Pioneer SX-1050 Re-Capping - Film or Electrolytic??
I have read numerous well-written technical articles on the attributes, both negative and positive, on the electrolytic and film capacitor debate. I am now re-capping a Pioneer SX-1050 receiver (37-years old). I want to make a sonic improvement in this receiver by using better capacitors. I realize that the caps made today are far better than what was used for production originally. Cost is not a consideration as I know film caps cost more. There are 3-PC boards I am focusing on:
1) Equalizer Amp Assembly (AWF-020) 2)
2) Flat Amp Assembly (AWG-040)
3) Control Amp Assembly (AWG-041)
These PC boards have a mix of film (mylar), tantalum, low-leakage EC and standard EC caps. I am going to replace all the tantalum and EC caps. The cap choices are:
1) Wima MKS2 PET Metallized Film
2) Nichicon KL Low-Leakage electrolytic
3) Nichicon KT, KZ, KA “Audio-grade”
4) Elna Silmic II electrolytic
Considerations:
1) The original “orange” and “pink” colored EC caps were low-leakage. If I stay with the low-leakage design constraint, then my choice is either Wima MKS2 or Nichicon KLs. The KLs are not “audio-grade” per se, but they are a replacement match for what was in the circuit. The Wimas are low-leakage as well and there is enough space on the boards for the larger case for the caps that are under 10uF. Which should I use??? Has anyone already used Wima MKS2 film caps and got a sonic improvement in this application????
2) The original tantalums need to be replaced, but with what? Low-leakage, Wima MKS2 film or “audio-grade” EC caps????
Any thoughts? Has anyone stuffed these board with all Wima MKS2s and how did it turn out?? I’d like to do things right the first time.
I have read numerous well-written technical articles on the attributes, both negative and positive, on the electrolytic and film capacitor debate. I am now re-capping a Pioneer SX-1050 receiver (37-years old). I want to make a sonic improvement in this receiver by using better capacitors. I realize that the caps made today are far better than what was used for production originally. Cost is not a consideration as I know film caps cost more. There are 3-PC boards I am focusing on:
1) Equalizer Amp Assembly (AWF-020) 2)
2) Flat Amp Assembly (AWG-040)
3) Control Amp Assembly (AWG-041)
These PC boards have a mix of film (mylar), tantalum, low-leakage EC and standard EC caps. I am going to replace all the tantalum and EC caps. The cap choices are:
1) Wima MKS2 PET Metallized Film
2) Nichicon KL Low-Leakage electrolytic
3) Nichicon KT, KZ, KA “Audio-grade”
4) Elna Silmic II electrolytic
Considerations:
1) The original “orange” and “pink” colored EC caps were low-leakage. If I stay with the low-leakage design constraint, then my choice is either Wima MKS2 or Nichicon KLs. The KLs are not “audio-grade” per se, but they are a replacement match for what was in the circuit. The Wimas are low-leakage as well and there is enough space on the boards for the larger case for the caps that are under 10uF. Which should I use??? Has anyone already used Wima MKS2 film caps and got a sonic improvement in this application????
2) The original tantalums need to be replaced, but with what? Low-leakage, Wima MKS2 film or “audio-grade” EC caps????
Any thoughts? Has anyone stuffed these board with all Wima MKS2s and how did it turn out?? I’d like to do things right the first time.
I would tackle the electrolytics in the power supply first -- you can get a lot more joules into the space today.
NP mentioned the Elna Silk caps several years ago and i tested them out with my AP distortion analyzer -- yes the measure better. I've used them in a couple of his designs (F3) and liked them. Muse have a lot of adherents.
I've worked on the Pioneer SX-838 and SX-1250, would like to see the schema for the 1050 which is going to be similar in several respects.
NP mentioned the Elna Silk caps several years ago and i tested them out with my AP distortion analyzer -- yes the measure better. I've used them in a couple of his designs (F3) and liked them. Muse have a lot of adherents.
I've worked on the Pioneer SX-838 and SX-1250, would like to see the schema for the 1050 which is going to be similar in several respects.
