LM3875s peak currrent capability

[AndrewT]

The peak current estimate, on starting and stopping transients, I have seen quoted by a few designers is
Ipeak~=Vpeak/Nominal impedance/0.35

For 19Vac and 6ohm, this comes out at ~13Apk.

[iqbalb]

The peak current and peak power dissipation are 2 different parameters when driving a loudpeaker.

On a typical class AB output stage the output voltage and current across the output transistors do not peak simultaneosly on a sine wave for purely resistive load.
For a class AB amp peak dissipation does not occur at peak currents for a resistive load.

If you have an inductive load like the loudspeaker, the output voltage and current would have significant phase shift around the mechanical resonant frequency of the speaker.

The peak current is then limited by the DC resistance, but the peak dissipation is a completely different story.



For an inductive load...
If the amp is designed to current limit, it could clamp the output much earlier in the cycle.

If the amp is designed to limit for peak dissipation, it would start clamping at lower output powers.

[AndrewT]

Originally posted by iqbalb
...........If the amp is designed to current limit, it could clamp the output much earlier in the cycle.

If the amp is designed to limit for peak dissipation, it would start clamping at lower output powers.

A protected chipamp probably does both and also adjusts the limits in line with temperature and with duration.
Using the recommended heatsink will take the chip temperatures up near maximum (on sustained peaks) and this is likely to result in the current limits being adjusted downwards by a large margin, well below the published maxima in the datasheet.

[Dxvideo]

Originally posted by dfdye
I quite like that solution for bridged operation vs. using two different power opamps. The argument is that the two "LM3886" chips in one LM4780 package have more consistent thermal tracking than two independent chips.

I've bought two parallelled 4780 PCBs from a body from US in 2005 for 15 just bugs (I guess) I dont remember his address but if I can recall then I'll write you.. Believe me Its not worth to make by yourself, too complex! Theyre still in my drawer may be I will make them..

Originally posted by AndrewT
A protected chipamp probably does both and also adjusts the limits in line with temperature and with duration.
Using the recommended heatsink will take the chip temperatures up near maximum (on sustained peaks) and this is likely to result in the current limits being adjusted downwards by a large margin, well below the published maxima in the datasheet.

So a final question;
Does just one 225VA toroid is enough for my project?
8 x MUR860
2 x 10,000uF/63v Rubycon
2 x LT1083 (fixed to 33v)
2 x 100uF Panasonic FM
+ Onboard
8 x 1,500uF Panasonic FM
4 x 100nF MKP
2 x LM3886T (with a huge heatsink)

Yes or no?

[dfdye]

Originally posted by Dxvideo
I've bought two parallelled 4780 PCBs from a body from US in 2005 for 15 just bugs (I guess)

I think the term you are looking for is "bucks" rather than "bugs"

As for the technical stuff, that parts list should definitely build an amplifier, assuming you have the appropriate resistors. The 225VA transformer should be more than sufficient!

That being said, some of the parts don't seem necessary. Could you post a schematic of what you were wanting to build? Specifically, if you are using the 10KuF caps on the rails, I am not sure what you are using the 1500uF caps for. Also, though regulated power supplies are definitely OK for these chips, I've never had an issue running them unregulated (which I do just for simplicity).

[Dxvideo]

Sure... Fatal mistake!
Anyway,
I must say, I am inspired from Carlos a bit. His comments on regulated psu are interesting. So bought two 30+30 toroids. Now I cannot use them wihout regulator (30x1,41= alot) so I have to use that regulators or change the chipamp (may be with a TDA7293)
I put 1500uF panasonics onboard for instantaneous current needs. Because as you can see (look Andrews comment) sometimes it will need more than 10A..
Its a classical gainclone circuit. Non inverted, 20/1=20v /v gain etc.
Thx.

[sr2002]

i have a question .. im planin to build a sub amp ... to power a 4ohm sub ....

So is it possible to parallel the lm3875's for this purpose ...?

What other solution can u suggest?

I also have the lm4731 ic's ... that are supposed to be 25wattsx2 .... can these be bridged and paralled and then connected?

Cuz i will atleast need 150+watts ...

Or can u suggest some other kind of circuit?

I also have the one opa541 and one opa549....

IS there a possiblity to add some power transistors in the output stage and boost the power?

Please send a schematic of such a combo....

Im a total newb to this ...

[AndrewT]

look up the datasheet to go with the National driver amp chips.