If the problem only occurs on the left channel, you most likely do not have a PSU issue. A PSU issue would affect both channels the same way. Your issue sounds to me like a broken DC-blocking cap on the left channel that passes some of the DC current it is supposed to block, and passes more current if it warmed up (that's a way to explain the "ramping up" effect you observe). If you are less lucky, it could also be caused by broken semiconductors that pass undesired current when they get warm.

Do you have the problem only on "speaker out", or also on "line out"? If you only have the problem on speaker out, the root cause of the problem is very likely fixable by replacing easily obtainable components using hobbyist's tools. You "just" need to find a hobbyist near you to do it for you.

Yeah this sounds very much like bad capacitors to me.

mkarcher wrote on 2021-06-24, 16:17:

If the problem only occurs on the left channel, you most likely do not have a PSU issue. A PSU issue would affect both channels the same way. Your issue sounds to me like a broken DC-blocking cap on the left channel that passes some of the DC current it is supposed to block, and passes more current if it warmed up (that's a way to explain the "ramping up" effect you observe). If you are less lucky, it could also be caused by broken semiconductors that pass undesired current when they get warm.

Do you have the problem only on "speaker out", or also on "line out"? If you only have the problem on speaker out, the root cause of the problem is very likely fixable by replacing easily obtainable components using hobbyist's tools. You "just" need to find a hobbyist near you to do it for you.

Okay, I see; thanks. It does seem like something is getting saturated over time, if it makes sense, so it may well be a busted cap. The problem is also present on the speaker out (I have tested the line out to begin with, while on spk it's predictably shrieking).

Another thing of note is that the issue gets worse as I work through the steps in the diagnose utility, it's like everything goes downhill by the time I'm getting to the audio test selection.

What do you suggest I should try to pin down the problem? Al I have on hand is a basic multimeter, would I be able to do any diagnosis at all with it? Or, should I just get someone to look at it (probably not really worth it, given the value of this thing)?

BeastOfSoda wrote on 2021-06-24, 17:29:

What do you suggest I should try to pin down the problem? Al I have on hand is a basic multimeter, would I be able to do any diagnosis at all with it? Or, should I just get someone to look at it (probably not really worth it, given the value of this thing)?

It's difficult to pin down the problem using just a meter on the isolated, as it only develops while the system is running. If it were my card, I would troubleshoot the problem just for the fun of it by reverse engineering the audio output section (up to the speaker amp, no need to reverse the speaker amp section itself), and check for correct DC bias by comparing the values a basic multimeter indicates in the "DC V" range on the corresponding points for the left and the right channel while the symptoms are ocurring in a running system. This means I plug the card in the leftmost or rightmost ISA slot of a board outside of the case, so I can access the card while the system is running. This includes the risk of generating a short circuit that causes even more damage to the card if I slip with the probe, so I would to it very carefully.

You might want to test whether the symptoms change when you adjust mixer channels (like Wave or synth) to further limit the part of the card that is probably broken, so the amount of reverse engineering can be limited. Issues like this are difficult to be troubleshot remotely through a forum like VOGONs, I considering the wide availability of SB16 cards, probably indeed not worth the trouble. Paying a professional to do that is very likely not worth the trouble, but if you happen to find an enthusiast locally, it might be worth a try.

mkarcher wrote on 2021-06-24, 17:59:

BeastOfSoda wrote on 2021-06-24, 17:29:

What do you suggest I should try to pin down the problem? Al I have on hand is a basic multimeter, would I be able to do any diagnosis at all with it? Or, should I just get someone to look at it (probably not really worth it, given the value of this thing)?

It's difficult to pin down the problem using just a meter on the isolated, as it only develops while the system is running. If it were my card, I would troubleshoot the problem just for the fun of it by reverse engineering the audio output section (up to the speaker amp, no need to reverse the speaker amp section itself), and check for correct DC bias by comparing the values a basic multimeter indicates in the "DC V" range on the corresponding points for the left and the right channel while the symptoms are ocurring in a running system. This means I plug the card in the leftmost or rightmost ISA slot of a board outside of the case, so I can access the card while the system is running. This includes the risk of generating a short circuit that causes even more damage to the card if I slip with the probe, so I would to it very carefully.