After more than a couple recap Pioneer jobs, I've learned a few things. There are two goals to recapping. Sonic improvement and dependability. If the receiver is working before you begin, sit down and really listen to it. Replay the same media after critical modifications to see if your ear can detect a difference and decide if it stays or not. Blanket recapping will not tell you exactly what you did to improve anything. I mentioned dependability because you can make your receiver sound fantastic with a good recap job but how long will it last? There are very good reasons why the factory used the components they did and you no doubt will realize that the transistors used back then can be greatly improved upon with modern semi's.(I should say 'forced' to realize). Also consider wire wrap elimination in critical areas. I know it sounds like bs but it has made a difference in many of the amps I've done. Good luck.
Found the schematic -- other than the power supply the only caps which should bear consideration are the coupling caps to the buffer amplifier.
Ditto the wire wrap if you have the patience. You might also want to see how the ground system is arranged. I reconfigured this for the SX-838 and it made a difference.
Ditto the wire wrap if you have the patience. You might also want to see how the ground system is arranged. I reconfigured this for the SX-838 and it made a difference.
I couldn't agree more. Use the new type caps but bear in mind why they used the types they did. Tant caps do tend to leak after many years. I wouldn't change any semiconductors unless they are faulty.
Suggest to go to audiokarma.org, where they have Pionner forum. They discuss this subject all the time. I do agree with what people have written here since they are speaking from experience.
As far as removing the w/w connections. I assume that you remove the sq posts and solder the solid wire directly to the pcb pads? This makes sense since I have read some issues with hum pickup after a repair/re-cap exercise, open ground returns from the PA pcb back to the common point ground.
I have a SX-950 from new, and have only changed o/p trans & volume/bal pots thus far! The switches could use a good cleaning with deox-it.
Rick
As far as removing the w/w connections. I assume that you remove the sq posts and solder the solid wire directly to the pcb pads? This makes sense since I have read some issues with hum pickup after a repair/re-cap exercise, open ground returns from the PA pcb back to the common point ground.
I have a SX-950 from new, and have only changed o/p trans & volume/bal pots thus far! The switches could use a good cleaning with deox-it.
Rick
The capacitors in the "Equalizer Assembly" are most likely mica or ceramic -- if they are mica your good to go, if ceramic you can replace with mica, polycarbonate, PP, or PET. The RIAA equalization makes a very significant difference in the quality of your LP listening experience, so you might want to look at the response over the audible freq range and adjust the values accordingly. Values drift over 40 years.
The capacitors in the "Control Amp" were electrolytic in my SX-838, but this receiver uses a single supply rail. In your case the values may be such that they can be replaced with PP or PET.
An externally hosted image should be here but it was not working when we last tested it.
In the schema I have at hand, I don't see any coupling caps on the OPS.
The pic you show is the simplified drawing of an older flat amp assy.
In the SX1050 flat amp assy (AWG-040). These are ecaps, C1-4 as well as C13,14. In the SX-1250, almost identical flat amp design they used 100V metallized mylar, so they new this was an area for improvement. To up grade the SX-1250 they use polypropylene such as MKP416.
The guys at audiokarma.org have a list of sub parts from Mouser for the SX-1250 which is very close, to the lesser SX1050.
The SM is avail from:
Main scan page
The phono eq use polystyro (c11-14) in the RIAA section. I would not mess with these.
In the SX1050 flat amp assy (AWG-040). These are ecaps, C1-4 as well as C13,14. In the SX-1250, almost identical flat amp design they used 100V metallized mylar, so they new this was an area for improvement. To up grade the SX-1250 they use polypropylene such as MKP416.
The guys at audiokarma.org have a list of sub parts from Mouser for the SX-1250 which is very close, to the lesser SX1050.
The SM is avail from:
Main scan page
The phono eq use polystyro (c11-14) in the RIAA section. I would not mess with these.
