Fresh-T, I need 1 amp to supply 1800W, but it won't work to make it with two bridged 900W amps because I need to bridge two of those 1800W into 6 ohm amps to make 3600W into 12 ohms for each channel of my amp. You can't bridge two briged amps. I will have to somehow build a single amp, or at least multiple amps that are in phase, in parallel, to put out 1800W. I shouldn't need to put a resistor at the output of each amp because the resistors for the output transistors should make sure that each amp does the same amount of work.
If I paralleled two 900W amps, each one would supply 900W into the load independant of the other, so I don't see why they would have the stability problems of one 1800W amp. Maybe they would though. Of course if one amp failed in any way, the other would be overloaded and would soon go with it.
Is there any other options for getting my 3600W per channel? I want to use bridged amps because the load would always be about the same for each side of the power supply.
If I paralleled two 900W amps, each one would supply 900W into the load independant of the other, so I don't see why they would have the stability problems of one 1800W amp. Maybe they would though. Of course if one amp failed in any way, the other would be overloaded and would soon go with it.
Is there any other options for getting my 3600W per channel? I want to use bridged amps because the load would always be about the same for each side of the power supply.
Hi Kilowatt,
How much are you planning on investing in this project? If you can get your hands on $1200 you might be better off just buying a QSC PLX-3402 amp. This amp will give you 3400 Watts of power into a 4 Ohm load and is a really killer amp for PA use (I have it's little brother the PLX-1602, it only 1600 Watts into a 4 Ohm load, but I manage to get by).
By the way, there is no audible difference in volume between a 3400W and a 3600W amp (fractions of a dB).
Phil
How much are you planning on investing in this project? If you can get your hands on $1200 you might be better off just buying a QSC PLX-3402 amp. This amp will give you 3400 Watts of power into a 4 Ohm load and is a really killer amp for PA use (I have it's little brother the PLX-1602, it only 1600 Watts into a 4 Ohm load, but I manage to get by).
By the way, there is no audible difference in volume between a 3400W and a 3600W amp (fractions of a dB).
Phil
very high power amp(s)
hi.
maybe im missing something but isnt the idea to power 8 speaker units ( each having a 2x6 ohm coil)?
why not use 8 amps then?? 8 amps of say 1200w is a lot easier to make/build/buy than one of 10.000w 🙂
class-d / pwm amplifiers are certainly the way to go with high power amplifiers , we have some experience with these (including full-bridge off-line design giving like 5000w +).
it was mentioned earlier they had bad freq. response , i dont agree unless you mean that 10hz to 22 khz is bad 🙂
off-line power is attractive (at least for 115v) but 8 transformers or 8 switch mode supplies should work pretty well too.....and give isolation and scaling of voltage to suit your needs.
for the not-so-experienced builder amplifier boards would certainly be worth looking at / for...
bye k madsen - cadaudio.dk
hi.
maybe im missing something but isnt the idea to power 8 speaker units ( each having a 2x6 ohm coil)?
why not use 8 amps then?? 8 amps of say 1200w is a lot easier to make/build/buy than one of 10.000w 🙂
class-d / pwm amplifiers are certainly the way to go with high power amplifiers , we have some experience with these (including full-bridge off-line design giving like 5000w +).
it was mentioned earlier they had bad freq. response , i dont agree unless you mean that 10hz to 22 khz is bad 🙂
off-line power is attractive (at least for 115v) but 8 transformers or 8 switch mode supplies should work pretty well too.....and give isolation and scaling of voltage to suit your needs.
for the not-so-experienced builder amplifier boards would certainly be worth looking at / for...
bye k madsen - cadaudio.dk
km, that's right, I will be powering 8 2x6 ohm speakers - for now. I want the capability to to use three 4 ohm speakers in series, or just one big 2x6 ohm also. There are many speaker options out there that would work with the amp I want to build. That's not the main reason that I want to do it like this though, I really want to use about +/-160V for a supply so I can make the power supply transformerless and therefore light and conveniently run from just one normal 240V 50A outlet.