You might want to test whether the symptoms change when you adjust mixer channels (like Wave or synth) to further limit the part of the card that is probably broken, so the amount of reverse engineering can be limited. Issues like this are difficult to be troubleshot remotely through a forum like VOGONs, I considering the wide availability of SB16 cards, probably indeed not worth the trouble. Paying a professional to do that is very likely not worth the trouble, but if you happen to find an enthusiast locally, it might be worth a try.

Hmm, I see. To be honest, I'm in way over my head, and I don't have the first idea of what to measure on the board. I am trying to study up on electronics, but it might take a good while before I know what I'm doing. The alternative would be to just replace every cap on the board, but I don't trust myself to accomplish that and it wouldn't get me any closer to doing things right.

Any chances I might find someone interested in taking a look at it on these forums?

BeastOfSoda wrote on 2021-06-24, 19:22:

Hmm, I see. To be honest, I'm in way over my head, and I don't have the first idea of what to measure on the board. I am trying to study up on electronics, but it might take a good while before I know that I'm doing. The alternative would be to just replace every cap on the board, but I don't trust myself to accomplish that and it wouldn't get me any closer to doing things right.

I agree. As you describe your knowledge, the job is likely above your skills right now.

Furthermore, the cause of your problem need not be an aged cap, it could also be ESD damage in an integrated circuit. So swapping all the electrolytic caps on the card is no guaranteed success, even if done properly.

Actually... Would I be right in assuming your card looks identical to this one? I mean, same model number, but you never know. Anyway, if it is the same, it looks they're all through-hole capacitors. Basically, the legs of the capacitor go through the board, and you soldering them on the opposite side. Those are a thousand times more easy to replace than surface mount (SMD) ones. There's a lot of people who get their start with soldering by replacing through-hole capacitors on old equipment just like this. I'd suggest looking up tutorials on youtube to see if you're up for it. If you are, there's a lot of tips we could give to help with the whole process.

This would be what people call the "shotgun" approach, where you just replace all of them. No need for measuring things, etc. Just a bit of soldering, checking your work, and maybe doing some continuity testing to make sure things aren't bridged anywhere.

Kyosho wrote on 2021-06-26, 13:58:

Actually... Would I be right in assuming your card looks identical to this one? I mean, same model number, but you never know. Anyway, if it is the same, it looks they're all through-hole capacitors. Basically, the legs of the capacitor go through the board, and you soldering them on the opposite side. Those are a thousand times more easy to replace than surface mount (SMD) ones. There's a lot of people who get their start with soldering by replacing through-hole capacitors on old equipment just like this. I'd suggest looking up tutorials on youtube to see if you're up for it. If you are, there's a lot of tips we could give to help with the whole process.

This would be what people call the "shotgun" approach, where you just replace all of them. No need for measuring things, etc. Just a bit of soldering, checking your work, and maybe doing some continuity testing to make sure things aren't bridged anywhere.

Yup, the very same; I do have some experience with through-hole soldering, although I am admittedly not very good at it. All I've done is simple jobs, and although this looks much more daunting I was just thinking about giving it a go. Thanks for your kind offer to help, I am sure I'll need quite a bit before I'm done with this.

Meanwhile, I have done some research: the history of this board is that it was pulled in a functioning state from its previous computer, and judging from the conditions I received it in, it seems to have been stored properly. This means that I can exclude the likelihood that it was damaged due to ESD or by moisture; plus, now it has started working even worse, further evidence pointing to a cap failure.

While attempting a fix, however, I still plan on doing some diagnosis first, in order to at least get better acquainted with the skills of the trade and transferring them to similar odd jobs on my appliances. I have read about using an ESR meter in order to determine if a cap is still good or not without desoldering it: I am wondering, what is your opinion about this particular tool? While I understand that testing a cap still on the board will give me skewed readings due to the surrounding components, I am under the impression that it will still be good enough to determine a go/no-go situation. I am still going to nuke everything, but I wish to have a better understanding of the situation.

I also could use more information on which capacitors to buy, exactly: I know they must have the same rating as the original ones and that I should buy from reputable brands, but many of them come in different series, so which ones would be right here? I am assuming they are differentiated by tolerance in their operating specs? Also, is there the possibility that the card came out of the factory with out of spec capacitors to begin with, or should I trust the values of the stock ones?

Thanks again.

