Question after reading this post from Backbones on the BA-3 sticky concerning P3 adjustment observations:
http://www.diyaudio.com/forums/pass-labs/195303-burning-amp-ba-3-a-11.html#post2817172
I use BA-3 preamp as well as BA-3 Amplifier. Should I leave Preamp P3 alone, and adjust P3 on Amplifier? Or vice versa, or is this tail chasing? I am inclined to adjust power amp, leaving pre alone.
One other question, for reference looking at BA-3 board as in BA-3 build thread: When Backbones says "turn towards R3" or "turn towards R4" I suppose you turn one clockwise and the other counterclockwise? Looking at board, and the way R3 or R4 are laid out, this would be necessary? I may need to PM Backbones on that one...
Thanks in advance for any clarity or suggestion,
Russellc
http://www.diyaudio.com/forums/pass-labs/195303-burning-amp-ba-3-a-11.html#post2817172
I use BA-3 preamp as well as BA-3 Amplifier. Should I leave Preamp P3 alone, and adjust P3 on Amplifier? Or vice versa, or is this tail chasing? I am inclined to adjust power amp, leaving pre alone.
One other question, for reference looking at BA-3 board as in BA-3 build thread: When Backbones says "turn towards R3" or "turn towards R4" I suppose you turn one clockwise and the other counterclockwise? Looking at board, and the way R3 or R4 are laid out, this would be necessary? I may need to PM Backbones on that one...
Thanks in advance for any clarity or suggestion,
Russellc
Last edited:
I have only used the BA-3 as a pre but I would be inclined to adjust the FE and OS as a group and the preamp by itself using a 47K resistor as load.
And the boards are mirrored which means the pots turn opposite of each other.
I've got my BA-3 FE set up for lowest distortion possible right now, 2nd harmonics are a bit below 3rd. I am not sure it sounds best this way but it is tight and fast and forward.
Next time I adjust P3 (using a FFT analyzer) I will set P3 so the 2nd harmonic is slightly higher than 3rd.
Also, setting P3 at midpoint seems to be completely random and not really an "average" or median setting. I probably depends on your jfet Idss and other components in the circuit. The only way to set P3 so the channels are even is with an analyzer or scope.
The interesting thing to me is that lowest distortion (slightly 3rd dominant) results in a narrower but taller soundstage. And it sounds "harder". 2nd dominant is wider but slower and not as detailed. Also, I have found that the bass is much tighter when the 2nd harmonic is nulled.
I't's probably subjective what someone likes. Depends on the room, speakers, etc.
And the boards are mirrored which means the pots turn opposite of each other.
I've got my BA-3 FE set up for lowest distortion possible right now, 2nd harmonics are a bit below 3rd. I am not sure it sounds best this way but it is tight and fast and forward.
Next time I adjust P3 (using a FFT analyzer) I will set P3 so the 2nd harmonic is slightly higher than 3rd.
Also, setting P3 at midpoint seems to be completely random and not really an "average" or median setting. I probably depends on your jfet Idss and other components in the circuit. The only way to set P3 so the channels are even is with an analyzer or scope.
The interesting thing to me is that lowest distortion (slightly 3rd dominant) results in a narrower but taller soundstage. And it sounds "harder". 2nd dominant is wider but slower and not as detailed. Also, I have found that the bass is much tighter when the 2nd harmonic is nulled.
I't's probably subjective what someone likes. Depends on the room, speakers, etc.
I'm inclined to agree with you. Thanks for clearing up the directional thing, I figured it must work that way. I thought about turning careful counts in the pre first, then going back to original position, and then trying the power amp next, then if any consensous, dialing the pre up again. I also agree it seem to be in a not so much 2nd degree area at the mid point.
I dont have an analyzer, so I am going to just keep careful count of the turns.
All my BA-3 devices, as well as the ones I'm going to build have Toshiba devices on the front end boards and Toshiba J fets around 8 ids. I will wait to see if Backbone responds, then will PM concerning how many turns he was experiencing. Of course if he uses the Fairchild devices or different ids J fets, it may or may not be the same...
Russellc
I don't know how many turns I have...I didn't really pay attention. I just kept cranking until the 2nd harmonic flattened out and started to go back up. Then I turned it back to lowest point and left it.
But I did notice that the channels were not "even" when set at midpoint. Which leads me to believe that either a jfet mismatch, mosfet mismatch or something else in the circuit causes variation. Also, a few turns(like 10) from midpoint may go toward 2nd harmonic and then back to 3rd. It is not a simple 2nd harmonic = clockwise, 3rd = counterclockwise.
