I disagree with this blanket statement. Peases' model for compensating precision integrators had no residual distortion proportional to the starting DA. His model was completely linear.
At the risk of beating the dead horse again, as Samuel Groner's measurments show yes there is a difference but film capacitors with ridiculously low distortion levels are readily available at a nominal cost premium.
There are different kinds of 'Tantalum' caps, and some are much better than others. The ones that I had the most trouble with were typical molded caps made for consumer use. The 'wet film' caps are much better in every respect. They are just very expensive and VERY polarized, not being able to tolerate any reverse voltage at all. This precludes them from being used as typical coupling caps.
In otherwords "boutique caps" ..... Scotty ..
Yes, if boutique means $2 rather than $0.5 fine, for the rest, well, consider the source. The boutique capacitor market was established to fleece unsuspecting marks from their hard earned cash.
Last edited:
No, I am not, Jan, don't be ridiculous.
I am saying outright that they are stuck in their old ways and are not moving on with the industry. I am saying what was good in say 1975 may no longer be good enough in 2014, 39 years later.
I am also saying that this inability and/or unwillingness to follow the market trends cost Germany a good part of its audio companies, such as Telefunken, Saba, Nordmende, Wega, etc.
I understand development, I understand they wanted to get the most out of their initial design, but EVERY design sooner or later reaches a dead end. Then, you either adapt, or you go under.
No, that's the common fate of all companies which rely too much on one man's authority, which by default cannot last forever.
And that's not true only of audio, that's a general truth. It has been common knowledge that the most critical time, when the vast majority of companies fail, is the step from a small to a medium size company. This implies a sharing of design, marketing and financial decisions, which the original owners somehow never want to do, and worse, which they never prepare for.
I don't remember the exact figures any more, but over 90% of start up companies fail at that time and for those reasons. Quite a few of them had viable products or product lines, and yet they still failed.
In my life, I've started three companies as an external consultant. While they listened to me, they prospered. Later on, two contiued to listen and are still alive, 25 years down the road, doing well, and one has prospered very handsomely. The third didn't listen, and went under some years ago, leaving behind a pile of debts and unpaid bills and a VERY bad rep of its owner, who now finds that nobody will loan him money to start up again.
It's really all about balance, you have to weigh carefully between several aspects to get it just right, and you have to be quick about it. No tricks, just my instence that the employees be married to their business results, relatively low pay cheque, but a wide open door for bonuses, and no boss/owner wincing when it turns out that some people had some fat cheques coming. Install that system in a small (1-30 people)/medium size (31-300) company and it will likely grow.
bonus systems can hard on engineers - easier for the sales/marketing guys to claim credit for their effort while the engineer is at the mercy of subjective and often uniformed evaluation of their "contribution"/meeting of goals set by others
telling marketing that their hopes for cost/performance is simply not physically possible may be interpreted as the engineer not being "a team player"
telling marketing that their hopes for cost/performance is simply not physically possible may be interpreted as the engineer not being "a team player"
Sounds like a sign of an "infestation":
http://www.amazon.com/Snakes-Suits-When-Psychopaths-Work/dp/0061147893
Tactics like this are often used by some people to climb the ladder.
True. And also true depending on actual type of company product.
But possible nevertheless, especially in small companies, where the company founder is usually also an engineer. Problems will start once he HAS to delegate some of his duties and hence decisions to other professional in leading a business.
That's why it is so important to impress upon the owner, usually of technical persuasion, that engineers are useless if the sales flop, but also that for sales to be good, you need good engineers.
A small company breaking through needs to be better than most of its compentition and cheaper than most of its competition. Engineers usually want more than you can afford at the moment, so you need to evaluate and weigh their demands with actual needs and financial capability. Sales people usually want massive ad campaigns you never have money for at that stage, so again, you need to weigh their needs against your actual capabilities.
I am not saying it easy, it isn't, in fact sometimes it can be a bitch of a job. and mistakes will be made, which is why you need to develop a very effective feedback system.
And, although an economist by trade, I have to say this - my experience is that it's almost always easier to find common ground with engineers than with the sales people. Engineers are usually more focused, sales people tend to drift and dream big dreams too easily. But once you make a deal, you better stick to it if you want to last. Start to wiggle and you lose credibility, lose that and the s. soon hits the fan. Stick to it, and sometimes, almost impossible things become possible.
A company I was working for once got flooded overnight, A pipe burst and in the morning, the whole machine shop was under a foot of water. Estimates said no less than 10 days to get working again; it was done in 4 days, all employees, including mysef, pulled unpaid double shifts four days in a row. A month later, my last official act was to fire myself, they didn't really need me any more, I was almost dead weight. At the time, they employed 22 people, today they employ 47 people, and everybody is happy - myself included. I go to check up on them every now and then. They throw some work at me every now and then.
Trust me, it works, no matter how hard it for the first 12 months.
Last edited:
I agree as a blanket statement it is broad. Maybe disagree with the source of cap's distortion and how it manifests itself. PP film are cheap enough and do a respectable job. Its what I use most of the time. better is that ever harder to find PStyrene dielectric. No need to spend a lot of money for them, IMO. Except for the Multicap by REL Cap, all others are pretty much the same except in quality of construction. I do not endorse Boutique caps. Just excellent quality PP and PS as film replacements.
