Sansui big capacitor replacement - disappointment

[Crazy]

Well yeah.........it sounds like wishful thinking to me too.

BUT------>

I'm not shy to ask.

Take your average scientist or doctor as an example. Both act like all they say is fact and they know the score...............but more times than not it's actually theory and they in fact are pretty much kinda clueless & regularly proved wrong even though you'll never get them to admit it. :yes:

You know.......like salt is bad..........then it's not.............then it is sometimes..................lol

So..........I'm always interested in all takes in order to debunk what is cold hard fact as opposed to what is theory.

 

[Morden2004]

No, a stable electrolytic (by this I'm assuming were saying that 'stable' means not bulging, not emitting smoke and not drawing so much current that other components are suffering) does not mean that the unit is operating within spec.

An electrolytic is, like my previous analogy of an engine oil filter, a component that can (and does) alter it's operational parameters as it ages. An oil filter does a poorer job as it get clogged with gunk, and an electrolytic provides less capacity (in farads - the C) and more resistance (in ohms - the R) as it ages.

The reason that electrolytics can (and do) alter the sound you hear is because all circuits behave exactly the same when presented with an audio (A/C) signal; they are RC circuits and if you know the formulas you can easily see where a change in the R and/or C values of a circuit affects the response curve of the circuit. Add more R or less C and you will drastically alter the linearity of the amplifier.

The big problem is that an old electrolytic changes in both the C and the R values as it ages whereas a resisitor usually only changes in the R. Those electrolytics cause double the problems of circuit dynamics.

Paul

 

[grateful]

Don't waste you breath or your bytes. Some people just don't want to learn.

 

[Kjell-B]

No, a stable electrolytic (by this I'm assuming were saying that 'stable' means not bulging, not emitting smoke and not drawing so much current that other components are suffering) does not mean that the unit is operating within spec.

An electrolytic is, like my previous analogy of an engine oil filter, a component that can (and does) alter it's operational parameters as it ages. An oil filter does a poorer job as it get clogged with gunk, and an electrolytic provides less capacity (in farads - the C) and more resistance (in ohms - the R) as it ages.

The reason that electrolytics can (and do) alter the sound you hear is because all circuits behave exactly the same when presented with an audio (A/C) signal; they are RC circuits and if you know the formulas you can easily see where a change in the R and/or C values of a circuit affects the response curve of the circuit. Add more R or less C and you will drastically alter the linearity of the amplifier.

The big problem is that an old electrolytic changes in both the C and the R values as it ages whereas a resisitor usually only changes in the R. Those electrolytics cause double the problems of circuit dynamics.

Paul

Then I would like to have an explanation why I measure +50 % capacity on many electrolytic capacitors after 60 years. That is, the manufacturer states that the capacitor may have a capacitance range from - 20 % to + 50 % in relation to nominal capacitance. When new, + 50 % would often be the normal value.

I never measure decreasing capacitance on the good old capacitors. Either they show full capacitance or they are (seldom) defective.

"Good old" capacitors in my hands are European 1950-70 and Japanese 1970-80.

Is is VERY EASY to check if a capacitor is within spec, there is no mystery about it. Since many vintage enthusiasts are DIYer's, a capacitance meter is a must. Then check for yourself and determine these big problems. I can promise the readers that differences in capacitor sound has relatively little to do with changes in R/C values.

 

[QuadBob]

Can't really say I was ever on the other side of the issue...I suppose that my unfamiliarity with your work, and a level of astonishment as to the number of caps in the TOTL Sansui quads caught me off guard. But...good post. Shows a lot more patience than I.

Hey Glenn..............you're right, that post actually looks a little dated now doesn't it!:thmbsp:

My main point in referencing it was simply to note that there is nothing wrong with skepticism, however too often we see folks who simply deny reality.

And, unfortunately it seems all to often that when it comes to "audio" either past or present, there is way too much "snake oil" most of the time. Being a formally trained engineer I like to consider my self pretty skeptical as well. That is why I didn't mind posting the info.

Scientific reality IS NOT snake oil. And, while the aging of our vintage equipment has not (and most likely will never be) studied in a totally scientific manner. I've approached it with that in mind for the last 20 years.

I think I also have the benefit as my first love was the Sansui QRX-9001 to have examined "up close" and personal over 60 of this exact model. I've tracked trends in failures by having the benefit of keeping in contact with the owners and also noting the history of specific units when I can.

