Thoughts On LM1875 Gainclones

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
ionomolo said:
...underrated ones to perform very badly.

Leaving those out, the point I was trying to make is that tweaking the sound of an amplifier by altering the voltage rating of a cap is bad science. There are a myriad of other parts and parameters to look at first. Of course, this has been posted all over this place many times and you are not the one that needs to be convinced it seems.

/Hugo
 
Netlist said:


Leaving those out, the point I was trying to make is that tweaking the sound of an amplifier by altering the voltage rating of a cap is bad science. There are a myriad of other parts and parameters to look at first. Of course, this has been posted all over this place many times and you are not the one that needs to be convinced it seems.

/Hugo

I'm not after the voltage rating. I'm after the canister size.

Too small, with weak materials all mashed close together. . . and so will be your audio signal.
Too big could harm the treble output.

Its needful to be aware of these factors, because a sufficient and inexpensive job at all of the voicing options will have a much better outcome than spending a mint at a single location.
 
Daniel,
What you are trying to tell us is plain wrong and not based on any scientific evidence.
As Anatech and many others have pointed out, the size of the cans have gone trough an evolution stage over the years. New materials and more refined production processes have made it possible to increase the size of ‘all’ parts we see in electronics today.
By claiming that the sound of an amplifier can be tweaked by the physical size of its capacitors, you are deluding yourself and other, less experienced members.

/Hugo
 
See proof attached
Photo is from the scale of a standard basic ESR meter, showing, beyond a doubt, one of the many ways that cap size can affect sound. This is seen at same capacitance and within the same capacitor model lineup.

From Nichicon's finer grade, a quote: "FG Series- larger capacity in a smaller package with a minimal decrease in electrical quality."
Note that there is, in fact, a documented decrease in quality in "modernsized" conveniently smaller capacitors.

In addition, a search here on diyaudio.com can turn up measurements taken with more advanced equipment.

Yet more proof is available in capacitor manufacturer provided spice modeling tools, and this lets you model the effect of various size capacitors for your amplifiers.

So, I'm saying that if a 50v capacitor selection doesn't suit you for application at Ci, then by all means try a 100v capacitor selection because it will most certainly sound different. The same is true between modernsized or traditional capacitors, according to Nichicon.

There's plenty of proof, and debating whether it exists is quite tiresome.
 

Attachments

  • basicesr.jpg
    basicesr.jpg
    75.7 KB · Views: 642
Disabled Account
Joined 2007
Netlist said:

By claiming that the sound of an amplifier can be tweaked by the physical size of its capacitors, you are deluding yourself and other, less experienced members.


Not much of a difference from the other widely believed claims floating around audio.
I'd list some, but we all know what they are.

If you keep up the good work Daniel, in about a year your capacitor theory will be accepted, established audiophoolery.

:)
 
MJL21193 said:
Not much of a difference from the other widely believed claims floating around audio.
I'd list some, but we all know what they are.

If you keep up the good work Daniel, in about a year your capacitor theory will be accepted, established audiophoolery.

:)

I guess that it was my terribly complicated way of debunking the "Ci must sound bad" myth. I did my best to illustrate that there are options, and not just one single possible result.

I didn't want anyone to avoid the protection because of assumptions. I like my LM1875 better with Ci in place, especially after tuning the bandwidth with Andrew's help, so the bass is very high resolution (and speakers are protected). ;)

EDIT: Sorry that I went "off the deep end" but the kits that I've purchased contained unseemly components for use at Ci, and most of what I was going on about, is specifics on how to make a difference--hopefully a pleasant difference.
 
I have some caps that are significantly diffirent porportions to each other....but same voltage and capacitance.

There is so much conjecture on this floating around it is not funny..

And every one seems to have diffirent results if any...

I don't realy fall for the special caps things.. at least not in the low dollar bracket us guys normally look at ... there are plenty $50 caps well worth their money...

As far as I am concerned most caps "sound" alike, but some sound like crap (and I experience this more as sibilance). Once in a while you run into a cap which adds a bit less colouration... and this, in my experience, were the "larger" caps for the voltage ones... there are however so many hundreds of difirent caps on the market, that the ones I have sampled can not begin to be used to form statistical opinions...

I think you need to start building some discrete amps now, and move away from the chase after the magic cap... I have only built a few now, but I would stake, my current one, built with pretty cheap components includeing the nastiest looking BP electrolytic as Cin (to prove a point to a freind) will hulimliate your chipamp in resolution detail, and the unmeasurable, musicality.
 
That sounds like fantastic fun! ;)

P.S. When nasty caps work just fine for input filters (Cin), that means that the rest of your amp is very, very good.
Impressive!

P.P.S. I wondered where all my diodes went. I seem to buy about 3x as many diodes as ever make it into power supplies. Mystery Solved!!! Just now, I saw Domino (kitten) taking them, and transporting them (at maximum speed) to underneath the furniture. ;) She takes the "N" so I have to use the MR/MUR. lol!
 
Originally posted by Nordic
As far as I am concerned most caps "sound" alike, but some sound like **** (and I experience this more as sibilance). Once in a while you run into a cap which adds a bit less colouration... and this, in my experience, were the "larger" caps for the voltage ones... there are however so many hundreds of difirent caps on the market, that the ones I have sampled can not begin to be used to form statistical opinions...

