Ive got 220W RMS into 8ohms continuously out of an amp i built a few years ago using bridged LM3886T's. It ran on 42-0-42 rails.
Heat was a huge problem, it took quite a lot of experementing to work out how i could get the heat away quick enough.
AFAIK 220W RMS is the most these amps can do in any configuration into 8ohms.
Photos of the beast:
http://www.overclockers.com.au/~mwp/gallery/thumbs.php?sub=My DIY/GainClone Amp
Heat was a huge problem, it took quite a lot of experementing to work out how i could get the heat away quick enough.
AFAIK 220W RMS is the most these amps can do in any configuration into 8ohms.
Photos of the beast:
http://www.overclockers.com.au/~mwp/gallery/thumbs.php?sub=My DIY/GainClone Amp
djk said:
Thats true to a certain extent but if you look at the FM curves you will see that what you said only applies to levels really below 80 and 70dB above that and you start to require more power.
If you say take a 200 watt amp into a 88dB speaker you get 111 dB out of the speaker before clipping. This is assuming the speaker is flat down to sub sonics. If you look at around 111dB on the curves you will notice that in the 30-50hz regions that now you require more like 7-8dB for an increase to be twice as loud.
As this high power malarky is directed at large SPL you have to look at the levels you will want to be producing on the curves. Yes at 20hz you require even less but what has 20hz besides HT stuff? nowt.
200W, I thought that you can only get the 100W mentioned above like the BP200 datasheet specifies.
Are they just connected by inverting one of the chips and keeping one not inverted?
Don't the chips only deliver 50W at 8 Ohms, then 2 would be only 100W.
Please reply.
Thank you very much in advance.
Best regards.
Be excellent to one another.
Are they just connected by inverting one of the chips and keeping one not inverted?
Don't the chips only deliver 50W at 8 Ohms, then 2 would be only 100W.
Please reply.
Thank you very much in advance.
Best regards.
Be excellent to one another.
"Thats true to a certain extent but if you look at the FM curves you will see that what you said only applies to levels really below 80 and 70dB above that and you start to require more power."
My curves show 80dB~90dB in the 30hz~40hz region are 3dB~4dB apart, but your point is well taken. Matter of fact, at very high levels it looks like they are more than 10dB apart.
"Ive got 220W RMS into 8ohms continuously out of an amp i built a few years ago using bridged LM3886T's. It ran on 42-0-42 rails."
With an unregulated supply you're right, about the most you will see is 200W~250W at 8R.
"Heat was a huge problem, it took quite a lot of experementing to work out how i could get the heat away quick enough."
Yes it is, that's why the National Semiconductor patent on this bridge amp is interesting, it runs class H for much higher efficency.
"AFAIK 220W RMS is the most these amps can do in any configuration into 8ohms."
But not if you have a regulated supply. 220W/8R is 59.32V peak, you started out with 84V, most of the loss is power supply sag.
800W/4R is only 80V peak, with the FET set to 84V it is possible to swing up to 80V across the load.
My curves show 80dB~90dB in the 30hz~40hz region are 3dB~4dB apart, but your point is well taken. Matter of fact, at very high levels it looks like they are more than 10dB apart.
"Ive got 220W RMS into 8ohms continuously out of an amp i built a few years ago using bridged LM3886T's. It ran on 42-0-42 rails."
With an unregulated supply you're right, about the most you will see is 200W~250W at 8R.
"Heat was a huge problem, it took quite a lot of experementing to work out how i could get the heat away quick enough."
Yes it is, that's why the National Semiconductor patent on this bridge amp is interesting, it runs class H for much higher efficency.
"AFAIK 220W RMS is the most these amps can do in any configuration into 8ohms."
But not if you have a regulated supply. 220W/8R is 59.32V peak, you started out with 84V, most of the loss is power supply sag.
800W/4R is only 80V peak, with the FET set to 84V it is possible to swing up to 80V across the load.
Hi guys,
I tried the overture design spreadsheet from NSC, and I tried to paralel 10x3886. To my amazement, The spreadsheet says I can get more than 1kW power if the supply rails is +/-35V.
The funny thing is, there is no warning text so that means that the design is workable. Common sense tells me this is not right, but the excel files says hey, no problem!!
Can this really be done?
I tried the overture design spreadsheet from NSC, and I tried to paralel 10x3886. To my amazement, The spreadsheet says I can get more than 1kW power if the supply rails is +/-35V.
The funny thing is, there is no warning text so that means that the design is workable. Common sense tells me this is not right, but the excel files says hey, no problem!!
Can this really be done?
no limit...
Hi,
When you parallel amps like this there is no technical limit to the number. You have the normal physical issues of distributing power amongst the devices, forcing them to share the output load etc, but the techniques that work for 2 will work for 20. The assumption of course is that you have a 0.8 ohm load to get the 1000+ watt output, ie huge current, the output voltage remains the same...better stock up on 4 gauge cables, the average current in the load for that scenario is over 30amps...
If you are planning on bridging, you can get twice the voltage, any more than this and you run the risk of destroying the output trannsistors, they simply cannot control the voltage. Twice the voltage will drive twice the current into the 8ohm load and require parallel amps to support the additional current, but will ultimately deliver 4 times the power into a 'normal' 8ohm load. Hence 4-6 amps are typically used to get 200w into 8 ohms or lower.
If you are into the car scene and have 1ohm loads by paralleling subs, the parallel amps might make sense, but beware, power supply issues with this much current are a problem...
Good luck
Stuart
Hi,
When you parallel amps like this there is no technical limit to the number. You have the normal physical issues of distributing power amongst the devices, forcing them to share the output load etc, but the techniques that work for 2 will work for 20. The assumption of course is that you have a 0.8 ohm load to get the 1000+ watt output, ie huge current, the output voltage remains the same...better stock up on 4 gauge cables, the average current in the load for that scenario is over 30amps...
If you are planning on bridging, you can get twice the voltage, any more than this and you run the risk of destroying the output trannsistors, they simply cannot control the voltage. Twice the voltage will drive twice the current into the 8ohm load and require parallel amps to support the additional current, but will ultimately deliver 4 times the power into a 'normal' 8ohm load. Hence 4-6 amps are typically used to get 200w into 8 ohms or lower.
If you are into the car scene and have 1ohm loads by paralleling subs, the parallel amps might make sense, but beware, power supply issues with this much current are a problem...
Good luck
Stuart
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.