Really, I think you need a analyzer to adjust this thing. Without one you are just chasing your tail. You may be able to hear the extremes of the settings but you won't be able to get identical settings channel to channel...also, it is extremely time consuming to do "by ear". I know before i used the fft analyzer I liked it a few turns from midpoint. Which direction, I am not sure (I am guessing it was towards the 3rd harmonic). The problem is 3 turns from mid might be 2nd harmonic, 6 turns might be 3rd, 10 turns might be 2nd. And the number of turns will vary with each setup and from channel to channel.
I think when talking about p3 settings it would be logical to say I like 2nd harmonic at XXdb and 3rd harmonic at XXdb. Right now I have my 2nd and 3rd about even which is as low as the 2nd harmonic would go.
For what it's worth i am just using a USB audio interface (m-audio) and faber software. I know next to nothing about analog scopes, testing but I figured it out. I don't think you could test an amp with this setup (might be to hot a signal) but a line level preamp, no problem. Literally, I just plugged the RCAs from the preamp into the USB interface and sent a 1kz test tone from the headphone out to the preamp. Not the best setup but it worked.
I had the usb interface from when I used to try to lay down some guitar to cakewalk (god, I am old) or garage band. Similar software to the faber toolkit can be found online for free or cheap. There are even some ipad/iphone apps that can do this. Or, if you are on windows there are some usb scopes/analyzers for about $100. It's pretty nice to send the test tone, monitor and measure all from your computer. I am sure you could put together something tat would be able to measure harmonics for less than $100.
But I did notice that the channels were not "even" when set at midpoint. Which leads me to believe that either a jfet mismatch, mosfet mismatch or something else in the circuit causes variation. Also, a few turns(like 10) from midpoint may go toward 2nd harmonic and then back to 3rd. It is not a simple 2nd harmonic = clockwise, 3rd = counterclockwise.
Really, I think you need a analyzer to adjust this thing. Without one you are just chasing your tail. You may be able to hear the extremes of the settings but you won't be able to get identical settings channel to channel...also, it is extremely time consuming to do "by ear". I know before i used the fft analyzer I liked it a few turns from midpoint. Which direction, I am not sure (I am guessing it was towards the 3rd harmonic). The problem is 3 turns from mid might be 2nd harmonic, 6 turns might be 3rd, 10 turns might be 2nd. And the number of turns will vary with each setup and from channel to channel.
I think when talking about p3 settings it would be logical to say I like 2nd harmonic at XXdb and 3rd harmonic at XXdb. Right now I have my 2nd and 3rd about even which is as low as the 2nd harmonic would go.
For what it's worth i am just using a USB audio interface (m-audio) and faber software. I know next to nothing about analog scopes, testing but I figured it out. I don't think you could test an amp with this setup (might be to hot a signal) but a line level preamp, no problem. Literally, I just plugged the RCAs from the preamp into the USB interface and sent a 1kz test tone from the headphone out to the preamp. Not the best setup but it worked.
I had the usb interface from when I used to try to lay down some guitar to cakewalk (god, I am old) or garage band. Similar software to the faber toolkit can be found online for free or cheap. There are even some ipad/iphone apps that can do this. Or, if you are on windows there are some usb scopes/analyzers for about $100. It's pretty nice to send the test tone, monitor and measure all from your computer. I am sure you could put together something tat would be able to measure harmonics for less than $100.
Thanks for the informative feedback. I get you on the two channels not being the same and all. I guess, as you say, they arent at mid setting either.
This is probably boring the heck out of the more advanced users here (likely most all more advanced than I) so if you wouldnt mind. PM me your setup dumbed down for a moron as myself can understand. I want to set this up! I dont mind buying stuff, many times finding freeware loaded with unwelcome stuff...I just dont know where to begin.
If any other beginners are interested, I will be glad to PM results. If all thinks its of value, post it here, I guess this is my thread, but for the respect of those that find this discussion boring, whatever!
Any help on what I need and how to hook it to amp much appreciated!
Russellc
My opinion on the original post after pondering much of Nelson's brain waves is to go for the absolute minimum distortion at all preceeding gain stages and leave harmonic spice additions to the end as much as possible.
That is try and null 2nd harmonic and mimimise distortion wherever possible at the preamp and only add 2nd Harmonic at the end assuming that is what you were chasing.
