| fedde |
Hi folks,
yesterday evening I did a new experiment. I own an old Technics amp, the SU-Z1:
http://www.vintagetechnics.com/tuners/stz1.htm
It has a broken STK module. So I thought it was a nice plan to build a Gainclone in it. It has a very nice 2x25 Vac shielded block transformer. (regrettably only 130VA, but well...). It is meant as a simple amplifier for my father's computer. I used cheapo parts and some parts I had in my drawer. (spent 5 euro in total).
I used:
2.2 uF ERO MKT cap
1000 uF Nichicon 50V
some standard carbon resistors
CAT5 wires
WIMA 0.22 uF 63V MKS4 and 1/2 W 1ohm metal for Zobel
OPA541 chips
Cheapo 250 k pot (inside of technics amp, with lots-a-contact spray!!! ;-)
I used this time a non-inverted configuration. (similar to the GC of Richard Murdeys page: http://www.geocities.com/rjm003.geo...dio/diy_gc.html)
I guess this config is very close to the original gaincard (2x 22k and 680 ohm).
I finished one channel yesterday. And I could not stop listening to finish the second :bigeyes:
For some reason this thingie sounds far too good. I will do some research lateron to find out what's the reason for that. (the NI config or the OPA chip). It seems more controlled, dynamic and faster than my inverted clone...
Though it is maybe too early to draw final conclusions...
Fedde |
|
|
| fedde |
In the beginning this new NI amp sounded more controlled but less natural than my other amp. After a night of playing, it sounds better now. I was just listening to some guitar tracks. There is an awesome level of detail and resolution, very nice resonances of the guitar. Quite crispy and full sound! I hope I'll have same time this evening to finish my second channel... (should be 30 minutes of work)
Fedde |
|
|
| GregGC |
Please do finish your other channel and keep us posted with the listening results. I still haven't decided on the configuration, though I'm leaning towards the NIGC for now.
Greg |
|
|
| ronc |
You know i have been building IGC for over a year now and havent even tried the NIGC.I have a couple of chips laying around and just may try it ,based on feedys reports.I take it the feedback loop remains the same?I should be able to use some 50K alps pots shunted to ground for volume control? No buffer necessary they way i see it?
Hay 47 labs must be doing something right, according to the reviews.
ron |
|
|
| fedde |
Exactly. Few people tried the non-inverted configuration. And even less people tried the (supposed) 47 Labs configuration (680 ohm, 2x22k)...
I always liked the sound of my IGC very much at low levels and less at higher levels (still not bad though!). I found the current enthousiasm for buffering somewhat disturbing; the fact that buffering is necessary is a bad sign.
Have a look at Richard Murdeys page for the schematic. Use a LM3875 instead and 2x 4.7 BG N anti-parallel for input coupling. The alps pot would work, but is not my favourite (the blue ones at least...). I use it without a buffer (and so does 47 Labs).
It's too quick to draw conclusions, but I'd guess the NI config is a better match for my latest DAC (similar to the Shigaraki DAC, to be revealed soon!).
Fedde |
|
|
| millwood |
| quote: | Originally posted by fedde
Exactly. Few people tried the non-inverted configuration.
Fedde |
I am one of those who have tried both inverting and non-inverting lm3875, with various decoupling / filtering caps.
I for one cannot tell them apart, in spite of all the "scientific" arguments of how much worse the NI version would be.
At least in my experience, topology did not matter. |
|
|
| vic |
Hey,
Isn't a non-inverted gainclone just a gainclone??? I thought the Gaincard is non-inverted.
Vic |
|
|
| ronc |
True, so actually the gainclone aint a clone.
Geee we have gone full circle it looks like LOL.
ron |
|
|
| fedde |
| quote: | Originally posted by vic
Isn't a non-inverted gainclone just a gainclone??? I thought the Gaincard is non-inverted.
|
Yes, the GC is NI. Gainclones are clones of the gaincard. The question is where you draw the border with the term clone. I think it is irrelevant. My personal definition of the term gainclone is opamp amplifiers built in the spirit of the gaincard (small capacitance, compact and few parts, both NI and I).
BTW: I finished the second channel. I have to say that the sound is quite... :bigeyes:
Very smooth&musical, I like it!
Fedde |
|
|
| millwood |
| quote: | Originally posted by fedde
Yes, the GC is NI.
Fedde |
Damn! Those guys at 47Lab screwed up big time. They had an opportunity to make their GC sound 10x better, had they used the (far?) superior inverting topology on the GC.
The same thing goes with krell, or most other power amp manufacturers.
Maybe they don't know how superior inverting configuration is and how much better their amp would have sounded in inverting configuration.
:) |
|
|
| GregGC |
fedde,
What output DC offset do you get for the NIGC with no NFB cap to GND?
Greg |
|
|
| fedde |
I did not accurately measure. I'd guess around 10 mV. I will measure that when I am home...
Fedde |
|
|
| platenspeler |
| quote: | Originally posted by millwood
I am one of those who have tried both inverting and non-inverting lm3875, with various decoupling / filtering caps.
I for one cannot tell them apart, in spite of all the "scientific" arguments of how much worse the NI version would be.
At least in my experience, topology did not matter. |
Same for me, I built a non-inverting and an inverting clone and I do think that they are both OK.
I like the sound of the inverted design better, but as the inverted design uses more expensive components this could play a role here too.
Still amazed about the sound qualities of this chip though (and the LM1875).
Maarten |
|
|
| fedde |
Well, the difference in my system is larger. Probably due to the large output impedance of my dac (between 2k and 3k). If a CD-player with a buffer or a preamp is used, the difference should be small...
Fedde |
|
|
| Peter Daniel |
The difference might be large because you use different transformer and basically the PS and chassis, as well some components may be different (Technics?) and this may play big role in causing different sound.
I tried non inverting topology yesterday, and the difference is not that very big, although it seems to be;) |
|
|
| fedde |
| quote: | Originally posted by Peter Daniel
The difference might be large because you use different transformer and basically the PS and chassis, as well some components may be different (Technics?) and this may play big role in causing different sound.
|
Very true.
| quote: | Originally posted by Peter Daniel
I tried non inverting topology yesterday, and the difference is not that very big, although it seems to be;) |
!?!?
Fedde |
|
|
| Peter Daniel |
| quote: | Originally posted by fedde
!?!?
|
That's exactly my impression;) At the moment it's hard to come to any specific conclusion. I have to consult some other people, what they think. But I can't say that I don't like that sound.
I also have to get the same resistor types to be more honest in my evaluation. |
|
|
| matjans |
hey guys,
i have also tried both inverting and non-inverting topologies but I couldn't really hear a difference. Both topologies are dead quiet and pretty musical.
