Help with Warmth vs Detail

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
The amps were originally design for JBL in the early 60's, as audio amps. I changed the driver stage load resistors from Caddock MP925s to some old Koolohm wirewound type. Now we are getting some where. I believe that I am now headed the right direction. More warmth without the loss of detail. Not sure about everyone else's system, but a resistor change is quite detectable. I also believe the room is a very big part of system response. The dyna's may be cool, but with the power of over 200 watts from 8 EL34s you can make the walls shake in this house. Lack of bass isn't and issue period. The wall/ceiling/floor coverings and tube traps allow room tuning and the coolness of the speakers can be altered through absorption and diffusion. I think many should consider the room as a component of the system that is easily modded and can dramatically affect the overall sound tone and soundstage. An accurate speaker should portray what is feed to it, warm or cool. I believe this is better that having a speaker that is only warm or cool. My opinion.
 
Jay, could you bring your speaker to my place. I live in a remote area in the mountains of Nevada/Cal. No one around to share equipment with. Would be a great Idea, just not applicable. It would be great to have that opportunity and would broaden my experience. It would not however change what sound I would like to have.
 
Speedskater, thanks for the link. What I read is that there is small deviations in the speaker response. Two interesting things. An upper end tilt, possibly adding to the brightness. However, an increase in upper bass which some here say would add to the warmth. Maybe offsetting!! I think that these effects has been completely made mute with the changing to the wirewound resistors. Listened last night for a couple of hours, and I believe this may have been at least a bullet(not the silver one) that has gotten the sound much closer to what I was hopping for. Much more warmth with very little loss in detail.
 
I see a lot of problems here.

1. looks like OP bought the speakers without hearing them, based on reviews and/or internet opinions. people have very differing definitions on warmth and detail and it's unrealistic to expect that it's easy to find two complete strangers that have the same definitions.

2. a lot of the usual "dogma". specialty resistors and caps etc. I should post a pic with the crossover of my Confidence 3's (a few times the price of the 1.3's when new). I think I counted more than 20 caps. Solens, nothing fancy. some are generic MKT/MKP and there's one tiny ceramic somewhere.

3. OP never tried EQ but knows for a fact that it is bad.

4. any speaker, and that includes the 1.3's has its limits. why does Dyna make the Evidence or Consequence line when the 1.3's are all there is? maybe it's all the 1.3's have to offer. I have listened to them in the past. they are detailed but with certain musical program they sound thin, boring, uninvolving, not musical. my first DIY speakers which cost me $300 in parts sounder better in some respects. no real soundstage, no detail but certainly better for music listening and not sound dissection.


but the string soundslike a hard pure tone without harmonics.
you are describing the typical Dynaudio sound. some call it accurate, some call it lifeless. some compensate for it with upstream electronics.

my guess? OP has fallen in a typical audiophile trap: buys equipment based on reviews and is going through a denial phase.
 
clutching at straws, IMO.
but, nevertheless, I have an audiophile Air Density Regenerator on sale for only $9999.

really, what you are hearing is the entry-mid level Dyna sound. and your tastes simply don't match it. can't comment on the upper level offerings as I haven't listened to them.

here's the Confidence 3 XO: http://photo.qip.ru/users/aspero2010/96610719/125446896/

the most basic caps, coils and resistors. you are deluding yourself thinking that a resistor in the amp is the problem. stop searching for the fix on online forums (hell, maybe I myself am full of it after all!) and better start shopping for new speakers, and only buy after auditioning in your room and setup.

LE: and one more thing. I haven't listened to those speakers and I'm speaking based on hear-say. the Sapphires seem to be something else. if I recall correctly the Stereophile review hinted about a sound differing from the typical Dynaudio one. I know people who heard them and said "these are something else". likely, they were talking about the typical "neutral" (note the quotes) Dynaudio sound which the Sapphires probably don't share.
 
Last edited:
IMD arises from opamps IME - especially in conjunction with digital stuff because that's putting out RF which opamps hate with a vengeance. I got a substantial lowering of my noise floor when I put a passive filter behind the I/V resistor of my NOS DAC. Only then did I learn first-hand what 'warmth' really means - it has led to a substantial improvement of tonality, especially on piano sustained notes.
Malcolm Hawksford talks about this in an interview. I have experienced it with a NOS TDA1543 + LM4542 output stage. and it's why I started taking all tweaking comments, especially when about DAC output stages with a grain of salt.
 
mr_push_pull , Doing a lot of speculation here. Not worth a comment. My 1.3 don't have 20caps, but yes solens are there. Please don't speculate what has or hasn't been done or experienced. Nuff said.