Last edited:
IF it's a coupling cap or in a filter, replace any electrolytic with a film cap (Your quoted Wima part is eminently suitable) if you can NORMALLY fit it into the allotted space (please don't shoe in huge caps intended for crossovers etc, I've had to pull these out and fix cracked joints and traces on so many 'audiophile' modifications of vintage equipment!).
MKP will also be better than mylar if you can fit them, but even modern mylar will be slightly better and use less space. The problem with mylar caps that were then used as standard is that they can be quite inductive for higher capacitance values, as it's a wound rather than stacked construction. In the case of small values and even in most other cases it's not a big deal but in some cases they were used as decouplers on various power rails, and this is not necessairly a good place for such a cap. Most modern PP, PET etc are non-inductive so it's a small bonus.
30 year ago film low leak long life electros were used in values from 0.1u up to a few uF because of space considerations. Today these would normally be miniaturized non-inductive film, and believe me, any electro is a far cry from any film cap in this sort of application.
For everything else, a good look at the service manual and a knowledge of circuitry that can tell you what the designers were trying to accomplish at least in general terms, is a must. As someone said, blanket cap replacements usually result in higher costs and no sonic benefit, and in fact, sometimes a sonic degradation or at least an unexpected change, not always to the better.
Someone mentioned installing larger electrolytics. DO NOT DO THIS unless you have a VERY good idea of the limits. Power transformers, rectifier and filter cap specs all depend on each other and just upping the capacitances usually means increasing ripple currents (sometimes signifficantly) for which the tansformer and rectifier will certainly not thank you and in fact if the amp is driven hard, may eventually die. That being said, some leeway is to be had:
1) Higher ripple current spec means larger cap for a given capacitance and voltage. Replacing a standard cap with a higher spec one in this regard normally gives you increased cap life. There are exceptions, however but they are very rare.
2) Higher voltage spec increases cap life. Capacitors age faster the closer the working voltage is to the maximum voltage. This is because applying a voltage actually slowly builds up the thickness of the dielectric. In the grand majority of cases old power supply electrolytics in vintage equipment show a marked reduction in capacitance compared to the spec, but not a proportional increase in ESR. This means the electrolite is just fine but the dielectric has built up over time resulting in capacitance reduction. This is especially the case in components where the voltage is close to maximum spec - keep in mind that mains voltage standards changed and now mains is typically higher (110 nominal became 115 in the US, 220 became 230 in Europe - only the UK saw a drop from 240 to 230). Many of these parts spent the last two decades of their life working right at their maximum spec.
3) Higher temperature rating increases longevity everything else being equal. BEcause it's not a linear relationship, this can get quite dramatic - and most caps in older equipment is rated 85 deg C. In the eman time electrolyte technology has advanced somewhat so now 105 C is practically considered standard, with 120 C emerging for switching PSUs and inverters. Higher temperature increases the rate of all chemical reactions, ones which degrade the electrolyte (water molecules escaping seals) and ones which grow oxide and result in lowering of the capacitance with time. Increased ESR is a special problem as it leads to self-heating of the cap, and becomes a self-reinfocing phenomenon. Cracked or exploded seals and leaked caps 'all of a sudden' are the results.
4) And here's where you can benefit regarding higher capacitance: today all electrolytic caps are made with pre-formed electrodes. This makes it possible to produce caps with very tight tolerance specs compared to times past. 30 years ago filter caps were usually specified as +50% -10% specified capacitance value, and actually in most cases were at least 30% over spec to account for ageing. When they drop to -10% of rated value, it's time for replacement. However, todays caps are routinely specced +-10% and in fact may even come a few % low from the factory. Because of this in most cases up to 50% over rated capacitance will be tolerated without problems, and this should really be considered as the upper limit unless you have some inside info on the actual equipment )such as 'same transformer was used in the next model up with larger caps').
MKP will also be better than mylar if you can fit them, but even modern mylar will be slightly better and use less space. The problem with mylar caps that were then used as standard is that they can be quite inductive for higher capacitance values, as it's a wound rather than stacked construction. In the case of small values and even in most other cases it's not a big deal but in some cases they were used as decouplers on various power rails, and this is not necessairly a good place for such a cap. Most modern PP, PET etc are non-inductive so it's a small bonus.
