IPB

Welcome Guest ( Log In | Register )

 
Reply to this topicStart new topic
replace electrolytic capacitors with solid state, Do i need to replace capacitors on a like for like basis?
john11
post Dec 21 2012, 06:22
Post #1





Group: Members
Posts: 43
Joined: 26-February 11
Member No.: 88511



Hi and thanks for reading this post.

Do you think it would be a good idea to replace the usual aluminium electrolytic capacitors on a Zero dac board for the better specs of a solid state (polymer) type capacitor.

I have a Zero dac which reproduces digital sound signals from my pc's usb port and it has 4 aluminium electrolytic's around the dac chip (47uf 35v) which are swollen and leaking and 4x 10uf 63v output coupling caps in the analog section, so 8 caps to replace in all, do you think a like for like replacement of the stock aluminium electrolytics for solid state polymer type caps would be okay in this situation.

I have done a google search and people have performed this on pc motherboards but there is not much on other types of equipment.

Many thanks in advance. John.
Go to the top of the page
+Quote Post
Speedskater
post Dec 21 2012, 23:09
Post #2





Group: Members
Posts: 139
Joined: 12-December 07
From: Cleveland, OH
Member No.: 49500



I don't think that many people are familiar with these new solid state polymer capacitors. Especially with regards to audio equipment. I never heard of them until today.

On second thought, I found references back to 2008 with some on audio forums.

This post has been edited by Speedskater: Dec 21 2012, 23:13


--------------------
Kevin
Go to the top of the page
+Quote Post
saratoga
post Dec 22 2012, 02:02
Post #3





Group: Members
Posts: 5003
Joined: 2-September 02
Member No.: 3264



The ESR won't be quite the same since the chemistry isn't the same. If they're part of a filter it will probably shift the center frequency very slightly. If they're just for the digital logic, I doubt it matters.
Go to the top of the page
+Quote Post
Dynamic
post Dec 22 2012, 04:26
Post #4





Group: Members
Posts: 824
Joined: 17-September 06
Member No.: 35307



I'd have thought they'd be power supply smoothing caps to cope with maintaining the DC supply voltage during current spikes. They're not particularly likely to be involved in the signal path. You could try to trace which pins of the chips they're connected to.

Those polymer alumin(i)um or tantalum caps might be worth a look. They'll probably have better Equivalent Series Resistance and high temperature performance, but probably check the data sheets versus the originals for no obvious spec shortfalls, but ensure you at least match or exceed the voltage specs. Also take care that you can fit them into the footprint and pads, which will presumably be through-hole rather than surface mount, but likely being smaller they might require longer legs to span the same distance, rather than sitting directly on the board. Also be sure to match the polarity of the caps you're replacing.


This post has been edited by Dynamic: Dec 22 2012, 04:30
Go to the top of the page
+Quote Post
Glenn Gundlach
post Dec 22 2012, 10:18
Post #5





Group: Members
Posts: 369
Joined: 19-April 08
From: LA
Member No.: 52914



QUOTE (Dynamic @ Dec 21 2012, 19:26) *
I'd have thought they'd be power supply smoothing caps to cope with maintaining the DC supply voltage during current spikes. They're not particularly likely to be involved in the signal path. You could try to trace which pins of the chips they're connected to.

Those polymer alumin(i)um or tantalum caps might be worth a look. They'll probably have better Equivalent Series Resistance and high temperature performance, but probably check the data sheets versus the originals for no obvious spec shortfalls, but ensure you at least match or exceed the voltage specs. Also take care that you can fit them into the footprint and pads, which will presumably be through-hole rather than surface mount, but likely being smaller they might require longer legs to span the same distance, rather than sitting directly on the board. Also be sure to match the polarity of the caps you're replacing.


The main reason polymer caps aren't used more frequently is they often cost 10 times more than plain old aluminum. Over the last 8 years I've personally replaced over 5000 caps and always buy buy the lowest ESR (highest ripple current) caps I can get that match the mechanical footprint. For through hole I prefer Panasonic FR, FM or FC in that order and for SMTs Panasonic FP and FK. When replacing caps on motherboards I find the tall skinny units from Nichicon at Mouser, again going for highest ripple current.