As it turns out, I found what has been plaguing the system for some time. I have been working with fellow DIY memebers on a positive path and have Isolated a major contributor to the brightness. There is a high freq suppression cap/resistor set on each output transformer. The caps are 60 year old 1000pf ceramics which apparently have failed. I replaced them with the only caps I had available, 2000pf silver/mica. BINGO!
Yes, warmth is here to stay. I have to retune the room and have already removed some panels as they are no longer useful. I like the new sound, with rich mids THAT I LIKE!
That is what this is all about anyway.
 
mr_push_pull , Doing a lot of speculation here. Not worth a comment. My 1.3 don't have 20caps, but yes solens are there. Please don't speculate what has or hasn't been done or experienced. Nuff said.
right.
you are deluding yourself. I own $8000 Dynaudios that, btw, are simply better compared to your speakers which I listened to. guess what? pretty much the same Dynaudio house sound.
you have explicitly stated that you bought them w/o auditioning. wake up and smell the coffee. you just don't buy stuff w/o listening to it and then try to fix it. go to audio shows, hi-fi dealers, listen to stuff, borrow equipment, listen to it, buy what you like.
period.
 
I have been working with fellow DIY memebers on a positive path and have Isolated a major contributor to the brightness. There is a high freq suppression cap/resistor set on each output transformer. The caps are 60 year old 1000pf ceramics which apparently have failed. I replaced them with the only caps I had available, 2000pf silver/mica. BINGO!
Yes, warmth is here to stay.

Would that be the Zobel network at the amp output by any chance? There are real experts on DIYAudio, but in case no one else is commenting on it, the idea of this is to cancel out some of the reactance of the load, which might otherwise cause high frequency instability in amps which use feedback. If so, and the caps had failed, that might indeed have been causing audible problems. Maybe you have found your silver bullet!
 
Here's a list of some possible esoteric options. These are all from memory.

Clean power:
From Keantoken: An RC of 50R trimmer+2u polyester cap set from ~ to ~ of the bridge rectifier(s) to snub the transformer secondary(s). This RC does not attach to 0v line and therefore can't ruin audio. But we might really enjoy decreased power noise.

Clean source power:
A K-Multiplier put before the regulator for a digital source.
Likewise, nested regulators.

Old fashioned overkill:
A high end old style inductor based equalizer.

Information extraction:
An extremely small dose of stereo expander, which, on a non-inverting amplifier, is much like an RC from speaker output of one channel to inverting input of other channel. A homeopathically mild, transparent, setting, is what I'm talking about. In a really high end system, such a damper has to be set at the make-break point of work or not--an existential stereo expander. When aligned to make-break of noticing or not, the circuit works dynamically in practice and if done carefully, could extract even more detail without causing a more forward tone.
Stereo expansion circuits like that can be used as compensation for the practice of using lead lag comp for acutance (audio version of a photo sharpening filter), which does cause a more forward tone as its main cost.

Current headroom:
(what a modern source doesn't have)
Modern sources with high profit works of omission direct driving power amplifiers typically results in poor bass plus forward mids. If this is irksome, a careful job with a good buffer project can level it out. To help a modern source, I would suggest trying a fine quality discrete parts buffer with adjustable regulator power and adjustable bias. You're simply installing the omitted parts onto the output of your modern source. For example, it helps the Creative Labs X-Fi finally make real bass instead of EQ reliance with insufficient outcome. Yes, if what you needed was current headroom, a digital EQ doesn't substitute.

Checking various PC sound cards reveals one, just one, interesting bit of information. The M-Audio Revolution 5.1 that Ostripper and I both like, has a Pair of Parallel stereo op-amps to cancel some crossover noise and enhance current headroom. With effective use of op-amps, this one sound card doesn't need a buffer.

Needed an entirely different output section: I recently tried an M-Audio Audiophile 2496, thinking it would be better but that awful thing was returned to its cardboard box in less than 20 minutes. Unfortunately the headache it gave me lasted the rest of the day. Maybe the computer power supply affected that reputable sound card, causing it to fail on quality. Perhaps it needed a coax connected external digital to analog converter run on linear+capmulti+regulated power, not computer power.

Impedance matching at input:
Such as with Lightspeed+Buffer, or BlareBuster, or Input Transfomer, to knock out the shout, not the details.

Decreased midrange compression:
Rail2Rail cap can remove some midrange power noise resulting in cleaner, quieter midrange, specifically upper midrange, ear sensitivity peak. This can be done at the power amp with 1 really cheap polyester tweeter cap, or an RCR, or a Cornell 250v SEK, and the typical approach is 1 cap or RCR of 1u~4.7u range from V+ to V- of the amplifier board. The added cap does Not connect to 0v.

Base stoppers:
Got em? Carbon film and carbon comp work nicely for base stopper locale. For example some high quality brands of 1/2w carbon film resistors work nicely for feedback-shunt resistor in non-inverting amplifiers. Improvement from using base stopper suitable parts is not dramatic but could be noticed in areas of a circuit that have gain.

Lastly:
Tuning by ear defines the need precisely, but fails to specify the location for repair/adjustment, and that has risk of applying more patches because of more patches, so we need to be a bit cautious there.
With a really high end system, fell swoops won't do, but multiple, transparently small, little steps work fine.

There's probably a lot more to it, but this is all I could remember at the moment.
 
