Sound
Brian,
Now that the 22 ufd cap is connected to 6 and 3, how is the sound compared to the connection of 2 and 6?
Looking at the board, I am not planning on connecting the cap at the LM318. I will connect from the pin of R6(R29) & R7(R30).
My trembling hands are not so good at soldering to those pins. It will not be so busy connecting up there.
My next project is to remove the 470 ufd cap I put in for C9 and C24. Going to use the correct 220 ufd. And bypass it with a small film cap.
I have read some much about the feedback resistor in all the NIGC threads that my 12K AB carbon comp is coming out too. A real expect said that they can increase distortion by 30dB when used in this application. He said he had measured this effect.
This will be easy, slap in some different ones and listen for less high level distortion. I have never heard any bad effects, noise, distortion, or fuzzy sound with carbon comps. But this is my first chip amp.
May do it tonight if I get a couple hours.
Thanks again for all the pics, I know now how to mod mine to Rev C status.
George
Brian,
Now that the 22 ufd cap is connected to 6 and 3, how is the sound compared to the connection of 2 and 6?
Looking at the board, I am not planning on connecting the cap at the LM318. I will connect from the pin of R6(R29) & R7(R30).
My trembling hands are not so good at soldering to those pins. It will not be so busy connecting up there.
My next project is to remove the 470 ufd cap I put in for C9 and C24. Going to use the correct 220 ufd. And bypass it with a small film cap.
I have read some much about the feedback resistor in all the NIGC threads that my 12K AB carbon comp is coming out too. A real expect said that they can increase distortion by 30dB when used in this application. He said he had measured this effect.
This will be easy, slap in some different ones and listen for less high level distortion. I have never heard any bad effects, noise, distortion, or fuzzy sound with carbon comps. But this is my first chip amp.
May do it tonight if I get a couple hours.
Thanks again for all the pics, I know now how to mod mine to Rev C status.
George
I haven't listened to it in it's current state, so I can't comment. I am going to be giving this one away to my brother, so I want to just get it in a proper state and send it off. I have my RevC monoblocks oin my main system right now and couldn't be happier.
I have yet to try Russ's bypass, and might replace the 270R with a 1R to see what happens. I need to make a decision about which to include in the kits. I might just put both in and let builders be builders.
I have yet to try Russ's bypass, and might replace the 270R with a 1R to see what happens. I need to make a decision about which to include in the kits. I might just put both in and let builders be builders.
, finally.
I did a bit more work on my Rev A amp and think its sounding even better. Its getting to the point that I'm not really motivated to experiment with changing it to a Rev C - I believe Mauro when he says that Rev A is the preferred version. Here's what I added to the stock amp kit provided by Russ and Brian:
--- changed C1, C2, C17, C18 (the LM3886 PS caps) from 220uf to 2200uf. I used Panasonic FC 35v parts since my power supply rails are running +-30v. These fit perfectly and are the exact size outlined on the board making me wonder if Mauro originally intended to use 2200uf caps. Could there have been a typo on the parts list?
--- added another 0.33uf bypass cap between V+ and V- right at the LM3886 pins (pin 4 and 5). I used Wima .33uf 100v polypro caps from Mouser with 5mm pin spacing which fit perfectly.
--- bypassed the power supply caps (C1, C2, C17, C18, C3, C8, C20, C28) with 0.1uf film caps. I don't recall the make of these caps - I had them left over from previous projects.
--- I added snubbers along the lines of what Carlosfm has advocated for his gainclone type amp. To make these fit on the board, I used 0.47ohm / 2W SMD resistors from Mouser and SMD COG ceramic caps so everything mounted nicely on the bottom of the board.
At this point it sounded very fine but the whole point of DIY is excess so I kept going.
I swapped out the $7 Parts Express transformer with a 400VA toroid I got off of ebay. This is definitely overkill but the price wasn't too bad. The seller has more if folks are interested (I'm not the seller or related in any way). Search for 22v toroid on ebay and you'll find the transformer I used.
I also replaced the power supply diode bridges with discrete bridges made from MUR1520 ultrafast/soft recovery diodes.