Small (?) note: the weakest link in my system is NOT the amp, it's my speakers. So... |
|
|
| Jennice |
Hi Guys,
Some talk about the LM1875 while othewrs refer to LM3875. Yet, both talk about a gainclone !?
What's the difference (if any)
:confused:
Jennice |
|
|
| matjans |
They are both power op-amps. THey can both be used in a GC.
The main difference is the difference in power output; whereas the 3875 delivers about 50W, the 1875 delivers a mere 20W. Fore more differences see the respective datasheets (link below).
I have never heard a 1875 so I can not comment on musical quality...
3875 datasheet versus 1875 datasheet |
|
|
| Jennice |
Thanks for the links. I'll study them ASAP ;-)
Jennice |
|
|
| rbroer |
| quote: | Originally posted by fedde
2x 4.7 BG N anti-parallel for input coupling.
|
IMHO the anti-parallel BG's sound nice (bit hifi-ish), but a bit veiled.
To be honest I didn't observe much change after some +400 hours, whereas others report tremendous changes due to breaking in :xeye:
Well I tried at least, but they're not my cup of tea. I'll stick to the good old KpSn caps. :cool:
As always YMMV... |
|
|
| indoubt |
Rudolf,
What brand kpsn are you using (intertechnic?). If those kpsn are sounding great too, or even better than black gates then I will go that route. They are easier to get here than black gates and waiting >400hrs for a cap to break in seems like an awfull lot to me.
:( |
|
|
| rbroer |
Indoubt,
they're like this:
I must admit that price/performance ratio is better for the BG's |
|
|
| zygibajt |
Then what is your non-inverted topology Peter?
Is is still 10K resistor in signal path ,220K (or 250K) feedback resistor and you just switched pins so now your "-" input is grounded?
Or maybe non-inverted topology looks diffrent?
Bartek |
|
|
| Peter Daniel |
I'm using 22k feedback resistor and 680 from - input to ground.
Signal is connected through 50k pot directly to + input. So it's 3 parts only and this supposed to be a setup used by GainCard. They also use RC network at the output (1 ohm, 0,22u) and a coupling cap with 22k shunt resistor at the input, but I didn't use them. If you have a chance, try it out, as I'm curious how you would compare it to the other circuit. I can't really complain about that non-inverting configuration. Seems to be sounding pretty good and my impression is that it has more clarity to the sound. |
|
|
| GregGC |
Peter,
What's the uotput offset voltage you get on the NIGC. Not using a cap to GND in the NFB means gain of 20 and more for the DC in signal. So I wonder if it makes output offset a bit too large.
I asked before but I didn't get clear answer.
Greg |
|
|
| Peter Daniel |
| It is smaller than in inverted amp, less than 10mV. |
|
|
| zygibajt |
Thanks Peter,
I will sure try it as soon as I order proper components values (no point to try this out with low quality ones).
I'm starting to think that this RC filter at the output is a good idea,as I'm getting ****ed of with capacitance oscilations in GC coused by high (not that very high) capacitance speaker cables.
I tried whole bunch of cables,some of them quite expensive ones,some factory made an DIY and most of them are simply cousing GC cracking.It seems like 300-400pF capacity is the limit not to exeed (at least with my speakers).My favorite ones are Thorsten's U-byte 2 speaker cables.They outperformed very expensive cables I had to compare.Mine are 560pF (2,5m long).
These are not qiute identical as I used thicker (10mm thick) sattelite cable.Orginals were based on CT100 cable.I will either have to shorten the cables,try to make them with CT100 (maybe less capacity) or put some RC at the output.I did one with low quality parts and it cured it but degraded the sound I think.Maybe better parts?
I also put some teflon tape around copper foil before puttting poliolefin heatshrink on them.You should try them if you haven't.
Any hints about that RC filter?
That 0,22R resistor changes nothing.
Bartek |
|
|
| Peter Daniel |
here's the basic schematic I tried
http://www.geocities.com/rjm003.geo...dio/diy_gc.html
and component values:
R1 ** C2 4.7 µF / 50V Black Gate N series
R2 22k C5 0.22 µF / 63V WIMA MKS
R3 680R C6 1000 µF / 35V United Chemicon LXV or similar
R4 22k C7 1000 µF / 35V United Chemicon LXV or similar
R5 1R U1 LM3875T
** 12 step ladder attenuator, use 22k, 4k7, 2k2, 1k5, 1k, 680R, 470R, 330R, 220R, 150R, 68R, 33R. A 50 kOhm volume potentiometer could be substituted. |
|
|
| fedde |
| quote: | Originally posted by zygibajt
Thanks Peter,
I will sure try it as soon as I order proper components values (no point to try this out with low quality ones).
|
Well, I think that the configuration is more important than the component quality. I use standard carbon resistors and it sounds quite fine. It can be bettered of course, but it is not too bad.
| quote: | Originally posted by zygibajt
I'm starting to think that this RC filter at the output is a good idea,as I'm getting ****ed of with capacitance oscilations in GC coused by high (not that very high) capacitance speaker cables.
|
Regrettably, I know all about the oscillation thingie from personal experience :(
BTW: I have 10.3 mV offset in the left channel and 20.3 mV offset in the right channel (!). This is with the OPA541
Fedde |
|
|
| fedde |
| quote: | Originally posted by Peter Daniel
I can't really complain about that non-inverting configuration. Seems to be sounding pretty good and my impression is that it has more clarity to the sound. |
Yep, nice crispy sound eh !? Acoustic guitars sound really nice. Full, but still fast. Try the Zobel once, I think it has a purpose...
Fedde |
|
|
| GregGC |
Fedde and Peter,
Thanks for the output offset levels!
Greg |
|
|
| zygibajt |
You mine the Zobel filter at the output ,which is 1R and 0,22uF to ground?
Experimenting with inverted GC I noticed that I prefered the sound with 0,22R at the output than without (I put two 0,47R 4W high qulity resistors from Intertechnik in parallel)
Bartek |
|
|
| tbla |
| one thing that really matters is the cap quality - use silmic 1000uF/50V.....;) |
|
|
| 2Bak |
OT: Which type of speaker cables goes best with a gainclone without a zobel?
I had oscilations in my gainclone with those heavy TNT FFRC cat 5 cables, then I switched to a very simple cable - two twisted wires from some phone cable - and no more oscilation, but I guess I could do better with some other cable types...
/Jan |
|
|
| tbla |
| quote: | | TNT FFRC cat 5 cables |
i tried this cable a year ago.....just "no good" - i reinstalled my cardas golden cross.....no competition.:yuck: |
|
|
| Nuuk |
| quote: | | I had oscilations in my gainclone with those heavy TNT FFRC cat 5 cables, then I switched to a very simple cable - two twisted wires from some phone cable - |
I'm using two twisted pairs of CAT5 solid core for each speaker and am surprised by the quality of sound. I did not have oscillations with any of the other cables I tried with the IGC (standard or valve-buffered). |
|
|
| Peter Daniel |
OK Guys, the moment of truth.;)
I had my GC at the store today, for some listening comparisons.
