Keyboard Shortcuts
ctrl + shift + ? :
Show all keyboard shortcuts
ctrl + g :
Navigate to a group
ctrl + shift + f :
Find
ctrl + / :
Quick actions
esc to dismiss
Likes
Search
UQRSS runs much too slow
|
Stefan Teuscher
Hello all,
first many thanks to Hans G0UPL for the quick delivery of the kits. The first one I have built up yesterday, but struggling with a problem: The MCU seems to run much too slow. This is indicated by the clock in the display, which changes the seconds after 6 or 7 seconds (and not every second which would be normal), and all transmissions are much too slow. E.g. the WSPR-transmission frame lasts not nearly 2 minutes but lasts about 12 minutes! I have no GPS (I have just ordered one). I have checked the oscillator, seemms to be ok (8 Vpp), also the signal at pin 9 of the MCU: TTL-level 10.140.200 Hz (after I have changed C5 from 22 pF to 4,7 pF otherwise I could not reach 10.140.200 - but that cannot be the reason for this behavior because it was the case just before the change of C5). The MCU has the version V1.05. Because I have ordered two kits I tried both MCUs - but they behave identical. Thats my config: Speed: 001 Frequency: 10,140,200 Callsign: DL4GCS Message: 73 Frame: 02 Start: 00 CW Ident: 00 Locator: JN47 Power: 23 Use GPS: off GPS Baud: 09600,01,1,8 AF Output: off AF Frequ: 1300 Inv. FSK off FSK (Hz): 4 FSK Adj: 1,300 Cursor: 4 Has anyone any idea? 73s Stefan, DL4GCS |
|
Gordon Kennedy
Just got mine last week- fired it up today and got the same situation- way too slow. Every thing physically looks good but is beaving exactly the same as yours. ? 73, Gord VE3GKN From: Stefan Teuscher
To: QRPLabs@... Sent: Sunday, September 16, 2012 5:11:04 PM Subject: [QRPLabs] UQRSS runs much too slow Hello all, first many thanks to Hans G0UPL for the quick delivery of the kits. The first one I have built up yesterday, but struggling with a problem: The MCU seems to run much too slow. This is indicated by the clock in the display, which changes the seconds after 6 or 7 seconds (and not every second which would be normal), and all transmissions are much too slow. E.g. the WSPR-transmission frame lasts not nearly 2 minutes but lasts about 12 minutes! I have no GPS (I have just ordered one). I have checked the oscillator, seemms to be ok (8 Vpp), also the signal at pin 9 of the MCU: TTL-level 10.140.200 Hz (after I have changed C5 from 22 pF to 4,7 pF otherwise I could not reach 10.140.200 - but that cannot be the reason for this behavior because it was the case just before the change of C5). The MCU has the version V1.05. Because I have ordered two kits I tried both MCUs - but they behave identical. Thats my config: Speed: 001 Frequency: 10,140,200 Callsign: DL4GCS Message: 73 Frame: 02 Start: 00 CW Ident: 00 Locator: JN47 Power: 23 Use GPS: off GPS Baud: 09600,01,1,8 AF Output: off AF Frequ: 1300 Inv. FSK off FSK (Hz): 4 FSK Adj: 1,300 Cursor: 4 Has anyone any idea? 73s Stefan, DL4GCS ------------------------------------ Yahoo! Groups Links <*> To visit your group on the web, go to: ? ? <*> Your email settings: ? ? Individual Email | Traditional <*> To change settings online go to: ? ? ? ? (Yahoo! ID required) <*> To change settings via email: ? ? QRPLabs-digest@... ? ? QRPLabs-fullfeatured@... <*> To unsubscribe from this group, send an email to: ? ? QRPLabs-unsubscribe@...m <*> Your use of Yahoo! Groups is subject to: ? ? |
|
Stephen Farthing
Hi Guys,
I'll wait for Hans to comment on this as he designed the hardware and wrote the software. I think he might be taking a short break so please be a little patient. If anyone else has a problem it would be helpful if you could quote :-
1. The version of the software - this is displayed on the LCD at start up. 2. The parameters you have entered into the kit.? I should be able to find some time later to see if I can duplicate the problem.
Apologies, Steve |
|
"rowsbyt"
Steve,
Same here! Received my kit last week. The clock as displayed on the LCD runs ~8 times slow. Using QRSS with speed 5 it takes 40 seconds to send a dot. In addition the FSK pin of the uP ((p15) has a continuous 20Hz 5v p-p square wave on it giving a rough tone - it is impossible to set up using ARGO as the (basic) bandwidth is too high! (The note is clean if I short out the LED!) V1.05 Settings: Mode: QRSS Speed: 005 Frequency: 10,140,00 Callsign: G4CQS Message: 5 Frame: 02 Start: 00 CW Ident: 00 Locator: IO82 Power: 23 Use GPS: off GPS Baud: 090600,01,1,8 AF Output: off AF Frequency: 0,800 Inv FSK: off FSK (Hz): 5 FSK Adjust: 00,400 Cursor: 4 Regards, Tony G4CQS |
|
Tony Volpe
That's a very strange thing to happen. I just built a new one (I had one already) and it is the same issue of the kit as yours, but it behaves perfectly.
toggle quoted message
Show quoted text
The settings look normal. I replicated Stefan's SPEED = 1, which was slower than either of my kits have ever been, and it made no difference.
