever, and are higher quality then ceramic caps. Polyester or polypropylene would work just fine. the circuit may already use those, but I don't know, as the schematic doesn't show type. If the 100nF are already a film cap (ceramics are always the little round, tan ones), I may just leave them be, as most film caps are really made equal...but that's just my opinion, worth looking around though, maybe a potted one would work :dunno:, not recommending it, per se', but it's an idea. (supposedly the potting [or plastic-dipped housing] would help isolate the cpa from interferance).

Nice Nichicons, by the way...heard alot of good things about that muse line, should try some out one of these days...:scratch2:, also on that note; very cool schematic, i'm still looking it over...

Thanks again!

I can't easily do the + to GND and - to GND as there's no true ground in the immediate area (and I'd like to keep this a "plug-in" for now, but I'll try bridging the + and - with a .1uf cap)
We'll see......

Go for it, just make sure the cap is rated for 30v or so minimum...most are, but better not to risk it. You'll probably still get results from that configuration:yes:

BTW I see that you're using EPI T/E 70s- I've got a pair of a100s and a a pair of a110s--LOVE 'EM! I've made a bit of a hobby of repairing/upgrading EPIs. They're all a tad bit bright sounding with solid state stuff, so the associated electronics are always a challenge; finding a balance between good detail and irritation....:music:

Ever since getting the 70's i've wanted to hear what the higher-end epi's sounded like, the a100's are near the top of the list of model's i'd love to hear. The 70's are pretty small, but sound great...as you said, a tad bright, but the pioneers I use are a little mellow, so it balances out nicely.

Go for it, just make sure the cap is rated for 30v or so minimum...most are, but better not to risk it. You'll probably still get results from that configuration:yes:.

Thanks!

I don't think I have a film cap <50V; most of the PP and PS I've got are 100-630V. I don't see a problem.....:D



Ever since getting the 70's i've wanted to hear what the higher-end epi's sounded like, the a100's are near the top of the list of model's i'd love to hear. The 70's are pretty small, but sound great...as you said, a tad bright, but the pioneers I use are a little mellow, so it balances out nicely.

The 70s are a 6" woofer, right?
The A100s and A110s are identical but for the enclosure size (gives another 10hz on the bottom end). I bought the 100s back in the 80's new and picked up the 110s on eBay for $38 shipped. Both sets have been refoamed by yours truly. I also replaced the NP electrolytics with some reasonable PP caps after replacing the internal wiring and the spring posts with binding posts. The 2 gurus on the EPI speakers are Tribute Audio and Human Speakers (a former Genesis employee)
I'm using a semi vintage Marantz SR5100 rec. It's a second generation unit (can't think about affording a 2200/2300/4200 series) that no longer has those great bipolar outputs-it uses amplifier modules instead. The bass is a bit better controlled but the treble gets a bit much at times.
These puppies, using the right source, can exhibit a truly concussive bass that has people swearing it's got a 10" or 12" bass driver instead of the measly 8". :yes:
Thanks again for the advice-I'll post the results-or a photo of the charred remains.....:sigh:

Thanks!

I don't think I have a film cap <50V; most of the PP and PS I've got are 100-630V. I don't see a problem.....:D

Ha! yeah, I don't think you'll have a problem:D



The 70s are a 6" woofer, right?
The A100s and A110s are identical but for the enclosure size (gives another 10hz on the bottom end). I bought the 100s back in the 80's new and picked up the 110s on eBay for $38 shipped. Both sets have been refoamed by yours truly. I also replaced the NP electrolytics with some reasonable PP caps after replacing the internal wiring and the spring posts with binding posts. The 2 gurus on the EPI speakers are Tribute Audio and Human Speakers (a former Genesis employee)
I'm using a semi vintage Marantz SR5100 rec. It's a second generation unit (can't think about affording a 2200/2300/4200 series) that no longer has those great bipolar outputs-it uses amplifier modules instead. The bass is a bit better controlled but the treble gets a bit much at times.
These puppies, using the right source, can exhibit a truly concussive bass that has people swearing it's got a 10" or 12" bass driver instead of the measly 8". :yes:
Thanks again for the advice-I'll post the results-or a photo of the charred remains.....:sigh:

:yes: 6" driver with a 12" sound, excellent speaker all around. I've heard smaller drivers actually have more controlled bass, because there is less inertia when they move. For the money, I don't think the epi's can be beat, since every pair i've seen is under $100, used. I've considered adding a brace in the cabniet to improve th bass, as the woofer sometimes shakes things up a little much.

And don't worry about the results, I think it'll work just fine:thmbsp:

Nice marantz, as well...the pioneers are things i've picked up pretty cheap, but seeing the prices of certain series, especially marantz is eye-opening to say the least:sigh:.

Let me know how it goes, I hope it cleans up the oscillation, I was thinking of the heat, and it can't be the voltage as long as its under maximum, so it has to be something else...adding a cap is the easiest way to try to control it, I actually have a project I may add a cap to, as it had somewhat related problems:scratch2:.

