With respect to restoring and/or upgrading of Solid State Audio Power Amps and Pre-Amps. ANY THOUGHTS AND OPINIONS for the following??
1. Replace all carbon film R's with metal film. Besides having tighter tolerances and less drift with time and heat what other advantages? Necessary? Do they really sound better?
2. Should I go out of my way to use higher wattage rating providing space isn't a problem i.e. 1/4watt to 1/2 watt. What is to benefit? I have "heard" this has an effect on sound.
3. ceramic caps in the audio circuit.... change to polystyrene? Where do YOU source from?
4. Electrolytics.... Realising any older unit is gonna benefit with newer electrolytics, what is the consensus on brands and quality. Where do YOU source from?
5. How common is it that polar electrolytics are used where no polarizing voltage is across it? Should the effort be made to replaced with a non-polar unit?
6. Thoughts on replacing caps with a higher voltage. Can you over do it in this regard?
Without dealing with specific designs and schematics are there any other general tweaks, upgrades, or improvements anyone can add? Although I am MAINLY concerned with sound any ideas are welcomed. For example, I have replace cheap VR's with bourns multi-turn trimmers in the past.
1. Replace all carbon film R's with metal film. Besides having tighter tolerances and less drift with time and heat what other advantages? Necessary? Do they really sound better?
2. Should I go out of my way to use higher wattage rating providing space isn't a problem i.e. 1/4watt to 1/2 watt. What is to benefit? I have "heard" this has an effect on sound.
3. ceramic caps in the audio circuit.... change to polystyrene? Where do YOU source from?
4. Electrolytics.... Realising any older unit is gonna benefit with newer electrolytics, what is the consensus on brands and quality. Where do YOU source from?
5. How common is it that polar electrolytics are used where no polarizing voltage is across it? Should the effort be made to replaced with a non-polar unit?
6. Thoughts on replacing caps with a higher voltage. Can you over do it in this regard?
Without dealing with specific designs and schematics are there any other general tweaks, upgrades, or improvements anyone can add? Although I am MAINLY concerned with sound any ideas are welcomed. For example, I have replace cheap VR's with bourns multi-turn trimmers in the past.
Yes. If you OVER do it then in time they could dry out.
Here's a posting that you should consider:
http://www.tnt-audio.com/clinica/bias_e.html
There are some changes you can make that won't make much difference, but will add reliability. There are some you need to do anyway unless you just like working on the same amp over and over.
Per your questions:
1. Metal film resistors are quieter than carbon film, but unless they are in a critical position you won't hear any difference. They are more reliable, though. Do it.
2. Higher wattage resistors would add reliability, but no difference in sound. If you want, do it, but I wouldn't.
3. Sometimes you want a ceramic cap. They make excellent snubbers for high frequency feedback, but they are too wide-band for audio signal path. Polypropylene are probably the most reliable of the film caps. Metal foil is better than metalized film. Oil & paper caps are the longest-lasting, but do you want an amp for 50 years? I use the Mouser catalog.
4. Electrolytics: I look for caps with low-ESR. Basically, there's more "cap" to them. Some cheaper ones with high ESR (equivalent series resistance) work like they have a resistor in series inside the cap. Not always good. But, on some classic tube amps, you actually can ruin the sound by going with too good a cap... but not on transistor amps. I use a lot of Sprague caps, but there's a lot of good brands. Look at the specs. Again, Mouser has specs on-line.
5. There is almost always SOME voltage across a cap. Exceptions are in cross-overs, tone control circuits, and other passive networks or AC circuits. Even at the inputs, if the input source is at ground, the bias of the input stage is probably at some positive voltage. But I would change caps in the audio signal path to film caps rather than electrolytic of any sort. Solen is good for film caps, but there's a lot of other good brands. See the Mouser catalog. See also http://electrochem.cwru.edu/ed/encycl/misc/c04-appguide.pdf for "too much" cap information from Cornell-Dubilier, a good cap manufacturer.
