Nice product 🙂 There's probably still a bit of scope for tinkering about, perhaps upgrading a few bits and pieces as you may imagine. Perhaps it would be better to start another thread drawing attention to the brand/model and draw some attention from the experts in Nakamichi mods and improvements.
I've always found it to suprise others with its sound.
It's not without it's faults..
I've previously moved the driver transistors onto a 'new' heatsink to stop them cooking the board further.
Why oh why did they have to use cheap bakelite rubbish.....
I shall indeed start a new thread for this unit.
And I will try to push this one back on track re foil/film in place of electro's..
So.
My other thought/concern is, if a small value (polarised) electro has been spec'd by the manufacturer, is there any issue with voltage polarity for the foil/film ?
Badly phrased, try again.
Are there any instances were the cap needs to be a polarised one ?
Extreme example, If you have a 10,000uf 50v power supply filter cap.
Could their be any improvement by using a 10,000/50 foil cap in its place ?
And yes, I am completely ignoring size/cost/insanity factors here.. 😀
It's not without it's faults..
I've previously moved the driver transistors onto a 'new' heatsink to stop them cooking the board further.
Why oh why did they have to use cheap bakelite rubbish.....
I shall indeed start a new thread for this unit.
And I will try to push this one back on track re foil/film in place of electro's..
So.
My other thought/concern is, if a small value (polarised) electro has been spec'd by the manufacturer, is there any issue with voltage polarity for the foil/film ?
Badly phrased, try again.
Are there any instances were the cap needs to be a polarised one ?
Extreme example, If you have a 10,000uf 50v power supply filter cap.
Could their be any improvement by using a 10,000/50 foil cap in its place ?
And yes, I am completely ignoring size/cost/insanity factors here.. 😀
Small lytics have enough damping and low enough inductance to damp small film capacitors and prevent resonance. If you remove them and start adding film caps you'll be up to your eyeballs in resonances. I would say leave them in and just add caps if you really want to try it.
There is the possibility that a resonance can improve the subjective quality rather than degrade it, but in such a case it can't be because the circuit itself is performing better.
There is the possibility that a resonance can improve the subjective quality rather than degrade it, but in such a case it can't be because the circuit itself is performing better.
This is where theory ends and magics begins. I truly believe there are things like this, things beyond theory and test equipments.
Maybe not so much with audiophile electronics, but I can tell you from designing guitar tube amps, after I do the rudimentary power up, check bias, all the voltages and look for oscillation and PS noise, I then run sine way through to quickly check for clipping. After that, I put my scope and generator away as it takes too much room and pretty much useless.
you can use a bigger plastic cap parallel to the 4,7uF electro.......0,47uF for that 0,047uF or something else, if there is enough space around.
Signal DC blocking caps (C220, C209...) can of course be replaced directly with film caps, there is no risk of resonance there unless you use multiple in parallel.
They aren't beyond measuring or anything like that, just no one cares until it's a "problem", and aren't willing to setup to look for them.
I've implemented some simple resonance dampening with amazing results. Granted I wasn't mixing a slew of capacitors.
Member
Joined 2009
Paid Member
I would agree with this, the exception is possibly where the designer was designing to a cost target lower than the DIYer's budget and had to scrimp on certain parts, or where modern components offer improvement in performance over what you might find in older equipment that simply wasn't available to rhe designer. But in any case you still want to understand what the original designer was about, you do want to have agood understanding of the circuit.
My approach to date has been to recap electros with electros, using larger values in many cases.
It was discussed that different types or values of cap would have resonances at differing frequencies.
If this is so, then wouldn't swapping the 0.047uF with a 0.47uF alter this resonance ?
Or am I over simplifying what I don't understand 🙂
With the above in mind, one part of this circuit uses a 0.047uF film across a 4.7uF electro.
Would the thought of replacing the parallel pair with a single film be an option to get away from resonance issues ?
Larger value electros are normally lower esr.
Would using 'low esr' caps of the original value not be a step in the right direction ?
One thing I'm always mindful of is going to far and placing higher stresses on other areas of the circuit.
Or is this not an issue ?
Or only a concern when going 'radically' larger ?
I'm guessing this folds back into the "better understanding than the original designer" 😀
The only reason I can see for using a 47nF decoupling cap across a 4.7uF lytic in a DC coupler is for some kind of voicing, or for making a notch filter at 5MHz or under, but the lytic is too lossy for that to do anything. I see no reason not to replace it with a 4.7uF film cap.
