| rs1026 |
| I have seen lot of discussion on this forum on GC using premium components but has anyone built a GC with all cheap/scrap components without bothering too much about using only metal film resistors or balck gate capacitors. was just curious :) |
|
|
| brianuk |
hi Igla,
Please comment on the results and whether you made a comparison to a commercial amp and/or a more expensive GC.
Thanks!
Brian |
|
|
| BobEllis |
I did too - results
At least in its "on the cheap" incarnation the gainclone is not my cup of tea. Not terrible, but missing the detail of some more traditional designs. It's a decent starting point, perfectly suitable for less than super critical applications. |
|
|
| Scribble |
ive built mine with el-cheapo components and it sound good to me. Very clear and detailed but only thing i dont like is the lack of bass.
But then again im proberly not as perdantic as some other members of the board ;) |
|
|
| psychosteve |
ive built tda 2030 chip amps with salvaged parts and that sounded fine to me. its ok if your sure of the values of components, a good multi meter, and preferably a capacitance meter. the trouble comes with resistors that look like 1watt, when they are only 0.2watt.
and what tollerance? 5%? 10%? 0.5%???? measure them, TWICE, then check again!!!
then run them as a dummy load, a simple battery circuit, with a volt and current meter to check they can handle the stress your putting on them.
see you later, steve.. .. |
|
|
| Igla |
Well....I can try with my limited english.
I have used no-name metal film resistors which
I had in my junk box (22k / 1k).
For the power suply I've used Nippon Chemi-Con capacitors (1000uF)-cheap and the only brand available at that time in our
local shop and one cube-like big 25A/600V cheap diode bridge.
At the input cheap Piher pot and cheap film cap (Arcotronics;
the gray one, I think MMK).All wirings made point-to-point
of CAT 5 FTP cable.
This was my first LM3875 NIGC and I must say that it sonicaly beat the s..t out of my AIWA 950.It also replaced my ZEN amp.
The first big improvement was, when I removed the input cap.
Then I replaced Piher pot with Alps blue-much more expensive
but sonicaly quite an improvement.
The next step was MUR860 double bridge-smoother mids and more defined bass.Then I replaced double bridge with just one.
I can't say why I liked it more.
The last improvement is different schematic (and same parts)made by Stefanobilliani-
something like mix-mode current amp-look for the last scheme:
http://www.diyaudio.com/forums/show...&threadid=41189).
My parts have different values but the basics stays the same.
I am still testing it.All I can say is, that this is totaly different soud
comming out of my full range speakers.
Igla |
|
|
| brianuk |
Thankyou for the comments.
Igla, your english is very good, much better than me trying ANY other language.:( |
|
|
| Igla |
Thanks Brianuk...you don't want to hear me speaking
german.:D
I must say that my really first GC was inverted design
by Thorsten schematics.
After a while I just had to try non inverted design from
LM3875 datasheet.I found it somehow better sounding.
Then all the modifications followed.
A few weaks ago I found an article Mr.Pass wrote about
current amps and full range drivers.
So I just had to convert my GC to current amp; well, some
sort of mix mode amp.(check Rod Elliot page-variable
impedance).
Mr.Pass wrote that a full range driver hooked to a solid state amp
sounds more like a kitchen radio.
I can confirm that after the modification.
Hard to describe the difference in sound....more like the difference between my old computer monitor with washed-out colours to my brand new one.
The sound of NIGC on my full range loudspeakers is just
washed-out comparing to the mix mode amp. |
|
|
| mateo88 |
| I've built several, all with very cheap components, and I do mean cheap. Just regular 10% carbon film resistors and some cheap 2200 uf capacitors. I've used both the lm3875 and lm3886, and I would strongly reccommend the lm3886 over the 3875. I just used the schematic from the data sheet, and there's much better bass response; it seems more detailed all around, as well. Really, I enjoy listening to my cheap gainclones with the lm3886 more than the one I have built using Brian's boards. You would be VERY surprised at what you can come up with for a few dollars. |
|
|
| thomas997 |
| quote: | Originally posted by mateo88
I've built several, all with very cheap components, and I do mean cheap. Just regular 10% carbon film resistors and some cheap 2200 uf capacitors. I've used both the lm3875 and lm3886, and I would strongly reccommend the lm3886 over the 3875. I just used the schematic from the data sheet, and there's much better bass response; it seems more detailed all around, as well. Really, I enjoy listening to my cheap gainclones with the lm3886 more than the one I have built using Brian's boards. You would be VERY surprised at what you can come up with for a few dollars. |
Yeah I cant bring myself to spend more than pennies on resistors..
Just get a bunch and a good multimeter, then you can pick out the ones within 1% if you want.
