I adjusted P1/P2, but one channel was more jumpy. Do you know why? Q1/Q2 are GR quads bought from diy audio store. Q3/Q4 are Fairchild.
Just post a few pics, but ensure when checking for jumpiness Iq on both chans need to be equal iot make a good assessment. Also, ensure audio ground wires are good and tightly connected, and that P3 is centered on both chans.
Otherwise, ensure «equal» torque to sinks for Q3-4.
If still jumpy, bobs your uncle or Vgs differs a bit or something else irrelevant.
Otherwise, ensure «equal» torque to sinks for Q3-4.
If still jumpy, bobs your uncle or Vgs differs a bit or something else irrelevant.
I too am working on my third BA3, 6 deep, but not having a distortion analyzer or an hi-res FFT, the P3 setting still remain a bit incomplete to my mind...
Why are you not fully satisfied? I am not sure if you are asking for help and/or stating that you do not like the sound of the amplifier and you think changing P3 may make you satisfied.
I found that a couple of turns on P3 made a very noticeable difference to the sound.
One way I prefer to the other.
Remember, you need to re-bias after adjustment.
This is based on actually listening to music. I know it will not satisfy those that spend more time measuring than listening.
One way I prefer to the other.
Remember, you need to re-bias after adjustment.
This is based on actually listening to music. I know it will not satisfy those that spend more time measuring than listening.
I just meant that, you know, not having instruments to measure what is the point of P3 and related harmonic contribution, that's not the best situation for a nerd 😉
This emphasises the difference between-
It sounds better to me---- and
It measures better therefore it must sound better
It sounds better to me---- and
It measures better therefore it must sound better
Ah!
This probably makes things worse, but... it is very fun and informative to adjust, listen, and measure. I understand now.
If you feel compelled, I think a REW-based solution with an existing computer and your choice of peripheral is an excellent starting point.
I hope you get the chance to do it some time in the future.
Anything wrong with just using two 50 ohm resistors in place of the trimmer?
Edit: or rather 60-ish ohm resistors in R3 and R4 leaving out the potentiometer.
^ At first glance, your edited version should be okay per schematic. I have not checked the boards.
With that said, the 'original' F5 w/o 'P3' simply had 10ohm source resistors for the JFETs as do many other amps. So, I'm not sure that the added 50R to sub in for the pot's equivalent value at midpoint is required at all. It was interesting (to me) that the 'P3 included' version of the F5 had a 200R pot and no equivalent to R5. I assume the BA-3 was... updated... idiot-proofed in a way by adding R5 to both limit the effect of P3 and ensure that one JFET could not have effectively 0 source resistance.
That's me throwing out something that is just on the edge of my stupidity... but making an effort to try. So, as always, a more skilled person should verify / contest / add insight to avoid ookups.
tl;dr - I'd try it at home, but maybe not everyone should. 😉
With that said, the 'original' F5 w/o 'P3' simply had 10ohm source resistors for the JFETs as do many other amps. So, I'm not sure that the added 50R to sub in for the pot's equivalent value at midpoint is required at all. It was interesting (to me) that the 'P3 included' version of the F5 had a 200R pot and no equivalent to R5. I assume the BA-3 was... updated... idiot-proofed in a way by adding R5 to both limit the effect of P3 and ensure that one JFET could not have effectively 0 source resistance.
That's me throwing out something that is just on the edge of my stupidity... but making an effort to try. So, as always, a more skilled person should verify / contest / add insight to avoid ookups.
tl;dr - I'd try it at home, but maybe not everyone should. 😉
I was looking at the schematic wrong. I was thinking of the resistance of R3, R4 and Each side of P3 in series when they are in parallel. So the resistance of 10 ohms and 50ohms in parallel is 8.3.
So maybe a couple of these:
https://www.digikey.com/en/products...oric-bc-components/MRS25000C8258FCT00/5064682
My thought is that people who don't have distortion analyzing equipment can simply skip installing P3 and use something close to 8.3 ohms. It would be cool if someone could find the sweet spot (like in 6l6's video) and measure each side of the trimmer pot. Then we would know the resistance that is likely going to give you the right sound.
