I saw ALC measurements of FT-897, done by Dr. Juergen Friedrich - OH4FR/DJ8FR : http://www.oh4fr.de/ , so I deside to check the ALC of my FT-857D. Since i don't have an oscilloscope with memory, I look at some sound card tools and found some of them useful for this task. Later I found also an wonderfull, free, audio osciloscope software by ON7YD, check this out http://www.qsl.net:80/on7yd/software.htm, there is also other great things there, thanks Rik! Many, if not all sound cards do not respond to DC at the input and I had to deal with this firstly. I have a cheap USB audio device($3 from ebay) with CM-108 chipset by C-Media, so I deside to modify the input. It was not hard, here is the result (click for zoom): 

I have made this experimental setup to measure the ALC behavior:

  (click for zoom)

The output of TRX is connected to 30dB 100W home made attenuator(inside: 30dB pad from ebay for 2$).

The coaxial crystal detector cost a little more, but you may build one easy, not so good, but useful up to 500MHz. My old device for that job is made with 5x2W resistors in paralel, another few for the internal attenuator and one schottky barrier diode BAS70-07(I took one from old mobile phone NOKIA 3310, where it is used to measure the output level of PA). My so called detector also have output contacts for standard multimeter probes.

Now the results, note that Y axis is calibrated and show the exact values in volts. The first table show the first CW dot at 60WPM(click for zoom):

This is CW dots at 60WPM, when the output power is 10W (click for zoom):

Here is measurements on start of FM transmission (click for zoom):

All measurements in single arhive:

http://sites.google.com/site/lz4thankyou/ftp/ALCFT-857D.rar?attredirects=0&d=1

73!