We tried non inverted amp (with Holco resistors no BG coupling caps), inverted amp (with Caddock and Rikens and BG coupling caps) and the S&B TVC preamp.
I had my GC venture partner listening to the amps as well. He definitely preffered the non inverted amp. The non inverted amp had more depth to the soundstage, sounded smoother and overall better. We tried adding transformers to this amp and there was further improvement in the soundstage but not as big as with inverting amp. The inverted amp sounded, OTOH, much better with the transformers, but according to my partner, not as good as non inverted amp on its own (I'm not sure about it yet).
So we made a full circle; after playing with inverted topology for a whole year, we coming back to basics: the non inverted topology.;)
I still cannot definitely tell one way or the other, but there is a strong possibility that non inverting amps might actually sound better. |
|
|
| Christer |
Do note, however, Peter that if the choice of passive components
can make such a big difference as many of you claim, maybe
the difference should primarily be attributed to these
differences rather than inverted vs. non-inverted. I am not
claiming that to be the case, but you have changed several
parameters at once. BTW, did you balance the impedances
looking out from the two inputs in the non-inverting config?
Edit: Since I eventually posted something in this thread, I
might as well add a remark wrt. to the thread subject I
have been tempted to say all from the start. I am happy
the hear that it is non-inverted fun. Inverted fun does not
sound very fun. :) |
|
|
| Peter Daniel |
| It makes difference too, and today I will build another amp with same brand of components. Also note that value of components is different as inverting amp is using 10k/250k resistors, while non inverting 680/22k. |
|
|
| Kuei Yang Wang |
Peter,
| quote: | Originally posted by Peter Daniel
I still cannot definitely tell one way or the other, but there is a strong possibility that non inverting amps might actually sound better. |
The first thing you should do is to loose the BG input coupling Cap's on the Inv GC, to make this a level playing field.... (BG's in the position where you put them are IMHO pretty dire a choice - but I feel that way about them pretty much everywhere).
Also, I hope that you have level matched the comparison, your NI GC will have more gain (34) than the IGC (22/25), not enough to be drastically noticable, but enough to force the ear to "preference" almost invariably. Also, the lowered NFB due to more gain plays in there....
You might also compensate by increasing the IGC NFB Resistor to 330K.
Sayonara |
|
|
| Peter Daniel |
I'm always "loosing" BG coupling caps in my personal use amps. One thing came obvious today. Non-inverted amp seems to have better depth of the soundstage.
I've already tried before 300k feedback resistor and the prefference was on 250k. |
|
|
| Christer |
Peter, consider matching the impedances on the two inputs
unless you have already done that,
since Jung seems to claim this will, in the ideal case I presume,
cancel the common mode distorsion, ususally plaguing the
non-inverting configuration.
Kuei, sorry for my persisting in this, as I asked you in another
thread recently, and didn't get any answer. What's your take
on this? |
|
|
| Kuei Yang Wang |
Konnichiwa,
| quote: | Originally posted by Christer
Kuei, sorry for my persisting in this, as I asked you in another
thread recently, and didn't get any answer. What's your take
on this? |
Matching the Impedances on both Inputs does nothing about the common mode induced (Early Effect IIRC, need to get that self book from the Attic) problems. It addresses (as I pointed out) a different issue.
Sayonara |
|
|
| Peter Daniel |
I don't want to match the impedances. With non inverted amp 10K is forced minimum, with non inverted we can go higher, so why would I like to make it less? I don't really care which configuration is superior. But if non inverted offers more flexibility (for instance fixed/shunt volume control) and doesn't sound worse, then why stay away from it?
I will make more accurate testing soon.
Kuei,
As to the PS electrolytics, I was searching here and there, but nobody could really recommend me a better electrolytic than BG. But the truth is, that I also didn't try that many caps in that application.
As to the coupling caps, I couldn't find smaller, better performing cap than BG N type. And I can't really afford to go with a larger size both moneywise and spacewise;) But if I stay with non inverted amps, I want be using any coupling caps. |
|
|
| Kuei Yang Wang |
Konnichiwa,
| quote: | Originally posted by Peter Daniel
I'm always "loosing" BG coupling caps in my personal use amps. One thing came obvious today. Non-inverted amp seems to have better depth of the soundstage.
I've already tried before 300k feedback resistor and the prefference was on 250k. |
So, we are then talking about a Amp in both cases 2 Resistors, no coupling cap and only the gain difference remains then.... Yes?
As for soundstage depth, I do not pay that much attention to "sound stage" as much of it is an artificial effect contained (usually) neither in the recording nor in the original event. It is in effect a form of distortion in many cases.
From a good seat in the concert hall I do get a good dpeth and with perception and a good idea where things are, but the Soundstage displayed by many audiophile systems is not like that at all, it tends to faked depth, rather than true separation of instruments.
So maybe what you call "better" I would on audition call "worse"... ;-)
Sayonara |
|
|
| Christer |
| quote: | Originally posted by Kuei Yang Wang
Konnichiwa,
Matching the Impedances on both Inputs does nothing about the common mode induced (Early Effect IIRC, need to get that self book from the Attic) problems. It addresses (as I pointed out) a different issue.
Sayonara |
I may have misunderstood Jung, but my impression is that
that is precisely what he claims, although he discusses the
distorsion due to the input capacitances, but maybe that
does not affect the common-mode induces distorsion.
Hm, I guess I have to think more about this. |
|
|
| Peter Daniel |
| quote: | Originally posted by Kuei Yang Wang
So, we are then talking about a Amp in both cases 2 Resistors, no coupling cap and only the gain difference remains then.... Yes?
As for soundstage depth, I do not pay that much attention to "sound stage" as much of it is an artificial effect contained (usually) neither in the recording nor in the original event. It is in effect a form of distortion in many cases.
From a good seat in the concert hall I do get a good dpeth and with perception and a good idea where things are, but the Soundstage displayed by many audiophile systems is not like that at all, it tends to faked depth, rather than true separation of instruments.
So maybe what you call "better" I would on audition call "worse"... ;-)
Sayonara |
Basically the 2 resistors in ea. amp, everything else stays the same. Personally, I'm also not that crazy about the soundstage (and maybe I don't really perceive it that well;)), but as I mentioned before, that was the main complaint from the people who tried the amps. They wanted a holographic space between speakers. Now, if you don't like it, or even think that it's "worse", you shouldn't also like TVCs, as they make the soundstage pretty enormous, especially with inverted GC ;-)
At the moment, I'm still confused about the actual differences, and more time is needed to come to any solid conclusions. |
|
|
| millwood |
| quote: | Originally posted by Christer
I may have misunderstood Jung, but my impression is that
that is precisely what he claims, |
given the number of laws of physics this forum has rewritten in the last 6 months, it wouldn't surprise me if folks come out with theories why Jung is wrong on that, :)
Seriously, I think Peter's experiment is a good one that proves, once again, the differences between inverting and non-inverting topologies are hard, if not impossible, to tell by our ears, in spite of repeated claims by the "theorists" that one is far superior to the other.