I understand the idea of the BA-3 is to promote experimenting so maybe this defeats the purpose.
So maybe a couple of these:
https://www.digikey.com/en/products...oric-bc-components/MRS25000C8258FCT00/5064682
My thought is that people who don't have distortion analyzing equipment can simply skip installing P3 and use something close to 8.3 ohms. It would be cool if someone could find the sweet spot (like in 6l6's video) and measure each side of the trimmer pot. Then we would know the resistance that is likely going to give you the right sound.
I understand the idea of the BA-3 is to promote experimenting so maybe this defeats the purpose.
I agree with / and like what you're proposing. Pots can always be added later. There have been enough 'what do I do with P3 / how do I set P3 to center if already installed?' types of posts to indicate a fixed resistance idea would be nice.
I'd just go with 10R like in other amps and call it a day vs. 8ish. Also... in hindsight 60R in R3 and R4 would likely not be advisable. You're correct, subbing out the pot in parallel with R3 and R4 shouldn't be added to R3 / R4. Enjoyed thinking it through.
I'd just go with 10R like in other amps and call it a day vs. 8ish. Also... in hindsight 60R in R3 and R4 would likely not be advisable. You're correct, subbing out the pot in parallel with R3 and R4 shouldn't be added to R3 / R4. Enjoyed thinking it through.
@wengtech -
Let's try to move things here... from the pre-amp thread. If the mods move some of the previous posts, great....
https://www.diyaudio.com/community/...eamp-build-guide.258022/page-146#post-7472461
You (best I can tell) only made a few very small errors re: dissipation. Someone should double check me also. ZM tried to teach me the "proper" way with then understanding of current through the vertical pair(s), but I still admit to cheating and think that I get the correct answer the wrong way.
First small thing - Let's say you have 65V rails 'loaded' => measured with the amp hooked up and fully biased. I'm switching it a little from what I typed earlier b/c you did the math on the output power to align with my SWAG.
Let's for fun bump up the loss at the MOSFETs (and other small losses) from ~4V to ~5V, just to make the math align with what you've already calculated. The dissipation is calculated on the 65V not the 60V. The output power is calculated on the 60. For most situations, it's not terribly critical, but in some situations people really care about those last few Watts of power and dissipation.
So... tl;dr - a better estimation would be to use 65Vp vs. 60Vp for the dissipation.
Then... you used the Vp to calculate your dissipation (you did Vp x Iq) vs. the Vrms.
So... tl;dr - the dissipation would be calculated off of 46Vrms x Iq vs. 65Vp.
Second small thing... as mentioned, you did it correctly for calculating the current required into your chosen load of 8R for output power. Now, as some folks alluded to... you can decide how much of that power you want to be in Class A into a particular load and set your Iq accordingly.
You don't have to set the Iq to get max power in "full Class-A". I doubt very seriously you'd ever output a 60Vp signal.... So, rather than have heatsinks the size of my room and heat the neighborhood, you can back off the Iq and use the Class A/B nature of the amp to let it "clunk" if you really, really, really need to get the cops called on you.
Note - I do this to continue my own learning too. 1/2 of what I've written could be wrong, but it's how I try to go about planning for how I'll choose my toroids / heatsinks etc. for a power target with my speakers.
Let's try to move things here... from the pre-amp thread. If the mods move some of the previous posts, great....
https://www.diyaudio.com/community/...eamp-build-guide.258022/page-146#post-7472461
You (best I can tell) only made a few very small errors re: dissipation. Someone should double check me also. ZM tried to teach me the "proper" way with then understanding of current through the vertical pair(s), but I still admit to cheating and think that I get the correct answer the wrong way.
First small thing - Let's say you have 65V rails 'loaded' => measured with the amp hooked up and fully biased. I'm switching it a little from what I typed earlier b/c you did the math on the output power to align with my SWAG.
Let's for fun bump up the loss at the MOSFETs (and other small losses) from ~4V to ~5V, just to make the math align with what you've already calculated. The dissipation is calculated on the 65V not the 60V. The output power is calculated on the 60. For most situations, it's not terribly critical, but in some situations people really care about those last few Watts of power and dissipation.