THx-RNMarsh
Last edited:
What one needs to do... and this applies to audio amplifiers as well, is to stop depending on sine waves exclusively for linearity testing of audio amps.... which have an average of zero. And then using test methods that are inherently averaging also. You will always get deceptively low Numbers. I suspect that with asymetric signal sources, evaluations of distortion will be more in line with listening evaluations.
[pls dont bother me with - sine waves can characterize everything in electronics perfectly story].
JC, if you recall, I found that DOW via the U. C. Berkeley library search and sent it to you... later I also found an excellent paper by TRW with the more complete model.
Besides DA, the DF et al which affects the phase angle Can be a major head-ache... though not a distortion artifact. It has audible consequences. This is especially true with EQ networks/filters. But again, if you can get the value and size of a good PP or PS, that isnt too much of a factor with the caps if used in the range well before the resonance. And, as I have shown before that phase and resonance is greatly affected by trace and lead length well inside the audio range with practicle circuits used for audio.
THx-RNMarsh
[pls dont bother me with - sine waves can characterize everything in electronics perfectly story].
JC, if you recall, I found that DOW via the U. C. Berkeley library search and sent it to you... later I also found an excellent paper by TRW with the more complete model.
Besides DA, the DF et al which affects the phase angle Can be a major head-ache... though not a distortion artifact. It has audible consequences. This is especially true with EQ networks/filters. But again, if you can get the value and size of a good PP or PS, that isnt too much of a factor with the caps if used in the range well before the resonance. And, as I have shown before that phase and resonance is greatly affected by trace and lead length well inside the audio range with practicle circuits used for audio.
THx-RNMarsh
Last edited:
Richard,
With your last statement about lead length and trace length that would seem to point to a superiority for audio purposes of surface mount components due to their physical size and lack of any real lead length and the much shorter trace distances required. Besides the difficulty that us old guys have with just seeing the damned things is there any reason not to go that route. I just look at the miniscule sized board and I think lateral mosfet devices that Lazy Cat uses for both of his amplifiers and it does seem if you can work with the devices there are real benefits to doing things that way. What is your "opinion" on this subject? I know there will be dissenting "opinions" and some will say use vacuum tubes but that just seems silly today, and thru hole devices do take a large real estate to use in today's ever smaller package sizes.
With your last statement about lead length and trace length that would seem to point to a superiority for audio purposes of surface mount components due to their physical size and lack of any real lead length and the much shorter trace distances required. Besides the difficulty that us old guys have with just seeing the damned things is there any reason not to go that route. I just look at the miniscule sized board and I think lateral mosfet devices that Lazy Cat uses for both of his amplifiers and it does seem if you can work with the devices there are real benefits to doing things that way. What is your "opinion" on this subject? I know there will be dissenting "opinions" and some will say use vacuum tubes but that just seems silly today, and thru hole devices do take a large real estate to use in today's ever smaller package sizes.
Here is an often over-looked example of cap quality affecting sound but distortion measurments show no problem... though listening can and often does show audible differences;
Two characteristics of caps are affected by current thru them..... series R (esr) and phase changes with freq -DF- and esr changes. You can run sine ways thru the speaker cross-over but read nil distortion from the caps in it. But internal temp increases cause the esr to increase. Freq changes cause the DF to change or loss tangent. How many Amps go thru a typical speaker cross-over cap? Large values tend to be Bipolar types with high esr and DF to start with. High currents thru thin metallized film and poor diectrics give rise to a dynamically shifting effective esr which then changes the phase angle and this affects the cross-over's desired characteristics. Going to a PP (if the cost and size/value is do-able) minimises these affects and this often is heard. It is sometimes described as a distortion has been removed.... but testing shows it isnt harmonic distortion. Its another one of those "as if" conditions.
THx-RNMarsh
Two characteristics of caps are affected by current thru them..... series R (esr) and phase changes with freq -DF- and esr changes. You can run sine ways thru the speaker cross-over but read nil distortion from the caps in it. But internal temp increases cause the esr to increase. Freq changes cause the DF to change or loss tangent. How many Amps go thru a typical speaker cross-over cap? Large values tend to be Bipolar types with high esr and DF to start with. High currents thru thin metallized film and poor diectrics give rise to a dynamically shifting effective esr which then changes the phase angle and this affects the cross-over's desired characteristics. Going to a PP (if the cost and size/value is do-able) minimises these affects and this often is heard. It is sometimes described as a distortion has been removed.... but testing shows it isnt harmonic distortion. Its another one of those "as if" conditions.
THx-RNMarsh
Last edited:
My opinion is that short trace and wiring lengths are more important than the part size... which is usually shorter than the trace lengths compared to any part package size. [If I have to build it and not a robot, i use thru hole only]
-RNM
Another relevant R that increases in this scenario is that of the voice coil heating up, might be a lot more than the increase of esr of even the worst crossover caps. Many reasons to avoid analog filters altogether.
Internal heating.... whether from PS ripple current or audio signal will raise the internal temp when those same speaker amps are flowing thru a speaker filter/cross-over cap. The popular use of metallized film makes it a guarentee the esr will rise significantly. I use film and (thicker) foil instead to keep the phase angle steady and predictable.
THx-RNMarsh
THx-RNMarsh
Last edited:
- Status
- Not open for further replies.