Sounds great to me. |
|
|
| mrskinny |
| Just read Igla's post about converting his amp to current feedback. Any chance you could post some details- schematic,parts values,etc? Unfortunately, the link in your post doesn't work- what thread does this refer to? |
|
|
| geewhizbang |
I have rather doubted that there are sonic effects from most fancy resistors. Most of the difference people "hear" is probably psychoacoustics.
You may find that a fancier part may make sense in a critical section where low level signals are involved. A resistor with a lot of stray inductance, such as a wirewound resistor, could have some ringing effects in such places. Even so the errors are pretty small relative to the desired properties, or otherwise it wouldn't even be fair to call it a resistor.
I wish that there was more engineering going on here that tried to eliminate sources of extra cost and not use the expensive stuff when it didn't matter. It seems like most people are doing the opposite and using the expensive stuff all of the time just because it is "better'.
If it then breaks your budget, it isn't better, especially when it doesn't matter.
So I say go for it. You probabaly can't hear a big difference, and you can have fun later trying to replace some of the cheap parts with more expensive ones and find out what difference it makes or not. |
|
|
| Igla |
Here is the link to Rod Elliott sound projects:
http://sound.westhost.com/index.html
Go to : Projects ; Power Amplifiers & Accessories and
there is article "Variable impedance".
I can't find stefanobillianis thread about his current amp
to attach the link.I think it was "Easy GC Current Amp"
in the Loudspeakers forum.
Here is anoter link with excelent article:
http://www.firstwatt.com/intro.htm
Go to Articles, and read
"Current Source Amplifiers and Sensitive / Full-Range Drivers".
I am trying to attach my schematic...hope it will work.
I have added R3 resistor by stefanobillianis advice and it
works great.
Don't forget-R2 is dissipating some power, so calculate the wattage you'll need.
Changing the resistors you can set the sound you like.For example-changing R3 from 4,7k to 1k and the mids and hights
starts to approach ordinary GC sound.
You must try for yourself and find the right combination.
I've tried different values for the resistors and found the
best results with values:
R1=220 to 330 ohm
R2=0.22 to 0.33ohm
R3=1k to 4.7k |
|
|
| Igla |
You are welcome.
I would like to try silicon pot on my GC...I've found
schematic on Pedja site :
http://users.verat.net/~pedjarogic/audio/
Pedja is quite positive on the sonic results.
The only thing is, that non of our stores hold DS1802 chip.
Whats a shame. |
|
|
| jcarr |
As some of you may know, a number of the audio products that I design retail for well over 10,000 dollars. And these products use relatively large quantites of exotic or custom components. Nonetheless, I find that it is possible to build very fine-sounding amplifiers without resorting to expensive components.
The key is to understand the topology and operating conditions of the circuit, and know what characteristics are required from individual components at each location in the circuit.
For instance, let's take resistors. For locations in a circuit that need good high frequency response, inductance in a resistor is undesireable. It can therefore be advantageous to use low-inductance resistors (in some cases, carbon composition resistors may be the ticket) for these locations. However, in locations subject to big voltage swing, a carbon-film resistor can lead to a 30dB increase in distortion, due to its temperature constant.
http://www.diyaudio.com/forums/show...0533#post240533
http://www.diyaudio.com/forums/show...0670#post240670
There are other instances where a resistor with inductive behavior can be an advantage (for example, in the degenerator resistors of an input stage), however, this is outside the scope of the Chip Amp forum.
To summarize, the characteristics that make a component suitable for one location may be totally misplaced for another. The real issue not how much money you spend (or for that matter, how little you spend), but rather, "Do you understand what you are doing?"
hth, jonathan carr |
|
|
| rnoble |
| quote: | Originally posted by jcarr
To summarize, the characteristics that make a component suitable for one location may be totally misplaced for another. The real issue not how much money you spend (or for that matter, how little you spend), but rather, "Do you understand what you are doing?"
hth, jonathan carr [/B] |
The original question, and the issue I have with your response, is in regards to GC amps. There are 4 resistors on a GC give or take - so enlighten us with WHICH ONES should be metal film or carbon or not that will improve the sound based on your ramblings above, please. |
|
|
| jcarr |
What?? And deprive you of the personal improvement and satisfaction to be gained from thinking through a problem for yourself? That would be most unkind of me. ;)
regards, jonathan carr |
|
|
| grege |
| quote: | Originally posted by rnoble
The original question, and the issue I have with your response, is in regards to GC amps. There are 4 resistors on a GC give or take - so enlighten us with WHICH ONES should be metal film or carbon or not that will improve the sound based on your ramblings above, please. |
Geez Ron, you sound like your dog just died. ;)
Jonathan was kind enough to offer some of his experience and you jump down his throat. I don't think the "please" at the end excuses your tone.
But... it would be nice if Jonathan could offer some suggestions if he has the time. :D
Thanks |
|
|
| jeff mai |
I've done this.