FWIW my Gigabyte boards (now 4 years old) have polymer caps in the power converters because of the much better life expectancy.

Be cautious with tantalums. It's much better to overrate the Voltage by about 3 because if they fail they turn into dead shorts. Keeping the max Voltage <33% will have nearly no failures.

Particularly on motherboards, a GOOD soldering iron is a must. I own 3 Metcals and after you use one of those everything else is a 'tusk'. We're getting an 80 Watt 5000 series at work as there are times the normal 40 Watter isn't quite enough.

G
Go to the top of the page
+Quote Post
Arnold B. Kruege...
post Dec 22 2012, 15:49
Post #6





Group: Members
Posts: 3822
Joined: 29-October 08
From: USA, 48236
Member No.: 61311



QUOTE
Do you think it would be a good idea to replace the usual aluminium electrolytic capacitors on a Zero dac board for the better specs of a solid state (polymer) type capacitor.


Rule number one of electronics maintenance: If it is not broken, don't fix it.

So, what's broken?

QUOTE
I have a Zero dac which reproduces digital sound signals from my pc's usb port and it has 4 aluminium electrolytic's around the dac chip (47uf 35v) which are swollen and leaking


Those are obviously broken, replace with close equivalents.

QUOTE
4x 10uf 63v output coupling caps in the analog section,


It it not unusual for the caps of a similar kind in a piece of gear to be on slightly different places in the same timetable. It is arguable that since 1/2 of the caps in the thing are broken, replace them all.

QUOTE
so 8 caps to replace in all, do you think a like for like replacement of the stock aluminium electrolytics for solid state polymer type caps would be okay in this situation.


Yes. Ordinarily electrolytic capacitors last for 20-40 years or more. When they fail within a year or two then they were defective when new. There was a rash of electrolytic capacitor failures starting about 5 years ago due to some stolen technology that was implemented incorrectly after being stolen. PC motherboards among other things had a period of time of ludicrously low reliability.

QUOTE
I have done a google search and people have performed this on pc motherboards but there is not much on other types of equipment.


Actually there is quite a bit of audiophile myth about capacitor upgrades. Here is the article that arguably kicked off this long-lasting but regrettable fad. Note it is a mixture of truth and untruth:

http://www.reliablecapacitors.com/pickcap.htm

**Warning! Warning! Again, do not take the above article at face value! Warning! Warning!**

Go to the top of the page
+Quote Post
Dynamic
post Dec 22 2012, 17:03
Post #7





Group: Members
Posts: 824
Joined: 17-September 06
Member No.: 35307



QUOTE (Glenn Gundlach @ Dec 22 2012, 09:18) *
[The main reason polymer caps aren't used more frequently is they often cost 10 times more than plain old aluminum. Over the last 8 years I've personally replaced over 5000 caps and always buy buy the lowest ESR (highest ripple current) caps I can get that match the mechanical footprint. For through hole I prefer Panasonic FR, FM or FC in that order and for SMTs Panasonic FP and FK. When replacing caps on motherboards I find the tall skinny units from Nichicon at Mouser, again going for highest ripple current.

FWIW my Gigabyte boards (now 4 years old) have polymer caps in the power converters because of the much better life expectancy.

Be cautious with tantalums. It's much better to overrate the Voltage by about 3 because if they fail they turn into dead shorts. Keeping the max Voltage <33% will have nearly no failures.

Particularly on motherboards, a GOOD soldering iron is a must. I own 3 Metcals and after you use one of those everything else is a 'tusk'. We're getting an 80 Watt 5000 series at work as there are times the normal 40 Watter isn't quite enough.

G


Glenn clearly has a lot of direct experience including reliability issues and failure modes. Certainly unit cost is the major factor in most consumer electronics providing reliability is acceptable. For repairs and small volume work, there's less to gain by going cheaper when you could go better. I'd take his advice and go for those Panasonics or Nichicon from a reputable supplier who hasn't bought a load of fake parts from "Wun-Hung-Lo" (to borrow Dave Jones' fictional Shenzhen factory name)

As Arnold says there's a lot of bunkum in audiophile circles around capacitors, but in power supply decoupling, conventional electrolytics and polymers are perfectly fine.