That is try and null 2nd harmonic and mimimise distortion wherever possible at the preamp and only add 2nd Harmonic at the end assuming that is what you were chasing.
Thanks for the informative feedback. I get you on the two channels not being the same and all. I guess, as you say, they arent at mid setting either.
This is probably boring the heck out of the more advanced users here (likely most all more advanced than I) so if you wouldnt mind. PM me your setup dumbed down for a moron as myself can understand. I want to set this up! I dont mind buying stuff, many times finding freeware loaded with unwelcome stuff...I just dont know where to begin.
If any other beginners are interested, I will be glad to PM results. If all thinks its of value, post it here, I guess this is my thread, but for the respect of those that find this discussion boring, whatever!
Any help on what I need and how to hook it to amp much appreciated!
Russellc
I don't think we need to PM, people smarter than us can chime in along the way, no offense taken. I don't know much about this stuff, more knowledgable people can probably give better advice.
Here's the software I used (mac only):
Electroacoustics Toolbox
Here is the USB interface:
M-Audio - M-Track II
The usb interface is good for line level signals (it's meant for microphones and guitars) but not the high level of a power amp. You would probably need a different interface with actual probes for that.
I used the fft analyzer in the software to see the harmonics after feeding the preamp a 1kz signal.
If you are on windows a better option might be picoscope (https://www.picotech.com/oscilloscope/2200/picoscope-2200-portable-oscilloscopes).
But it seems alot of the computer based test solutions seem to be focused on the ipad and iphone. Probably because of portability. Something like this looks very interesting:
https://www.oscium.com/oscilloscopes
Honestly, that's about the extent of my knowledge. All the software requires that you know at least something about how a conventional scope or test equipment works. The interfaces seem to be digital versions of analog scopes. I don't know how to use an analog scope so it took me a long time to get what I want. I watched a lot of youtube videos and things. The learning curve was steep. It is similar to photoshop (just trying to find an analogy some of you might know)...the tools and functions of photoshop are kind of digital versions of what you would find in a traditional darkroom but just becasue you know how to print pictures in a real darkroom it doesn't mean you know photoshop. But the learning curve will be easier.
Nothing boring here in Pass Labs. Plenty of dummies too, no need to feel like you're alone in that department.
The experts are just gonna tell us we need $20K Audio Precision analyzers anyways (you don't, at least not to see the effect of adjusting P3...for other things, yes...maybe).
From what I understand, the bandwidth and sensitivity of these digital scope solutions are not as good as dedicated stuff like AP makes. You are always limited by the self-generated noise, hash, whatever of the PC. So the ability to measure really low level noise, etc. is compromised. Basically, anytime you try to measure below what the PC is introducing to the signal you are screwed. For things like adjusting P3 it's not an issue. For other things it is.
From what I understand, the bandwidth and sensitivity of these digital scope solutions are not as good as dedicated stuff like AP makes. You are always limited by the self-generated noise, hash, whatever of the PC. So the ability to measure really low level noise, etc. is compromised. Basically, anytime you try to measure below what the PC is introducing to the signal you are screwed. For things like adjusting P3 it's not an issue. For other things it is.
What is most interesting to me about the whole P3 thing is that minimum distortion might not be preferred to some. I now realize that the sound of P3 nulled is more similar to some of the more "hifi" sounding gear I have listened to. I used to wonder why hi-end gear in stores didn't do it for me...it was because years ago I was listening to amps/preamps with a lot of 2nd harmonic that produced a big, diffuse soundstage. Now, with better amps and speakers I am hearing a more focused sound with more fidelity but less of the big, wet soundstage. I think this is the appeal of some of the SET tube amps with gross 2nd harmonic. I big and lush soundstage. You pay for it with image height and bass though. People like what they are used to. I am trying to play with P3 to find out what I like, subjectively.
Right now I've got the 2nd harmonic nulled. It's as fast and cleaned as it will get and the bass is awesome. Te speakers I have now (KEF r700) are very revealing to changes to P3. The Focals I had before not so much. I love the detail, image height and bass. But the soundstage is narrower. If I could leave it like this but have the soundstage width of a 2nd harmonic dominant P3 setup I would be happy.
Anyways, my f5 is all taken apart (drilling and tapping heatsinks for more mosfets) so experimentation will have to wait.
