Hey guys,
I've noticed that I have amassed quite a few TIP3055, TIP2955, 2N3055, MJ2955 transistors throughout the years. These are the standard 60V/15A variety.
I would like to build a power amp that can put out approximately 300-500W into a 4 ohm load with +/- 30V rails. My main problem is the lack of a proven/suitable design to work off of.
Does anyone have any ideas?
I've noticed that I have amassed quite a few TIP3055, TIP2955, 2N3055, MJ2955 transistors throughout the years. These are the standard 60V/15A variety.
I would like to build a power amp that can put out approximately 300-500W into a 4 ohm load with +/- 30V rails. My main problem is the lack of a proven/suitable design to work off of.
Does anyone have any ideas?
Hmmm. With 30V rails, your output will be some 28V peak which is 20V RMS, over 4 ohms that's 100W. So you will be needing two amps bridged to get to 300-400W. Each amp will 'see' 2 ohms load. To avoid 'gain droop' at high currents you probably want to have at least two output pairs in each amp.
Other than that, there's probably a ton of amp circuits around.
jd
You could look to some of the "golden oldies" commercial amps for inspiration.
There's a collection of them growing here: http://www.diyaudio.com/forums/soli...-models-quasi-complementary-power-output.html
There's probably schematics for some of them on Jan's site here: Welcome to Audio Circuit Denmark
There's a collection of them growing here: http://www.diyaudio.com/forums/soli...-models-quasi-complementary-power-output.html
There's probably schematics for some of them on Jan's site here: Welcome to Audio Circuit Denmark
Build a multi-rail supply and you can use up all your 3055/2955 parts.
Use one pair of MJ21195/96 on ±42V, use MJE15032/33 for drivers.
Use three pair of the plastic TIP3055/2955 in parallel, with another pair as drivers. They run on the ±84V rail.
Use whatever front end you like, use the basic Carver circuit to drive the high-side outputs.
Should do 500W4R with a suitable transformer and heatsink, uses up 8pcs of 3055/2955 per channel.
Use one pair of MJ21195/96 on ±42V, use MJE15032/33 for drivers.
Use three pair of the plastic TIP3055/2955 in parallel, with another pair as drivers. They run on the ±84V rail.
Use whatever front end you like, use the basic Carver circuit to drive the high-side outputs.
Should do 500W4R with a suitable transformer and heatsink, uses up 8pcs of 3055/2955 per channel.
Last edited:
Hi,
as suggested earlier, a pair of 150W into 2r0 amplifiers connected as bridged will give 300W into 4r0.
This can be achieved reliably with a 4pair output stage on +-30Vdc. That requires 16devices to get your 300W.
You will probably need another pair of devices as drivers due to the low gain of the 3055. Your total is now 18off 3055 and 2off 2955. There are better ways to do 300W.
Neither 400W nor 500W is possible from a +-30V supply.
as suggested earlier, a pair of 150W into 2r0 amplifiers connected as bridged will give 300W into 4r0.
This can be achieved reliably with a 4pair output stage on +-30Vdc. That requires 16devices to get your 300W.
You will probably need another pair of devices as drivers due to the low gain of the 3055. Your total is now 18off 3055 and 2off 2955. There are better ways to do 300W.
Neither 400W nor 500W is possible from a +-30V supply.
Famous Amplifier "Hadley 622" Schematic
I have start this thread:
http://www.diyaudio.com/forums/soli...-models-quasi-complementary-power-output.html
look about post#22 and #23 for an interesting circuit topology for you.
I recommend highly to avoid true-complentary buffer topologies, because the
counterpart - the MJ2955/MJE2955/TIP2955 - have not mirror-symmetric behaviour !!!
Please note, that halving of output power means only a loss of 3db by max undistorted SPL !
I have start this thread:
http://www.diyaudio.com/forums/soli...-models-quasi-complementary-power-output.html
look about post#22 and #23 for an interesting circuit topology for you.
I recommend highly to avoid true-complentary buffer topologies, because the
counterpart - the MJ2955/MJE2955/TIP2955 - have not mirror-symmetric behaviour !!!
Please note, that halving of output power means only a loss of 3db by max undistorted SPL !
