Hello all,
I've just built a new GC 🙂 which sounds very open and clean but is still breaking in. However, it seems excessively bright and harsh particularly where vocals and percussion are concerened. Is this just something that will calm down with time or should I try to remedy the situation? More bass would be good also.
From reading previous threads I'm lead to believe that a zobel network may help if the amp is unstable but I'm not sure what 'unstable' really means. I'm using dual monoblock LM3875s GCs on 36V rails.
Thanks.
I've just built a new GC 🙂 which sounds very open and clean but is still breaking in. However, it seems excessively bright and harsh particularly where vocals and percussion are concerened. Is this just something that will calm down with time or should I try to remedy the situation? More bass would be good also.
From reading previous threads I'm lead to believe that a zobel network may help if the amp is unstable but I'm not sure what 'unstable' really means. I'm using dual monoblock LM3875s GCs on 36V rails.
Thanks.
The sound of the amp will be influenced by parts and type of enclosure that you are using. Any mor info on that. You might also try low pass filter and maybe even a buffer, as this seems to be a cure fot bright sounding amps.
Thanks Peter, It's built into a sturdy 10mm ply box with two external PSUs, using 2 1000uF decoupling caps per channel, 4.7 uF Black Gate N input caps, 10K input resistor and a stepped attenuator using RC55Cs.
I tried some QED silver anniversary bi wire cable with it last night and that has helped bring the sound together a lot. I will try the buffer some time in the future! 🙂
Piero
I tried some QED silver anniversary bi wire cable with it last night and that has helped bring the sound together a lot. I will try the buffer some time in the future! 🙂
Piero
Gain Clone is bright sounding
I have just looked inside the chip PDF on the National site of the LM3875 . The input stage is excellent . The compensation of the voltage amplifier stage ( VAS ) a modest 10 pF . I have one doubt , the biasing of the output transistors is not state of the art . I suspect that there is crossover distortion of a higher order than we expect these days . The 10 pF of the VAS indicated to me that crossover has been reduced by the crudist of methods , massive negatve feedback .
I think taking the output/ loudspeaker terminal to the negative rail via a resistor might might work as a cure . If about 1 watt was disipated the crossover distortion would be offset . The resistor will see the rail voltage as if connected to ground . I guess a higher value could be tried . 5 watts is as far as I would go .
This trick is used as many will know with op amps , in that case a 10 K resistor is taken to the negative or positive rail ( not both ) . I will not calculate the value for this amplifier as I feel someone who has a working example and the skills should try it first . Having read these pages these skills are not in doubt .
I have just looked inside the chip PDF on the National site of the LM3875 . The input stage is excellent . The compensation of the voltage amplifier stage ( VAS ) a modest 10 pF . I have one doubt , the biasing of the output transistors is not state of the art . I suspect that there is crossover distortion of a higher order than we expect these days . The 10 pF of the VAS indicated to me that crossover has been reduced by the crudist of methods , massive negatve feedback .
I think taking the output/ loudspeaker terminal to the negative rail via a resistor might might work as a cure . If about 1 watt was disipated the crossover distortion would be offset . The resistor will see the rail voltage as if connected to ground . I guess a higher value could be tried . 5 watts is as far as I would go .
This trick is used as many will know with op amps , in that case a 10 K resistor is taken to the negative or positive rail ( not both ) . I will not calculate the value for this amplifier as I feel someone who has a working example and the skills should try it first . Having read these pages these skills are not in doubt .
Hi,
could the brightness be, in part, due to oscillation and/or overshoot as the stability margin becomes less?
Is there a way to identify this without an oscilloscope?
Does Ropie need a Zobel (series R+C //load) or a full Thiel network (series R+C & R//L in series with the load) to effect a cure?
or
something completely different?
I'd be interested in hearing suggestions for improving the lack of bass.
Break in? I don't believe it. Electronics should be designed to settle into full replay mode within seconds of start-up. Some would say mS, but junction temperatures may have a small effect on the operation.
I think folk that talk about break-in, particularly over many weeks, are simply allowing familiarity to confuse the subjective assessment of what they are hearing
. (Oh dear, I fear the tinder box is dry 
)
could the brightness be, in part, due to oscillation and/or overshoot as the stability margin becomes less?
Is there a way to identify this without an oscilloscope?
Does Ropie need a Zobel (series R+C //load) or a full Thiel network (series R+C & R//L in series with the load) to effect a cure?
or
something completely different?
I'd be interested in hearing suggestions for improving the lack of bass.
Break in? I don't believe it. Electronics should be designed to settle into full replay mode within seconds of start-up. Some would say mS, but junction temperatures may have a small effect on the operation.