Right now I've got the 2nd harmonic nulled. It's as fast and cleaned as it will get and the bass is awesome. Te speakers I have now (KEF r700) are very revealing to changes to P3. The Focals I had before not so much. I love the detail, image height and bass. But the soundstage is narrower. If I could leave it like this but have the soundstage width of a 2nd harmonic dominant P3 setup I would be happy.
Anyways, my f5 is all taken apart (drilling and tapping heatsinks for more mosfets) so experimentation will have to wait.
IMO it's better to get as close to identical channel match than worry about absolute levels of distortion, they are low enough to not care.
Any decent audio interface and the free RMAA software can do this. In a pinch, people have used laptop sound cards - the crappy inbuilt ones and while they're rubbish for absolute measurements, for relative measurements like these they are acceptable.
You have to use a small attenuator to bring down the signal level from the amplifier output to something that won't kill your soundcard. A simple two-resistor divider, or a pot if you want it to be adjustable. Inverse of gain ratio is fine for a fixed divider, for adjustable you set it at minimum and move it till you get get something on the screen. I also advise a small series resistor and cap before the pot to reduce current and block DC, a cap after the pot to ensure it doesn't interfere with any DC-coupled stages afterward and a pair of back to back zeners to protect the soundcard from user goof-ups. So a total of six components and you have a little amp measuring jig.
I normally use just the FFT calibration screen from RMAA for P3 adjustment. Target mildly higher H2 on one channel, note down the average level, then adjust second channel for the same result. Readjust offset which would have moved slightly, and away you go. Put a test load on the amp, or it isn't a valid result.
Download the software and play around, it doesn't cost anything.
Any decent audio interface and the free RMAA software can do this. In a pinch, people have used laptop sound cards - the crappy inbuilt ones and while they're rubbish for absolute measurements, for relative measurements like these they are acceptable.
You have to use a small attenuator to bring down the signal level from the amplifier output to something that won't kill your soundcard. A simple two-resistor divider, or a pot if you want it to be adjustable. Inverse of gain ratio is fine for a fixed divider, for adjustable you set it at minimum and move it till you get get something on the screen. I also advise a small series resistor and cap before the pot to reduce current and block DC, a cap after the pot to ensure it doesn't interfere with any DC-coupled stages afterward and a pair of back to back zeners to protect the soundcard from user goof-ups. So a total of six components and you have a little amp measuring jig.
I normally use just the FFT calibration screen from RMAA for P3 adjustment. Target mildly higher H2 on one channel, note down the average level, then adjust second channel for the same result. Readjust offset which would have moved slightly, and away you go. Put a test load on the amp, or it isn't a valid result.
Download the software and play around, it doesn't cost anything.
You can measure the wiper to ends resistance when the amp is off.
This allows you to set the wiper exactly to centre point.
But as mentioned previously this does not equate to exact minimum distortion.
The slight difference in gm between N & P requires the Rs for each to be slightly different to equalise the effective gm (is that degenerated gm?).
This is not an Idss matching problem. The Idss could be identical, but the gm can still be different.
Last edited:
Here's a screenshot of a my FFT analyzer.
So far, with P3 experimentation this has sounded best to me in my system. More 2nd harmonic gets too vague and fuzzy and vocals aren't as well isolated, although the soundstage gets a bit wider (and shorter) with more 2nd harmonic. Bass also gets a bit fuzzier with more 2nd harmonic.
I primarily listen to vinyl and through my vinyl setup (Rega P5/Ortofon 2M bronze, Salas phono) it sounds great. Through a ESS 9023 Dac it sounds a bit forward and sterile. I probably would dial in more 2nd harmonic if I was just using digital.
For what it's worth, I don't know much about this stuff. So if anyone sees something I m doing wrong, software or measurement-wise feel free to comment.
So far, with P3 experimentation this has sounded best to me in my system. More 2nd harmonic gets too vague and fuzzy and vocals aren't as well isolated, although the soundstage gets a bit wider (and shorter) with more 2nd harmonic. Bass also gets a bit fuzzier with more 2nd harmonic.
I primarily listen to vinyl and through my vinyl setup (Rega P5/Ortofon 2M bronze, Salas phono) it sounds great. Through a ESS 9023 Dac it sounds a bit forward and sterile. I probably would dial in more 2nd harmonic if I was just using digital.
For what it's worth, I don't know much about this stuff. So if anyone sees something I m doing wrong, software or measurement-wise feel free to comment.
Attachments
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.