As far as class D goes, I know they're very efficient, but It would be too hard to design, especially if I want it to sound good.
If I can't do it any other way, I will use 8 amps, but I'd really rather not.
Haldor, is there any way I can get a schematic for that QSC PLX-3402? It couldn't be better! - class AB+B, bipolar outputs, sufficient power - GREAT! I want to build it myself so I can substitute parts for higher rail voltage, I may have to put the output transistors in series/parallel which would slighlty alter the design. Also, I'll be able to say I built it, and I might be able to do it for less than $1200.
THANKS!
As far as class D goes, I know they're very efficient, but It would be too hard to design, especially if I want it to sound good.
If I can't do it any other way, I will use 8 amps, but I'd really rather not.
Haldor, is there any way I can get a schematic for that QSC PLX-3402? It couldn't be better! - class AB+B, bipolar outputs, sufficient power - GREAT! I want to build it myself so I can substitute parts for higher rail voltage, I may have to put the output transistors in series/parallel which would slighlty alter the design. Also, I'll be able to say I built it, and I might be able to do it for less than $1200.
THANKS!
No problem Kilowatt. Here is the URL. Just clip on the "PDF's' tag then look in 'Schematics and Manuals' The full schematics can be downloaded from their website.
http://www.qscaudio.com/
QSC is one of the top three sound reinforcement amp makers in the world (Crown and Crest being the other two). These amps are used in concert settings by all the top acts around the world. You may find the schematics pretty daunting (5 pages, 11" x 17"), but there is lots of top notch engineering in this product.
This is not an audiophile amp (no single ended triods or tweaker sensibilities here), but if you are looking for massive power in a lightweight, portable package there isn't a better amp available in my opinion.
Phil
http://www.qscaudio.com/
QSC is one of the top three sound reinforcement amp makers in the world (Crown and Crest being the other two). These amps are used in concert settings by all the top acts around the world. You may find the schematics pretty daunting (5 pages, 11" x 17"), but there is lots of top notch engineering in this product.
This is not an audiophile amp (no single ended triods or tweaker sensibilities here), but if you are looking for massive power in a lightweight, portable package there isn't a better amp available in my opinion.
Phil
Thanks haldor, you're right about it being complicated, but it's pretty good. I'm sure it won't be easy to make it do what I want though. It would be nice if there was a PCB layout for it. 🙂
On second thought, what's with the output section on this thing? 4 x 2SA1302+2SC3281 wouldn't do 1600W RMS! Why do they have it like that?
[Edited by Kilowatt on 11-28-2001 at 01:17 AM]
On second thought, what's with the output section on this thing? 4 x 2SA1302+2SC3281 wouldn't do 1600W RMS! Why do they have it like that?
[Edited by Kilowatt on 11-28-2001 at 01:17 AM]
I'm afraid the PLX3402 won't work for what I need without just as much modification as the Leach SuperAmp. I could make 8 Leach amps that put out 900W with my original rail voltage and bridge every two of them to get 1800W into 24 ohms (two of my speakers in series). What differece would it make if the outputs of those 1800W sections were in parallel to give me 3600W into 12 ohms per channel, which was my original specs for the amp? Each of the 8 leech amps would have the same load as they would for 1800W into 24 ohms bridged. Would that make the amp less stable? It wouldn't be that bad just to leave it with 1800W into 24 ohms for each of 4 channels, but I'd rather not.
[Edited by Kilowatt on 11-30-2001 at 03:14 PM]
[Edited by Kilowatt on 11-30-2001 at 03:14 PM]
Kilowatt ..... here's another thought ..... if you built a suitably beefed up front end on your amps, then you could probably just 'bolt' multiple output stages onto it till you get the power you require ...... i dont see using parellel amps or output stages as a problem as long as the output devices are extremely well matched between the 2.... generally speaking, the less current you require each output device to supply, the less distortion it will produce and the less stress will be placed upon it so if you half the load impedance putting 2 amps or output stages in parellel is a good idea subject to device matching.
