Sorry to bother you again Pinkmouse,but one more question. I wondered if you use a resistor to ground on the + input? If I remember correctly,the original Gainclone schematic showed the + input connected directly to signal ground,but then was changed to a 20k resistor,and eventually to a 220k.
Hello,
As the conversation about OPA549 is up I add my opinion to the soup.
I have recently built lm3875 based regulated NIGC with all the fancy resistors and capacitors based on the BrianGT:s schematic. The sound is nice but...
My second gainclone was OPA549 based IGC with regulated supply and 1000uF per rail after regulation all made of standard industrial parts.
This is the best amp that I have heard so far especially the mids and highs are excellent. But the bass section is better with LM3875.
I´ve found out that the input cap can make a huge difference to the sound of the OPA549 amp. After extensive testing I ended up using 2.2uF MKP.
I still have to let the lm3875 amp to burn in for a week and I hopefully get much better results and the I have to begin modifying it with input caps and different zobel values.
As the conversation about OPA549 is up I add my opinion to the soup.
I have recently built lm3875 based regulated NIGC with all the fancy resistors and capacitors based on the BrianGT:s schematic. The sound is nice but...
My second gainclone was OPA549 based IGC with regulated supply and 1000uF per rail after regulation all made of standard industrial parts.
This is the best amp that I have heard so far especially the mids and highs are excellent. But the bass section is better with LM3875.
I´ve found out that the input cap can make a huge difference to the sound of the OPA549 amp. After extensive testing I ended up using 2.2uF MKP.
I still have to let the lm3875 amp to burn in for a week and I hopefully get much better results and the I have to begin modifying it with input caps and different zobel values.
Thank you for the reply Pinkmouse. Jarkaa mentioned using 1000uf after the regulators. I did try this myself (pair of 1500uf), but found that this made the sound worse,and was better with just the 33uf on the chips.Something that I hadn't thought about changing,though,is the input cap.I tried a number of different input caps with the LM1875 amp,and found some 4uf caps that had the best sound,and have used these since. 2.2uf you say? Very interesting....
pinkmouse said:My favorite is still the 549, after two years or so of mucking around with chip amps. However, I do have a regulated LM4780 implementation nearly ready for testing, so maybe that will change my opinion, who knows!
Hi Pinkmouse!
My sample OPA549 arrived and I am interested in testing it in inverted mode. Would you be kind enough to point me to your implemented schematic diagram?
Many thanks!
Finally picked up the parts to convert the OPA549 amp to inverted. I decided to change as few parts as possible to keep the comparison between it and the NI version as fair as possible. Changed the feedback resistor to 220k,removed the 22k from the + input and connected it directly to ground, removed the 680R from the - input and added the 22k in series. Everything else remained unchanged. First thing that I noticed was the reduction in the amount of hiss through the speakers.Not as quiet as the LM1875 amp,but pretty close now.Next thing that I noticed was the reduction in output level (probably from the additional 22k on the input)- 8db less than before.However,this did improve the sound quality. Not a night and day difference,but definitely clearer sounding than before. Still surprised at the poor bass response from this chip. For a chip that is supposed to be capable of a 10A output,you would never know it. Just replaced the 4uf input caps with 2.2uf ones and am letting them burn in now. Will see what difference this makes later in the day. Now (drum roll please) its time for the Stupid Question of the Day! Sorry, no EE here. I wondered if there was any value of feedback resistor that could be used so that the resistor in series with - input could be omitted entirely? Anyone out there who has tried this?
I am comparing these two chips because- 1) The only chip that any of the electronic stores near me stock is the LM1875,and won't order just a couple of chips for me to try.OPA549's were from TI's sample program.Tried National's sample program,but no luck. 2) They both use the same supply voltage. Unfortunately,few surplus transformers around here,and are quite expensive new- the two toroids that I am using now cost $160! 3) As I have efficient speakers,the power output from the chips doesn't really mean much either way. My 8w/ch SE amp will play these speakers more than loud enough.
aHobbit said:
Can't post pics at the moment due to only getting internet access from work, but it is basically the original Thorsten circuit, just swap out the 3875 for a 549, (noting pin-out differences, of course), and put a trimmer pot in the place of the 18k to ground.
I'm sure there is a copy of my circuit somewhere on the board, but I couldn't find it searching just now, maybe you will have a little more luck!
You know, I really haven't listened to it as a full range amplifier. As a subwoofer amplifier (mine is bridged), it's pretty good. It's got a lot of power, but I still think the 38xx chips sound better, but they lack the power the 549 has.
The other problem could be the very high gain I have the chip at - this chip doesn't get along with high gain too well.
Maybe if a 3886 or 3875 were in a B/P config, they would have the advantage over the 549.
Oh, the other benefit of the 549 is it can handle lower impedences better than national's chips - I hooked up a 4-ohm speaker to my bridged 549 amp, and although it ran quite hot, the chip didn't complain at all - a 3886 would start making this horrible popping sound when it was given a 2-ohm load.
edit: You mention the 3886 is too bright sounding. How much capacitance do you have on it? I have a hi-fi LM3875 amp I built and it sounds very warm, actually, it sounds better than my very good yamaha 5.1 amp - all except for less power. Maybe add some more capacitance and your issue will get better.
The other problem could be the very high gain I have the chip at - this chip doesn't get along with high gain too well.
Maybe if a 3886 or 3875 were in a B/P config, they would have the advantage over the 549.
Oh, the other benefit of the 549 is it can handle lower impedences better than national's chips - I hooked up a 4-ohm speaker to my bridged 549 amp, and although it ran quite hot, the chip didn't complain at all - a 3886 would start making this horrible popping sound when it was given a 2-ohm load.
edit: You mention the 3886 is too bright sounding. How much capacitance do you have on it? I have a hi-fi LM3875 amp I built and it sounds very warm, actually, it sounds better than my very good yamaha 5.1 amp - all except for less power. Maybe add some more capacitance and your issue will get better.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.