It does seem that most amp designers know what they are doing, contrary to what you would have been led to believe by those theorists (arm-chair amp designers?). |
|
|
| Peter Daniel |
Here's what I find the biggest advantage of non inverted topology. One can use modufied volume control, where only a single, high quality resistor is in series with a signal path and (if you want to simplify matters) a pot shunting to the ground. In this case I used 10k S102 Vishay and 50K log Nobel pot. The control range is pretty good and I'm loosing only 1.6dB of gain at full volume. I believe that this type of arrangement is even better than a law faking cermet linear pot, and as far as I am aware, this arrangement was not possible with inverted amp (as the series 10k resistor sets the gain).
BTW, I was pushing the amp pretty hard, without the complete chassis, and although rear panel (2" X 8.5") got hot, it was still fine. I also found that such "open" setup sounds better than a completely assembled chassis. It also seems like the amp doesn't really need much in terms of heatsinking. |
|
|
| Peter Daniel |
So I made another non inverting GC, this time with a Caddock resistor in a feedback. My first impression was that it sounds a bit more brighter, harsher and less smooth than Holco, but also a bit more refined and definitely more involving. After listening for a while, I decided to go even further and modified volume control (by doing what was described in a previous post). Now, this brought things back on track and "cleaned the top end", as Vishays always seem to soften high registers.
Also, this improved the resolution and tonal balance of the amp. I still didn't do any comparisons, as I want the amp to break in, but so far it sounds very good. It is much more involving that the one I built yesterday (with Holcos and conventional use of the pot).
Other things I noticed so far is that the bass notes are better defined and overall it seems like the amp has more control. It sounds cleaner, with more coherence. There is more separation between the notes with what seems like a better pace and rhytm, as well a better soundstage. I just spent an hour listening to familiar tracks and I found myself totally involved with music, which is always a good sign. I listened to both classical material and the industrial classics and both sounded very satisfying. On some previous occasions, I noticed that depending on the tuning, the amp may sound good with one kind of music, but somehow may seem out of place with radically different tunes. It wasn't the case this time.
In here, I used the words 'more' and 'better' quite frequently, but you must understand that after one hour of listening nobody can be shure of that, so approach it as just my initial impressions only.;) |
|
|
| Peter Daniel |
| This is my current setup: |
|
|
| Tor M |
Hi!
This is interresting!
Fedde seems like you started a revolution;)
Just when I have convinced all my friends that inverting is the way to go (without trying non-inverting:cannotbe: ) this may not be true anymore! But hey, my friends don`t know what the hell I`m talking about anyway, they`re just nodding and thinking "Freek!";)
Peter: In your shunt volume control do you connect the "lower" pin of the pot to ground as well? Or just leave it unconnected like in the schematic?
If so It wouldn`t make any difference if you use a log pot or a linaer pot?
Fedde: What did you do with the current sense pin of the opa541?
Keep on the good work!
Rgds
Tor Martin:) |
|
|
| platenspeler |
This is the schema I used for Geenkloon a year ago. If a little more gain is desired then use a 680 Ohms for R1 and R5 instead of a 1k on the negative input. And I still like the 12-step attenuator (R4 in the schema, for lack of my drawing skills), which in material cost is lower than a quality pot.
DC on the output is just 1 mV whereas it's 22 or so for the inverted clone (Cyclone).
Funny that suddenly everybody is building non-inverted clones, whereas the schema's are in the datasheets of both the LM1875 and 3875.
Maarten
More on http://gainclone.platenspeler.com, also how to build the 12-step |
|
|
| moamps |
| quote: | Originally posted by platenspeler
[B]
Funny that suddenly everybody is building non-inverted clones, whereas the schema's are in the datasheets of both the LM1875 and 3875.
|
Hi,
Looks like "back to the roots" .
Your excel att. calc is great.
Regards |
|
|
| platenspeler |
Thanks,
I'll publish an update one of these days. I have a break-before-make available (for cheap switches) and a rather complicated for inverted designs which I want to simplify first.
Maarten |
|
|
| Peter Daniel |
| quote: | Originally posted by Tor M
Peter: In your shunt volume control do you connect the "lower" pin of the pot to ground as well? Or just leave it unconnected like in the schematic?
If so It wouldn`t make any difference if you use a log pot or a linaer pot?
|
I leave the lower pin unconnected.
You can also use lin pot, but I'm not sure how well it will work in this arrangement. Here's the link to a simple calculator and you can compare different resistor values and see what works the best
http://homepages.tcp.co.uk/~nroberts/shunt.html |
|
|
| Tor M |
Hi Peter!
Maybe I have misunderstod something, but I think I have read somewhere, probably on Rod Elliots pages, that most log pots are made up from two linaer sections with different resistance gradient. Or that some log pots are linaer pots witch is law faked inside with an internal resistor.
Then I thought that if you only use the upper section it wouldn`t matter if you use a log or linaer pot because this section is linaer in both types of pots.
But as I said, I might have misunderstood the whole thing.;)
Keep on the good work!
Tor Martin |
|
|
| Peter Daniel |
Hi, This are values of a shunt resistor when series resistor is 10K. As you see this is not a linear function and from what I know the log curve is more appropriate here. There are 24 steps listed and step 12 is using 368 shunt resistor. For linear pot it would be half of maximum value, or approx 25000ohms. As you see it would be way off the (linear) line;)
Step 1, Attenuation = Infinity dB, Rx = 10000 ohms, Ry = 0 ohms.
Step 2, Attenuation = 75 dB, Rx = 10000 ohms, Ry = 2 ohms.
Step 3, Attenuation = 69 dB, Rx = 10000 ohms, Ry = 4 ohms.
Step 4, Attenuation = 63 dB, Rx = 10000 ohms, Ry = 7 ohms.
Step 5, Attenuation = 58 dB, Rx = 10000 ohms, Ry = 13 ohms.
Step 6, Attenuation = 53 dB, Rx = 10000 ohms, Ry = 22 ohms.
Step 7, Attenuation = 48 dB, Rx = 10000 ohms, Ry = 40 ohms.
Step 8, Attenuation = 43 dB, Rx = 10000 ohms, Ry = 71 ohms.
Step 9, Attenuation = 39 dB, Rx = 10000 ohms, Ry = 113 ohms.
Step 10, Attenuation = 35 dB, Rx = 10000 ohms, Ry = 181 ohms.
