Adaptec 29160 + Fujitsu MAS3367NC - cannot force U160

[admin at cordef.net.pl]

Justin T. Gibbs napisał(a):
> The jumpers, when populated, force termination to be enabled.  You 
> shouldn't need
> them:
>
> http://adaptec-tic.adaptec.com/cgi-bin/adaptec_tic.cfg/php/enduser/std_adp.php?p_faqid=1092&p_created=949587020&p_sid=oPDooumi&p_lva=&p_sp=cF9zcmNoPTEmcF9zb3J0X2J5PSZwX2dyaWRzb3J0PSZwX3Jvd19jbnQ9MjImcF9wcm9kcz0wJnBfY2F0cz0wJnBfcHY9JnBfY3Y9JnBfc2VhcmNoX3R5cGU9YW5zd2Vycy5zZWFyY2hfbmwmcF9wYWdlPTEmcF9zZWFyY2hfdGV4dD0yOTE2MCBqdW1wZXI*&p_li=&p_topview=1 
>
>
> The only other guess I have is that the "force-se" jumper is populated 
> on both of your
> drives.  You could verify that the 29160 believes the bus is in SE 
> mode by instrumenting
> aic7xxx.c:ahc_devlimited_syncrate().  The first if block is where the 
> driver checks the
> operating mode of the transceiver and limits the negotiation accordingly.
>
> -- 
> Justin
Well, I don't think devices are forced to SE, at lest not because of 
jumper setting. I checked - on IBM drive this jumper is opened and 
Fujitsu doesn't have one at all. The only one source of problems I can 
see is adapter. It has the DIFFSNS signal left not connected. 
Theoretically it means device is set to differential. But what if 
controller forces it to low? Tomorrow I'll try to connect it to the +5V 
and see what will happen.
Justin, let me ask you one more question: I wonder where is the setting 
"GOAL" taken from? I understand that "USER" is a maximum possible 
transfer for the interface and  "CURR" is maximum possible transfer for 
the device (calculated at DV stage). Is this setting connected with 
state of bus - i.e. if bus is SE then GOAL is forced to be 40.0 instead 
of 160.0? Or it is set somewhere in device EEPROM?