The behaviour seems to be as if you had a bad crystal that ?was operating on a lower frequency and you were tuning in on a higher harmonic of the 'actual' frequency it was wrongly operating on when you checked on your receiver. This may be impossible. I think the crystal is supposed to be a fundamental type, but IF that was happening, the MCU would be using a slower clock frequency than it ought to be on and miscalculating the time. I can't think of anything else, and I am sure you have both made meticulous checks under magnification to see that there are no bad joints or solder bridges on the board around the oscillator and the MCU.?
If my kit was doing this, I'd tune lower on a receiver and see if I could find a strong output at a lower frequency such as half of 10.140, or one third of it and so on. This may be a total red herring - if so, I am sorry, but what else but a slow clock frequency could make the system operate at this slow rate?
Good luck anyway, and I hope my guesswork doesn't lead you up the wrong path. G0BZB Tony On 17 September 2012 07:12, Stephen Farthing <squirrox@...> wrote:
|
|
"rowsbyt"
Tony,
toggle quoted message
Show quoted text
Tried that! Using a 'scope & counter the xtal oscillator is definitely running at 10MHz! Tony G4CQS --- In QRPLabs@..., Tony Volpe <tony.volpe.1951@...> wrote:
|
|
Tony Volpe
Hi Tony,
toggle quoted message
Show quoted text
Ok - looking at the one eighth speed your kit is running at, for my suggestion to be right the oscillator would have to be on?1.2675mhz , which it obviously isn't. That would leave a software problem as the issue as far as I can see, but I and other people have V1.05 that works normally. Unless there are different versions of v1.05, I'm stuck.
tony G0bzb On 17 September 2012 09:03, rowsbyt <g4cqs@...> wrote:
|
|
"David Barber"
Perhaps an obvious question but did you check the signal at the CLK pin of
toggle quoted message
Show quoted text
the processor? (Pin 9 I believe) *** ________________________________________ From: QRPLabs@... [mailto:QRPLabs@...] On Behalf Of Tony Volpe Sent: 17 September 2012 09:15 To: QRPLabs@... Subject: Re: [QRPLabs] Re: UQRSS runs much too slow Hi Tony, Ok - looking at the one eighth speed your kit is running at, for my suggestion to be right the oscillator would have to be on?1.2675mhz , which it obviously isn't. That would leave a software problem as the issue as far as I can see, but I and other people have V1.05 that works normally. Unless there are different versions of v1.05, I'm stuck. tony G0bzb On 17 September 2012 09:03, rowsbyt <g4cqs@...-online.co.uk>
wrote: ? Tony, Tried that! Using a 'scope & counter the xtal oscillator is definitely running at 10MHz! Tony G4CQS |
|
"lowemmission"
I am running 2 UQRSS, ver 1.04 & 1.05, both have the same parameters and are configured for 30m WSPR. ver 1.05 has now been running for 4 days with no problem, both with GPS. After reading these posts I changed 1 parameter in v 1.05 of Callsign: DL4GCS to Message: 73
toggle quoted message
Show quoted text
Frame: 02 and it ran slow. GPS disabled. Maybe consider your message statements. Only an observation. lowemmission@ --- In QRPLabs@..., Gordon Kennedy <ve3gkn@...> wrote:
|
|
cvest@...
My latest order arrived today , I'll try to get it put together tonight and see how it behaves . I have 2 others to compare it to .
toggle quoted message
Show quoted text
Charlie , W5COV ---- Stephen Farthing <squirrox@...> wrote: Guys, |
|
Stephen Farthing
Guys,
toggle quoted message
Show quoted text
My theory is that I have set the clock divide by 8 fuse accidentally when I programmed a batch of chips. This would account for the fact that some versions of 1.05 work and some have this timing error. Unfortunately I have never found a way to get AVR studio to save the programming parameters so I have to set it up manually each time I program a batch of chips.?
So what you are ?seeing might be the results of such an error. I need to run some tests to verify this but I don't have the time to do this for a couple of days.? If this is the case I'll replace your chips with ones that have been programmed correctly FOC.