Last edited:
As it turns out, the most significant contributor to an enhancement for MY systems midrange was to change out the conductive plastic volume pot on the amp to a tantalum resistor divider control. I tried quality MF resistors, which improved the definition/sound stage etc. but left the same edge/harshness and sibilance I wished to eliminate. The tantalums gave the smooth and warm mids while eliminating the edge and excess sibilance. Since the detail and the highs are coming through although somewhat decrease from the past I assume this process is not because of a filtering effect by the tantalums, but the need for me to adjust my years of hearing excessive and edgy systems. Time will tell, but I like this new sound very much. Funny how such a small component change can make such a big difference regardless if it is for better or worse. I don't understand how others could say that this couldn't happen. Again, years of trial and error have convince me not to be skeptical until I hear for myself.
 
Not funny at all, this is bloody serious ...!! :D The closer one gets to optimum sound, the more significant the most minor alteration has to the subjective perception of the sound - this can become a nightmare, and did for me many years ago!

I'm a little concerned that you're losing detail, of course if this is false exaggeration of subtleties then this is not a bad thing, but one has to be careful that what one is doing is not just adding syrup to the sound.

How have you implemented the tantalum resistor divider control? And, have you considered or tried something like the Lightspeed optical pot?
 
A very recent experience has reinforced to me the importance of seemingly obscure but in my experience critical setup issue. One of my monoblock power amps recently met an unfortunate demise after I was doing some soldering on my speaker crossovers and overlooked turning off the amp. A grounded iron touching a positive leg of the crossover resulted in me letting out the smoke on one output transistor; predictable but infuriatingly stupid...

Anyway my small backup LM 3875 chip amps went into the system. They sounded detailed but thin and unappealing. On a whim I measured the leakage currents on the chassis. They measured approximately 100v AC. I therefore inverted the primaries of the transformers and the measured voltage lowered significantly. Most importantly the sound improved in an extraordinary way. Suddenly there was more ease and the decay of notes was much more audible and natural.

So the same amp with correctly oriented mains primary connections sounded significantly "warmer" and more natural.

If you haven't attended to this fundamental issue, you don't know what you are missing. Best of all it is free.
 
Leakage currents? 100v AC?? :confused:

But, yes, that is precisely the sort of thing that becomes crucial! Unfortunately, it's easy to go the other way ... you start with natural and full sound, but feel you can go a bit further, so you implement a change which is fully rational, makes perfect sense, but after a while realise that the sound is gravitating towards "thin and unappealing" ... :eek:. If you think then that in fact you're getting closer to the raw, undiluted, recorded sound, well, you're wrong: the change made was a wrong turning, or requires other elements in place also to function correctly. The whole process can become very complex, because everything is interrelated - there are no easy answers ... at least, at the moment, :)
 
Leakage currents? 100v AC?? :confused:

But, yes, that is precisely the sort of thing that becomes crucial! Unfortunately, it's easy to go the other way ... you start with natural and full sound, but feel you can go a bit further, so you implement a change which is fully rational, makes perfect sense, but after a while realise that the sound is gravitating towards "thin and unappealing" ... :eek:. If you think then that in fact you're getting closer to the raw, undiluted, recorded sound, well, you're wrong: the change made was a wrong turning, or requires other elements in place also to function correctly. The whole process can become very complex, because everything is interrelated - there are no easy answers ... at least, at the moment, :)

Given that the AC voltage is capacitively coupled to the chassis there is nothing safety wise you need worry about and those sort of measurements are surprisingly common in 230 v AC land. That said I think spacing the transformer further from the chassis will help lower the leakage voltage further.
 
Last edited:
Implementation of the tantalum voltage divider is a directly soldered to the input. Only one setting at the moment, and I will be adding a six position switch, shunt style control later. The change in the input impedance should be percentage wise quite small as the overall input impedance is 250k and the input side of this divider will be at most 38k of the total 250k. In other words, the range of input impedance will only vary from 200k to 238k over the six settings.
............................

I'm a little concerned that you're losing detail,

I too though this, but with additional listening, is seems that more mid warmth was added, the edge gone with the sibilance in female voices still coming through. Its like now the highs coming through the warmth, without their presents all the time. Just nice!
................................

I too have gone through the AC polarity issues. My house does not carry the "ground" to outlets. Only the hot and neutral are present. However, I too found that polarity of power cords must be investigated. I found this important in the soundstage center area, where it was better defined and precisely held the position. Never measured voltage between components took any measurements. Just went by sound. Maybe less of a problem with only two components! Ok, four, cd player, two mono blocks and speakers, ops, five components.
..................................

have you considered or tried something like the Lightspeed optical pot?

Not on my budget as of yet. The tantulums resistors are from a 60's HP signal generator. High quality parts at reasonable prices, free, sans the silver solder.
..................................

The whole process can become very complex, because everything is interrelated - there are no easy answers .

Couldn't have said it better! For such a simple system everything affect the sound directly, even a deaf man could hear the changes.......................
 
Last edited:
I don't know if this works with LM3875; however. . .
LM1875 sounds like a decent and rather high resolution chip amplifier run solo; however, it sounds like about $5000 worth of "out of the black" voiced audiophile amplifier if run Parallel. The solo chip did not do both warm and detailed but rather one or the other; however, the Parallel gladly does both warm And detailed. Perhaps, therein, is a clue, or two.
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.