BeastOfSoda wrote on 2021-06-26, 16:54:

While attempting a fix, however, I still plan on doing some diagnosis first, in order to at least get better acquainted with the skills of the trade and transferring them to similar odd jobs on my appliances. I have read about using an ESR meter in order to determine if a cap is still good or not without desoldering it: I am wondering, what is your opinion about this particular tool? While I understand that testing a cap still on the board will give me skewed readings due to the surrounding components, I am under the impression that it will still be good enough to determine a go/no-go situation. I am still going to nuke everything, but I wish to have a better understanding of the situation.

Oppinions on using ESR-meters in circuit are mixed. My stance is that it generally works, but is of limited use in certain circumstances. Most notably, if multiple caps are used in parallel at different positions on the board, the frequency of the ESR meter isn't high enough to locally measure mostly the cap you point it to, but it will measure all the caps. The resistance of surrounding components is usually so much higher than the impedance of a capacitor, especially if it is an low-ESR capacitor, that the measurement isn't skewed a lot.

On the other hand, ESR is just one parameter of an (electrolytic) cap. It is the most important parameter in switch-mode power supplies. In the case of sound cards, especially for DC-blocking caps in the audio path, the leakage current is more important. The issue you have with the card sounds much more like leakage current than too high ESR. A high ESR might be a general sign for a degraded cap, but most ESR meters can not differentiate between caps with extremely high leakage (i.e. nearly short circuit) and caps with perfect ESR. Testing for leakage in circuit is way more delicate than testing for ESR in circuit, so using an ESR meter on your sound card might point you to the core issue, or it might be nearly pointless.

Most audio circuits on the sound card exist twice: Once for the left channel and once for the right channel. If you manage to trace the circuit well enough to identify the corresponding components, performing the same measurements (mostly DC resistance and ESR) on corresponding points, every significant mismatch points to a potential issue. As your fault is located on the left channel only (better make sure that only the left channel misbehaves, and it's not a broken right channel on your speakers!), doing this kind of "differential analysis" might be a very effective tool.

mkarcher wrote on 2021-06-26, 18:15:

Oppinions on using ESR-meters in circuit are mixed. My stance is that it generally works, but is of limited use in certain circu […]
Show full quote

BeastOfSoda wrote on 2021-06-26, 16:54:

While attempting a fix, however, I still plan on doing some diagnosis first, in order to at least get better acquainted with the skills of the trade and transferring them to similar odd jobs on my appliances. I have read about using an ESR meter in order to determine if a cap is still good or not without desoldering it: I am wondering, what is your opinion about this particular tool? While I understand that testing a cap still on the board will give me skewed readings due to the surrounding components, I am under the impression that it will still be good enough to determine a go/no-go situation. I am still going to nuke everything, but I wish to have a better understanding of the situation.

Oppinions on using ESR-meters in circuit are mixed. My stance is that it generally works, but is of limited use in certain circumstances. Most notably, if multiple caps are used in parallel at different positions on the board, the frequency of the ESR meter isn't high enough to locally measure mostly the cap you point it to, but it will measure all the caps. The resistance of surrounding components is usually so much higher than the impedance of a capacitor, especially if it is an low-ESR capacitor, that the measurement isn't skewed a lot.

On the other hand, ESR is just one parameter of an (electrolytic) cap. It is the most important parameter in switch-mode power supplies. In the case of sound cards, especially for DC-blocking caps in the audio path, the leakage current is more important. The issue you have with the card sounds much more like leakage current than too high ESR. A high ESR might be a general sign for a degraded cap, but most ESR meters can not differentiate between caps with extremely high leakage (i.e. nearly short circuit) and caps with perfect ESR. Testing for leakage in circuit is way more delicate than testing for ESR in circuit, so using an ESR meter on your sound card might point you to the core issue, or it might be nearly pointless.

Most audio circuits on the sound card exist twice: Once for the left channel and once for the right channel. If you manage to trace the circuit well enough to identify the corresponding components, performing the same measurements (mostly DC resistance and ESR) on corresponding points, every significant mismatch points to a potential issue. As your fault is located on the left channel only (better make sure that only the left channel misbehaves, and it's not a broken right channel on your speakers!), doing this kind of "differential analysis" might be a very effective tool.