Step 11, Attenuation = 32 dB, Rx = 10000 ohms, Ry = 258 ohms.
Step 12, Attenuation = 29 dB, Rx = 10000 ohms, Ry = 368 ohms.
Step 13, Attenuation = 26 dB, Rx = 10000 ohms, Ry = 528 ohms.
Step 14, Attenuation = 23 dB, Rx = 10000 ohms, Ry = 762 ohms.
Step 15, Attenuation = 20 dB, Rx = 10000 ohms, Ry = 1111 ohms.
Step 16, Attenuation = 17 dB, Rx = 10000 ohms, Ry = 1645 ohms.
Step 17, Attenuation = 14 dB, Rx = 10000 ohms, Ry = 2493 ohms.
Step 18, Attenuation = 12 dB, Rx = 10000 ohms, Ry = 3354 ohms.
Step 19, Attenuation = 10 dB, Rx = 10000 ohms, Ry = 4625 ohms.
Step 20, Attenuation = 8 dB, Rx = 10000 ohms, Ry = 6614 ohms.
Step 21, Attenuation = 6 dB, Rx = 10000 ohms, Ry = 10048 ohms.
Step 22, Attenuation = 4 dB, Rx = 10000 ohms, Ry = 17097 ohms.
Step 23, Attenuation = 2 dB, Rx = 10000 ohms, Ry = 38621 ohms.
Step 24, Attenuation = 0 dB, Rx = 10000 ohms, Ry = Infinity ohms. |
|
|
| Tor M |
Yup.
Peter:
You are using the top section of a pot for the shunt and not a stepped attenuator made after theese calculations?
I know there is a difference between log and linaer, but is it in the top section of the log pot you are using compared with the top section of a similar linaer pot?
If the viper and the "bottom" pin were joined together there definenly would be a difference.
I`m asking because I don`t know, nothing to argue about. ;)
Tor Martin |
|
|
| Peter Daniel |
The wiper devides the pot into two sections. The direction of the turn defines which section you can use for shunt operation. If you use the wrong section, you would be turning the knob clockwise and volume will decrease. This is not what you would like to do, right?
You can connect the wiper pin with the other unused pin together, but this influences nothing. It is sometimes done in this way, so when the wiper looses contact, there will still be max resistance between input and ground so you won't blow anything. But with quality pots this should not be a problem. |
|
|
| lohk |
You guys really astonish me all the time.
After posting thousands posts dealing with the inverted topology you now are starting to listen carefully to the "old" and "classic" non-inverted chip amps. Many things were discovered about parts quality and construction influences, all very valuable stuff. But everybody at the end thought that the inverted amp was the only version which was worth to discuss. Now the change!
I stopped posting in the chipamp forums after getting only bad comments on my experiences with both versions, like "now you try the inverted version, that is much better".
I made several versions of chip amps in the last two years, still my personal reference version is "a version of a copy" - so to say - of the "orginal gaincard", as far as the images were correct. That meant:
Non-inverted LM3875
- 2x1000yF HQ electrolytics from Frako
- 680/22k FB with (subminiature) carbon resistors (important)
- 10yF (or 47yF) nichicon MUSE bipolar electrolytic (no film cap!)
- no FB capacity (with the drawback of some more DC offset, but the sound was classes better that with the cap in place)
- 1R/220nF Zobel (HQ-MKT, 1W carbon resistor)
- 20k carbon input pots; 22k + input to ground
2x24V 250VA torodial with two full bridges for both channels seperated
And all done on slightly more than a square centimeter in p2p wiring (only one star ground) between the chip and the both psu caps.
This little amp has slam and finesse, is highly listenable and completely noise and hum free. Although I am not sure if all details are the best possible (like the psu/bridge version) I still keep this version, listen to music with it and compare all other versions to it. It sounds GREAT!
All my inverted versions sounded good (I made three yet), but were terribly dependent on the source, sounded different on different levels, some sounded kind of dependent on the musical material too. More below.
Klaus / who still loves those tiny amps ;)
I am to lazy to search for my old posts so here are excerpts from the draft files (from 2002 and 2003):
...
My new inverted chip-amp sounded very, very good from the start, and I must admit that I really was a bit sceptic before and I did not expect much compared to the "normal" non-inverted circuit. I am still testing, and I am a bit unsure to write down clearly defined expressions about the sound yet, because it sounds so much different, in a way, to more classical high-quality hifi amplifiers, like my Naims.
Polite? Not in the way I would call something polite, more neutral, a bit like a perfect butler with no own opinion.
Not tiring ? Yes, definitely. I heard so many of this very clean sounding amps, which tired me to death (like Krells, ML, etc. - sorry folks).
Bass shy ? I thought that at first glance but after putting on records with lots of real bass I could not confirm it. Bass is maybe a bit more right ? My inverted chip-amp can play really loud (24V secondaries), and, although running hot then, shows no signs of giving up, and even the bass is very stable. This fact amazes me most, because the PSU is a slap in the face of a normal textbook one.
...
The musical flow is there and well preserved but the presentation of details is quite different. But this is still just a test and not a defintive setup. I will try different input caps, an impedance smoothing resistor (.22 Ohm, like suggested), and other mods. But this amp is much "faster" than many others, the sound starts and stops in an breathtaking both secure and lively way, it is truly amazing.
I can confirm your observations about the sensitivness of the inverted version to mains noise. Although my amp is as quiet as its brothers (or sisters) and completetly noise and hum-free, it picks up mains noise quite heavily like the switching noise of the *******ing fridge from the kitchen nearby.
I read in another thread here on this forum about a very possible reason: In the inverted version the input (output of the source and cable) is a part of the feedback loop and therefore very sensitiv to picking up all kind of noise. In every mixing desk the most sensitive point is the input of the mixing amp connected to the mixing rails (usually a virtual ground inverted amplifier). Here normally every possible care is taken to sreen everything properly. Very expensive mixing desks therfore even have symmetric (balanced) mixing rails to get rid of the least bit of RFI noise picked up at this parts of the circuit.
...
...
Yesterday I switched off my new inverted chipamp. Finally it failed to impress me.
It definitely is bass shy, bass is there and deep, but somehow too "thin" or "dry". And voices are subdued, not like a filter, more like being put to the backside. All details are there, but the overall presentation is no satisfying at least. The musical flow is there, but somehow lacklustre, hard to describe. Dynamics are a kind of exaggerated, but strangely this makes it sound less dramatic. This resembles to some tube amps, but otherwise it does not sound like tube amps at all. It tries to expand those heavy compressed music (most pop or rock music), but this does not make sense to me.
Don't get me wrong: It is a really good amplifier. But to justify all the fuss and the hype I has to be better. At the moment my other non-inverting version sound better to me.
