LM4780 clipping ?

[rgrayton]

I put together a "Brian's Kit" with parallel configuration. I guess I
do not understand something, because I expected to get over 100W with 4 ohm load (per app notes) -- only got 80W, before clipping the output.

Here is the setup and measurements:

power supply: toroidal 25v, 800VA

2ndary out: 28Vrms under load/no load

rails: 33v under 4.5A load/around 35 no-load

Load: 4x1ohm 50W

input: 1KHZ sin

measure output across load: 18Vrms(I think that is 4.5A -- 81W)
Above 18V, the output would clip.

heat sink: 4.5x7.5x1.5 (did not leave on more than 5 minutes at 80W)

I did not check the rectifier waveform, yet, but I got the same clipping driving an 8ohm load with 50W/2.5A and the rail load/no-load voltages are pretty close. What am I not understanding?
Thanks for any inputs.
-BG

[Nordic]

Been working on a 3875, but I think the princicple is the same ...
I think on using 25v secondaries one would expect rails in excess of 35V...

Also looking at the datasheet, it seems only guarantted up to 50W in either 8, 6 or 4 ohms, and 30W seems like a more reasonable load

[rgrayton]

Been working on a 3875, but I think the princicple is the same ...
I think on using 25v secondaries one would expect rails in excess of 35V...

(BG)Me, too, but that is what they measure, unless my DVM is really bad cal, but so what, the app notes for a 33v supply say 120W(unless I do not know how to read them)

Also looking at the datasheet, it seems only guarantted up to 50W in either 8, 6 or 4 ohms, and 30W seems like a more reasonable load

(BG)Which data sheet?

see page 13
http://www.national.com/ds/LM/LM4780.pdf#page=1

[Nordic]

Check the table at page 3...
under Po.... Although I must say I looked at minimum values...

@25v 4ohm -55W
@30V 6ohm -55W
@35V 8ohm 60W

[rgrayton]

wouldn't that be for each 3886? else, what configuration, parallel, bridged, single...

how do you read the 4ohm graph output vs supply for parallel config on pg13?

[Nordic]

I think thas for both, especialy if you look at the THD VS output power table, see it climbs near verticaly past 100W

If you look at the first table on page 10 it does the same at about 50W, so I would asume that graph would represent 1 chanel

[rgrayton]

Nordic,
pg 10 if for per channel, as you state.
I am using 2 in parallel, through .1ohm, so I am not sure if you can double that, but I would think the last chart on pg 13, output pwr vs supply, parallel mode for 4 ohm load would apply.
But now that we are talking about it, what if the .1 were not matched all that well, (and I would bet the difference is squared, too.) then one would get something in between.
I should be able to measure the drop across them under load and see. THX,
-BG

[carlosfm]

Originally posted by rgrayton
I guess I do not understand something, because I expected to get over 100W with 4 ohm load (per app notes) -- only got 80W, before clipping the output.

What PSU are you using?
Is it just 1.000~1.500uf caps?

[rgrayton]

Hi Carlos,
It is Brian's kit:
http://www.chipamp.com/lm4780.shtml

If I remember correctly, the last caps in the list are for the rectifier (10uf):
LM4780 Dual Mono Kit Contents
1;PCB set, consisting of 2 amplifier and 2 rectifier boards.
2;LM4780
6;22 kohm 0.5w compact metal film resistors
4;1 kohm 0.5w compact metal film resistors
4;680 ohm 0.5w compact metal film resistors
2;10 kohm 0.5w compact metal film resistors (mute)
4;0.1 ohm 3w Dale LVR3 power resistors (output)
4;2.7 ohm 2w Panasonic metal film resistors (zobel)
4;0.1 uf BC polypropylene cap (zobel)
4;1500 uF 50v Panasonic FC Capacitors
16;MUR860 On Semiconductor Diodes
4;10 uF Panasonic FC Capacitors

THX,
-BG

[carlosfm]

BG, with that low capacitance I don't think you can ever meet the power figures that are on the datasheet.
PSU voltage drops too much...