30 year ago film low leak long life electros were used in values from 0.1u up to a few uF because of space considerations. Today these would normally be miniaturized non-inductive film, and believe me, any electro is a far cry from any film cap in this sort of application.
For everything else, a good look at the service manual and a knowledge of circuitry that can tell you what the designers were trying to accomplish at least in general terms, is a must. As someone said, blanket cap replacements usually result in higher costs and no sonic benefit, and in fact, sometimes a sonic degradation or at least an unexpected change, not always to the better.
Someone mentioned installing larger electrolytics. DO NOT DO THIS unless you have a VERY good idea of the limits. Power transformers, rectifier and filter cap specs all depend on each other and just upping the capacitances usually means increasing ripple currents (sometimes signifficantly) for which the tansformer and rectifier will certainly not thank you and in fact if the amp is driven hard, may eventually die. That being said, some leeway is to be had:
1) Higher ripple current spec means larger cap for a given capacitance and voltage. Replacing a standard cap with a higher spec one in this regard normally gives you increased cap life. There are exceptions, however but they are very rare.
2) Higher voltage spec increases cap life. Capacitors age faster the closer the working voltage is to the maximum voltage. This is because applying a voltage actually slowly builds up the thickness of the dielectric. In the grand majority of cases old power supply electrolytics in vintage equipment show a marked reduction in capacitance compared to the spec, but not a proportional increase in ESR. This means the electrolite is just fine but the dielectric has built up over time resulting in capacitance reduction. This is especially the case in components where the voltage is close to maximum spec - keep in mind that mains voltage standards changed and now mains is typically higher (110 nominal became 115 in the US, 220 became 230 in Europe - only the UK saw a drop from 240 to 230). Many of these parts spent the last two decades of their life working right at their maximum spec.
3) Higher temperature rating increases longevity everything else being equal. BEcause it's not a linear relationship, this can get quite dramatic - and most caps in older equipment is rated 85 deg C. In the eman time electrolyte technology has advanced somewhat so now 105 C is practically considered standard, with 120 C emerging for switching PSUs and inverters. Higher temperature increases the rate of all chemical reactions, ones which degrade the electrolyte (water molecules escaping seals) and ones which grow oxide and result in lowering of the capacitance with time. Increased ESR is a special problem as it leads to self-heating of the cap, and becomes a self-reinfocing phenomenon. Cracked or exploded seals and leaked caps 'all of a sudden' are the results.
4) And here's where you can benefit regarding higher capacitance: today all electrolytic caps are made with pre-formed electrodes. This makes it possible to produce caps with very tight tolerance specs compared to times past. 30 years ago filter caps were usually specified as +50% -10% specified capacitance value, and actually in most cases were at least 30% over spec to account for ageing. When they drop to -10% of rated value, it's time for replacement. However, todays caps are routinely specced +-10% and in fact may even come a few % low from the factory. Because of this in most cases up to 50% over rated capacitance will be tolerated without problems, and this should really be considered as the upper limit unless you have some inside info on the actual equipment )such as 'same transformer was used in the next model up with larger caps').
In a high power Pioneer receiver of this era it shouldn't be an issue. I had 2x 8,000uF/50V -- replaced with 2x 39,000uF/50V.
An externally hosted image should be here but it was not working when we last tested it.
While we're busily knocking electrolytics in coupling applications, everyone should take a deep breath and read Douglas Self's new preamplifier article in Volume 5 of Linear Audio. (or take a look at the First Watt F3)
thx for the explanation but indeed when you think logical, then the chinese capas in my case which are allready terrible has to work the hardest, since we have 230 v 50/60 Hz . So the increase from 4700u to 6800u ain't a problem. And since it's a normal brand, i'm waiting for the better ones, which FIT... allways a ptoblem. The rectifiers etc can take it easy. Good info.
gr richard
gr richard
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.