[hpsdr] Penelope PTT

Gerd Loch g.loch at nt-electronics.de
Tue Oct 14 08:30:18 PDT 2008


Phil and Bill,

I have meanwhile installed Windows XP but I still have the audio pops with
the same frequency as before.
I am using a Motherboard Asrock AM2NF3-VSTA with an AMD Athlon 2400.
Any further ideas how to eliminate the pops?

73, Gerd
DJ8AY

-----Original Message-----
From: Phil Harman [mailto:phil at pharman.ohirg] 
Sent: Wednesday, August 20, 2008 3:14 AM
To: g.loch at nt-electronics.de
Cc: 'Bill Tracey'; hpsdr at hpsdr.org
Subject: Re: AW: [hpsdr] Penelope PTT



Gerd,

You need to fix your popping audio on receive problem. Whilst a timeout 
in the Penelope code will paper over the PTT problem you can't provide 
a similar solution for dropped USB data on transmit - this will be a 
real problem on CW.

Find the reason for the dropped data and you don't need the "fix" in 
Penelope. I've found that running PowerSDR under XP in RealTime on an 
old slow PC I can use 192kHz sampling without missing USB data.

73's Phil...VK6APH




Quoting Gerd Loch <g.loch at nt-electronics.de>:

> Hi Bill,
>
> I really like your objective answer.
> I already supposed that your intention was to protect the power 
> amplifier which is of course reasonable. However I believe that a 
> moderate timeout of about a second would not harm the hardware in case 
> of ...
>
> Probably I will follow your advice and change to XP. I have it here 
> for about a year but I hate to reinstall a lot of programs and 
> utilities with which it always ends up when changing the os. I rely on 
> your statement that the USB2 stack has a much better performance with 
> XP.
>
> 73, Gerd
> DJ8AY
>
>
>
> -----Ursprüngliche Nachricht-----
> Von: Bill Tracey [mailto:bill at ewjt.com]
> Gesendet: Dienstag, 19. August 2008 20:58
> An: g.loch at nt-electronics.de; 'Phil Harman'
> Cc: hpsdr at hpsdr.org
> Betreff: Re: [hpsdr] Penelope PTT
>
>
> You're not going to like my answer, but I'd say use Windows XP.  The 
> USB2 stack in Win2K is a bolt on and it's performance is miserable.  
> During the original development of Ozy and Janus I ended up moving to 
> XP to get away from perf issues in the Win 2k USB2 stack.  Moving to 
> XP made things much better.
>
> The reason for making PTT dependent on HAVE_SYNC was to avoid leaving 
> an external amp keyed down in the case of the PC hanging up and going 
> away.  We're trying to have the hardware do the right/safe thing even 
> if the PC blows up.  That being said, I don't think  we need to 
> declare the PC broken after 1 lost sync, think adding some sort of 
> moderate timeout in there to deal with a occasional dropped frame 
> would be a good thing.
>
> One other suggestion - try running PowerSDR at higher than normal or 
> real time priority (there's a drop down for priority in the setup 
> pages).
>
> Regards,
>
> Bill (kd5tfd)
>
> At 01:55 AM 8/19/2008, Gerd Loch wrote:
>> ***** High Performance Software Defined Radio Discussion List *****
>>
>> Hi Phil,
>>
>> Of course you are right that this will only cure the issue but I 
>> doubt that you can totally avoid dropouts in the system Windows-USB. 
>> I am running W2k and my cpu load ist somewhere about 25% and I am not 
>> willing to run PowerSDR exclusively on this PC. I always had poppings 
>> and I do not know how I could avoid them in practice. Maybe you can 
>> reduce the number per timeunit. USB is not realtime. On the other 
>> side what is the practical use of making PTT dependent on HAVE_SYNC 
>> so that this workaround would not make sense?
>>
>> Can you offer a concept how to totally avoid the fifo running empty?
>>
>> 73, Gerd
>> DJ8AY
>
>
>



----------------------------------------------------------------
This message was sent using IMP, the Internet Messaging Program.




More information about the Hpsdr mailing list