These chip amps - to contradict somebody here - seem to be VERY much dependent on parts quality and selection, the less parts the more. How can it be that carbon resistors actually do sound better here ? And do not tell me that they are better in general.
The inverted version is definitely much more source sensitive and, although Thorsten's idea with the linear pot is good, a buffer will probably be mandatory.
...
;) |
|
|
| millwood |
| quote: | Originally posted by lohk
I stopped posting in the chipamp forums after getting only bad comments on my experiences with both versions, like "now you try the inverted version, that is much better".
;) |
that's why I take what I read here with a grain of salt.
It is very unfortunately that sometimes we sound more like snake oil salesmen, arguing for that 0.000000000001% improvement in THD, with no regard to reality. The topology discussion is one such perfect example. |
|
|
| Tor M |
| quote: | Originally posted by Peter Daniel
The wiper devides the pot into two sections. The direction of the turn defines which section you can use for shunt operation. If you use the wrong section, you would be turning the knob clockwise and volume will decrease. This is not what you would like to do, right?
You can connect the wiper pin with the other unused pin together, but this influences nothing. It is sometimes done in this way, so when the wiper looses contact, there will still be max resistance between input and ground so you won't blow anything. But with quality pots this should not be a problem. |
Yep, you are right, my mistake!
My head wasn`t thinking stright. I guess working to many nighshifts and to little sleep this week made it misbehave a little.:rolleyes:
Tor Martin |
|
|
| ronc |
Very true millwood.Its been my experience that the general public cant tell the difference in the last 10% anyway.
I have had many ppl come over and listen to my system and have had varing results.
When we talk about the tiny diff between NIGC and IGC even TBIGC and it takes a trained ear to dectect the difference then i find very little desire to try several diff config.
Now the going thing looks to be NIGC, no buffer just pots on the front.
Sounds like the orig gaincard to me.Hay maybe they knew what they (47 labs) were doing.
ron |
|
|
| lohk |
Well that is - ahem - what I could have said.
Now please let us start to find answers to some of the still unsolved questions like the best topology of the psu (rectification and transformer).
Klaus
ps: A wise old white haired Japanese Master was sitting in the background all the time watching and wondering, silently laughing |
|
|
| tbla |
| please all you inverted cloners - try the "figur 1 model" in the datasheet....1K, 1K, 20K - noninverted......! this is IMHO the way to go.......:) |
|
|
| millwood |
| quote: | Originally posted by tbla
try the "figur 1 model" in the datasheet....1K, 1K, 20K - noninverted......! this is IMHO the way to go.......:) |
that is how I run my clones (two of them so far). My half deaf ears couldn't tell them apart from the "far superior" inverted gainclones. |
|
|
| tbla |
| quote: | | how I run my clones |
hehe, after a few liters of a very nice pinot noir (argentine) you're my hero...... |
|
|
| Kuei Yang Wang |
Konnichiwa,
| quote: | Originally posted by millwood
given the number of laws of physics this forum has rewritten in the last 6 months, it wouldn't surprise me if folks come out with theories why Jung is wrong on that, :)
|
Well, sometimes it helps to actually know waht is being talked about when contributing to a thread.
The passage in question which has Christer so badly misinterpreting Mr. Jung and where he insists on clinging on to his misinterpretation in the face of whatever is being presented reads as follows (I have copies of the EDN Articles):
"With an equal source impedance condition, the two sets of distortion components generated by the nonlinear capacitances match, or V1 = V2. Since this distortion is CM to the op amp (not differential), it is rejected. "
What Mr. Jung claims is exactly the same as I have written, namely that making the effective input impedances the same converts the error from the input capacitance (and incidentally base current modulation) into a COMMON MODE signal, which is then rejected by the comparably poor CMRR of the Op-Amp, which is a shame of course, as in inverting mode circuit eliminates any CM signals and thus any issues related to the CMRR and it's rapid decrease towards high frequencies.
Now subjective sound is a very different and interesting issue, but IF we take the "wire with gain" view the Inverting Mode circuit performs better and if we compare the circuit against a " wire bypass" it is less audible if you daisychain two inverting sections (to preserve polarity) than one non-inverting section.
Of course, the NI Op-Amp circuit may make the signal going through it sound to some people BETTER than the wire, but that is personal preference. My tastes in listening to music and in engineering tend to be fairly (not excessively) literally minded.
Sayonara |
|
|
| tbla |
| quote: | | Now subjective sound is a very different and interesting issue, but IF we take the "wire with gain" view the Inverting Mode circuit performs better and if we compare the circuit against a " wire bypass" it is less audible if you daisychain two inverting sections (to preserve polarity) than one non-inverting section. |
i thought that "some" people just had a listening session - and found that differencies between inv and non inv were very small.....and noninv were the winner....??????
:judge: |
|
|
| ronc |
Lets face it, we are talking subjective opinion here.
I am building a NIGC at the moment and using a TDA1524a pre chip (on both) and will do an A/B test.Both will have Batt P/S and i have hooked up a series of switches that allows me to switch back and forth from my listening postion, using the remote on the CD player i can even back up and listen to the same passage on different amps within seconds.
This is the same method i have been using for over 1 year and its easy (sometimes) to tell any differences.I adjust the volume (on both)on the A/B test by a signal generator set at 1Khz and an SPL meter, at my listening position, to equal.
This has been an easy way to test out different amp/component/config. and i believe its a fair test.When i get the actual A/B (subjective) test done i invite at least 5 different ppl over at different times and do a blind test (free beer helps).
Some of the tests show an obvious winner, some dont.
ron |
|
|
| ronc |
Found some of my old notes on A/B tests and decided to share.The ppl doing the grading were 2 females ages 24 and 32 the males range in age from 18 to 46.Only 1 ,the oldest male, had any exerience with higher end systems.
#1.(first test with the A/B setup) modified marantz 2238/homebrew 6bq5 PP amp.Surprising that both females perferred the tube sound and the males the SS sound.
#2.(taken at the same time as #2a with a little time to re-wire) Marantz2238,6BQ5 amps/IGC.IGC hands down the winner.
#2a.Marantz 2238 (pre only) IGC/IGC with 50k alps pots on the front.All liked the pre better but the older male stated" the amp with the pots(no pre) sounded slightly cleaner but less body"
#3.IGC with metal film feedback resistors/IGC with carbon feedback.4 out of 5 perferred the carbon as they all stated it sounds cleaner and more natural.The only diff was the youngest male who appeared to like the brighter sound of the metal film.
#4.TDA1524a pre chip IGC/IGC 50K alps pots on the front.All perferred the chip pre.
Test #1,2,2a were taken with RS-1354a TQWT with silk dome tweets rolled in @ 10Khz.
Test 3,4 with Fostex fe-103a rear loaded front firing horns tuned to 57 Hz..