Okay, I understand; so there is no good way to test them in circuit after all. I guess that's why people normally decide to replace everything in one go, as opposed to taking them out and testing the good from the bunk ones. By the way: I believe I was unclear about my situation, but the left channel is the one that's still somewhat working until the board warms up and they both completely stop functioning. Your suggestion remains solid, but since they're both busted anyway I have been successfully convinced that I might as well go nuclear on it.

That said, I have done some more research and I seem to understand that I want to go with Japanese brands for the most part (Rubycon, Panasonic, Sanyo and the like), but I still don't know much about which specific ones to get (at equal rating, that is). I am imagining that I need to stick with electrolytic, as for example solid state might have an adverse effect on functionality, right? Also, which series are most recommended, and where can I source them to begin with?

BeastOfSoda wrote on 2021-06-27, 19:42:

That said, I have done some more research and I seem to understand that I want to go with Japanese brands for the most part (Rubycon, Panasonic, Sanyo and the like), but I still don't know much about which specific ones to get (at equal rating, that is). I am imagining that I need to stick with electrolytic, as for example solid state might have an adverse effect on functionality, right? Also, which series are most recommended, and where can I source them to begin with?

I wouldn't put much thought into choosing capacitors for a soundblaster card. Take any kind of capacitor (preferably a recognized brand like Rubycon, Panasonic, Nippon Chemicon) that is somehow marked as "low ESR", and it should be good enough for that purpose. The differences between the brands and series matter if you build high-power switch mode converters, where the capacitors are carrying high frequency loads (up to several amps at several 100kHz) all the time, but they don't matter at audio frequencies. Audiophiles might possibly hear differences between different low-ESR caps, but Soundblaster 16 cards aren't anywhere near to audiophile grade anyway,.

mkarcher wrote on 2021-06-27, 20:38:

BeastOfSoda wrote on 2021-06-27, 19:42:

That said, I have done some more research and I seem to understand that I want to go with Japanese brands for the most part (Rubycon, Panasonic, Sanyo and the like), but I still don't know much about which specific ones to get (at equal rating, that is). I am imagining that I need to stick with electrolytic, as for example solid state might have an adverse effect on functionality, right? Also, which series are most recommended, and where can I source them to begin with?

I wouldn't put much thought into choosing capacitors for a soundblaster card. Take any kind of capacitor (preferably a recognized brand like Rubycon, Panasonic, Nippon Chemicon) that is somehow marked as "low ESR", and it should be good enough for that purpose. The differences between the brands and series matter if you build high-power switch mode converters, where the capacitors are carrying high frequency loads (up to several amps at several 100kHz) all the time, but they don't matter at audio frequencies. Audiophiles might possibly hear differences between different low-ESR caps, but Soundblaster 16 cards aren't anywhere near to audiophile grade anyway,.

Okay, thanks. While I do not consider myself an audiophile, I still did go out of my way to get a model known for its low output noise and with a sound signature close to the one I had back in '97; I am afraid I belong with the guys concerned that the wrong caps may alter the sound profile, even though all I want out of it is just OPL (which is why I was wondering about solid state caps, as I seem to understand they have a faster release and therefore might saturate differently). Would you have some pointers to that effect?

Also, I have read that if I can't find caps with the same exact specs, it's preferable to go for higher capacitance than lower; is this a good rule of thumb, or are there other considerations to make there?

BeastOfSoda wrote on 2021-06-27, 21:04:

Also, I have read that if I can't find caps with the same exact specs, it's preferable to go for higher capacitance than lower; is this a good rule of thumb, or are there other considerations to make there?

Both the capacity and the ESR of caps used inside the audio path have an effect on the signal. My suggestion was to "get rid of the ESR effect, which is most likely not intended by the sound card designers", but if you are after authentic sound, things are a bit different:

You must keep the capacitance the same, as capacity can easily have a noticable effect on sound.

Also, for sound authenticity, you shouldn't go for "best caps" or lowest ESR, but for "similar ESR as the original", so no way around tossing lots of data sheet of the original caps and finding caps with similar specification. Often, general purpose electrolytics don't specify "ESR", but they might specify the loss ration "tan delta". You can convert ESR into "tan delta" and vice versa for a given capacity, but I expect that data sheets for comparable capacitors either both specify ESR or both specify "tan delta".