[Edited by AudioFreak on 11-30-2001 at 02:59 PM]
[Edited by AudioFreak on 11-30-2001 at 02:59 PM]
Ah, so I could put two slightly beefed up Leach amps in parallel for each 1800W amp. I'll probably start by trying to get 900W out of one, I can't lose too many outputs at once then. I'll have the output of each 900W amp limited so if one fails, It would be less likely to take the other with it.
Maybe I could have the output monitor cut the input amplitude in half if one 900W section failes and is cut off from the rest of the circuit so the channel would keep running if that happened, but just at 1800W instead of 3600W.
[Edited by Kilowatt on 11-30-2001 at 03:34 PM]
Maybe I could have the output monitor cut the input amplitude in half if one 900W section failes and is cut off from the rest of the circuit so the channel would keep running if that happened, but just at 1800W instead of 3600W.
[Edited by Kilowatt on 11-30-2001 at 03:34 PM]
Kilowatt,
Class D or T amps are not so hard to design as you think. See the Tripath site and, I think these amps have good sound quality.
Regards,
Class D or T amps are not so hard to design as you think. See the Tripath site and, I think these amps have good sound quality.
Regards,
My point is Kilowatt that if you do the job right, you will only need to build 2 front ends per channel (1 for the in-phase signal and one for the out of phase signal) .... 4 for 2 channels...... then you've only gotta build multiple output stages which will save you a pretty penny and can be just as reliable as complete amps if done correctly.
AudioFreak, you say that multiple output stages from one voltage/driver stage is just as reliable as two complete amps? That would make things easier 🙂
Some have said it would take a lot of experimenting to get 1800W from a Leach amp The parts for the voltage amp and driver stage and PC boards and stuff to make two 900W amps and parallel them for each 1800W section would be much less than the money I'd go through blowing stuff up trying to use just 4 front ends.
How about stability? Fresh-T said it wouldn't help, but AudioFreak implied that it would be just fine. Come on, what is it really? It seems like that would be the way to go (2 amps/1800W), but I could probably be convinced otherwise. It seems like it would be more reliable too, with the thingy I described in my last post. It would really suck to lose a whole channel of this amp!
Class T is pretty neat, I must admit, but yes, it would be hard to design and build. Making a class T amp myself (I bet no one makes an 1800W RMS chip!) would be a little like making a PC motherboard from scratch- etching a rediculously intricate PCB and soldering countless thousands of pins with machine-like precision. Sorry, not my idea of an easy project.
[Edited by Kilowatt on 12-03-2001 at 01:25 AM]
Some have said it would take a lot of experimenting to get 1800W from a Leach amp The parts for the voltage amp and driver stage and PC boards and stuff to make two 900W amps and parallel them for each 1800W section would be much less than the money I'd go through blowing stuff up trying to use just 4 front ends.
How about stability? Fresh-T said it wouldn't help, but AudioFreak implied that it would be just fine. Come on, what is it really? It seems like that would be the way to go (2 amps/1800W), but I could probably be convinced otherwise. It seems like it would be more reliable too, with the thingy I described in my last post. It would really suck to lose a whole channel of this amp!
Class T is pretty neat, I must admit, but yes, it would be hard to design and build. Making a class T amp myself (I bet no one makes an 1800W RMS chip!) would be a little like making a PC motherboard from scratch- etching a rediculously intricate PCB and soldering countless thousands of pins with machine-like precision. Sorry, not my idea of an easy project.
[Edited by Kilowatt on 12-03-2001 at 01:25 AM]
You've sure got that right Kilowatt.......
In my experience, the main problem encounted when using multiple output stages is the same problem encounted when using multiple complete amps ...... output stage device/impedance matching ...... this will need to be very close to avoid uneven power sharing .........
As far as using multiple output stages i will make a post later regarding the necessary mods that need to be made to allow this ......