So... tl;dr - a better estimation would be to use 65Vp vs. 60Vp for the dissipation.
Then... you used the Vp to calculate your dissipation (you did Vp x Iq) vs. the Vrms.
So... tl;dr - the dissipation would be calculated off of 46Vrms x Iq vs. 65Vp.
Second small thing... as mentioned, you did it correctly for calculating the current required into your chosen load of 8R for output power. Now, as some folks alluded to... you can decide how much of that power you want to be in Class A into a particular load and set your Iq accordingly.
You don't have to set the Iq to get max power in "full Class-A". I doubt very seriously you'd ever output a 60Vp signal.... So, rather than have heatsinks the size of my room and heat the neighborhood, you can back off the Iq and use the Class A/B nature of the amp to let it "clunk" if you really, really, really need to get the cops called on you.
Note - I do this to continue my own learning too. 1/2 of what I've written could be wrong, but it's how I try to go about planning for how I'll choose my toroids / heatsinks etc. for a power target with my speakers.
Perhaps I missed it but I didn't see any mention if the source/preamp considered will give the requisite voltage swing to drive this high power follower to full output.
Thanks for moving the thread.@wengtech -
Let's try to move things here... from the pre-amp thread. If the mods move some of the previous posts, great....
https://www.diyaudio.com/community/...eamp-build-guide.258022/page-146#post-7472461
You (best I can tell) only made a few very small errors re: dissipation. Someone should double check me also. ZM tried to teach me the "proper" way with then understanding of current through the vertical pair(s), but I still admit to cheating and think that I get the correct answer the wrong way.
First small thing - Let's say you have 65V rails 'loaded' => measured with the amp hooked up and fully biased. I'm switching it a little from what I typed earlier b/c you did the math on the output power to align with my SWAG.
Let's for fun bump up the loss at the MOSFETs (and other small losses) from ~4V to ~5V, just to make the math align with what you've already calculated. The dissipation is calculated on the 65V not the 60V. The output power is calculated on the 60. For most situations, it's not terribly critical, but in some situations people really care about those last few Watts of power and dissipation.
So... tl;dr - a better estimation would be to use 65Vp vs. 60Vp for the dissipation.
Then... you used the Vp to calculate your dissipation (you did Vp x Iq) vs. the Vrms.
So... tl;dr - the dissipation would be calculated off of 46Vrms x Iq vs. 65Vp.
Second small thing... as mentioned, you did it correctly for calculating the current required into your chosen load of 8R for output power. Now, as some folks alluded to... you can decide how much of that power you want to be in Class A into a particular load and set your Iq accordingly.
You don't have to set the Iq to get max power in "full Class-A". I doubt very seriously you'd ever output a 60Vp signal.... So, rather than have heatsinks the size of my room and heat the neighborhood, you can back off the Iq and use the Class A/B nature of the amp to let it "clunk" if you really, really, really need to get the cops called on you.
Note - I do this to continue my own learning too. 1/2 of what I've written could be wrong, but it's how I try to go about planning for how I'll choose my toroids / heatsinks etc. for a power target with my speakers.
i always forgot about the class a/ab thingies...
That's why it's look like crazy because I demand 300watt into class A 😀
So how to bias this amp into class ab??
Nothing is impossible for those passlabs guys to make this happens 😀
(But those chassis dimension is still possible for crazy diy guys also 😀 )
Are there any pictures of inside this XS300 amp??
I wondering why this amp using 2 chassis, is that one chassis is for PSU? Why they also put heatsink in the psu chassis??
😀
Last edited:
Remember that the Burning Amp series of amplifiers have no global feedback, and get their linearity from running the output stage biased very heavily into Class-A.
Lowering the bias may still sound good, but it's going to have some kind of decreased objective performance. You'll have to try and see. Will likely work just fine.
Yes, XS series has PSU in separate chassis. Chassis looks neat with heatsinks. It may even be regulated, you'll have to ask Nelson.
Lowering the bias may still sound good, but it's going to have some kind of decreased objective performance. You'll have to try and see. Will likely work just fine.
Yes, XS series has PSU in separate chassis. Chassis looks neat with heatsinks. It may even be regulated, you'll have to ask Nelson.