We're not talking about the 11% error but the concept of rms Watts which I agree is meaningless. More useful maybe would be continuous power vs. crest factor.
I have a question for all you intelligent people here. While most amplifiers are single ended and rated for usually 4 or 8 ohms when you run bridged they are usually rated for 4 or 2 ohms, Now when you are going to a PBLT configuration does that mean it would be rated for 1 ohm? How do you figure the power output in the PBLT configuration if the actual load stays at 8 ohms for an 8 ohms speaker load?
You have that the wrong way around.
Typical stereo amplifiers are rated for 8/4 ohms loading single ended.
When such amplifiers are rated for BTL operation, the spec is 8 ohms loading.
The reason is that in BTL mode, with actual 8 ohms loading, each amplifier effectively sees 4 ohms loading.
There are Pro-Audio amplifiers that are rated down to 1 ohm single ended, and 2 ohms BTL operation, but these amplifiers are arguably 'overbuilt' for 8/4 ohms single ended loading.
The problems/solutions include power supply capability and output stage device SOA.
Dan.
Last edited:
Thanks Max I guess I have been misunderstanding that. But a bridged parallel amplifier wouldn't be rated for 16 ohm loads would it?
Taken from IR page for the TPA 3251
Minimum supported single ended load 3 ohms
Minimum supported bridged load 4 ohms
Minimum supported parallel bridged tied load 2 ohms.
Taken from IR page for the TPA 3251
Minimum supported single ended load 3 ohms
Minimum supported bridged load 4 ohms
Minimum supported parallel bridged tied load 2 ohms.
Since efficiency is near max at max power ---- all you need to do is look at the rear panel power consumption (max) label. Or labeled Amps given at max. And calculate the actual total power output.
THx-RNMarsh
Ok, for linear amplifiers, the ratings are actually describing the output stage current/thermal capability (and power supply capability).
That means the amplifier is capable of delivering current to a load of specified resistance long term.
Increasing the load resistance value reduces the delivered current.
This is perfectly safe in the case of linear SS amplifiers.....indeed an output stage can be loaded with a line level input (say 2000 ohms) perfectly safely.
PBTL doubles the current capability, which iow means that the load resistance value can be halved.
Also same deal, higher load resistance value (eg 16 ohms) is perfectly safe.
Things change with switching amplifier stages, because the output devices are switching and incur next to nil thermal losses.
The load resistance limiting factor is the current capability of the switching output devices/bonding wires more than thermal considerations.
Could you please give me a link to the data sheet you refer to and I will explain further.
Dan.
Hmmm, I have always found those name plate ratings to be essentially meaningless/useless.
Dan.
http://www.ti.com/lit/ds/symlink/tpa3251d2.pdf
A question is what would the power output be with the bridged parallel tied load be with a 8 ohm load?
A question is what would the power output be with the bridged parallel tied load be with a 8 ohm load?
Last edited:
2/8 x 285W = 71W.
In practice, the average parasitic losses in the output stage/filter stage reduce with reduced average output current.
This means that the expected output power would be more than the 71W calculated, but not by much, probably by a couple of percent at most.
Another way to think through this question is to consider the 2 ohm total load as four paralleled 8 ohm loads.
Each of the 8 ohm loads will dissipate 1/4 of the available power.
Hope this clears things up.
Dan.
BTW - The 285W figure is I suspect wrong and should actually be 280W.
Last edited:
I guess what confuses me is that with the single ended output you have what appears to be the same output with the four sections tied together divided by the 8 ohm load. I would think there was an increase in output with the four tied outputs vs the single output from that class-d amp. Where am I making my mistake?
For single ended mode, adding multiple paralleled output stages increases the total current capability but not the total voltage...ie the output voltage remains constant.
When running single BTL the effective output voltage doubles, but not the current capability.
PBTL doubles effective output voltage and doubles current capability.
I hope I am not confusing you now.
Dan.
Agree - if the power consumption labeling complies with electrical specification standards this will indicate the max output power.
So i would like to know, if it is true, that the impedance of a speaker drops down when feed with a short and strong impulse.
We have a swiss company, Rowen, who claims this and a guy from Dynaudio told me this also once. They talk abouth 10 % of the nominal impedance, but i don't understand enough what happens.
As Dan pointed out, adding more power devices at the output will decrerase ouput impedance (thus improving the damping factor) and will uincrease output current capability ba sharing the workload among more power devices. After a point, your amp will turn into an ideal voltage source, once the realistically available power output goes over the actual necessary current delivery. This mean the voltage will remain at its ideal nominal maximum irrespective of the work load impedance.
Also, expected thermal variances will tend to decrease as more power devices will demand bigger heat sinks, but how much bigger still mostly depends on you. Since more transistors mean less per device power dissipation, your overall THD will also tend to go down a bit, nothing spectacular, but still down.
All true and correct for linear amplifiers.
In this case Steve's confusion is with a Class-D amplifier IC - ti.com/tpa3251d2.pdf.
Dan.
labelling of Max power consumed is required by UL and other safety agencies. The amplifier max output Power cant be greater than what is drawn from the wall. useful for a reality check on mfr's claimed power output of amps...... and to size power needs/loading of the ac utility source.
-RNM
Last edited:
- Status
- Not open for further replies.