And I replaced the input cap with a 1uf Russian paper in oil cap - another ebay special. Search for 1uf 200v and you'll turn these up. One seller labels them as type K40Y9 and another K40U-9 but both look similar. I bought mine from KWTUBES and they are the type K40Y9. These are physically quite big and don't fit on the board - instead I just wire them directly between the board and the RCA jacks.
This is the combination I'm listening to and it sounds very good. In terms of bang for the buck, I'd say that the first set of changes to the power supply caps and bypassing was most significant. Changing the diodes and input caps was a more subtle but still worthwhile upgrade.
---Gary
--- changed C1, C2, C17, C18 (the LM3886 PS caps) from 220uf to 2200uf. I used Panasonic FC 35v parts since my power supply rails are running +-30v. These fit perfectly and are the exact size outlined on the board making me wonder if Mauro originally intended to use 2200uf caps. Could there have been a typo on the parts list?
--- added another 0.33uf bypass cap between V+ and V- right at the LM3886 pins (pin 4 and 5). I used Wima .33uf 100v polypro caps from Mouser with 5mm pin spacing which fit perfectly.
--- bypassed the power supply caps (C1, C2, C17, C18, C3, C8, C20, C28) with 0.1uf film caps. I don't recall the make of these caps - I had them left over from previous projects.
--- I added snubbers along the lines of what Carlosfm has advocated for his gainclone type amp. To make these fit on the board, I used 0.47ohm / 2W SMD resistors from Mouser and SMD COG ceramic caps so everything mounted nicely on the bottom of the board.
At this point it sounded very fine but the whole point of DIY is excess so I kept going.
I swapped out the $7 Parts Express transformer with a 400VA toroid I got off of ebay. This is definitely overkill but the price wasn't too bad. The seller has more if folks are interested (I'm not the seller or related in any way). Search for 22v toroid on ebay and you'll find the transformer I used.
I also replaced the power supply diode bridges with discrete bridges made from MUR1520 ultrafast/soft recovery diodes.
And I replaced the input cap with a 1uf Russian paper in oil cap - another ebay special. Search for 1uf 200v and you'll turn these up. One seller labels them as type K40Y9 and another K40U-9 but both look similar. I bought mine from KWTUBES and they are the type K40Y9. These are physically quite big and don't fit on the board - instead I just wire them directly between the board and the RCA jacks.
This is the combination I'm listening to and it sounds very good. In terms of bang for the buck, I'd say that the first set of changes to the power supply caps and bypassing was most significant. Changing the diodes and input caps was a more subtle but still worthwhile upgrade.
---Gary
A couple of notes...
Gary,
The spots for the electrolytic bypass caps on the PCBs are large to accomodate LOW ESR caps which are almost always larger than standard. No other reason.
A larger bypass cap may not be (and usually will not be) better. The reason is the purpose of that cap. The purpose is not so much a power resevoir as a power bypass. The real resevoir is the 4 10Kuf caps. More important here is the qaulity of the cap and its ability to provide a low impedance bypass .
Now, on the merrits of REV C I would not say it is "better" than REV A . I would just say that "I" generally like it better. The reason is that I percieve tha the soundstange is improved by the better phase tracking in my setup. Some people may not even notice this. I Love REV A as well, and in some cases I use it, and in some REV C.
Having both is the best situation.
You would likely only notice the change in a Direct A/C test.
The bottom line is A,B or C nobody will be dissapointed with a "My REF" amp in either the stereo config or the monobloc.
Cheers!
Russ
Gary,
The spots for the electrolytic bypass caps on the PCBs are large to accomodate LOW ESR caps which are almost always larger than standard. No other reason.
A larger bypass cap may not be (and usually will not be) better. The reason is the purpose of that cap. The purpose is not so much a power resevoir as a power bypass. The real resevoir is the 4 10Kuf caps. More important here is the qaulity of the cap and its ability to provide a low impedance bypass .
Now, on the merrits of REV C I would not say it is "better" than REV A . I would just say that "I" generally like it better. The reason is that I percieve tha the soundstange is improved by the better phase tracking in my setup. Some people may not even notice this. I Love REV A as well, and in some cases I use it, and in some REV C.