This post has been edited by Dynamic: Dec 22 2012, 17:07
Go to the top of the page
+Quote Post
john11
post Dec 24 2012, 09:16
Post #8





Group: Members
Posts: 43
Joined: 26-February 11
Member No.: 88511



Hi. Thanks for the replies.

You are absolutely right concerning the panasonic fm series. I placed a few in a power supply last year and they were brilliant. The only problem is that they only go up to 50v.

About the panasonic fp and fk series i have found a few on mouser and farnel, in the values i need the esr is 0.18 ohms and cost 50 pence each. The nichicon rhu1h470mdn1ph is $2.50, okay a little more expensive but the esr is 0.02 ohms which is about a tenth of the panasonics. The panasonics do sound a little high in esr to me for a solid state capacitor so i am torn between the two deciding which to buy.

Talking about things to buy can anyone recomend a good desoldering iron for this job. I purchased this:

http://www.ebay.co.uk/itm/2-In-1-Welding-E...=item46019b4760

Thing is it melts the solder but the suction is very weak and it does not pick up any of the solder at all, can anyone recommend a 2-in-1 tool like this but with better suction. I don't want to spend a lot of money on a desoldering station for a one off job.

Many thanks. John.

Go to the top of the page
+Quote Post
Glenn Gundlach
post Dec 24 2012, 10:53
Post #9





Group: Members
Posts: 369
Joined: 19-April 08
From: LA
Member No.: 52914



QUOTE (john11 @ Dec 24 2012, 00:16) *
Hi. Thanks for the replies.

You are absolutely right concerning the panasonic fm series. I placed a few in a power supply last year and they were brilliant. The only problem is that they only go up to 50v.

About the panasonic fp and fk series i have found a few on mouser and farnel, in the values i need the esr is 0.18 ohms and cost 50 pence each. The nichicon rhu1h470mdn1ph is $2.50, okay a little more expensive but the esr is 0.02 ohms which is about a tenth of the panasonics. The panasonics do sound a little high in esr to me for a solid state capacitor so i am torn between the two deciding which to buy.

Talking about things to buy can anyone recomend a good desoldering iron for this job. I purchased this:

http://www.ebay.co.uk/itm/2-In-1-Welding-E...=item46019b4760

Thing is it melts the solder but the suction is very weak and it does not pick up any of the solder at all, can anyone recommend a 2-in-1 tool like this but with better suction. I don't want to spend a lot of money on a desoldering station for a one off job.

Many thanks. John.


The main issue with a soldering iron is getting plenty of heat when needed but NOT going over temperature meaning you want a tool smart enough to be low Wattage when idling or doing small joints but be high power on the larger connections. The old Weller Magnastat units were built starting in the mid/late '60s and are still built. I own 2 though one needs repair. The problem with them is the limited RATE of power supplied meaning the tip cools significantly when touching the work. The thermostat then turns on the power and it begins to recover. The Weller is around 25 Watts when pushed but idles at the right temperature. The Metcals work in a different fashion as the heating element is also the thermostat. The 40 Watt power unit is actually a 13.5 MHz transmitter that puts out a constant current. When the tip is below temp the resistance goes up, the Voltage and power goes up. When it reaches temperature the resistance drops, the Voltage and power drops and it holds at the proper temperature. The operational difference is difficult to believe. If you never used a soldering tool you wouldn't be impressed as it does what it should. If you spent a lot of time with less impressive tools you'll get it. While I own the Metcal vacuum desoldering tool, I most ofter just use an old Edsyn Soldapullt. Keep it clean with a good tip and it works well enough particularly when using a Metcal sttc-126 tip. The small tip with the 30 bend fits nicely in the hole and backing if out a few thousandths and the tripping the Edsyn will clean nearly any hole and because the Metcal doesn't overheat, very rarely lifts a pad. End sales pitch. BTW I have no affiliation with Metcal other than being a very satisfied customer. Between 2 jobs and my own units I've caused 15 units to be purchased. All came from eBay and only one was actually new. When operating properly (if not it's usually a bad tip) they'll be melting solder in 6-7 seconds from power up.