I think folk that talk about break-in, particularly over many weeks, are simply allowing familiarity to confuse the subjective assessment of what they are hearing
. (Oh dear, I fear the tinder box is dry )If I understand you correctly you are actually saying that the sound will be influenced by the enclosure?
The sound can be influenced by a choice in parts.
The enclosure has nothing to do with the sound quality of the gainclown or any other amplifier.
The physical layout and grounding will play a part in the absence of hum but that is as far as it goes.
Hi Burned,
I come down on your side, but I feel inclined to point out that the sound output will be influenced by the stability margin.
A high margin will give a very damped effect on transient signals and a small margin will tend towards spittyness and other audible artefacts brought on, or exagerated by, the overshoot that a stable amp must produce when run with a small margin.
The margin will be affected by the cable layout and I suspect the quality of the ground and the PSU.
Based on the above, the sound could and probably will vary with any cause that results in a change in the stability of the amp.
Summarising my view, electrical changes can change the quality of the sound output, non electrical changes will have no effect on the sound output.
edit;
how could he agree completely with my bad spelling (now corrected)?
I come down on your side, but I feel inclined to point out that the sound output will be influenced by the stability margin.
A high margin will give a very damped effect on transient signals and a small margin will tend towards spittyness and other audible artefacts brought on, or exagerated by, the overshoot that a stable amp must produce when run with a small margin.
The margin will be affected by the cable layout and I suspect the quality of the ground and the PSU.
Based on the above, the sound could and probably will vary with any cause that results in a change in the stability of the amp.
Summarising my view, electrical changes can change the quality of the sound output, non electrical changes will have no effect on the sound output.
edit;
how could he agree completely with my bad spelling (now corrected)?
Hi Burned,
I am only too pleased to hear agreement, but what if we now introduce a metal enclosure?
Could it change the sound output?
As an aside, I always find that I can never get a metal encased amplifier to perform as well as an uncased one!!!!
I am only too pleased to hear agreement, but what if we now introduce a metal enclosure?
Could it change the sound output?
As an aside, I always find that I can never get a metal encased amplifier to perform as well as an uncased one!!!!
Have you guys noticed when this thread started? I suspect Ropie has moved on from this project. 😉
burnedfingers said:The enclosure has nothing to do with the sound quality of the gainclown or any other amplifier.
And how did you come to that conclusion?
Futher on brightness
Instability will cause brightness . I havn't looked at all the variations of Zobel circuit so forgive any repetition if this has been covered . 0.1 uF 100 V film type and 4R7 3 watt is a good generic type for all amplifiers of the modern op amp layout . I would say if it needs more the amplifier is not a good design to persist with .
If no oscilloscope is available a finger test of the 4 R7 resistor will tell you something . If hot there is a problem ( can be very hot ! ) . Often the earthing is the problem . Look on the web for various PCB layouts recomended by the chip designers . Copy the layout as best you can in the hardwired form .
I once had this problem with a TDA2030 , a cheap PCB from RS in the UK cured it . Try as best I could I couldn't get a stip board layout to work . The TDA 2030 never was very stable . It wouldn't drive a motor without great bursts of distortion . A five transistor MOS FET amp of the simplest type had no trouble .
I have tried to add power transistors to op amps ( NE 5534 etc ) . Again not as easy as some people make out . Try driving a motor to see the real problems . I suspect there is no finer test .
I have also found an AM / FM radio useful . If it becomes noisy when the amplifier is on there is a problem . The bridge rectifier needs typically 0.01 uF capacitors on each diode as this will cause radio noise . About 20 cm from the amp is about right .
Be careful when using osciloscopes . I had a nice sine wave on the scope one day from a new amp . It was RF from a low energy light bulb in the room getting in .
The bass might be part of the same problem .
Instability will cause brightness . I havn't looked at all the variations of Zobel circuit so forgive any repetition if this has been covered . 0.1 uF 100 V film type and 4R7 3 watt is a good generic type for all amplifiers of the modern op amp layout . I would say if it needs more the amplifier is not a good design to persist with .
If no oscilloscope is available a finger test of the 4 R7 resistor will tell you something . If hot there is a problem ( can be very hot ! ) . Often the earthing is the problem . Look on the web for various PCB layouts recomended by the chip designers . Copy the layout as best you can in the hardwired form .
I once had this problem with a TDA2030 , a cheap PCB from RS in the UK cured it . Try as best I could I couldn't get a stip board layout to work . The TDA 2030 never was very stable . It wouldn't drive a motor without great bursts of distortion . A five transistor MOS FET amp of the simplest type had no trouble .
I have tried to add power transistors to op amps ( NE 5534 etc ) . Again not as easy as some people make out . Try driving a motor to see the real problems . I suspect there is no finer test .