It brings to mind a factory service school I went to years ago while in the Navy. The training was on a dedicated Chest Xray system made by Dupont. We had 3 of them in use here in San Diego.....one at the Marine Corps Recruit Depot, one at the Navy boot camp, and one at the Armed Forces induction center where they were used for Chest Xray screening for TB, etc.
Anyway, in the service school the trainers mentioned parts that "wear" out in the units over time. Those of us from the Navy questioned other parts that we saw failing routinely. At first the instructor tried to tell us the parts we saw failing were an aberration.....until we finally "compared notes". Dupont had studied wear on the units based upon 20,000 exams per year over a two year period. We then calmly informed them that are units averaged 5,000 exams PER MONTH or 60,000 per year!! They then began picking our brains for design changes!!!:scratch2:

The main point of which is simply that while making comparisons based upon individual units is certainly valid.........being able to "sample" and evaluate a larger group is always much more telling and true trends can be noted.

QB

 

[QuadBob]

Unfortunately I now have to decide if vintage stereo equipment is still going to give me the " big bang for the buck" I thought I was getting. ( I don't need another "expensive" hobby as building choppers covers that)

3 - How about prices buying this stuff since we have to take into account the cost of total recaps to preserve the equipment and prevent further damage.
Are some of these selling prices we see really fair in context??? Would people pay these prices if they realized how much money & time was needed to truly bring the equipment up to snuff? I mean let's face it.......in context....all this vintage stuff bought on ebay and elsewhere is actually in need of repair if not already restored.

In my opinion............these are "NOW" some of the most important questions when it comes to vintage Hi-Fi equipment.

This really is opening up a big can of worms.:scratch2:

Thanks Crazy, you've covered a lot of ground in this post. And, I'm going to save the "quality" questions for a post specific to that subject.

But, I'll address the two points above now.

"Bang for the buck" is certainly subjective..........How much bucks you got Man? But, overall I'd say it certainly is "if" you're rational about it!

First of all.........the quality of construction and "discrete" components of our vintage equipment bests most any "consumer grade" equipment made today. You can certainly find as good both in design and components in todays "audiophile" equipment.....but then you're most likely looking at individual units.....amps, preamps, phono stages, that can easily cost $5k and up today.

And, just have many have noted the "trend" to cheaper construction over Sansui's production period from 1970-1985. Today there seems to be a tremendous gap in quality of todays equipment. Basically there is no "mid-fi"....its either BPC or expensive audiophile equipment.......with a slight "tip of the hat" to companys like Outlaw audio who really do seem to be trying to fill in the gap.

But there is no doubt.........there is no "free ride". And, certainly not everything that was produced in the '70s is worth the investment of a total restoration. Many of us do however feel that the "top of the line" and best designs from then are certainly worth the time and expense to keep them running and sounding as great as they possibly can!

As to the prices being paid for this equipment today.......who's to say really? New highs are being seen all the time now........and as more and more of them go to the graveyard, they'll most likely just become more rare and expensive in the future. I do tell those that I restore for that it takes just as much parts and labor to properly restore the QRX-9001 that they paid $100 for as the one that goes for $1,000 with the original box.....and my recommendation is to look for an honest seller that reveals a unit does have problems (therefore sells cheaper!) but is in the best physical condition they can find. And, they factor in the cost of restoration with their purchases so when it is all said and done........they end up with a beautiful piece of audio history that both looks and sounds great!

QB

 

[QuadBob]

As I continued to recap all the units that came in my possession I got lazy and started skipping the large main power filter caps in units that they appeared the most difficult to do so (Yamaha CR-1000, CA-1000).

Having replaced every other cap in these units, is it worth going back and replacing these large can caps? I was moderately frightened by QBs massively informative and appreciated post.

From a strictly technical viewpoint...........yes.

However, it can be said also that those large main caps (with their 20% tolerance) might still survive at least for awhile longer. Depending on the actual values in a specific unit........I tend to look first to see how its doing, measure ESR....measure "underload" voltage and determine whether it is still adequate for the load its driving. While I fully expect some amount of "roll-off" the good news is that these designs are fairly tolerant and robust.

