Good to hear Bob.
I don't have a BA-3 FE, but according to the article Nelson says... "voltage appearing at the Gate of Q3 is amplified by something less than the ratio of R10 divided by R8..."
So maybe those values (R8,R10...R9,R11) could be altered slightly to get the channels to match. Maybe someone with some experience with it will offer up some good advice on how to proceed.
I don't have a BA-3 FE, but according to the article Nelson says... "voltage appearing at the Gate of Q3 is amplified by something less than the ratio of R10 divided by R8..."
So maybe those values (R8,R10...R9,R11) could be altered slightly to get the channels to match. Maybe someone with some experience with it will offer up some good advice on how to proceed.
Gain of Q3/Q4 of left channel is slightly different than gain of Q3/Q4 of right channel, due to unmatched devices I assume. JFET input stage is unity gain.
So, put in matched output devices for Q3 and Q4 (L-L,R-R) or tweak resistor values. On my BA-1 FE, I was lucky to have a matched pair of J109V's, and matched K2013's I purchased from you buzz (thanks!), image is locked dead center.
So, put in matched output devices for Q3 and Q4 (L-L,R-R) or tweak resistor values. On my BA-1 FE, I was lucky to have a matched pair of J109V's, and matched K2013's I purchased from you buzz (thanks!), image is locked dead center.
Check voltage from D to S on the fet showing 0V. If looking at the face of the fet, it is right two pins. Then check Vgs, by probing first and last pin. Be careful. Dont short the legs. Right now, you have one fet bearing all the current in the output stage. Makes me think that you have a shorted fet. Unplug just to see if anything changes, but I dont expect it to unless you have some mis-wired connection.
All's well in the Kingdom!!
The amp has been playing on the Sunflower speakers for several hours and it sounds great. It'a a PC with JRiver Media Studio 17 feeding a Mini2496 DAC (with D. Inserra's upgrade) - through a DIY Alps Blue pot. The speaker cables are DIY Cat 6 braids.
I honestly expected a lot more volume, but it is close to both the MyRef FE and the Honey Badger. I'm suspecting when a pre with gain is added I'll hear the same super clean sound at higher SPL. I'll probably gain some body - not that there's much lacking now - with the second 500/18 trani, but the transformer sag will probably return when the extra output boards are added.
The heat is also going to be much more manageable than I had expected. You guys had me thinking there should be a fire extinguisher nearby just in case.
I put regular (quiet) 80 mm fans under the heatsink and have 60 -62 C on both the aluminum bars and the heatsink. I'll still start with the automotive style radiator for the liquid cooled build, but I think a smaller unit might be enough.
Am now in the process of ordering what's needed for a full dual mono build and look forward to major progress by the weekend.
The amp has been playing on the Sunflower speakers for several hours and it sounds great. It'a a PC with JRiver Media Studio 17 feeding a Mini2496 DAC (with D. Inserra's upgrade) - through a DIY Alps Blue pot. The speaker cables are DIY Cat 6 braids.
I honestly expected a lot more volume, but it is close to both the MyRef FE and the Honey Badger. I'm suspecting when a pre with gain is added I'll hear the same super clean sound at higher SPL. I'll probably gain some body - not that there's much lacking now - with the second 500/18 trani, but the transformer sag will probably return when the extra output boards are added.
The heat is also going to be much more manageable than I had expected. You guys had me thinking there should be a fire extinguisher nearby just in case.
I put regular (quiet) 80 mm fans under the heatsink and have 60 -62 C on both the aluminum bars and the heatsink. I'll still start with the automotive style radiator for the liquid cooled build, but I think a smaller unit might be enough.Am now in the process of ordering what's needed for a full dual mono build and look forward to major progress by the weekend.
Attachments
please , differentiate mHz and MHz
when occasional typo , it's OK ......
Hi Zen,
"A typo"It was late and I wuz getting sleepie, The ripple has gone now I also changed the bridge,I had made this one from single diodes and changed it to a 35A all in one ,added a few small caps and it's good enough for me,lol
Thanks fugly Boy,
NS
All depends how you set it up. The BA-1 article shows a 6 deep configuration with a total bias current of 3.5 amps. My own experience tells me this requires some serious hardware to deal with the heat of a monster like that.
So if I'm doing Ohm's law correctly, I'm running at .6 amps. What is the best way to change that - making the resistors smaller or increasing the potential difference with higher voltage - or both? And if done, would that require a higher VA rated transformer?
I'd enjoy doing a little research if someone knows of some good articles. I re-read some of the Pass documents but he assumes I know more than I do.
Been playing classical music all afternoon and the amp sounds very smooth and dynamic. I'd probably prefer just a little more "color" but it's really too early to tell. There are so few caps that I'm not expecting a lot of burn-in change, but I'll know more at the end of the week.
I'd enjoy doing a little research if someone knows of some good articles. I re-read some of the Pass documents but he assumes I know more than I do.