All perferred the fostex horns over the TQWT even though the horns didnt go as low in freqency as the pipes.
ron |
|
|
| GregGC |
You are “the man” Ron. Looking forward to the tests to follow.
Greg |
|
|
| Peter Daniel |
Millwood,
You seem to be overjoyed by new NIGC trend. So far nobody stated that both config. sound the same (except for you) and nobody says that NI config is better. It just happens that some prefer it one way and others, the other way. I still didn't make my mind, for instance. |
|
|
| ronc |
As stated , a lot of this is subjective opinion as PD stated.Without objective measurements (damn hard to do and prove anything,look at SETs)the only thing we can do is put it out to the public and note responses.
One of the odd things i have noted is the ladies IMHO have a better understanding of the actual music and its finer points than the guys do.The ladies appear to have a better HF hearing and they will note a harsh sounding system sooner than the guys,which is why i went with the TDA1524a pre as it kinda sorta tames the louder(greater volume) slightly harsh sound of the IGC.
ron |
|
|
| millwood |
| quote: | Originally posted by Peter Daniel
So far nobody stated that both config. sound the same (except for you) and nobody says that NI config is better. It |
I suppose that most of the (real) amp designers voting with their designs (and money) doesn't count. |
|
|
| Peter Daniel |
| quote: | Originally posted by millwood
I suppose that most of the (real) amp designers voting with their designs (and money) doesn't count. |
How do you know if real amps designers tried subjectively both configurations and chose the better sounding one? Maybe they share similar POV as yours and don't give a second thought about it. |
|
|
| Pedja |
| quote: | Originally posted by Peter Daniel
How do you know if real amps designers tried subjectively both configurations and chose the better sounding one? |
The final sonic result of the bulk of the equipment around is an objective proof that they in fact mostly neither do that nor they do many other things that can result in good sounding equipment. In the majority of the cases they are not even paid to choose better sounding options.
Of course, this doesn’t apply to Mr. Kimura which tried more than a hundred of configurations (inverted, non-inverted… and all the others :clown: ) before he settled exactly on that one he uses in his amp (and which is only by accident the same one suggested by the manufacturer of the chip).
Pedja :wave2: |
|
|
| Pedja |
| quote: | Originally posted by Pedja
...only by accident... [/i] |
It just occurred to me: it might me the cosmic law!
Pedja :cloud9: |
|
|
| millwood |
| quote: | Originally posted by Peter Daniel
How do you know if real amps designers tried subjectively both configurations and chose the better sounding one? |
I don't know. I guess there might be some of real amp designers who have survived cut-throat competion by designing and selling horrible sounding amps so they hope one day they might get lucky and lose all of their investment and then some. |
|
|
| GregGC |
It's really simple. Pick the one YOU BELIEVE is BETTER and LIKE the best and build it. Obviously it's a very close call. Splitting hair. If it takes a few days of auditioning to decide on a version, most likely the differences are more a matter of nuance rather than "better" or "worst". And for people that build both versions, you guys are for real. I think you are the ones that can express any opinion on the subject. And it's again dependent on specifics of the design (components, power supply, layout and so much more. I like Peter's approach. Build it, listen to it and tweak it until you like it and you can sleep well. I'm sure there are poorly and very well build/performing IGCs and NIGCs. Lets just give it a rest to which one is better.
I for example want to build the one with the least amount of components and I LIKE the NIGC topology. If build properly it'll perform very well. If not it'll suck big time. I shouldn't blame the topology but my skills and knowledge. Otherwise the real Gaincard wouldn't be selling for that much money if it doesn't sound good.
So, let's build those great amps and share the experience no matter IGC or NIGC or whatever GC. They are all verry good.
:grouphug:
Thanks for your attention. |
|
|
| Colin |
Reference Ron's last post. I think he's hit on the way to test - leave the amp playing and get your Significant Other Not Into HiFi to comment, preferably without being put under any pressure to notice anything particular. Mine is irritatingly good at this and is capable of spotting cable changes whilst I'm still trying to hear differences.
Our perception is very good at picking up unattended inputs - which is why it's often easier to become aware of differences when you're not concentrating on hearing them.
Okay, I'll keep quiet now - I haven't even built a GC but the discussion is fascinating.
Colin |
|
|
| millwood |
| quote: | Originally posted by GregGC
If build properly it'll perform very well. |
Those chip amps can perform well, regardless of how you built them. I have toyed around with both components, value and topologies for a while, and they just seem to work no matter how "primitive" they are. for example, I had one version with just two (carbon) resistors and it had no trouble working (just as well as other more elaborate ones).
Somehow, there seem to be a concerted effort to complicate such a simple design / project. |
|
|
| ted |
| Then there were the manufactures that make one channel inverting and the other non-inverting. (Carver M-500t) I guess that was another school of though. :) |
|
|
| GregGC |
| What a good approach. This way both sides will be happy (or maybe both sides will be disappointed). |
|
|
| Terry Demol |
| quote: | Originally posted by Kuei Yang Wang
Konnichiwa,
What Mr. Jung claims is exactly the same as I have written, namely that making the effective input impedances the same converts the error from the input capacitance (and incidentally base current modulation) into a COMMON MODE signal, which is then rejected by the comparably poor CMRR of the Op-Amp, which is a shame of course, as in inverting mode circuit eliminates any CM signals and thus any issues related to the CMRR and it's rapid decrease towards high frequencies.
snip
Sayonara |
Kuei,
Just a quick note for some here who are unaware, when looking
at impedances into each IP, the OP is considered a ground.
So FB resistor and resistor to gnd are effectively in parallel.
Example:
Simple OPA non-inverting with gain of 2.
1k FB R and 1k to gnd. The correct balancing R to go inseries with
+IP is 500R.
Cheers,
Terry |
|
|
| Terry Demol |
| quote: | Originally posted by lohk
You guys really astonish me all the time.
After posting thousands posts dealing with the inverted topology you now are starting to listen carefully to the "old" and "classic" non-inverted chip amps. Many things were discovered about parts quality and construction influences, all very valuable stuff. But everybody at the end thought that the inverted amp was the only version which was worth to discuss. Now the change!
I stopped posting in the chipamp forums after getting only bad comments on my experiences with both versions, like "now you try the inverted version, that is much better".
I made several versions of chip amps in the last two years, still my personal reference version is "a version of a copy" - so to say - of the "orginal gaincard", as far as the images were correct. That meant:
Non-inverted LM3875
- 2x1000yF HQ electrolytics from Frako
- 680/22k FB with (subminiature) carbon resistors (important)
- 10yF (or 47yF) nichicon MUSE bipolar electrolytic (no film cap!)
