By looking at the BA-3 schematic, I assume the board as is will be sufficient to drive 6 pairs, as that is what the schematic shows, no? My real question begins with more than 6 pairs...how many pairs will the boards drive above this if any?
I am thinking of an amplifier that would be on similar scale as the F-5 Turbo or Aleph 2. I know there will be rail limitations without cascode, and since the boards dont provide for this it may be impossible.
I also know that doubling up on the J-fets will drive more, but I do not know if this allows for higher (over 32 volts) of rail voltage. What is the best, if any exists to get to a high amperage 100 watts plus without Xing it....or is this the only way?
I really like my BA-3 amp! Some time ago I was collecting parts for Aleph 2. Then F-5 Turbo came along, and I thought that was the way to go. Its "twitchiness" is a bit more than I want to deal with. I know people have done the Balanced BA-3, but I would really like a straight up two channel amp. Maybe I'm wrong here...
Russellc
I am thinking of an amplifier that would be on similar scale as the F-5 Turbo or Aleph 2. I know there will be rail limitations without cascode, and since the boards dont provide for this it may be impossible.
I also know that doubling up on the J-fets will drive more, but I do not know if this allows for higher (over 32 volts) of rail voltage. What is the best, if any exists to get to a high amperage 100 watts plus without Xing it....or is this the only way?
I really like my BA-3 amp! Some time ago I was collecting parts for Aleph 2. Then F-5 Turbo came along, and I thought that was the way to go. Its "twitchiness" is a bit more than I want to deal with. I know people have done the Balanced BA-3, but I would really like a straight up two channel amp. Maybe I'm wrong here...
Russellc
Thanks for the response Zen Mod, good to know it can drive more pairs, tops I would do is 12...more likely 6, but I would like to scale up. I dont want to mess up board with kludgy cascode, any pics or tips on doing it cleanly on stock boards? To get to 100 watts I'm sure I'll need more than 32 volts certainly, likely 40 or a little more judging from what folks are getting from T-F-5.
Thinking of Aleph 2, I guess I could use single ended output boards, same number of pairs as Aleph 2 and same/similar rail voltage...I need to figure out how to kludge on a cascode circuit, as well as calculation of values needed.
Got some learn'in to do before this point. Guess I could get one up and running on lesser rails then scale up...Maybe just do Aleph 2.
Russellc
Thinking of Aleph 2, I guess I could use single ended output boards, same number of pairs as Aleph 2 and same/similar rail voltage...I need to figure out how to kludge on a cascode circuit, as well as calculation of values needed.
Got some learn'in to do before this point. Guess I could get one up and running on lesser rails then scale up...Maybe just do Aleph 2.
Russellc
no free lunch
though , you'll get one for free ...... with 40Vdc rails , you'll soon (or immediately) have JFet launch
so - cascode them , or fry them or launch them , depending of voltage ..... they're going south either from dissipation or sheer voltage
I have no doubt that cascode is absolutely necessary, I just am not prepared yet to be able to do it on a board that does not have ready made places to do so.
When I say I will get it up and going and upgrade later, I mean for now, I will use safe voltage levels for J fets, then once I figure out how to cascode this board ,then and only then I will scale up the voltage.
Seems like I remember someone doing this in the earlier F-5X threads. I will search around and see if I can find that. I just dont want a sloppy kludge, but a neat one!
Russellc
What do you mean by "twitchiness" of F5T?
If you want 100W total, make a ~25W class A amp that goes into class AB for the final 100W?
What I mean is the number of amps that cook, most if which Ive seen are the high powered ones with the diodes. Somd have had good luck, dome not so much. I just want something a little more stable. Just my opinion, not bufning bush or anything. I have a kit I bought in a group buy, assembled boards, just not what I want is all
Thats the initial obvious path, but this for a system that will only have single ended sources, and I only have single ended preamplification. Main source is vinyl. Will I still keep these aspects you mention if using single ended RCA inputs on this Balanced BA-3? Assuming that is possible, I will need two sets of 4 matched J-fets (Which I happen to have) right? I bought these back when F-5X was beginning and figured I better get them while I could.