6. No problem with, say, a 63 volt in place of a 50 volt cap, but more voltage carries with it other problems, like higher ESR and larger size caps. Newer caps are generally smaller anyway, but I would not use 500 volt caps throughout a solid-state amp. Aim at 50% over maximum voltage the cap sees, for safety. Never go under what the OEM put in there.
Lead dress and shielding is important. All inputs should be shielded. Very-high-gain inputs probably could use better shielding. Most of it is only 70% shielded.
Run wires along the metal chassis to prevent them from radiating or picking up signals.
Good grounding is lost from bare-metal grounds getting oxidized. Use star washers to bite through the paint, oxide, etc.
Ventilation on some amps seems to be limited to the heat-sinked output devices. The inside of an amp can get REAL warm. You can string two 12v (computer) fans in series and they are real quiet.
I'd use also 105 degree caps, touch up the soldering on all the boards, and check stuff like big resistors heating up electrolytics next to them. Dried out caps are no good for anyone.
I repair amps for a living, so I see a lot of things that most often need done.
Mouser is here in Texas, so I get my parts in two days max, and pay for UPS ground. And their prices are good.
Mike.
Per your questions:
1. Metal film resistors are quieter than carbon film, but unless they are in a critical position you won't hear any difference. They are more reliable, though. Do it.
2. Higher wattage resistors would add reliability, but no difference in sound. If you want, do it, but I wouldn't.
3. Sometimes you want a ceramic cap. They make excellent snubbers for high frequency feedback, but they are too wide-band for audio signal path. Polypropylene are probably the most reliable of the film caps. Metal foil is better than metalized film. Oil & paper caps are the longest-lasting, but do you want an amp for 50 years? I use the Mouser catalog.
4. Electrolytics: I look for caps with low-ESR. Basically, there's more "cap" to them. Some cheaper ones with high ESR (equivalent series resistance) work like they have a resistor in series inside the cap. Not always good. But, on some classic tube amps, you actually can ruin the sound by going with too good a cap... but not on transistor amps. I use a lot of Sprague caps, but there's a lot of good brands. Look at the specs. Again, Mouser has specs on-line.
5. There is almost always SOME voltage across a cap. Exceptions are in cross-overs, tone control circuits, and other passive networks or AC circuits. Even at the inputs, if the input source is at ground, the bias of the input stage is probably at some positive voltage. But I would change caps in the audio signal path to film caps rather than electrolytic of any sort. Solen is good for film caps, but there's a lot of other good brands. See the Mouser catalog. See also http://electrochem.cwru.edu/ed/encycl/misc/c04-appguide.pdf for "too much" cap information from Cornell-Dubilier, a good cap manufacturer.
6. No problem with, say, a 63 volt in place of a 50 volt cap, but more voltage carries with it other problems, like higher ESR and larger size caps. Newer caps are generally smaller anyway, but I would not use 500 volt caps throughout a solid-state amp. Aim at 50% over maximum voltage the cap sees, for safety. Never go under what the OEM put in there.
Lead dress and shielding is important. All inputs should be shielded. Very-high-gain inputs probably could use better shielding. Most of it is only 70% shielded.
Run wires along the metal chassis to prevent them from radiating or picking up signals.
Good grounding is lost from bare-metal grounds getting oxidized. Use star washers to bite through the paint, oxide, etc.
Ventilation on some amps seems to be limited to the heat-sinked output devices. The inside of an amp can get REAL warm. You can string two 12v (computer) fans in series and they are real quiet.
I'd use also 105 degree caps, touch up the soldering on all the boards, and check stuff like big resistors heating up electrolytics next to them. Dried out caps are no good for anyone.
I repair amps for a living, so I see a lot of things that most often need done.
Mouser is here in Texas, so I get my parts in two days max, and pay for UPS ground. And their prices are good.
Mike.
Slow down now fellas!
Firstly, I hate that darn article that ppfred posted. It has more emotion than fact. Please ignore it because I don't want to dissect it line by line. Use common sense - please!
One thing I will say. There is nothing wrong, or cheap about most of the 270° trim pots used for adjustments. Multiturn controls may actually be worse in these locations.