Merry Christmas

G

This post has been edited by Glenn Gundlach: Dec 24 2012, 10:54
Go to the top of the page
+Quote Post
Arnold B. Kruege...
post Dec 24 2012, 13:49
Post #10





Group: Members
Posts: 3822
Joined: 29-October 08
From: USA, 48236
Member No.: 61311



QUOTE (john11 @ Dec 24 2012, 03:16) *
Hi. Thanks for the replies.

You are absolutely right concerning the panasonic fm series. I placed a few in a power supply last year and they were brilliant. The only problem is that they only go up to 50v.

About the panasonic fp and fk series i have found a few on mouser and farnel, in the values i need the esr is 0.18 ohms and cost 50 pence each. The nichicon rhu1h470mdn1ph is $2.50, okay a little more expensive but the esr is 0.02 ohms which is about a tenth of the panasonics. The panasonics do sound a little high in esr to me for a solid state capacitor so i am torn between the two deciding which to buy.

Talking about things to buy can anyone recomend a good desoldering iron for this job. I purchased this:

http://www.ebay.co.uk/itm/2-In-1-Welding-E...=item46019b4760

Thing is it melts the solder but the suction is very weak and it does not pick up any of the solder at all, can anyone recommend a 2-in-1 tool like this but with better suction. I don't want to spend a lot of money on a desoldering station for a one off job.


A great deal of work has been done over the years by many technicans with a standard soldering iron and a dedicated solder vacuuming tool. It is 2 pieces and keeps both hands busy, but it works!

The traditional brand name has been the Edsyn SOLDAPULLT. AFAIK they are still sold for competitive prices. My plain old plastic one is maybe 20 years old and stll gets the job done.

As far as soldering irons go I have gone through two temperature controlled soldering stations and currently do most of my work with a old-fashioned soldering iron with just a handpiece.

I think that if I was doing heavy assembly work that required soldering I would go back to a higher end 2-piece station, but that's not where the work is. I do a lot more crimping these days...

The trick to soldering is getting the temperature right for the solder and joint at hand. I've been soldering for over 55 years since I was 8 and doing it right is intuitive for me so I'm probably a bad source of advice about these things. That doesn't mean that I still don't occasionally screw up.

I think that for newbies the key is practice. Learn what a good joint looks like and solder up nonsense projects and then small non-critical projects until you are competent enough to be comformatble with more complex projects. Anybody who has never soldered and then picks up a project on a modern piece of elecronics with SMT parts is shall we say wildly optimistic about their learning curve. ;-)

As far as heat goes, IME you can do well with simple tools if there is the right balance between temperature, application of new heat, and use of heat stored in the metal parts of the soldering iron's tip. A larger soldering iron element (30-40 watts) with a fine tip is generally better than a smaller heating element.

This post has been edited by Arnold B. Krueger: Dec 24 2012, 13:57
Go to the top of the page
+Quote Post
Glenn Gundlach
post Dec 25 2012, 08:20
Post #11





Group: Members
Posts: 369
Joined: 19-April 08
From: LA
Member No.: 52914



Arny, try a Metcal and I ASURE you you'll have a change of heart.

Merry Christmas everyone

G
Go to the top of the page
+Quote Post
john11
post Dec 25 2012, 08:37
Post #12





Group: Members
Posts: 43
Joined: 26-February 11
Member No.: 88511



Hi. Thanks for the replies.

Can anyone support Glenn with his preference of the panasonic fp and fk series.

47uf panasonic fk esr=0.2 ohms @11 pence each (smt)
47uf panasonic fm esr=0.3 ohms @ 50pence each (aluminium electrolytic)
47uf Nichicon RNU1H47 esr= 0.02 ohms @ $2.50 each (smt)

It appears the nichicon has ten times less esr than the panasonic fk, but looking at it another way the esr difference is 0.18 ohms is this something significant something the human ear can detect.