Having both is the best situation.You would likely only notice the change in a Direct A/C test.
The bottom line is A,B or C nobody will be dissapointed with a "My REF" amp in either the stereo config or the monobloc.
Cheers!
Russ
ESR
Hi Russ,
I understand your comments that a bigger capacitor may not always be better but in this case I've tried it and think it actually is better. Feel free to disagree, but people shouldn't let theoretical considerations stop them from trying it. In terms of low ESR, both the Nichicon UPW caps provided in the original kit and the Panasonic FC caps that I used are low ESR caps. According to datasheets available on the web, the Nichicon 220uf UPW cap has an impedance at 100khz of ~0.1ohms, which is a good measure of ESR. The Panasonic 2200uf FC has an impedance at 100khz of ~0.02ohms, so if ESR is what really matters then the larger cap should be better. Now I'm sure this isn't the only parameter thats relevant. For example larger capacitors can have other parasitics such as inductance that could cause problems. I think thats what snubbers are there for - to try and compensate for the inductance of larger capacitors. Thats why I suggest using the larger cap and snubbers togethers. The exact calculation of what makes a good snubber is pretty empirical which is why I went with the tried and true values which seemed to work in other projects.
Happy listening.
---Gary
Hi Russ,
I understand your comments that a bigger capacitor may not always be better but in this case I've tried it and think it actually is better. Feel free to disagree, but people shouldn't let theoretical considerations stop them from trying it. In terms of low ESR, both the Nichicon UPW caps provided in the original kit and the Panasonic FC caps that I used are low ESR caps. According to datasheets available on the web, the Nichicon 220uf UPW cap has an impedance at 100khz of ~0.1ohms, which is a good measure of ESR. The Panasonic 2200uf FC has an impedance at 100khz of ~0.02ohms, so if ESR is what really matters then the larger cap should be better. Now I'm sure this isn't the only parameter thats relevant. For example larger capacitors can have other parasitics such as inductance that could cause problems. I think thats what snubbers are there for - to try and compensate for the inductance of larger capacitors. Thats why I suggest using the larger cap and snubbers togethers. The exact calculation of what makes a good snubber is pretty empirical which is why I went with the tried and true values which seemed to work in other projects.
Happy listening.
---Gary
Better transformer
I suspect the better transformer is helping also. Once I read the current draw with no load, I knew the 7.00 specials are 7.00 transformers. I nned to source something else, those ebay ones are too big to fit in my box.
Did you connect these 43.00 transformers up using the SIP diodes to measure output. The 7.00 transformers give 34.5 volt rails with a 8 amp 600 volt SIP diode bridge. The discrete diodes you are using should give a little higher output than the SIP package ones.
Sourced all my own parts, used 1000 and 220 ufd FM caps in first board, the same values but Rubycon ZL in second. These were 50 volt on the LM3886, and 25 volt across the zeners. Suspect there may be an advantage to slightly larger caps, but like mentioned the inductance goes up fast as the caps get larger in value and phyical size.
If you bypass those 10K caps you may get some switching noise. Every case where I bypassed the first filter caps has been noisier than just big electrolytics alone.
Glad you are so happy with the amp. These are nice, took mine to someones house and connected it to his big ugly speakers. He loved it. His normal amp was something like 350 watts.
George
I suspect the better transformer is helping also. Once I read the current draw with no load, I knew the 7.00 specials are 7.00 transformers. I nned to source something else, those ebay ones are too big to fit in my box.
Did you connect these 43.00 transformers up using the SIP diodes to measure output. The 7.00 transformers give 34.5 volt rails with a 8 amp 600 volt SIP diode bridge. The discrete diodes you are using should give a little higher output than the SIP package ones.
Sourced all my own parts, used 1000 and 220 ufd FM caps in first board, the same values but Rubycon ZL in second. These were 50 volt on the LM3886, and 25 volt across the zeners. Suspect there may be an advantage to slightly larger caps, but like mentioned the inductance goes up fast as the caps get larger in value and phyical size.
If you bypass those 10K caps you may get some switching noise. Every case where I bypassed the first filter caps has been noisier than just big electrolytics alone.