Metal Oxide resistors are more reliable than some film types. I have measured some fairly inductive metal film parts over the years.
So, unless you have the schematic (service manual is much better), you should not be swapping out resistors. Beware of the maximum voltage ratings for some resistors.
So, with that in mind ....
1. Don't waste your time with this. There are a couple that might improve the S/N ratio. I doubt it though.
2. There is no effect on sound. If you see some overheating, use a higher wattage that still fits and space them up some.
3. Avoid oil and paper caps. They can be very unreliable. Most mica and film caps will be better than ceramic types. NPO's may be okay.
4. Some older units used great capacitors. On average, new current electrolytics are better than older types. The larger cans are usually better than the smaller can sizes.
5. Often there may only be a few mV across some electrolytics. Use polarized in those cases. A non-polarized cap will be too large.
6. No problem. I often use higher voltage caps because the leads will fit in the holes. The size is closer to the original part. Higher voltage rated caps will last longer. They will not "dry out" any faster and may actually last much longer.
Use common sense.
-Chris
Firstly, I hate that darn article that ppfred posted. It has more emotion than fact. Please ignore it because I don't want to dissect it line by line. Use common sense - please!
One thing I will say. There is nothing wrong, or cheap about most of the 270° trim pots used for adjustments. Multiturn controls may actually be worse in these locations.
Metal Oxide resistors are more reliable than some film types. I have measured some fairly inductive metal film parts over the years.
So, unless you have the schematic (service manual is much better), you should not be swapping out resistors. Beware of the maximum voltage ratings for some resistors.
So, with that in mind ....
1. Don't waste your time with this. There are a couple that might improve the S/N ratio. I doubt it though.
2. There is no effect on sound. If you see some overheating, use a higher wattage that still fits and space them up some.
3. Avoid oil and paper caps. They can be very unreliable. Most mica and film caps will be better than ceramic types. NPO's may be okay.
4. Some older units used great capacitors. On average, new current electrolytics are better than older types. The larger cans are usually better than the smaller can sizes.
5. Often there may only be a few mV across some electrolytics. Use polarized in those cases. A non-polarized cap will be too large.
6. No problem. I often use higher voltage caps because the leads will fit in the holes. The size is closer to the original part. Higher voltage rated caps will last longer. They will not "dry out" any faster and may actually last much longer.
Use common sense.
-Chris
Re: Slow down now fellas!
AS always your wisdom is appreciated Chris. What would be your concerns regarding the trimmers? I might be missing something here but I thought a variable resistor is just that, a variable resistor. I fail to see why replacing single turn trimmers with sealed multi-turn of the same values being of any concern. I see nothing but benefits like no more dust/dirt problems with less chance of drifting and less sensitivity with precise adjustment. The only thing I can think of is the initial setting of the new VR. If for example I'm changing out a bias adj. pot I make sure its set correctly and then measure the pot resistance. The new multi-turn is then preset and replaced and tweaked if needed.
anatech said:
AS always your wisdom is appreciated Chris. What would be your concerns regarding the trimmers? I might be missing something here but I thought a variable resistor is just that, a variable resistor. I fail to see why replacing single turn trimmers with sealed multi-turn of the same values being of any concern. I see nothing but benefits like no more dust/dirt problems with less chance of drifting and less sensitivity with precise adjustment. The only thing I can think of is the initial setting of the new VR. If for example I'm changing out a bias adj. pot I make sure its set correctly and then measure the pot resistance. The new multi-turn is then preset and replaced and tweaked if needed.
Hi Bigred,
Well, for one, the element and wiper are smaller on multiturns. That means less contact area and possibly less pressure. When you turn the screw, you are actuating a geared down wiper. This means that it is easier to set a value using an incremental change, but this is of little use for bias level setting. A false sense of improvement.
The problem is that a multiturn control is not designed to pass current on average and the contact will not put up with any oxide at all. The 270° types are easy to set if the design engineer did his job properly. They can easily dissipate 1/4 watt and pass some current as the moving contact is larger and is under more pressure. I think the 270° control is well suited to the job at hand. Even if the wipers and wattages were the same in comparison, a multiturn type sells you things like a reduction gear drive the is of no advantage to you.