Been playing classical music all afternoon and the amp sounds very smooth and dynamic. I'd probably prefer just a little more "color" but it's really too early to tell. There are so few caps that I'm not expecting a lot of burn-in change, but I'll know more at the end of the week.
Last edited:
That's the identical product I'm using but only 3.25 inches with two 80 mm fans. I'll shoot for at least .8 amps with a total of 24 devices on the liquid build and see how much trouble I can get myself in.
What technique/procedure are you using to bump up the bias?
What technique/procedure are you using to bump up the bias?
Attachments
Last edited:
bcmbob- Your 3 deep configuration is .6 amps X 3, so 1.8 amps total. My own variation of this amplifier uses a "2 deep" setup with power JFET's/IRFP240's with .34 ohm source resistors. Bias current is just under 3 amps total after tweaking (R219 on your pcb). And my heatsinks/enclosure are definitely put to work here. My outputs are run hard, but I feel the dissipation is still reasonable.
Nelson has a good article here talking about bias: https://passlabs.com/articles/leaving-class-a
I think Figure 3 is quite interesting...
Nelson has a good article here talking about bias: https://passlabs.com/articles/leaving-class-a
I think Figure 3 is quite interesting...
Well that's a new way for me to think about configurations. So you are saying the output of an F5 would be considered 1 deep - and each additional 4 devices creates another depth -correct?
I was reading articles on the PassDiy site yesterday, but the Pass Labs has a better presentation. Some of the DIY site articles have sections/paragraphs out of sequence.
I left R219 open as it was described as optional. My entry into the BA universe was initially as a heat management project, but I like the sound enough to cycle back and try to optimize the output of the amp. I ordered some of these to try in C3 on the front end. Buzz gave me some suggestions on other caps and resistors choices I'll investigate later. Have a roll of 5/8" Kapton on the way. Even with shipping it is less expensive that the Mouser item mentioned earlier. Minimum order was 38 yards, so if anyone wants to try some, PM me for a free chunk of the remaining 430 inches. Also got an informative link on thermal interface from the ACA thread.
I was reading articles on the PassDiy site yesterday, but the Pass Labs has a better presentation. Some of the DIY site articles have sections/paragraphs out of sequence.
I left R219 open as it was described as optional. My entry into the BA universe was initially as a heat management project, but I like the sound enough to cycle back and try to optimize the output of the amp. I ordered some of these to try in C3 on the front end. Buzz gave me some suggestions on other caps and resistors choices I'll investigate later. Have a roll of 5/8" Kapton on the way. Even with shipping it is less expensive that the Mouser item mentioned earlier. Minimum order was 38 yards, so if anyone wants to try some, PM me for a free chunk of the remaining 430 inches.
Also got an informative link on thermal interface from the ACA thread.Sil Pad varies from a best of 0.32Cin²/W (silpad2000) to >1Cin²/W
None of these are as good as thin Mica.
And none get anywhere close to the current best.
F5 is a 1pair output stage. Two output devices. That what you are calling 1deep.
The F5Turbos use a variety of paralleled output devices. Two pair (4devices) is 2deep. Three pair (6devices) is 3deep. Four pair (8devices) is 4deep.
The total output stage bias current is the current through ONE device times the number of paralleled pairs used.
A ClassAB push pull amplifier has a ClassA current limit equal to the total output bias current times ~2.
A 2pr output stage with the bias set to 0.6A gives a ClassA limit of 2.4Apk
3pr gives 3.6Apk
None of these are as good as thin Mica.
And none get anywhere close to the current best.
F5 is a 1pair output stage. Two output devices. That what you are calling 1deep.
The F5Turbos use a variety of paralleled output devices. Two pair (4devices) is 2deep. Three pair (6devices) is 3deep. Four pair (8devices) is 4deep.
The total output stage bias current is the current through ONE device times the number of paralleled pairs used.
A ClassAB push pull amplifier has a ClassA current limit equal to the total output bias current times ~2.
A 2pr output stage with the bias set to 0.6A gives a ClassA limit of 2.4Apk
3pr gives 3.6Apk
The "Deep" term came from the first Burning Amp article (BA-1) and had to do with the chassis that Nelson gave away at the Burning Amp (event) a few years ago.
There were 2 chassis configurations, one longer than the other, and they held either 4 pair of devices, or 6 pair of devices. Hence "4 deep" or "6 deep". That article speaks to the differences in how the users would utilize said chassis, and gives some pointers and suggestions.
http://www.diyaudio.com/forums/diyaudio-com-articles/154777-burning-amplifier-ba-1-a.html
There were 2 chassis configurations, one longer than the other, and they held either 4 pair of devices, or 6 pair of devices. Hence "4 deep" or "6 deep". That article speaks to the differences in how the users would utilize said chassis, and gives some pointers and suggestions.
http://www.diyaudio.com/forums/diyaudio-com-articles/154777-burning-amplifier-ba-1-a.html