In my LM3875 prototype amp I used cheap metal film resistors, Panasonic FC caps, a cheap bridge and a toroid and heatsink from Jaycar. The IEC socket, input jacks and speaker terminals are attached with 1" nylon standoffs and woodscrews. Very easy.
All of the components except the bridges and toroid are soldered directly to the chip pins. It's a pain, but I believe this is *key* to getting good results.
It doesn't have the magic of my now dismantled valve amp, but it is a valve-like, relaxed sound. It was certainly worth the effort.
I'm currently building a version with "better" parts, but it will be a few weeks before I can compare. |
|
|
| geewhizbang |
My degree is in Mechanical rather than electrical engineering so I don't understand why some find it necessary to make posts, but then not answer basic questions. I certainly understand basic electrical theory, but why not answer the question that someone asked a bit further back in this thread.
It is pretty damn insulting to suggest that you are "doing them a favor" by not answering. The person asking the question did not know, and may not even know the theory you are talking about.
Why NOT enlighten them.
If you know the answer, then I don't see how it is necessary to mete it out in small doses as if that will somehow make you run out of good ideas ...
In my own chosen field I have never been afraid of sharing my knowledge because I'm pretty sure my brain is still working and will have other good ideas tomorrow. I like the idea of a good idea being shared so it can do the most good.
So again, what is the problem with answering an honest question?
Of the four resistors in a gainclone, what would you suggest and why? Then we can get an idea of the theory involved, which makes the reading of the thread worth the time. Otherwise your non-answer is a waste of electrons. |
|
|
| rnoble |
| quote: | Originally posted by grege
Geez Ron, you sound like your dog just died. ;)
Jonathan was kind enough to offer some of his experience and you jump down his throat. I don't think the "please" at the end excuses your tone.
But... it would be nice if Jonathan could offer some suggestions if he has the time. :D
Thanks |
Hi Greg yes it was a bit harsh but it was just so .... esoteric.... :P
I'll leave the rest. |
|
|
| jeff mai |
| quote: | Originally posted by geewhizbang
Why NOT enlighten them. |
Not time?
It's all out there to be discovered. Motivated people find their own answers. I'm certainly appreciative when someone takes the time to explain things to me, but I don't *expect* it.
Jonathan took the time to say what he had to say in his first post. There was information in that post, but apparently it wasn't specific enough for some. |
|
|
| geewhizbang |
Why post at all then? There was no need for the insulting "doing you a favor" non-answer.
It took about as long to non-answer the question as it would have to post a simple answer. |
|
|
| jcarr |
Greg: No problem. The way that Ron came across did a wonderful job of de-fanging any harshness which he may have otherwise intended. Indeed I found Ron's post to be dreadfully hilarious.
But lest I be accused of deriving pleasure from toying with others :D, may I point out that the second post that I linked to (which you did read, Ron, didn't you?) specifically described an experiment involving the global NFB resistor. And the last time I looked, global NFB resistors were definitely considered to be part of GainClone topology (both non-inverting and inverting).
At any rate, to rehash the points that I was trying to make in my old post, I would definitely suggest avoiding carbon resistors for this purpose.
Given the poor tempco and hygroscopic characteristics of carbon-composition, I must consider these to be manifestly unsuitable. Now, even though I previously found carbon-film resistors to be unsuitable for a global NFB resistor, it _may_ be possible to obtain satisfactory results - if you use a carbon resistor with a higher-than-normal power rating. But I could be wrong (after seeing the initial results from the carbon-film resistors, I didn't attempt any deeper analyses). The outcome would depend on what the distortion mechanism is. If the cause is the power being applied across the resistor, a resistor of higher power rating (or using multiple resistors in parallel) should be most worthwhile. However, if the distortion is due to a voltage effect, carbon may not be a good idea after all.
In my experience, the other resistors don't have nearly the same impact on the performance as the NFB resistor. For the input loading (which I find to rank second in its impact on the performance), input and NFB node-to-ground (or NFB node-to-input in the case of inverting topology) resistors, low-inductance and low-noise behaviour appear to be the most important characteristics.
I also prefer to avoid resistors (whether metal-film or carbon-film) that exhibit magnetic properties (I keep a magnet handy when I go component-shopping), because my experience has been that such resistors frequently don't sound to my liking. But this is subjective rather than being something that is consistently measureable (in terms of distortion).
hth, jonathan carr
PS. I have no problem answering questions that are asked in good faith, but when I see responses that suggest that the poster isn't trying to think (or even read linked material), this quickly kills any desire that I may have to answer. If the poster isn't willing to at least try to help himself, why should I bother?
Besides, I subscribe to the philosophy that you ultimately do people a disservice if you only give them the answers without encouraging them to think for themselves. You know the saying, "give a man a fish and he'll eat for a day." |
|
|
|