Glad you are so happy with the amp. These are nice, took mine to someones house and connected it to his big ugly speakers. He loved it. His normal amp was something like 350 watts.
George
its the motion (of the electrons)
But I listened to the amp after changing just the LM3886 PS caps and adding the bypass caps / snubbers without changing the transformer. And as I tried to say in my original message, this change had a larger effect than just changing the transformer - which I did later.
---Gary
But I listened to the amp after changing just the LM3886 PS caps and adding the bypass caps / snubbers without changing the transformer. And as I tried to say in my original message, this change had a larger effect than just changing the transformer - which I did later.
---Gary
Re: Better transformer
George,
My $7 special transformer gave lower rails - I measured +-30v when running the amp with the SIP bridge.
The $43 ebay transformer was listed as a 22v transformer but when it showed up it was actually labelled as a 20v transformer. In circuit, it gave +-28.75v rails with the stock SIP bridge and +- 29.3v rails with the MUR1520 discrete bridges.
How do these compare to other folks using the $7 specials? I'm wondering if one of my LM3886 chips could be drawing a lot of current thus loading the transformer more.
---Gary
Panelhead said:
George,
My $7 special transformer gave lower rails - I measured +-30v when running the amp with the SIP bridge.
The $43 ebay transformer was listed as a 22v transformer but when it showed up it was actually labelled as a 20v transformer. In circuit, it gave +-28.75v rails with the stock SIP bridge and +- 29.3v rails with the MUR1520 discrete bridges.
How do these compare to other folks using the $7 specials? I'm wondering if one of my LM3886 chips could be drawing a lot of current thus loading the transformer more.
---Gary
No Carlos has not even built the amp, in fact he has said in a previous post he does not intend to. So I am not sure (since he does not share it) what practical basis his comments have at all.
I am very open to new ideas, but the one with the ideas should present them in a way that they can be objectively analyzed, and repeated.
So his purpose in posting is dubious in my mind until it is better clarified as Carlos has no stake in the design as he (apparently) never intends to build it.
Silly, promote it in another thread Carlos.
Cheers!
Russ
I am very open to new ideas, but the one with the ideas should present them in a way that they can be objectively analyzed, and repeated.
So his purpose in posting is dubious in my mind until it is better clarified as Carlos has no stake in the design as he (apparently) never intends to build it.
Silly, promote it in another thread Carlos.
Cheers!
Russ
why can't we be friends?
In the infamous words of Rodney King - can't we all just get along?
I must say I'm honestly perplexed by the hostility shown in the last few posts.
Whether or not Carlos has built these amps isn't relevant. I built them, I tried out some ideas related to power supply bypassing, I liked what I heard, and reported the results.
The response seems to have been that I'm wrong - it can't be an improvement. And when I suggest reasons why I think it might be an improvement and when Carlos jumps and in agrees - then he's attacked.
Seems like I've hit a nerve with some folks - which wasn't my intention. I just wanted to pass along some tips on what worked for me.
Regards to all.
---Gary
In the infamous words of Rodney King - can't we all just get along?
I must say I'm honestly perplexed by the hostility shown in the last few posts.
Whether or not Carlos has built these amps isn't relevant. I built them, I tried out some ideas related to power supply bypassing, I liked what I heard, and reported the results.
The response seems to have been that I'm wrong - it can't be an improvement. And when I suggest reasons why I think it might be an improvement and when Carlos jumps and in agrees - then he's attacked.
Seems like I've hit a nerve with some folks - which wasn't my intention. I just wanted to pass along some tips on what worked for me.
Regards to all.
---Gary
OK - how about something less controversial.
Russ - did you see my post about rail voltages with the $7 Parts Express transformer? If you get a chance, could you grab a meter and let me know what you measure? Do you get the +- 34v mentioned by George or the +-30v that I measure?
Thanks,
---Gary
Russ - did you see my post about rail voltages with the $7 Parts Express transformer? If you get a chance, could you grab a meter and let me know what you measure? Do you get the +- 34v mentioned by George or the +-30v that I measure?
Thanks,
---Gary