On the farnell website when the caps are listed in order of esr, the lowest esr ones are the most expensive is there something to that.

Many thanks. John.
Go to the top of the page
+Quote Post
Glenn Gundlach
post Dec 25 2012, 09:18
Post #13





Group: Members
Posts: 369
Joined: 19-April 08
From: LA
Member No.: 52914



QUOTE (john11 @ Dec 24 2012, 23:37) *
Hi. Thanks for the replies.

Can anyone support Glenn with his preference of the panasonic fp and fk series.

47uf panasonic fk esr=0.2 ohms @11 pence each (smt)
47uf panasonic fm esr=0.3 ohms @ 50pence each (aluminium electrolytic)
47uf Nichicon RNU1H47 esr= 0.02 ohms @ $2.50 each (smt)

It appears the nichicon has ten times less esr than the panasonic fk, but looking at it another way the esr difference is 0.18 ohms is this something significant something the human ear can detect.

On the farnell website when the caps are listed in order of esr, the lowest esr ones are the most expensive is there something to that.

Many thanks. John.


The Nichicons are organic polymers and the unit price at Mouser is $3.95 Where one would use this kind of cap is in switching power supplies where normally 'excellent' capacitors become only 'adequate'. You need really good caps in switchers and high dynamic loads like video cards, computer motherboards and high speed digital processing. Analog audio use with linear supplies is comparatively non critical.

In all engineering there are trade offs. You'd like infinite life and Voltage, 0 ESR, low cost, small size but you have to trade. Since I replace things I'd prefer not to do it too often per machine so long life is more a priority. ESR / ripple current translates directly to internal heating. Heat kills components so reduce temperature and internal heating.

When I buy SMT caps I get the highest Voltage for a given capacitance in the standard size packages with highest ripple current so I can keep a smaller number of parts that cover a larger range of replacements. Looking at Digikey just now I found the TP and FT series I was unaware of so I'll be checking those out.

I'm sure there are many other caps but when I get good results I tend to stick with a brand until they burn me. So far good results.

Merry Christmas

G
Go to the top of the page
+Quote Post
Glenn Gundlach
post Dec 25 2012, 09:20
Post #14





Group: Members
Posts: 369
Joined: 19-April 08
From: LA
Member No.: 52914



Sorry about the extra post. How do you delete?

This post has been edited by Glenn Gundlach: Dec 25 2012, 09:22
Go to the top of the page
+Quote Post
Arnold B. Kruege...
post Dec 25 2012, 13:59
Post #15





Group: Members
Posts: 3822
Joined: 29-October 08
From: USA, 48236
Member No.: 61311



QUOTE (Glenn Gundlach @ Dec 25 2012, 02:20) *
Arny, try a Metcal and I ASURE you you'll have a change of heart.

Merry Christmas everyone

G


At $600-900 per station I would need a change of life's situation and occupation. ;-)
Go to the top of the page
+Quote Post
Glenn Gundlach
post Dec 25 2012, 22:46
Post #16





Group: Members
Posts: 369
Joined: 19-April 08
From: LA
Member No.: 52914



QUOTE (Arnold B. Krueger @ Dec 25 2012, 04:59) *
QUOTE (Glenn Gundlach @ Dec 25 2012, 02:20) *
Arny, try a Metcal and I ASURE you you'll have a change of heart.

Merry Christmas everyone

G


At $600-900 per station I would need a change of life's situation and occupation. ;-)


Good grief. Don't pay that ! That's why you buy on eBay. You should be able to get one $150 to $250 though you may have to buy it 'parted out'. I bought 2 broken power supplies and repaired them and still have one in the garage. Make an offer. If you're too high I'll tell you.

Merry Christmas

G
Go to the top of the page
+Quote Post
john11
post Dec 28 2012, 04:05
Post #17





Group: Members
Posts: 43
Joined: 26-February 11
Member No.: 88511



Hi. Thanks for the replies.