You may substitute caps with caps rated for a higher voltage. So if you don't find a 10µF / 16V cap, a 10µF/25V cap will do as well.

mkarcher wrote on 2021-06-27, 21:27:

Both the capacity and the ESR of caps used inside the audio path have an effect on the signal. My suggestion was to "get rid of […]
Show full quote

BeastOfSoda wrote on 2021-06-27, 21:04:

Also, I have read that if I can't find caps with the same exact specs, it's preferable to go for higher capacitance than lower; is this a good rule of thumb, or are there other considerations to make there?

Both the capacity and the ESR of caps used inside the audio path have an effect on the signal. My suggestion was to "get rid of the ESR effect, which is most likely not intended by the sound card designers", but if you are after authentic sound, things are a bit different:

You must keep the capacitance the same, as capacity can easily have a noticable effect on sound.

Also, for sound authenticity, you shouldn't go for "best caps" or lowest ESR, but for "similar ESR as the original", so no way around tossing lots of data sheet of the original caps and finding caps with similar specification. Often, general purpose electrolytics don't specify "ESR", but they might specify the loss ration "tan delta". You can convert ESR into "tan delta" and vice versa for a given capacity, but I expect that data sheets for comparable capacitors either both specify ESR or both specify "tan delta".

You may substitute caps with caps rated for a higher voltage. So if you don't find a 10µF / 16V cap, a 10µF/25V cap will do as well.

Alright, thanks. This is one deep rabbit hole; for now I am writing down what's on there.

The easiest ones are what I also assume to be the least reliable, as they have the brand name Elgen written on the side (which I remember to be one of the nasty brands); another bunch of those have SRG written on them, which according to my research should be Nippon Chemicons.

The ones I'm stumped about are a bunch of blue ones, all of which are rated as 50v 4.7uf and which only show an "R" logo, what looks like a "CE" just to the right and what look like model numbers 1M9440 and 1M9436 (which I haven't found online):

(also, it looks like I can now add "capacitor photography" to my ever-growing list of silly skills.)

Do these caps ring any bell to you?

Can't really help you with the brand identification there.

But otherwise, all good suggestions so far on buying replacements. I wouldn't worry too much about size (I know that's something you asked about earlier). So long as they're radial through-hole, electrolytic, and of the same specs (or a little higher voltage) they'll work. If they show up, and they're larger or smaller than the originals, it won't matter, since they come with incredibly long legs. Which means you can get them to fit in there regardless. Whether you have to bend the pins closer together to get them through the holes if it is a bigger cap than the original, or the opposite and bend the legs so you can reach holes that are further apart if the new caps are smaller.

Another aspect not mentioned when it comes to buying capacitors is their tolerance. This is given as a percentage. The percentage is the amount the capacitor can be different from the labeled rating. Lets say you're buying a 100uf cap that has a 20% tolerance. That means, if you were to test it, it could actually be anywhere from 80uf to 120uf. At least, that's always been my understanding of it. Pretty much all cheap caps are going to be 10 or 20%. Might be worth spending a little more for 5%, but any further is probably not really worthwhile. It's up to you really. Personally, I'd spend more on getting a good brand that has a reputation for a long life before I'd spend it on getting really low tolerance caps.

Edit: Oh, the numbers on the caps there (and many components are similar) likely tell what manufacturing plant they were made in, and the date they were made. So the last bit of it, "9440," for instance, likely means the 40th week of 1994. And I think (it's hard to tell from the pic) the CE is likely the CE certification logo that you'll see on pretty much any reputable piece of electronics.

Alright, I think I have completed my research. I was able to find almost everything, except for those R branded blue capacitors; it is possible that they were manufactured by a rebranded company (for example, I had a time finding data on those Elgen caps because meanwhile I understand that it had been rebranded as Laube). Either way, many datasheets on similar caps seemed to float between 0.10 and 0.12 TAN delta; since I don't have an exact spec in hand, I am just going to cheat a little and assume it is the better one, as that's directly in the audio signal path.

I have triple checked the diagram against the parts, and I do not think I have made any mistakes. It will still be a while before I can do the job, though, as I need to buy the caps as well as proper desoldering equipment (I refuse to go at it with braid wick and hand pump, it's a good excuse as any to upgrade to something more professional).