Keep in mind that they all oxidize over time. I've had to replace 10 turn trimmers as well in test gear. I think someone who doesn't understand electronics liked the finer control and higher price tag of the 10 turn types and decided arbitrarily that the 270° type was used to reduce costs. This is pure silliness of course, the 10 turn control is not well suited to these jobs to begin with.
-Chris
Edit: If you have a dust and dirt problem, there are covered 270° types, but this would not be your biggest problem. So dust and dirt is really a non-issue.
Well, for one, the element and wiper are smaller on multiturns. That means less contact area and possibly less pressure. When you turn the screw, you are actuating a geared down wiper. This means that it is easier to set a value using an incremental change, but this is of little use for bias level setting. A false sense of improvement.
The problem is that a multiturn control is not designed to pass current on average and the contact will not put up with any oxide at all. The 270° types are easy to set if the design engineer did his job properly. They can easily dissipate 1/4 watt and pass some current as the moving contact is larger and is under more pressure. I think the 270° control is well suited to the job at hand. Even if the wipers and wattages were the same in comparison, a multiturn type sells you things like a reduction gear drive the is of no advantage to you.
Keep in mind that they all oxidize over time. I've had to replace 10 turn trimmers as well in test gear. I think someone who doesn't understand electronics liked the finer control and higher price tag of the 10 turn types and decided arbitrarily that the 270° type was used to reduce costs. This is pure silliness of course, the 10 turn control is not well suited to these jobs to begin with.
-Chris
Edit: If you have a dust and dirt problem, there are covered 270° types, but this would not be your biggest problem. So dust and dirt is really a non-issue.
Bigred said:
I would not bother, except perhaps in some critical areas - but even then it takes a lot of knowledge of circuit topology to identify them. As to what sounds better, that's a whole different can of worms...
I doubt that the same type resistor of higher wttage will have a discernible effect on sound. Keep in mind that some higher wattage resistor technologies differ from what you are replacing, so the sound will likely be different (if it is different - depending enormously on where the actual resistor is used), this will be the reason. I would not bother with such replacements unless there is evidence of thermal stress. In that case, Chris has given excellent advice to which it is hard to add anything.
Be aware that when you say 'ceramic caps' you are referring to a whole bunch of very different types. In general, small ceramics are NPO type and should be fine. It is more likely they will suffer enviromental problems (some wax covered ones can by hygroscopic and develop internal oxidation). This being said, there can be signifficant advantage in replacing ceramic caps in some strategic places. The primary merit of a ceramic cap is very low series inductance, so if you replace it with anything else, this is what you need to take care of. Polystirene can be inductive - take note of the construction. Places where it is worth replacing ceramics would be those where the voltage across the cap is not nearly constant. Even the best ceramics have some form of capacitance dependance on voltage (NPO have the least). Places like compensation caps (especially miller caps) are good spots to do a ceramic cap replacement. My favorite here would be silver mica. They generally have a very low inductance construction. Use epoxy dipped ones, some older types used wax and also turned out to have moisture problems. And now, THAT being said, there are places where ceramic is the best, one would be decoupling of HF circuits - for instance decoupling of DA converter chips. Very often a surface mount NPO cap is the right one to use.
This is an area where it is easily possible to burn money without any need to do so. Nearly every named capacitor manufacturer (as oposed to no-name stuff mostly from china, then no-one can get data for) make some sort of good cap or caps. Datasheets can be quite helpful.
The primary agent that shortens electrolytic cap life is temperature. One of the secondary agents is very low DC voltage. It is often the case that a vintage component will have several dried out coupling electrolitics that have little, no, or reverse voltage across them, but the same unit will have perfectly good power supply filter caps even though these are subjected to considerable stress. The reason is that they are practically always polarized with a nontrivial voltage. That being said, modern caps are FAR better at this than they used to be in olden times. Datasheets help - one from Nippon Chemi states that most modern constructions actually act as bipolars up to about 1V reverse polarisation.