It often becomes a matter of for example a 9090db that is rated at 125wpc....and can easily bench to 160wpc or more, are you really going to be pushing it so hard that you'll notice or strain those main caps trying to push it to its limits.

It is really much more important to protect those "low powered" circuits where the most sensitive components reside, as well as the actual signal path where final sound quality is determined.

QB

 

[QuadBob]

I also want to thank QB for bringing a LOT to the mix with his detailed reasoning on the subject. I've read his contributions elsewhere and he never fails to add clarity to whatever he discusses.

I was going to start a new thread with this following question, but I think it might fit in here so I ask those who know:

I have a gorgeous AU-777A that's definitely worthy of restoring. However, it has the typical "weak channel" on one side.
My question is this; should I diagnose the problem and fix that before going any further with restoration, or should I just go ahead and replace all the caps and suspect resistors first, and THEN go looking for the bad channel problem.
I'm really just asking for an opinion on recommended procedure of repair and restoration in order to do it in the most efficient way.

This question actually applies to ANY unit that has a particular "problem", but one that is also in line for a full re-cap restoration anyway.

TIA

I think the answer to your question Mark really depends on whether you are a "hobbiest" with limited technical ability or an experienced "pro" who knows how to dig yourself out of a bad "bug".

For most hobbiests.........fixing any original "hard failure" problems first would be the way to go. First off.........if you can't resolve the original problem, you probably should get it to someone else's bench. But, more importantly....if you start recapping a "working unit" and then hit problems, you KNOW its something that you did in the process and you can then concentrate on finding your mistake.

Paul made a good point as well.......about working from the output back, again just a means to keep you organized and able to verify your work.

For myself........I don't usually even bother to "turn-on" a unit that is undergoing restoration. I totally re-cap it first.......then troubleshoot any "bugs" that remain afterward. It's simply more efficient, and it allows me to "debug" a unit ONCE before its finished.

Things like doing the amp calibration.......are often tricky "unrestored" and will have to be done once its recapped anyway, so I simply eliminate the caps as a problem area before doing it. In a related vein for instance....the fuseable resistors on the driver pcbs of most Sansui amps have drifted way out by now in unrestored units. We replace all of them immediately during a restoration.....I don't even bother measuring them anymore. That way I start off with "new" resistors there before even powering it up for the first time.

It's a pretty subjective area.......so others here like EchoWars might have something to add.

QB

 

[Morden2004]

Then I would like to have an explanation why I measure +50 % capacity on many electrolytic capacitors after 60 years. That is, the manufacturer states that the capacitor may have a capacitance range from - 20 % to + 50 % in relation to nominal capacitance. When new, + 50 % would often be the normal value.

I never measure decreasing capacitance on the good old capacitors. Either they show full capacitance or they are (seldom) defective.

"Good old" capacitors in my hands are European 1950-70 and Japanese 1970-80.

Is is VERY EASY to check if a capacitor is within spec, there is no mystery about it. Since many vintage enthusiasts are DIYer's, a capacitance meter is a must. Then check for yourself and determine these big problems. I can promise the readers that differences in capacitor sound has relatively little to do with changes in R/C values.

The important point that I was hoping to make in my comments is that a capacitor is BOTH a capacitor and a resistor - a fact that is often overlooked in discussions about how an old electrolytic affects the circuit dynamics.

Over the years I learned (slowly) that just because an electrolytic measures (farad-wise) within the manufacturers specs is simply not enough to quantify a capacitor as "well". You must also know what the R factor is (often refered to as the ESR) because that value will absolutely affect the circuit parameters.

Getting real techincal, even a "new" electrolytic has a specific ESR and, if the circuit is frequency sensitive, you could actually worsen the case with a new improperly-spec'd electrolytic -- usually it's the operative voltage rating that makes the difference. Of course this is truly an obscure possibility -- but possible never-the-less. :scratch2:

Food for thought.

Paul

 

[xoaphexox]

usually it's the operative voltage rating that makes the difference

So if I am reading this correctly, the higher the voltage rating of a cap, the higher the ESR? What is the maximum margin you would increase a caps voltage by? (When recapping I usually use the next-higher-up voltage when I cannot find one of a certain value)

I think the 160V caps I replaced in a crossover were replaced by Solen 400V polypropylene caps. Is ESR not an issue as much with polys as it is with lytics?