I got the PSU parts and the last caps this afternoon, so i can start building.

I was a little "impressed" by the size of the coupling caps on the output.

From left to right: Muse BP-S 22uF 50V, Muse BP 22uF 50V, Panasonic 33uF 35V, DIL8 opamp.
That muse BP-S is fucking huge. I have no idea how i will be able to get it to fit with all the big parts im using already. the current analog parts is 1x4 inch or something like that, and besides the much bigger resistors and some new bigger caps i also have those huge burson op amps. Not sure if i will be using this cap. Could not find to much info about it. Anyone here with some good info of the Muse BP-S line?

http://www.audiokarma.org/forums/attachment.php?attachmentid=193687&stc=1&d=1265424839

I got the PSU parts and the last caps this afternoon, so i can start building.

I was a little "impressed" by the size of the coupling caps on the output.

From left to right: Muse BP-S 22uF 50V, Muse BP 22uF 50V, Panasonic 33uF 35V, DIL8 opamp.
That muse BP-S is fucking huge. I have no idea how i will be able to get it to fit with all the big parts im using already. the current analog parts is 1x4 inch or something like that, and besides the much bigger resistors and some new bigger caps i also have those huge burson op amps. Not sure if i will be using this cap. Could not find to much info about it. Anyone here with some good info of the Muse BP-S line?

Wow...those are something ...where will they be going? or which c are they on the schematic?...just wondering:scratch2:...

If I remember correctly, the Muse caps are Nichicon's high line audio caps. They seem to be used in quite a few of the Chinese-made audio products sold on eBay.
Compared to some Panasonic FM series low ESR caps I've used, they are pretty good for the money. Not Black Gates, but pretty good.
Also, FWIW, the OPA2604 is a pretty nice sounding op amp in the right circuit. You might also want to check out the OPA1642. It's a purpose-designed Audio Device, and sounds excellent, again, in the right circuit/application. This, and a pair of OPA1641s mounted on a dual single to single DIP adapter are current faves of mine in the analog outputs of a tda1541a-based DAC.:thmbsp:
I've heard the different Bursons in a few different circuits. I have to say that their sound varies widely and seem to be more circuit/application sensitive than I'd have expected.
Beautifully designed and built, though.:beerchug:

Dan

Wow...those are something ...where will they be going? or which c are they on the schematic?...just wondering:scratch2:...

like i said a few posts back. they will be the coupling the output. they are c514, 515 in the schematic.

Here is a picture of the oritional analog part.

http://www.audiokarma.org/forums/attachment.php?attachmentid=193858&stc=1&d=1265492335

Those small caps next to that relay are the output caps.

Anyway, these thing are huge. they scare me:P The ebay item description has some info about them. It sounded good when i ordered them, but i forgot to look at their size. :P
http://cgi.ebay.com.sg/Nichicon-BP-S-MUSE-Electrolytic-capacitor-22uF-50V-1pc_W0QQitemZ320346085290QQcmdZViewItemQQptZLH_Def aultDomain_0?hash=item4a961d57aa

Had similar issues using Pansonic FMs in the analog stages of the many Magnavix/Phillips CD players I've modded over the years.....
They were easily 5X larger than the originals.
I made a practice of using short lengths of teflon tubing over the leads to allow me to offset or even raise the caps' positions.
I know that it's good to keep lead lengths short, but.......:dunno:

haha, i would only wish they were 5 times as big :P But i got some of that teflon with my burson's. Im also planning on routing some of the components on the other side of the print. I dont see any problem with doing this with resistors and some smaller caps.

Quite so.

So long as there's adequate clearance, it shouldn't matter at all...Also, while you're there, bypassing those 22uf 50V caps with some .01-.1uf PP or PS caps of equal or higher voltage wouldn't hurt either if space allows.
Good luck, and please post photos of the final product.:thmbsp:

Dan

Had similar issues using Pansonic FMs in the analog stages of the many Magnavix/Phillips CD players I've modded over the years.....
They were easily 5X larger than the originals.
I made a practice of using short lengths of teflon tubing over the leads to allow me to offset or even raise the caps' positions.
I know that it's good to keep lead lengths short, but.......:dunno:

:yes:, when space is tight, and a component needs to be raised or such, I strip a little bit of wire (standard copper stuff), and use the casing from the wire as a sheath for the components leads.

Quite so.

So long as there's adequate clearance, it shouldn't matter at all...Also, while you're there, bypassing those 22uf 50V caps with some .01-.1uf PP or PS caps of equal or higher voltage wouldn't hurt either if space allows.
Good luck, and please post photos of the final product.:thmbsp:

Dan

I'd love to see it too...neat looking board to start with, the end results will probably look great:thmbsp:.

i got some SMD 100n caps to solder beween the pins of the caps. When ill start modding i might make a thead about it.