[hpsdr] Mercury 2.6

Mon Feb 16 19:49:23 PST 2009

Hi Phil, Bill, All,

I made two strange observations:
I was testing at a sample rate of 192 Ks/s, buffer size 1024.
A lot of distortion.
After having set the buffer size to 2048 it was even more distorted.

I set the process priority to the LOWEST setting "low" (!).

Guess what: the distortion was A LOT LESS than at higher process priority
settings!?

I also found that after changing the sample rate or buffer size, the process
priority changed back to the default "high" setting?!
Is this OK/meant to be?
It also works this way when changing VAC sample rate settings.

I wonder if this is the case with others or just in my setup.

I am now working at the lowest sample rate of 48K / buffer 2048 and at a
process priority of "low". 
Processor load is 6 - 12% on an AMD 2000+ / 512 KB, WinXP Home, SP3 (!)
It sure is a lot more quiet now but I have to do more listening tests.
It seems that other processes loading the CPU are interrupting PowerSDR more
though (I guess because of lower process priority)

Henry.

> -----Original Message-----
> From: hpsdr-bounces at lists.hpsdr.org 
> [mailto:hpsdr-bounces at lists.hpsdr.org] On Behalf Of Phil Harman
> Sent: dinsdag 17 februari 2009 1:51
> To: Doug Bade
> Cc: hpsdr at lists.hpsdr.org
> Subject: Re: [hpsdr] Mercury 2.6
> 
> 
> ***** High Performance Software Defined Radio Discussion List *****
> 
> 
> Hi Doug,
> 
> Thanks for the offer but we are close to a new version of code that 
> should fix the problem.
> 
> With the Atom 330 does the USB controller share an interrupt with 
> another device?
> 
> Phil....
> 
> 
> 
> Quoting Doug Bade <kd8b at thebades.net>:
> 
> > The Atom 330 intel mini itx boards all seem to do it....at least 
> > those of us running the HPSDR this way who have compared notes find 
> > this.. Even 96K at 2048 no longer works it distorts as he 
> describes.. 
> > interestingly the CPU load is only about 12-18 % even when it is 
> > popping and distorting... I have had to drop to 96k 1024 to 
> get it to 
> > work 99% as it still pops once in a while.. 96k 512 is good 
> audio but 
> > cpu load starts climbing..... Not complaining but it has 
> gotten worse 
> > with later releases of code. Would it help to have one of these 
> > boards shipped to one of you for tests??? they are not real 
> > expensive....and small enough to ship cheap..
> >
> > I would volunteer mine as you have your own Mercury etc... :-) The 
> > board runs on 160w ps so about anything you have laying around runs 
> > it...
> >
> > Doug
> > KD8B
> >
> >
> > At 07:05 PM 2/16/2009, Phil Harman wrote:
> >> ***** High Performance Software Defined Radio Discussion List *****
> >>
> >>
> >> Hi Jeff,
> >>
> >> Thanks for your tests.  We are working hard on tracking down the 
> >> reason for the popping at 192k and 2048 buffer. Neither Bill, 
> >> KD5TFD, nor I can reproduce the problem which makes fixing it a 
> >> little harder.
> >>
> >> I've never been able to run at 192k with 256 buffers with any 
> >> version of the code.
> >>
> >> 73's Phil...VK6APH
> >>
> >>
> >>
> >>
> >> Quoting Jeff Cook <jeffrie at talktalk.net>:
> >>
> >>> ***** High Performance Software Defined Radio Discussion 
> List *****
> >>>
> >>> These are my observations regarding Mercury 2.6
> >>>
> >>>                                              _Primary_
> >>>
> >>> Sample Rate             Buffer Size                    Result
> >>>
> >>> 192,000                         2048
> >>>      Popping
> >>>
> >>> 192,000                         1024
> >>>      Good
> >>>
> >>> 192,000                         512
> >>>       Good
> >>>
> >>> 192,000                         256
> >>>       Break up
> >>>
> >>>
> >>>                                                _VAC_
> >>>
> >>> 192,000                         2048
> >>>      Gross Distortion
> >>>
> >>> 192,000                         1024
> >>>      Popping
> >>>
> >>> 192,000                         512
> >>>       popping
> >>>
> >>> 192,000                         256
> >>>       Gross Breakup
> >>>
> >>> 96,000                           Any
> >>>       Good
> >>>
> >>> In my view HPSDR is the best thing to happen in ham radio 
> for years.
> >>> It, together with the software offers a degree of control 
> that I've 
> >>> never experienced in a radio before.
> >>>
> >>> Jeff Cook G0AFQ.
> >>>
> >>> _______________________________________________
> >>> HPSDR Discussion List
> >>> To post msg: hpsdr at hpsdr.org
> >>> Subscription help: 
> http://lists.hpsdr.org/listinfo.cgi/hpsdr-hpsdr.org
> >>> HPSDR web page: http://hpsdr.org
> >>> Archives: http://lists.hpsdr.org/pipermail/hpsdr-hpsdr.org/
> >>
> >>
> >>
> >> ----------------------------------------------------------------
> >> This message was sent using IMP, the Internet Messaging Program.
> >>
> >> _______________________________________________
> >> HPSDR Discussion List
> >> To post msg: hpsdr at hpsdr.org
> >> Subscription help: 
> http://lists.hpsdr.org/listinfo.cgi/hpsdr-hpsdr.org
> >> HPSDR web page: http://hpsdr.org
> >> Archives: http://lists.hpsdr.org/pipermail/hpsdr-hpsdr.org/
> >
> >
> 
> 
> 
> ----------------------------------------------------------------
> This message was sent using IMP, the Internet Messaging Program.
> 
> _______________________________________________
> HPSDR Discussion List
> To post msg: hpsdr at hpsdr.org
> Subscription help: http://lists.hpsdr.org/listinfo.cgi/hpsdr-hpsdr.org
> HPSDR web page: http://hpsdr.org
> Archives: http://lists.hpsdr.org/pipermail/hpsdr-hpsdr.org/
> 

 1234842563.0