I have also found an AM / FM radio useful . If it becomes noisy when the amplifier is on there is a problem . The bridge rectifier needs typically 0.01 uF capacitors on each diode as this will cause radio noise . About 20 cm from the amp is about right .
Be careful when using osciloscopes . I had a nice sine wave on the scope one day from a new amp . It was RF from a low energy light bulb in the room getting in .
The bass might be part of the same problem .
Metal boxes
Wood boxes do sound good . With careful design no problems should arrise . With class D even the heat sink isn't required . Class D might be the death of these gain clone amps .
I wrote today as I felt there was still room to improve this thread . Sorry I came so late , I have read it from time to time . Too lazy to join in up to now .
Wood boxes do sound good . With careful design no problems should arrise . With class D even the heat sink isn't required . Class D might be the death of these gain clone amps .
I wrote today as I felt there was still room to improve this thread . Sorry I came so late , I have read it from time to time . Too lazy to join in up to now .
Mr. Peter how did you arrive at the conclusion that an enclosure made a sonic difference? Is it because you happen to market a brand of Gainclown? Granted you make an impressive piece I will say that. Whats next? Do copper heatsinks make the gainclown sound better? Maybe a different brand of heatsink grease will make a difference. Maybe a different type of wood is in order.
What do you say you take a good long look at reality again?
burnedfingers said:
I can only confirm that conclusion after putting exactly same circuits into different enclosures for last 3 years.
When I posted the quoted by you statement, I was not marketing GC amps yet.
How did you arrived to your conclusion?
burnedfingers said:
Actually the copper in AMP-2 monoblocks didn't sound that good and we switched to aluminum.
But the cork side panels that I used in the original prototype influenced the sound of the amp much less than the production acrylic sides, that my partner insisted on using.
Its amazing how we can arrive at conclusions here on this forum. We think we are going to hear a difference therefore we do. This line of reasoning is applied to capacitors, resistors, volume controls, even the type of wire we use. Granted there are very very small differences between different components and one must rationalize whether it is worth the expenditure of the additional funds. In a blind listening test 9 times out of 10 the average person will not be able to tell any sonic difference. However, if you introduce the idea that a certain part will result in a sonic change that sonic difference will be heard.
I arrived at my decision as a result of over 30 years playing with amplifiers. I have conducted blind listening tests with about everything I have built. In addition I have used numerous test equipment at my disposal.
I recently built two identical gainclowns. One is in a aluminum enclosure and one is attached to a board. They both test the same and they both sound the same according to the people that auditioned them.
It is unfortunate that in todays world anything goes. Products are marketed with questionable promises.
Sorry Mr. Peter, I try to view things with an impartial nature. Unless you or someone else on this forum can provide information that will prove beyond a shadow of a doubt that there are indeed sonic differences that are a result of the types of cabinets used I will continue to fly the BS flag in this matter.
I arrived at my decision as a result of over 30 years playing with amplifiers. I have conducted blind listening tests with about everything I have built. In addition I have used numerous test equipment at my disposal.
I recently built two identical gainclowns. One is in a aluminum enclosure and one is attached to a board. They both test the same and they both sound the same according to the people that auditioned them.
It is unfortunate that in todays world anything goes. Products are marketed with questionable promises.
Sorry Mr. Peter, I try to view things with an impartial nature. Unless you or someone else on this forum can provide information that will prove beyond a shadow of a doubt that there are indeed sonic differences that are a result of the types of cabinets used I will continue to fly the BS flag in this matter.
Since you mentioned blind listening tests, I can only send you to that post: http://www.goodsoundclub.com/Forums/ShowPost.aspx?postID=2734#2734 as myself I have no desire to discuss those matters.
Getting back to topic of "Bright GC", I'd like to point out that the non insulated chip has tendency to sound actually brighter than the insulated version. After I observed that, and it was also confirmed by people listening to both versions (in a same amp chassis), I never went back to the non insulated chip package.
Getting back to topic of "Bright GC", I'd like to point out that the non insulated chip has tendency to sound actually brighter than the insulated version. After I observed that, and it was also confirmed by people listening to both versions (in a same amp chassis), I never went back to the non insulated chip package.
Properly implemented the blind listening tests have merit. Unfortunately your reference to the post above has none. I'm still looking for proof from you and unfortunately it seems that your lacking.
How was this test conducted? Was this a blind test Peter?
Same chip with the exception that the uninsulated chip requires an insulator. Any difference in circuit design? No there isn't. Any reason to believe that a coating on the outside of the chip would alter the sonics? No there isn't.
I guess we really don't have to have any proof now do we Peter?
- Status
- Not open for further replies.