Mmmmmm. So many questions.
Why put the film across the electro when most would just use the electro.
As it seems the use of the combo may make more problems than it fixes.
Is there some sort of design flaw that requires the need for a notch.
How much effect could a 0.047 have to change the overall sonic character.
I can understand the use of an electro over a film for cost/size constraints.
But as it's been pointed out, "only the designer would know".
Why put the film across the electro when most would just use the electro.
As it seems the use of the combo may make more problems than it fixes.
Is there some sort of design flaw that requires the need for a notch.
How much effect could a 0.047 have to change the overall sonic character.
I can understand the use of an electro over a film for cost/size constraints.
But as it's been pointed out, "only the designer would know".
Member
Joined 2009
Paid Member
I have usually taken the approach that electrolytics are not an impediment to good sound so I havent made my own experiments by comparing small electros with film caps. But when usung electros I prefer larger values in order to greatly reduce the ac / signal voltage across the cap since this is expected (for any type of capacitor) to push any distortion products way down in frequeny where they no longer matter. The ESR performancehasn't been a consideration for caps in the signal path, it's something I consider when I want to minimize self-heating in high ripple current power supplies.
In tube circuits I tend to favour non-electros because they are more readily available from my parts bin in high voltage ratings and impedances are high so small values are fine.
Edit: IIRC somewhere on Youtube is a clip where a film cap is compared with an electro or bypassed and the author could find no benefit to using a film cap whatsoever using a high sensitivity distortion measurement or his ears. I don't have the link handy.
Last edited:
For 1uf and 0.47uf, you can simply use Nichicon ES caps. No bypass work needed.
Well, when I checked those, it was possible to bypass with 4n7 polyester dip cap. But that is unlikely to be needed.
Well, when I checked those, it was possible to bypass with 4n7 polyester dip cap. But that is unlikely to be needed.
You are misquoting what I have said.
I usually suggest about a decade beyond the required passband.
If you want a passband of 20Hz to 20kHz, then setting the passive single pole input filters to ~2Hz and ~200kHz gives the following receiver an easier job in processing the wanted signal with less interference. And keeps the wanted passband virtually flat.
If you want a narrower passband, then you are free to use that, but don't include me in the same sentence.
from post154 in http://www.diyaudio.com/forums/solid-state/5107-capacitor-distortion-8.html#post2801993
and read posts 147 & 153
Last edited:
Cap bypassing has some similarities with matching a tweeter to a woofer, in that a random selection has significant risk of useless output; OR, the parts can be well-matched for a more useful result.
So much panic so little experience. I've done this. It has worked, if you can cram the film caps in. If you can't I use 50 V COG ceramic caps (horrors) on 2 v signals. Successful projects include a 1970's Reader's digest AM/FM/shortwave radio, two 1966-68 Hammond organs, the RIAA sections of a Sony TC250 tape deck used as a preamp after the rubber tape parts went to ****. I tried buying a new radio, (a Sony PLL unit) it had the sensitivity of the filling in your back tooth and howled into the mag phono cartridge. The organs sound like wet kazoos when I buy them, then sound like an instrument costing more than a convertible pony car with all the options when I'm done. My objection to electrolytics? I've put electrolytics 4 times in my ST70 and PAS2 hifi equipment, much too much work for 6 or 8 years benefit.
You know your radio or preamp needs new electros if the tuning point changes as the room temperature swings from 50 F to 80 F, or the sensitivity fades out a low temperatures but is fine at 70 F, or the sound is fuzzy. Screw meters, if a piano source sounds like a real piano at the end of your chain, your circuit is performing admirably, as this wasn't even possible for <$2000 until the seventies, and in speakers, IMHO, wasn't possible until the 2005 era. I use a Steinway console as an ear calibrator.
Electros under 15 uf are, in my book, obsolete. Go ahead, professionals, design systems that will have to be repaired or thrown away in 8 years. I'm hoping this stuff I repair with film caps lasts until I'm uncomfortably installed in a wooden box. Without opening the covers and replacing the electros again. The PowerSupply e-caps? I'm very fond of caps rated 10000 hours service life at 105 deg C. If I can't get that, I never buy anything under 2000 hours service life, and really look hard to see if I can't at least get 3000 hours service life. I'll pay an additional $12 freight charge to get something cap might last a while, over some generic e-cap that will need looking at again when I am in my 80's.