The Leach Superamp may not be suitable even for this because of the power supply rail voltage you will be using .....
I will post later with more details.
Regards,
AudioFreak
p.s. now watch the inevitable torrent begin 🙂
In my experience, the main problem encounted when using multiple output stages is the same problem encounted when using multiple complete amps ...... output stage device/impedance matching ...... this will need to be very close to avoid uneven power sharing .........
As far as using multiple output stages i will make a post later regarding the necessary mods that need to be made to allow this ......
The Leach Superamp may not be suitable even for this because of the power supply rail voltage you will be using .....
I will post later with more details.
Regards,
AudioFreak
p.s. now watch the inevitable torrent begin 🙂
Kilowatt,
Go to the Tripath site, they have the complete project, kits (including PCB - not too big, by the way), pcb drawings, specifications and details for mosfet selections and, believe me, I think it would easier (a lot) than all these things you´re thinking about.
If you prefer to implement to whole design (complex mechanical mounting and full of details, in your case), go ahead with class AB or B. I think it will be a great choice too.
Regards
Go to the Tripath site, they have the complete project, kits (including PCB - not too big, by the way), pcb drawings, specifications and details for mosfet selections and, believe me, I think it would easier (a lot) than all these things you´re thinking about.
If you prefer to implement to whole design (complex mechanical mounting and full of details, in your case), go ahead with class AB or B. I think it will be a great choice too.
Regards
blmn, I didn't find anything too useful on the Tripath site, maybe I just don't know where to look. You do know that I want 1800W into 6 ohms(about +/-150V) from each of 4 amps right? Tripath's biggest, baddest, highest voltage amplifier driver, the TA0104A, is rated 2 x 500 or 1 x 1000W and +/-92V. It drives 4 ohms. In brief, it won't work for my amp the way I want it. I guess I'd go either class AB or T, but class AB looks like the easier way to go. Class T is just too digital.
For device matching, would I have to get very low tolerance output devices and/or emitter resistors? How is ON Semiconductor for device tolerances? Is there anything I can do at the bases of the output devices to account for any slight differences in the devices, like a precision resistor of some sort, maybe variable?
About how much capacitance should I have in my power supply caps on each side? It's the same design, rectified 220V mains, center tapped.
[Edited by Kilowatt on 12-05-2001 at 06:38 PM]
About how much capacitance should I have in my power supply caps on each side? It's the same design, rectified 220V mains, center tapped.
[Edited by Kilowatt on 12-05-2001 at 06:38 PM]
Kilowatt,
The baddest IC of Tripath allows +/- 100V (92 volts working in the kit) , so, with two bridged outputs over 6 ohms you can reach near 200V peak over the load or 200/1.414= 141V RMS or 141*141/6=3300W continuous (rounded). With more than 90% efficiency, you can put more than 3000W over each 6 ohms loads. I don't remember if you need 2 1800W units to put at bridged mode or not, but, if you need just 1800W per unit, you will reach this power with much less voltage than 92V. At the site you have instructions about paralleling mosfets to achieve more power than 2x500W and/or work with lower impedance loads, maybe using the kit and adapting it for more power.
However, as I told you, the AB or B way is a very interesting way. Go ahead and be happy.
Regards
The baddest IC of Tripath allows +/- 100V (92 volts working in the kit) , so, with two bridged outputs over 6 ohms you can reach near 200V peak over the load or 200/1.414= 141V RMS or 141*141/6=3300W continuous (rounded). With more than 90% efficiency, you can put more than 3000W over each 6 ohms loads. I don't remember if you need 2 1800W units to put at bridged mode or not, but, if you need just 1800W per unit, you will reach this power with much less voltage than 92V. At the site you have instructions about paralleling mosfets to achieve more power than 2x500W and/or work with lower impedance loads, maybe using the kit and adapting it for more power.
However, as I told you, the AB or B way is a very interesting way. Go ahead and be happy.
Regards
- Status
- Not open for further replies.