Also, if this BBA-3 requires higher rail voltage, I will still have the same cascode problem, or are these done at "normal" (not requiring cascode) rail voltages?
Does BBA-3 use existing store boards or are they sourced from members or what?
I may be dead wrong, but I like the idea of a massive two channel amp...like I say, I may be dead wrong.
Russellc
Last edited:
Hi Russellc,
My understanding is that the BBA3 is a bridged amp running at
'normal' rail voltage.
You need one store front end board per channel
and the mod is basically the addition of one resistor connecting
the two halves together; each half has its own output stage.
Dennis
BA-3B requires normal voltage rails. (25-32ish)
BA-3B can be driven from SE sources.
BA-3B uses normal BA-3 PCB (just doubled up, kinda) Instead of one FE board for 2 channels, each channel needs one whole FE board. And a Stereo amp's worth of output PCB becomes one channel.
http://www.diyaudio.com/forums/pass-labs/201281-burning-amp-ba-3b-balanced.html
I'll be helping a local guy here in Denver make some BA-3B monoblocks soon.
BA-3B can be driven from SE sources.
BA-3B uses normal BA-3 PCB (just doubled up, kinda) Instead of one FE board for 2 channels, each channel needs one whole FE board. And a Stereo amp's worth of output PCB becomes one channel.
http://www.diyaudio.com/forums/pass-labs/201281-burning-amp-ba-3b-balanced.html
I'll be helping a local guy here in Denver make some BA-3B monoblocks soon.
It looks like it should be fairly straightforward to convert a stereo BA3
amp into a mono BA3-b. If I'm reading it correctly, just connect the
R5 (100 ohm) resistors of the two channels together and you now have
a mono BA3-b. I guess you can even put a stereo to mono-bridged switch
there.
Dennis
amp into a mono BA3-b. If I'm reading it correctly, just connect the
R5 (100 ohm) resistors of the two channels together and you now have
a mono BA3-b. I guess you can even put a stereo to mono-bridged switch
there.
Dennis
Yes!
a 25W into 8ohms amplifier that is genuinely capable for driving a 4ohms load to 45W will deliver 90W into 8ohms when assembled as a balanced pair.
The 45W into 4ohms amp will run from approximately +-24Vdc supplies.
It will be biased to give a maximum ClassA output of your choosing. That depends on the number of devices you choose and the size of the heatsink you choose. The bias current does not depend on the figure you might want to quote to your impressionable friends.
Supplies of +-24Vdc should not need any cascoding of the jFET input stage (= simpler)
The output stage emitter resistors of an EF pair will not need voltage saving diodes (= simpler).
I'm guessing with 32 volt rails, you can cleanly swing to 27 volts or more.
Let's say 27 volts. So each 'half' gives enough voltage swing for around
91W into 4 ohms so when bridged you get twice that into 8 ohms.
Dennis
Why after 12years as a Member are you still not calculating a prediction for maximum power?
The Jfet don't have to be a quad. (Though having the pairs close-ish would be nice.)
An amp with 32V rails will, as Dennis said, cleanly swing 27V or so, making 54V pk-pk at the output.
Now with balanced you get to magically double that, bringing it up to 108V. Let's say there's a bit more losses and you can get 100V pk-pk
Output power is Vrms^2/load(in ohm)
So let's assume an 8 ohm speaker.
100*.3535 = 35.35 Vrms
35.35^2 = 1249
1249/8 = 156W into 8ohm
That is more thsn big enough for me. Not only are 2 channel amps simpler in operation and build, 1/2 the parts, but the math is less confusing as well!
That said, looks like I need another board, another transformer, and another 5U case. I may just keep my current BA-3 as is, and build the mono blocks as a separate project.
Ok, let the parts collection begin!
Russellc
That said, looks like I need another board, another transformer, and another 5U case. I may just keep my current BA-3 as is, and build the mono blocks as a separate project.
Ok, let the parts collection begin!
Russellc
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.