Last edited:
Bridged in 4 ohm, just about 400W. Only just with 28V out. Not sure it would work with 3055 on 30V, so you may be right.
jd
But are there better ways to use up a bunch of low voltage transistors? Unless you plan on building a dozen (or two) gel-cell chargers, they'll sit for years while working on more exciting things.
How 'bout this...
Cascode S-Leach style output using two series connected banks of 3x 2N3055driven by a TIP31 (and a 60V-selected 2N3904 predriver for a triple). PNP's would make the minus side easy, but of course you'd need an equal number of 2955's MJ802's, etc. Heck, mix-n-match them - just make sure parallel banks are the same type/date code. We're not looking at .00007% distortion anymore anyway. You could try an all-NPN but don't say I didn't warn you about sziklai's not tolerating non-zero rail impedance (the cascode will oscillate). They may work for the slaves, but not the masters! Use up that stash of old video outputs for the VAS's. Then bridge it. Three 24V high current surplus transformers in series to power it - 100 volt single rail. Even with the rail dropping to 65V under load you'll still get 400 watts at 4 ohms.
Thanks for the responses guys!
I realize that the TIP and the 2N transistors are "old technology" but I have 25+ pairs of each type
; I also figured that a power amp with an array of retro transistors would also be kinda interesting. Are there any easily modifiable designs out there that I could build? This is my first discrete amp and I don't feel confident enough to design my own at the moment...
Any info would be appreciated!
I realize that the TIP and the 2N transistors are "old technology" but I have 25+ pairs of each type
; I also figured that a power amp with an array of retro transistors would also be kinda interesting. Are there any easily modifiable designs out there that I could build? This is my first discrete amp and I don't feel confident enough to design my own at the moment...Any info would be appreciated!
Sziklai pairs hold an advantage here. It can hide both the base current
and the base voltage swings of these old dino-transistors into the total.
Slap a Schottky and .1R in series with the effective emitter of this pair.
Gives a nice Squarelaw curvature, independant the clunker's curves and
thermals. Wether or not an MJ and 2N behave as good compliments, two
Schottkys far away from the heat will.
and the base voltage swings of these old dino-transistors into the total.
Slap a Schottky and .1R in series with the effective emitter of this pair.
Gives a nice Squarelaw curvature, independant the clunker's curves and
thermals. Wether or not an MJ and 2N behave as good compliments, two
Schottkys far away from the heat will.
Quality not quantity!!! Use all your XX2955s and XX3055s to build dozens of JLH 10 watt class A amps! Theycan be bridged, built in complimentary form etc.......the transitors can be used in the power supply.......and build a more sensitive pair of speakers.
Edit. just noticed you're in Canada......the JLH design also double as room heaters. Its a win-win!
And we've enjoyed the Winter Olynpics too.....Australia as usual boxing above its weight! TWO Gold!
Jonathan
Edit. just noticed you're in Canada......the JLH design also double as room heaters. Its a win-win!
And we've enjoyed the Winter Olynpics too.....Australia as usual boxing above its weight! TWO Gold!
Jonathan
Last edited:
Sorta what I meant... Take the 2N3055 base currents and thermals
out of the decision making process.
And this may look like JLH split current drive. But due to big leaky
bypass resistors, behaves more like folded concertina, split voltage.
Important distinction if you want the Schottky square law voltage
drops to have a say in the output currents.
Also big leaky base to emitter resistor, help put the brakes on that
2NLedSled.
out of the decision making process.
And this may look like JLH split current drive. But due to big leaky
bypass resistors, behaves more like folded concertina, split voltage.
Important distinction if you want the Schottky square law voltage
drops to have a say in the output currents.
Also big leaky base to emitter resistor, help put the brakes on that
2NLedSled.
Attachments
IIRC, the forum software does not "allow" usernames with characters other than letters and numbers; I'm not sure why I was even allowed to sign up with the username. I guess to forum seems to be linking my profile to anyone with "db" in their profile name. I sent a PM to the admin a long while back and I never got a response; either ways I'm not too concerned...
As for the amplifier, I would really prefer something that uses both the xx3055 and it's compliment, the xx2955. I have looked into building several JLH amps and bridging them but the amount and cost of parts required is more than I want to pay...
I am looking into using the LM4702 chip to drive 14-20 pairs of transistors. Since the voltage I intend to use is very close to the lm4702's minimum, will this cause any problems? Should I use a dual rail supply instead?
Any and all comments/suggestions are greatly appreciated!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.