Hopefully this will be a good starting point to carry out a conservative recap job; I know that some of these caps are actually out of spec (for example, I seem to understand that the caps used around the tea2025b are all wrong, with the specs demanding bipolar input/output caps instead of polarized and requiring 100uf on the feedback lines instead of the Wincrap 47uf ones, but as mentioned, I'm going for the stock sound signature, even though it's admittedly not as crisp as it could be).

Am I missing anything obvious, or should I be good to go?

Sounds good. That image will be handy to others in the future. Nicely done. As far as getting a proper desolding station, that's up to you. But for a simple cap replacement job like this, typically I would hold onto the cap on its side of the board with needle-nose pliers, and on the other side add fresh solder to the pins - to get the old solder flowing - and just pull the cap back through and off the board. If you do it that way, solder wick or a solder sucker is good enough if there's too much solder left over (or if you're determined to completely clean off all the old solder, and only want fresh solder on the board).

Oh, and having flux is always handy. If the caps have been leaking, the solder might be hard to heat up properly, even when applying fresh solder. Flux would help with that. Hell, flux helps with all soldering in general, really. I'd suggest "No-Clean" flux if you can get it. It basically evaporates after a minute or two, since a good portion of it is IPA. Paste flux is messy and requires cleanup after, so I always go for No-Clean if I can.

Kyosho wrote on 2021-06-28, 15:00:

Sounds good. That image will be handy to others in the future. Nicely done. As far as getting a proper desolding station, that's up to you. But for a simple cap replacement job like this, typically I would hold onto the cap on its side of the board with needle-nose pliers, and on the other side add fresh solder to the pins - to get the old solder flowing - and just pull the cap back through and off the board. If you do it that way, solder wick or a solder sucker is good enough if there's too much solder left over (or if you're determined to completely clean off all the old solder, and only want fresh solder on the board).

Oh, and having flux is always handy. If the caps have been leaking, the solder might be hard to heat up properly, even when applying fresh solder. Flux would help with that. Hell, flux helps with all soldering in general, really. I'd suggest "No-Clean" flux if you can get it. It basically evaporates after a minute or two, since a good portion of it is IPA. Paste flux is messy and requires cleanup after, so I always go for No-Clean if I can.

Yup, I've got an off-brand soldering station and I have some basic knowledge about removing parts manually; it's just that I'm looking at 86 solder points to tackle, and the manual method is really time consuming. Also, every time I get crazy nervous about heating up the board and the traces for too long, and it seems that I'm just too clumsy to do a decent job in a reasonable amount of time (and this is after using flux and keeping the iron's head tinned/cleaned). It's clear that I need more time with the iron in hand to begin with, but I'll gladly take any help I can get in making the job easier. Plus, I am planning to take on other projects as well (namely a couple of consoles, including a Sega Genesis which needs a slot replacement and had been sitting there waiting for someone to recap it -- might as well be me).

I have seen solder sucker guns around, I was interested in something like this: https://www.youtube.com/watch?v=nVkXeOJSsck I've heard only good things about them.

About flux, right now I only have a flux pen, which looks like a marker; it doesn't seem to leave much at all on the board, though, so it's a bit disappointing. Do you recommend buying proper paste, then?

Sorry for the double post, I am going to condense in layman's terms something else I've learned over the last couple of days.

I have gone ahead and converted the TAN delta values to ESR, with the following results:

Although the values look strange, especially for the 1uf caps, they are correct: since the TAN delta is expressed as a proportion of the capacitance, the ESR value in ohm looks proportionally large for lower values.

Now, I did encounter issues in finding low tolerance caps with these sorts of high ESR values, but I also learned that these characteristics may or may not matter depending on the application. For example, it looks like to me like the lower left cap on the board, which is connected to the +5v rail, and two 16v 100uf on the top right (connected to the power regulators right there) are used for power delivery, and there may be other ones, so in these cases a different ESR will not affect the sound profile: it's probably more important to use tighter specs and tolerances and lower ESR here in order to deliver clean power.

the places where I feel like faithful specifications are mandatory are wherever a cap is directly in the signal's path, like basically the entire right-hand area of the board. As I understand, altering the ESR here at equal capacitance will have the most direct effect on the sound, from changing how "fast" the audio is (in terms of attack) to the attenuation applied by lowpass filters. I feel like this is where I have to more strictly adhere to specs, as the rounded sound profile and filtering are in my mind what give this board its character. Therefore, I think I will take the trade-off of looser tolerances if it means getting the same ESR.