The first effort that needs to be done is polarising the cap correctly to whatever small voltage is present. Take particular notice of DC blocking caps in amp feedback networks. It is amazing how often they are polarized wrongly.
The next effort would be investigating if the electrolytic can be replaced by a foil cap. In many cases electrolytic caps up to 3.3 or 4.7uF were used simply because the foil caps available then were too large and/or expensive. Today you can get non-inductive metalized film caps that are very small, and will work well instead of many small electros.
The final effort would be investigating wether you can get a nonpolar cap as a replacement. This can be a problem as mostly you can easily get only those made for crossover networks. Even though thiy will often offer exemplary performance, they will be far too large. It is likely you will have to order these special.
Usually you can't overdo it. A very large voltage cap will simply not fit, even though it is often possible to fit a modern cap with 2-3 times the working voltage of the original, as it will be the same size. A HV cap will definitely NOT 'dry out' if it is used on lower voltages. This happens when we are talking milivolts - and HV caps are only a little more susceptible than lower voltages.
What you do need to look up in the manufacturer datasheets is the ESR. For a single type of cap this can vary considerably depending on what the voltage rating is, and there is a definite minimum. Use this to your advantage.
Regarding that tnt-audio site that was linked by ppfred, I have to agree with Anatech/Chris. There are painfully few instances where any of what they say in that article is true, and many instances where it will be downright dangerous. People that end up with damaged amps as a result of the 'advice' on that site should hold them liable and ask for damages, then hopefully it will be taken down as it should be.
If you do decide to upgrade any components, you can go to the usual sources:
Digikey www.digikey.com
Mouser www.mouser.com
Allied Elec www.alliedelec.com
or for some more esoteric components (or higher end depending upon your view), Parts Connexion: www.partsconnexion.com
I wouldn't go swapping parts for the sake of it. If you are building from scratch, that is another issue...
Digikey www.digikey.com
Mouser www.mouser.com
Allied Elec www.alliedelec.com
or for some more esoteric components (or higher end depending upon your view), Parts Connexion: www.partsconnexion.com
I wouldn't go swapping parts for the sake of it. If you are building from scratch, that is another issue...
I don't like to work blind, and think too many caps are replaced for no good reason. Get any sort of cap meter or bridge that can measure both value and dissipation factor. If it can do leakage, better yet. I have a lot of 50 year old equipment with perfectly good filter caps, but have also replaced bad caps in receivers 5 or 10 years old. In general, I replace most resistors in the signal path with metal films, particularly in the feedback network. Having had bad experiences with MOX and thick film resisters in critical locations where linearity was important (not audio stuff), I avoid them, but again, measurements are the only way to know what's good and what's bad. Matching between channels is a good idea, so I check every part and select the closest matches, as opposed to the closest to nominal. I suggest (almost) never touching caps in RIAA networks. Chances are they're already polystyrene and you'll likely do more harm than good. Coupling caps are another story, and getting polypropylene types in there will likely improve things. IMO, you can accomplish even more if you can curve trace the semiconductors and check matching. This is of great importance with things like current mirrors, and some types of outputs. Even if the "curve tracer" is nothing more than some gain measurements on a crudely wired protoboard, it's way better than guessing. Go after things that really affect the way the amp works!
Hi Conrad,
I couldn't have said this better myself. I do the same things that you do, close enough anyway.
-Chris
I couldn't have said this better myself. I do the same things that you do, close enough anyway.
Definitely! I test most caps first with an HP 4263A, but in power supplies I simply look at the rectified waveform for "pips" on the leading edge of the ripple. This is a very accurate test for filter caps. That horribly expensive meter I have is a complete luxury. Just change the smaller electrolytics with new ones from a known source and brand name. No need to spend that kind of loot.
I have found metal oxides to be fine.
Just match the beta at the expected current level. That is more than close enough. I did build a diff pair circuit with no feedback to match pairs. It works great and is very inexpensive. Using a curve tracer takes a lot of time to get use to using the instrument and also to interpret the display for what you want to know about the part. But, if you can - great.
-Chris
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.