- no FB capacity (with the drawback of some more DC offset, but the sound was classes better that with the cap in place)
- 1R/220nF Zobel (HQ-MKT, 1W carbon resistor)
- 20k carbon input pots; 22k + input to ground
2x24V 250VA torodial with two full bridges for both channels seperated
And all done on slightly more than a square centimeter in p2p wiring (only one star ground) between the chip and the both psu caps.
Long post snipped
;) |
Hi Lohk,
Good post.
Yes, certain people seem to have more credibility here than
others but that's the way it goes.
To treat forums like this as a main resource is probably a bit
dangerous, however as a supplement amongst many resources,
it's excellent.
Cheers,
Terry |
|
|
| zygibajt |
| supplement of own expirences! |
|
|
| Christer |
| quote: | Originally posted by Kuei Yang Wang
Well, sometimes it helps to actually know waht is being talked about when contributing to a thread.
|
Of course, if one "contributes" in the way of asking a question,
one usually does so because there is something one doesn't
know and wants to learn more about.
| quote: |
The passage in question which has Christer so badly misinterpreting Mr. Jung and where he insists on clinging on to his misinterpretation in the face of whatever is being presented reads as follows (I have copies of the EDN Articles):
|
My insisting may, of course, have been caused by there being
no clear convincing arguments or any pointing out what I had
misunderstood before you posted the text quoted below.
| quote: |
"With an equal source impedance condition, the two sets of distortion components generated by the nonlinear capacitances match, or V1 = V2. Since this distortion is CM to the op amp (not differential), it is rejected. "
|
This explains my misunderstanding. Call it sloppy reading if
you wish, but I probably interpreted the word "rejected" as
some kind of cancellation, rather than common-mode rejection.
Thanks for the explanantion. |
|
|
| Peter Daniel |
| quote: | Originally posted by Peter Daniel
This is my current setup: |
As good as this setup worked
http://www.diyaudio.com/forums/atta...=&postid=257590
it appears that it had substantial error. It seems like feedback resistor (22K) defines the input imp of a circuit and anything above 22k increases the DC offset. For instance, without the input to ground resistor (just 50k pot at input) the offset went so high as 200mV - 400mV (depending on a chip). For some reason I didn't notice it in the beginning;)
So to get low offset (still around 60mV at max vol) I had to place 20k resistor to ground at the input. This prevents me from using series fixed resistor, as 10k (I had before) lowers the gain by 4.5dB in current setup. There is no free lunch, it seems;) |
|
|
| tbla |
peter - please, .......try this - noninverted 1k, 1k and 20k......and let us know what you think.
best regards, troels |
|
|
| Peter Daniel |
| I need more gain than that. I'm not using preamp and passive stage on TDA1543 isn't very high in output. |
|
|
| tbla |
| ok - then ajust for more gain, but keep those 2 other resistors at the same value......:) |
|
|
| GregGC |
| quote: | Originally posted by Peter Daniel
As good as this setup worked
http://www.diyaudio.com/forums/atta...=&postid=257590
it appears that it had substantial error. It seems like feedback resistor (22K) defines the input imp of a circuit and anything above 22k increases the DC offset. For instance, without the input to ground resistor (just 50k pot at input) the offset went so high as 200mV - 400mV (depending on a chip). For some reason I didn't notice it in the beginning;)
So to get low offset (still around 60mV at max vol) I had to place 20k resistor to ground at the input. This prevents me from using series fixed resistor, as 10k (I had before) lowers the gain by 4.5dB in current setup. There is no free lunch, it seems;) |
Peter,
You have to use input cap if you don’t want to get those offset changes dependant on the POT position, but it contradicts with the simplicity theory to which I subscribe too. That why I came up to this version:
http://www.diyaudio.com/forums/show...&threadid=22287
Your comment would be appreciated. |
|
|
| Peter Daniel |
So eventually you are using a coupling cap. For now I decided to not use a coupling cap. I have 50k pot at the input and 22k resistor from + input to ground and no series resistor. The offset never goes above 60mV in this way and when pot is turned down it's even less. It works for me this way, as I don't want to use a coupling cap.
Because I don't have a series input resistor at max volume (no series resistance from pot) the offset jumps a bit (around 200mV) but as long as I don't use max pos on a pot it works well again. |
|
|
| GregGC |
| quote: | Originally posted by Peter Daniel
So eventually you are using a coupling cap. For now I decided to not use a coupling cap. I have 50k pot at the input and 22k resistor from + input to ground and no series resistor. The offset never goes above 60mV in this way and when pot is turned down it's even less. It works for me this way, as I don't want to use a coupling cap.
Because I don't have a series input resistor at max volume (no series resistance from pot) the offset jumps a bit (around 200mV) but as long as I don't use max pos on a pot it works well again. |
I think the ser. Res in your case has nothing to do with the offset (unless your sorce has a low DC resistance to GND, less than tens of kOhms). It just present an additional divider at the input and protects the input of the GC. Actually you can use a resistor for protection if you wish. As long as it is smaller than 2k will not eat away too much of your input signal. The reason the offset of your amp went low is because the extra 22k path to GND. Before you had 50k to GND now 22k||50K. This way is even better as the POT may not be perfect and not connect all the time you wont get any DC fluctuations on the input/output (good insurance). I was going to suggest 22k pot and no 22k resistor to GND, if that works for you. But connecting the slider of the pot to the other end will give you that DC "insurance" thing.
Peter, I admire your work! Great creations come out of your hands, man! Thanks for the inspiration. Maybe it's a bit too much. Having yourGC's in mind I cant decide on the housing of my GC. I feel unworthy.
:bawling: |
|
|
| GregGC |
| quote: | Originally posted by GregGC
I think the ser. Res in your case has nothing to do with the offset (unless your sorce has a low DC resistance to GND, less than tens of kOhms). It just present an additional divider at the input and protects the input of the GC. Actually you can use a resistor for protection if you wish. As long as it is smaller than 2k will not eat away too much of your input signal. The reason the offset of your amp went low is because the extra 22k path to GND. Before you had 50k to GND now 22k||50K. This way is even better as the POT may not be perfect and not connect all the time you wont get any DC fluctuations on the input/output (good insurance). I was going to suggest 22k pot and no 22k resistor to GND, if that works for you. But connecting the slider of the pot to the other end will give you that DC "insurance" thing.
: |
It should say "I was going to suggest 22k pot and no 22k resistor to GND, if that works for you. But used as a standard POT not a shunt attenuator". The worst case (max DC) will be when you are at max volume the best case, close to min. volume.
Sorry. And ofcourse in your schematic the ser. resistor with the POT is used as an voltage devider. I should stop talking too much. Nonsense comes out of your mouth/writing.:rolleyes:
Anyway, check this sch. out. Tell me what you think. R1 not need it if you are not affraid of killing the IC. In that case one res. less. And also the lower the output DC impedance of the source the better. It means at max levels (or close to it) the DC may start going down again. |
|
|
|