I totally agree with your appreciation of capacitor distortion, but (sorry for saying it again) I started with diy amps reading at the headphone community forums and being told one time and another that if you don't use hard to find parts whose properties are unreleated to any objective measurement you will get terrible results made me think that this was more like witchcraft than science and that i wouldn't be able to make anything that sounded good.

I have built many amplifiers that sound well and don't use exotic parts.
 
All I'd like to see is reproducable, inexpensive, and available options. The $12 LM1875 K50 kit comes with a 22uf Nanya, but its NFB is 180k with 10k (a workaround). Actually, that kit sounds rather good. But I'm picky, so I like options.
And, I'm curious, so I like to explore all of the options. My second favorite input filter cap is 17 cents. ;) The favorite is homemade.

I also think that audio equipment is for pleasing ears, so that means whatever you like IS the best. ;)

P.S. I got my diodes back (temporarily).
 
Originally posted by danielwritesbac
Photo is from the scale of a standard basic ESR meter, showing, beyond a doubt, one of the many ways that cap size can affect sound. This is seen at same capacitance and within the same capacitor model lineup.

What is the effect of, say 10 ohms ESR, in series with a 10 KOhm resistor with 1% tolerance (+/- 100 Ohms) ?
 
ionomolo said:
What is the effect of, say 10 ohms ESR, in series with a 10 KOhm resistor with 1% tolerance (+/- 100 Ohms) ?

10ESR would be an average resistance of 10 ohms, but its not so nice as that. That's just the average.
Some of the cap will have "perfect" (very hard sound) response, and some of it will have much more resistance (very soft sound) for a lesser response. Although ESR and resolution aren't generally comparable, there is frequency beyond which no capacitive effect is seen. This varies.

The bypass cap practice is all about decreasing ESR--right on time. The larger value capacitance can lose effect at higher frequencies. Those can be "patched" with the bypass cap practice and either a lot of highly-questionable formulas or a small bit of luck. Since Ci is pretty big, it doesn't take much of either formulas or luck. Attempt values from 0.0047uF, 0.01uF and 0.022uF.
Try smallest first, and stop at the value where results are seen. Going beyond necessary results in distortion.
Match soft to soft (50v-100v polyester) or hard to hard (Polypro/MKP) for sound.
Just don't overdo it--They're only a few cents. ;)

Its not necessary to replace the cap or bypass it when you can change the band by choosing different resistors, such as the LM1875's available 20k, 33k, 50k, 100k, 120k, 180k, 220k options.

Please beware of spending more than $1.50 for a cap at Ci, because most of the decent examples are $0.85 (85 cents) and lower--with plenty of variety to choose from. You can get more fun varieties from choosing the size/voltage of your cap than you can from its price.


Hey, but now I've got a question for you! ;) How do I make a Schumann pulse / Theta wave generator from LM1875? The place where I work is made of metal and I don't always like to wait until 1:30 a.m. to hear as clearly as possible. I think that it would be very cool to make such a high end device from a $12 amp. ;)
 
More realistic values give 0.02 dB assuming that ESR changes from zero to 2 at all audio frequencies. This change is exagerated so the effect won't be noticeable.

I can't help you with your question since i've not studied anything about bioelectronics, but i must say i'm skeptical about it.
 
ionomolo said:
More realistic values give 0.02 dB assuming that ESR changes from zero to 2 at all audio frequencies. This change is exagerated so the effect won't be noticeable.

Oh good point! With the K50 kits 180k and 10k, there's very little ESR effect seen at Ci (because of the 10k additional). I think that the smaller of the two NFB values would have to drop below 3k (approximiation!) before you would begin to have a more prominent role from Ci.

I wonder what (component, value, setup etc. . .) makes the difference between the audio presentation observed?:
Ahead of a driver (stock k50 LM1875 kit)
Behind a driver (laid back)
Center stage (both at once)
A clue on this would be really wonderful.
 
From the bowels of DIY Audio comes - someone thinking about this 10+ years after...

Every circuit I've seen has that capacitor to ground in the negative feedback loop, which promotes safer operation by reducing the gain to 1 at DC. Alright and good, however this cap definitely effects SQ and IMHO if your going to use it, should be the best quality cap in the system. I've eliminated mine - am I riding on borrowed time by doing so, before one day the whole thing blows up when switched on? I'd think the 1875 should operate OK, DC included, like any other Op-Amp. Wrong thinking?

The two diodes in the +/- supply rails, I cant for the life of me find an explanation for how these work. Manufacturers spec sheet is just a list of the different voltages vs diode type for the surge protection devices. With current flow from anode to cathode, is it just a diode? A diode that has 7.5V drop? Is the 7.5V level reversed - cathode to anode - like a zener? If someone could 'splain "How It Works" I'd love to hear.

I've read where this amp has the ability to sound "tube-like". With two caps in the signal path, I think I can understand why, as most common tube amplifiers have such a capacitor coupled arrangement by design. In a LM1875 amp design, are these caps what is chiefly responsible for "voicing" the amplifier?

I've read where other voicing techniques - such as placing a 4.7 uf from + to - rails - but specific brand / vintage electrolytic - can improve the midrange sound. Could someone explain how this works? From my understanding of capacitors and sound, I'm much more comfortable with such an arrangement than an electrolytic directly in the signal path - with AC across it. Thanks for any enligtenment on these topics!
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.