Jcat launches the new, 'Femto' USB card

[Summit]

2 hours ago, One and a half said:

The Jcat USB Femto is fitted to the current server build. There's a nice lift in clarity and removing grunge using the card, a very worthwhile addition and highly recommended. I did notice a variation in potential of the metal bracket and the 0V of the ATX power supply of about 12mV. Running a small braid from the metal bracket to the case of the ATX power suppl fixed that, now that voltage is <0.001mV.

Using the ATX power at the moment, just wondering if an external 5V is worthwhile considering the ATX supply has (usually) very good regulation and low impedance (it's a 900W unit, came with the case). The 0V is grounded to the incoming ground wire from the AC, but like the metal bracket, other parts like the mobo and SSD were at different potentials, braids to the frame fixed that. I don't have a scope or meter sensitive enough to see if this would affect the output, but the ears like it for sure. Simple to do.

 

ATX power are much better than many here give it credit for. To use a LPS of good quality that are separated from the rest of the computer are beneficial thou, not only because the LPS itself is better but because it prevent all the noise, emi etc from the mother board to pollute the USB card and USB signal.   

[Summit]

33 minutes ago, One and a half said:

The other issue is that are standards for ATX power supplies are built to require them to maintain voltage levels within tolerances, and for values at 5V at high currents or 1.3V, that's not so simple. For the JCAT card, the 5V is not a major problem, since the current demand is small. Wasn't there a problem with the SOtM tX-USBexp that if the 5V external failed, the card would smoke?

From Sotm

"The external power input voltage is +6.5V  +9Vdc, and the input current is 2A max.

While using the external power supply, please connect the internal 4 pin IDE power connector. This is to protect in any case if the external power get fails."

 

 

 

developer-specs-atx2_2.PDF

 

I would hope that all good PSUs are made to maintain voltage levels within tolerances. If one doesn’t it not very good in my book.  

 

Sorry I don’t know anything about the SOtM tX-USBexp and its tolerance or intolerance.

[Summit]

17 hours ago, jabbr said:

Tight voltage regulation may or may not be important in many situations. Capacitance multipliers, for example, can greatly reduce output impedance yet not involve voltage regulation. 

 

Am not sure I understand your remark to my post. Do you think that’s NOT important for a good PSUs to maintain voltage levels within stated tolerances?

[Summit]

25 minutes ago, jabbr said:

Perhaps within stated tolerance, but a tight stated tolerance is not nesessarily the mark of a good PSU.

 

ATX specifies a voltage tolerance of +/- 5% which is very reasonable. That doesn’t mean a PSU with +/- 0.00005 % is “better”

 

Depends entirely on the application. Low leakage current and low wide range output impedance, for example,  may be far more important. It all depends.

 

Yes of course other bounds and factors are impotent too, and some are of even more significance, but now the topic was PSU’s ability to maintain voltage levels within stated tolerances just like ATX power. And I stand by my original comment that “I would hope that all good PSUs are made to maintain voltage levels within tolerances. If one doesn’t it not very good in my book.”