 

[Kjell-B]

The important point that I was hoping to make in my comments is that a capacitor is BOTH a capacitor and a resistor - a fact that is often overlooked in discussions about how an old electrolytic affects the circuit dynamics.

Over the years I learned (slowly) that just because an electrolytic measures (farad-wise) within the manufacturers specs is simply not enough to quantify a capacitor as "well". You must also know what the R factor is (often refered to as the ESR) because that value will absolutely affect the circuit parameters.

Getting real techincal, even a "new" electrolytic has a specific ESR and, if the circuit is frequency sensitive, you could actually worsen the case with a new improperly-spec'd electrolytic -- usually it's the operative voltage rating that makes the difference. Of course this is truly an obscure possibility -- but possible never-the-less. :scratch2:

Food for thought.

Paul

Then please supply the following information:

1) What was the ESR value of vintage capacitors when they were new. And how large variations can be permitted ? A few examples will be sufficient.

2) How often have you (or others) found old capacitors with original good capacitance (say + 50 % value), no leakage, and still a defective value of high ESR ?


As a general information people should know that ESR measurements have come as a response to the high failure rate of modern electrolytic capacitors in modern computer circuits (for extended battery life) or switch mode power supplies where an extremely low ESR just is part of the design. We are then talking of ESR values much lower than was necessary (or wanted) in the good vintage audio amplifiers.

In measuring ESR a 100 kHz signal is used, just for the application on computer service and similar. I am afraid the relevance to vintage audio circuits is limited - but the pure existence of simple ESR meters make measurements easy to do and publish. Another interesting story. I am awaiting an explanation on the relevance of ESR, when we already easily can measure leakage current and capacitance. And yes, I know that ESR and leakage are not connected.

 

[bolly]

Kjell-B, I bet your Turberrys sound excellent!

 

[Kjell-B]

Kjell-B, I bet your Turberrys sound excellent!

Yes, they are great. And they have everything, bottom, midrange, dynamics, control, details. I plan on adding a home-made super-tweeter (Pioneer vintage horn) - have one working now. Still I would love to have a comparison to the vintage Tannoys.

I have to admit that my house is filled with vintage stuff ......

 

[Crazy]

Thanks to all...............

WOW-----> one thing for sure....this thread is one of the most interesting I've seen in a while here. I LOVE IT! :yes:

Quadbob, Morden2004 and others have submitted much food for thought.

Especially THANKS to Kjell-B --> :thmbsp: --> who has injected a great amount to consider.

I only hope that others here answer Kjell-B take on the subject.

With my being one of the most clueless members here, I eagerly await more with this thread as I do find it even more fascinating as it proceeds and love to expand of my knowledge on this very important subject.Sometimes being clueless is an advantage as it makes you more open to seeing the truth as a clueless person has no die hard preconceived notions to get out of the way.

There is a big learning curve here for me.............but I think it's worth the effort as I would like to not only enjoy my equipment forever, but I'd also like to do only what really needs done to preserve it in the process and therefore get the most bang for the buck. I ain't looking for a free ride.....but I would like a low cost ride if possible as that is part of the enjoyment for me.

Guess I need to shop for a capacitance meter, some books on the subject.

Again I only hope that all the theories here are tackled in full as everyone will only benefit from it.

Thanks again for everyones time in this thread.:thmbsp:

 

[Morden2004]

So if I am reading this correctly, the higher the voltage rating of a cap, the higher the ESR? What is the maximum margin you would increase a caps voltage by? (When recapping I usually use the next-higher-up voltage when I cannot find one of a certain value)

I think the 160V caps I replaced in a crossover were replaced by Solen 400V polypropylene caps. Is ESR not an issue as much with polys as it is with lytics?

To the best of my knowledge the ESR changes over time only in an electrolytic -- hence the changing dynamics of any circuit with an old(er) electrolytic.

Although all capacitors exhibit a R factor (it's the law :yes: ) it's the electrolytics that change as they age.

Paul

 

[Morden2004]

Then please supply the following information:

1) What was the ESR value of vintage capacitors when they were new. And how large variations can be permitted ? A few examples will be sufficient.

I have here in my hand two recent removals; both are 33uF 25VDC and should measure about 2-4 ohms ESR. One measures 12 ohms ESR and the other measures 16 ohms ESR.