73s Steve On 17 September 2012 19:42, <cvest@...> wrote: My latest order arrived today , I'll try to get it put together tonight and see how it behaves . I have 2 others to compare it to . |
|
Stefan Teuscher
Hello Steve,
toggle quoted message
Show quoted text
hello all, many thanks, Steve, to detect this error that fast. Yes, this could be the case. I didn't measure the time exactly how slow the MCU runs, but factor 8 could match. Another observation I've made: Sometimes the MCU forgets its time settings, starting with 00:00 when stopping the transmission by pressing the key S2. But sometimes the time goes on - slowly, but it goes on. Also the effect that Tony (G4CQS) reported was exactly the same here. There were lots of spurious signals with a distance of multpies of about 20 Hz left and right of the main signal. That's why 20 Hz is the fundamental frequency of the PWM-signal at pin 15 of my MCU. For such a low frequency the low pass filter (R2, C2) is much to small. So it does not smoothing enough, modulating sidelines of +- n * 20 Hz. I think, if the MCU runs correct, i.e. factor 8 faster, this sidelines would be around +- n * 160 Hz, but filtered out by the low pass filter (R2, C2). And so these sidelines are only another effect of a much too slow running MCU. Tony, G0BZB: Thanks for your tries. The same thoughts that I had yesterday. But with the scope I could see that the oscillator is really running right and the signals are ok. And I really also checked all the frequency band for other strong signals - but nothing found. So the clock signal for the MCU was probably correct. @lowemission: I don't think that it is the callsign or the message. I also tried other messages, nothing changed. Perhaps it works correct with a GPS. Unfortunately I cannot test it because at the moment I have none. So I think it is probable like Steve wrote, a little lapse when programming the MCU. And looking forward for a correction. But, Steve and Hans, don't hurry. It's a hobby, but a nice one. I would like to pay the postage at least. 73s, Stefan, DL4GCS Am 17.09.2012 21:19, schrieb Stephen Farthing: Guys, |
|
Gordon Kennedy
Just tried the second unit I ppurchased at the same time and same slow speed.
?
73, Gord
From: Stefan Teuscher
To: QRPLabs@... Sent: Monday, September 17, 2012 5:52:21 PM Subject: Re: [QRPLabs] Re: UQRSS runs much too slow ?
Hello Steve, hello all, many thanks, Steve, to detect this error that fast. Yes, this could be the case. I didn't measure the time exactly how slow the MCU runs, but factor 8 could match. Another observation I've made: Sometimes the MCU forgets its time settings, starting with 00:00 when stopping the transmission by pressing the key S2. But sometimes the time goes on - slowly, but it goes on. Also the effect that Tony (G4CQS) reported was exactly the same here. There were lots of spurious signals with a distance of multpies of about 20 Hz left and right of the main signal. That's why 20 Hz is the fundamental frequency of the PWM-signal at pin 15 of my MCU. For such a low frequency the low pass filter (R2, C2) is much to small. So it does not smoothing enough, modulating sidelines of +- n * 20 Hz. I think, if the MCU runs correct, i.e. factor 8 faster, this sidelines would be around +- n * 160 Hz, but filtered out by the low pass filter (R2, C2). And so these sidelines are only another effect of a much too slow running MCU. Tony, G0BZB: Thanks for your tries. The same thoughts that I had yesterday. But with the scope I could see that the oscillator is really running right and the signals are ok. And I really also checked all the frequency band for other strong signals - but nothing found. So the clock signal for the MCU was probably correct. @lowemission: I don't think that it is the callsign or the message. I also tried other messages, nothing changed. Perhaps it works correct with a GPS. Unfortunately I cannot test it because at the moment I have none. So I think it is probable like Steve wrote, a little lapse when programming the MCU. And looking forward for a correction. But, Steve and Hans, don't hurry. It's a hobby, but a nice one. I would like to pay the postage at least. 73s, Stefan, DL4GCS Am 17.09.2012 21:19, schrieb Stephen Farthing: > Guys, > > My theory is that I have set the clock divide by 8 fuse accidentally when I > programmed a batch of chips. This would account for the fact that some > versions of 1.05 work and some have this timing error. Unfortunately I have > never found a way to get AVR studio to save the programming parameters so I > have to set it up manually each time I program a batch of chips. > > So what you are seeing might be the results of such an error. I need to > run some tests to verify this but I don't have the time to do this for a > couple of days. > > If this is the case I'll replace your chips with ones that have been > programmed correctly FOC. > > 73s Steve > > |
|
Stefan Teuscher
Hello Steve,
hello Hans, many thanks for the replacement of the too slow chips. In the meantime I have got a GPS also, so there is only an antenna missing. The new chip seems running correct. But I had to remove capacitor C5 completely. Otherwise with "GPS on" some strange things happen when the chip can't tune to the frequency set in the parameters, but tries and tries and tries. Without C5 I can tune with C8 in test mode (GPS on) nearly exactly to 10.240.200 displayed and then the frequency locked loop (I think it is a sort of it) locks correctly. Now I have one UQRSS running in test but only via coax cable fed directly into the receiver (Rohde+Schwarz ESV) which measures exactly 128 dB???V at 50 Ohms = 21 dBm = 126 mW (5 Volt power supply also for the PA). I want to run it first for a while in this mode and in the meantime thinking about a suitable antenne. Many thanks Stefan DL4GCS |
to navigate to use esc to dismiss