I saw on the amp board quite a few diodes probably for biasing voltages and was wondering if these are worth replacing

with something better, fast diodes, schottky?

Have you had any experience of these.

Many thanks in advance. John.
Go to the top of the page
+Quote Post
Glenn Gundlach
post Dec 28 2012, 04:24
Post #18





Group: Members
Posts: 369
Joined: 19-April 08
From: LA
Member No.: 52914



QUOTE (john11 @ Dec 27 2012, 19:05) *
Hi. Thanks for the replies.

I saw on the amp board quite a few diodes probably for biasing voltages and was wondering if these are worth replacing

with something better, fast diodes, schottky?

Have you had any experience of these.

Many thanks in advance. John.


Biasing diodes have specific Voltage vs. temperature characteristics. There is nothing 'better', just different which could require altering other components to compensate for the changes. They are not run in reverse and do not switch so there is nothing to be gained by changing them - unless they have failed.

Sometimes quieter / faster opamps CAN make improvements but not always and can even sometimes degrade performance. I've changed AN6552 amps to OPA2134 units in an old Technics receiver with no downside. The question is, does it sound better? In all honesty I can't say one way or the other. I 'feel' better about it and it measures a little better. Changing the center frequencies of the lower 2 bands of the built in a 5 band graphic equalizer was indeed an improvement. it was 80 Hz / 320 Hz and it's now 25 Hz / 180 Hz. It's very easy to retune gyrator circuits.

G
Go to the top of the page
+Quote Post
Arnold B. Kruege...
post Dec 28 2012, 15:05
Post #19





Group: Members
Posts: 3822
Joined: 29-October 08
From: USA, 48236
Member No.: 61311



QUOTE (john11 @ Dec 27 2012, 22:05) *
Hi. Thanks for the replies.

I saw on the amp board quite a few diodes probably for biasing voltages and was wondering if these are worth replacing

with something better, fast diodes, schottky?

Have you had any experience of these.


You really need to understand the logic behind "If its not broke, don't fix it".

Biasing diodes are generally highly reliable. They need to have the same forward voltage drop as the origionals which Schottky diodes don't have.

All you seem to be doing is getting training about soldering, because the sound quality impact of what you are trying is generally zero.

If you want to figure out what equipment needs to be up to snuff, first learn how to measure its performance.

A good starting point might be the Audio Rightmark Program - RMAA freeware.
Go to the top of the page
+Quote Post
Soap
post Dec 28 2012, 16:19
Post #20





Group: Members
Posts: 1016
Joined: 19-November 06
Member No.: 37767



QUOTE (Glenn Gundlach @ Dec 27 2012, 22:24) *
Changing the center frequencies of the lower 2 bands of the built in a 5 band graphic equalizer was indeed an improvement. it was 80 Hz / 320 Hz and it's now 25 Hz / 180 Hz. I


Holy smokes! Not to drag this off topic but you moved the center to 25Hz? I rather suspect I'd like whatever you have as a sub!


--------------------
Creature of habit.
Go to the top of the page
+Quote Post
Glenn Gundlach
post Dec 28 2012, 20:11
Post #21





Group: Members
Posts: 369
Joined: 19-April 08
From: LA
Member No.: 52914



QUOTE (Soap @ Dec 28 2012, 07:19) *
QUOTE (Glenn Gundlach @ Dec 27 2012, 22:24) *
Changing the center frequencies of the lower 2 bands of the built in a 5 band graphic equalizer was indeed an improvement. it was 80 Hz / 320 Hz and it's now 25 Hz / 180 Hz. I


Holy smokes! Not to drag this off topic but you moved the center to 25Hz? I rather suspect I'd like whatever you have as a sub!


Electrostatic headphones and 32 foot stops on pipe organs.

Happy New Year

G
Go to the top of the page
+Quote Post

Reply to this topicStart new topic
1 User(s) are reading this topic (1 Guests and 0 Anonymous Users)
0 Members:

 



RSS Lo-Fi Version Time is now: 2nd October 2014 - 18:49