The film caps I've been using have been roderstein mylars in most cases, the ceramic caps I used were Aerovox gold 50 v series. Waaay better sound than the tantalum caps Hammond and Dynakit used, too. **** tantalums came from the store with popcorn noise, I don't care what samples you company owners get in the mail, here in the middle if you go to the TV rapair store you get **** grade, everybody elses's rejects. At least Newark started selling directly to me after debit cards were invented; for years they wouldn't ship to a home address, leaving me at the mercy of TV repair supply houses for my supply of **** and 500 hour service life parts.
As far as switcher howl coming into your circuit, put it in a grounded steel box, with a toroid on the AC coming in and filter caps on the analog inputs, to keep the RF out. The treatment stops CB radio nuts driving by from interfering from your ethereal music , too. Switcher supply in your box? - shame on you, you deserve all the howl you get. Use a linear transformer, E frame if you can get it. This is diyaudio, you're not designing to a price point here. And there are plenty of dead PC power supplies laying around to salvage RF suppression parts from, at least around my house.
Last edited:
I'll often replace small e-caps with films. Look at what's in series with it. Or sometimes in parallel. There's no point in worrying about ESR and resonances if the circuit already has at least some small amount of series resistance that completely dominates the contribution of the cap.
Parallel caps make sense for HF/RF bypass, but tend to be misunderstood at audio frequencies.
True or false- Adding a small film cap across a large aluminum electrolytic will always increase the total capacitance.
True or false- Adding a small film cap across a large aluminum electrolytic will improve its performance by a large amount (in terms of ESR) at the highest audio frequencies.
True or false- Aluminum electrolytic capacitors are useless and have high impedance above their self-resonant frequencies since they become inductors.
Enjoy!
Parallel caps make sense for HF/RF bypass, but tend to be misunderstood at audio frequencies.
True or false- Adding a small film cap across a large aluminum electrolytic will always increase the total capacitance.
True or false- Adding a small film cap across a large aluminum electrolytic will improve its performance by a large amount (in terms of ESR) at the highest audio frequencies.
True or false- Aluminum electrolytic capacitors are useless and have high impedance above their self-resonant frequencies since they become inductors.
Enjoy!
True
Parallel caps can be good for quality (especially true if pairs of exact same model cap, yielding a combination of ease and higher quality).
True doubly
. . . and it is true that there is much documented misunderstanding of parallel caps for audio frequencies. Notably, the easiest and/or least problematic ways of doing this are the least often seen.
False
Going by the 20% (or rarely 10%) tolerances of the larger caps, adding a bypass makes no *significant* change in total capacitance. . . because it does not change the total beyond the larger cap's specs.
But, if you had asked if the effective inductance reduces beyond the larger caps's tolerance specs, then my answer would have to be quite different.
False in too many cases, because the matter takes more care than that description mentions.
Directly paralleling a random selection of different size caps without ballast resistance in-between those caps, is usually destructive to performance.
In audio circuits, this leads to wrong location fixes applied and without due care, comes with significant risk of a permanent shortfall of performance, since the audio appliance becomes more difficult to work on, usually resulting in the need of somewhat greater compensation which then doesn't work as fully as expected, with the locale of the shortfall becoming progressively more difficult to identify. All it takes is one mistake, such as omission of ballast resistance between different size caps, to begin a snowball effect of wrong-location fixes applied and then more applied to fix the added problems those cause, eventually resulting in a really bad mess that never had the original simple problem fixed. The resulting epic mess CAN perform well, although the construction labor is excessive in that case.
This relates to your other question about misunderstood at audio frequencies.
The deal is that if different size caps are directly parallel, then LF roll-off of the smaller cap has to serve as ballast resistance. And therefore I can claim that most bypass caps are too big, in which case wishful thinking has not guaranteed better performance.
False
Aluminum electrolytic caps are useful and the HF roll-off depends mainly on the capacitance value, a feature also shared by plastic caps as well.
False doubly
At large signal power needs, the inductance or ordinary/power grade caps can be employed as a portion of a useful filter, and that is also true of larger size plastic caps as well.
Thanks! And, good morning.
Last edited:
Member
Joined 2009
Paid Member
This is the kind of post that makes DIYAudio valuable - real experience, no hyperbole. Some of it may not be applicable to *my* application but there's a responsibility on behalf of the reader to learn enough to know what they can apply.
- Status
- Not open for further replies.