2) How often have you (or others) found old capacitors with original good capacitance (say + 50 % value), no leakage, and still a defective value of high ESR ?

Almost 100% of the time for 30 year old gear.

Do all electrolytics with a bad ESR cause problems in vintage equipment? Probably not, but series resistance could, IMHO, be a factor in changing the frequency response of circuits where the capacitors are not strictly used as filters. It's all theory, :scratch2: but these devices are cheap.

Paul

 

[Kjell-B]

I have here in my hand two recent removals; both are 33uF 25VDC and should measure about 2-4 ohms ESR. One measures 12 ohms ESR and the other measures 16 ohms ESR.

Almost 100% of the time for 30 year old gear.

Do all electrolytics with a bad ESR cause problems in vintage equipment? Probably not, but series resistance could, IMHO, be a factor in changing the frequency response of circuits where the capacitors are not strictly used as filters. It's all theory, :scratch2: but these devices are cheap.

Paul

Thanks for the interesting information. It is important to know some figures in a debate like this. Still I think the ESR values you show are of relatively little relevance (sorry ).

And you may have changed your argument somewhat - I site your earlier comment:
"An electrolytic is, like my previous analogy of an engine oil filter, a component that can (and does) alter it's operational parameters as it ages. An oil filter does a poorer job as it get clogged with gunk, and an electrolytic provides less capacity (in farads - the C) and more resistance (in ohms - the R) as it ages."

So now we can agree that old capacitors often hold their capacitance well. Your point is the decreasing R value. I will proceed on that part of the debate.

Your capacitor 33uF 25VDC may typically be placed in a circuit with impedance 1000 - 10 000 ohms or more, and if the capacitor holds an internal resistance of 4 ohms or 16 ohms will not change the circuit function. Replacing it will however change the sound from a typically vintage sound to a sound directed by modern principles of making computer capacitors.

So even if a new capacitor is cheap I myself would be very cautious to change the caps. This if the original sound is a requisite. If not, the circuit will surely work very well using the new caps.

And your ESR figures are values obtained at 100 kHz, that may also reduce the importance.

 

[Morden2004]

Thanks for the interesting information. It is important to know some figures in a debate like this. Still I think the ESR values you show are of relatively little relevance (sorry ).

And you may have changed your argument somewhat - I site your earlier comment:
"An electrolytic is, like my previous analogy of an engine oil filter, a component that can (and does) alter it's operational parameters as it ages. An oil filter does a poorer job as it get clogged with gunk, and an electrolytic provides less capacity (in farads - the C) and more resistance (in ohms - the R) as it ages."

So now we can agree that old capacitors often hold their capacitance well. Your point is the decreasing R value. I will proceed on that part of the debate.

Your capacitor 33uF 25VDC may typically be placed in a circuit with impedance 1000 - 10 000 ohms or more, and if the capacitor holds an internal resistance of 4 ohms or 16 ohms will not change the circuit function. Replacing it will however change the sound from a typically vintage sound to a sound directed by modern principles of making computer capacitors.

So even if a new capacitor is cheap I myself would be very cautious to change the caps. This if the original sound is a requisite. If not, the circuit will surely work very well using the new caps.

And your ESR figures are values obtained at 100 kHz, that may also reduce the importance.

As I said, it's strictly my opinion based upon theory. I have not swept a circuit with old electrolytics that had ESR problems and then swept it again after replacement. I don't have any hard evidence other than the real fact that the ESR has drifted well off in old electrolytics from 'usual values' as compared to a new electrolytic. I'm predicting that it must count for something. I'm not a big believer that one 'brand' of electrolytic can color a musical (electrical) signal to some magical degree over others.

And I won't accept material gathered from the internet for obvious reasons. If someone has any real hard evidence one way or the other it would be nice to review it here. :thmbsp:

Paul

 

[Morden2004]

Further, I don't agree that old electrolytics "keep their capacitance" as a rule. I do agree that many will measure farads values within tolerance with a capacitance tester -- but that isn't under the stress of working voltages and temperatures.

Again, these devices do age according to many who know far more than I. It's a chemical problem in electrolytics that is hard to downplay. Heat and time degrade the chemicals that do the work.

Paul

 

[Morden2004]

Excellent thread

Yes, we have differing opinions but to parrot several others in the thread -- this has been an excellent discussion!

Paul