|
I found that when I I feed a higher amplitude 20 hz sine wave into the FM mod input with the coupling mode set to AC, I get a warning message stating that A6A1 has a transient failure F3. I turn down the audio level and the message goes away after I clear it. At lower amplitudes this doesn't seem to be a problem but as I increase the amplitude again, that message appears every time. At first I thought I had a module failing but I found when I switched to DC coupling mode this problem goes away, instead of flashing A6A1 transient failure I get just a HI indicator at 20 HZ. Higher frequencies in AC coupled mode doesn't seem to reproduce this problem. Perhaps if anyone could reproduce this on a 8642B that happens to be sitting by.
|
I think what you are seeing is that in AC mode, the FM PLL is being asked to generate too large of a phase deviation which knocks the loop out of lock causing the transient error. ?Switching to DC opens the PLL and just FMs the VCO. ?Since the peak phase deviation is equal to the peak frequency deviation divided by the modulating frequency, reducing the peak frequency deviation or increasing the modulating frequency should eliminate the transient error. If the specifications indicate you should be able to achieve the deviation you are setting at 20Hz in AC mode then perhaps the FM VCO center frequency has shifted.
Jim
toggle quoted message
Show quoted text
On Nov 3, 2018, at 8:12 PM, wilson2115@... wrote: I found that when I I feed a higher amplitude 20 hz sine wave into the FM mod input with the coupling mode set to AC, I get a warning message stating that A6A1 has a transient failure F3. I turn down the audio level and the message goes away after I clear it. At lower amplitudes this doesn't seem to be a problem but as I increase the amplitude again, that message appears every time. At first I thought I had a module failing but I found when I switched to DC coupling mode this problem goes away, instead of flashing A6A1 transient failure I get just a HI indicator at 20 HZ. Higher frequencies in AC coupled mode doesn't seem to reproduce this problem. Perhaps if anyone could reproduce this on a 8642B that happens to be sitting by.
|
Thanks for the very informative reply, I did some additional testing and included some schematics of the A6A1 A6A2 module If I calibrate my sine wave generator (iPad) for an amplitude that is shown neither high or low on the 8642, in FM AC coupled at 50 kHz deviation at 100 MHz there is no issues at higher frequencies >= 100Hz, but when I start feeding lower and lower frequencies the issue seems to occur around <= 60 Hz when I get a transient failure. I have increased the deviation from 50 kHz to 500 kHz but there is no change in the symptoms. I tried setting the generator to 1 GHz and 700 Mhz but symptoms still appear the same. I am not to terribly concerned about this symptom but curious if its just a operational characteristic of this generator or perhaps the FM VCO has shifted off. But I find it strange if the FM VCO might be off shouldn't I get a transient failure at any other frequency I feed into the FM mod input at the same amplitude that causes the transient failure at 60 Hz or below? Thanks  
|
The transient failure occurs when the phase deviation between the reference signal and the divided VCO signal exceeds the range of the loop phase detector. ?If you have the schematic you can look for the circuit which generates the out of lock indication. ?This might be a window comparator on the phase detector output or the loop integrator input. ?Since I don't have a schematic I am making assumptions here. ?Anyway, the modulating signal that is shown being applied to the input of the integrator is there to cancel the phase signal being generated by the phase detector. ?The FM signal applied directly to the VCO should ideally generate the modulation so the PLL doesn't have to do any correction via the PLL tune port on the VCO. ?If the signal wasn't applied at the integrator, the loop would see the modulation as an error and try to remove it. ?This would create a high-pass response to the modulation applied to the VCO with the corner frequency determined by the PLL bandwidth. ?Anyway, if the phase deviation at the phase detector exceeds the detector's range, the loop will lose lock on the phase deviation peaks. ?At low modulating frequencies the divided VCO frequency will differ from the reference frequency for longer periods causing more phase difference to build between them. ?
Here is something to try to see if this is what is going on. ?Go to 100MHz at 50kHz deviation and reduce the modulating frequency until the transient error just begins to occur. ?Then cut the deviation to 25kHz. ?The error should go away until you reach about half the mod frequency where the error started at 50kHz deviation. ?There is one thing that could upset this experiment. The 8642 can generate 100MHz by dividing or mixing down to the output frequency. ?If the firmware decides to change the method based on the deviation setting the actual FM loop deviation will change. ?You could avoid this by going to a main band output frequency like 700 MHz.
The VCO center frequency drift could cause problems if the loop integrator output is operating too close to its rails to compensate and if the loop is trying to correct for gain errors between the inside and outside loop BW modulation paths.
Jim
toggle quoted message
Show quoted text
On Nov 4, 2018, at 8:46 AM, wilson2115@... wrote: Thanks for the very informative reply, I did some additional testing and included some schematics of the A6A1 A6A2 module If I calibrate my sine wave generator (iPad) for an amplitude that is shown neither high or low on the 8642, in FM AC coupled at 50 kHz deviation at 100 MHz there is no issues at higher frequencies >= 100Hz, but when I start feeding lower and lower frequencies the issue seems to occur around <= 60 Hz when I get a transient failure. I have increased the deviation from 50 kHz to 500 kHz but there is no change in the symptoms. I tried setting the generator to 1 GHz and 700 Mhz but symptoms still appear the same. I am not to terribly concerned about this symptom but curious if its just a operational characteristic of this generator or perhaps the FM VCO has shifted off. But I find it strange if the FM VCO might be off shouldn't I get a transient failure at any other frequency I feed into the FM mod input at the same amplitude that causes the transient failure at 60 Hz or below? Thanks <A6A2.jpg><A6A1.jpg>
|
I am running my test with this setup
100Mhz -140dB RF output not used
mod output connected to fm mod input AC coupled?
mod output 1.00VL at 60 hz
FM deviation 70 kHz
if I drop FM deviation to 50kHz screen flashes transient failure A6A1, if I increase mod frequency to 65 hz it’s right on the threshold of detecting alarm 70 hz is fine, I switch operating frequency to 700 MHz it is running fine, drop mod frequency back to 65 Hz still ok, 60 hz fine, drop to 20 hz fine, 13 hv is threshold, 10 he triggers transient failure. Interesting?
|
700 MHz -140dB Mod out 1.00VL at 10Hz in FM mod in AC coupled?
I am playing around with FM deviation, with test setup above 33khz is the threshold which it is about to throw an alarm, I can reduce the FM deviation to 1 kHz and below no problem, seems any fm deviation above 33 kHz deviation causes an alarm even to 1.5 MHz. Interesting
|
I was part of the design team on the 8642 but I’m not sure I remember the FM modulation details well enough. Is the error truly transient? Is the deviation correct after the error message? ?
toggle quoted message
Show quoted text
From: [email protected] [mailto:[email protected]] On Behalf Of wilson2115@...
Sent: Sunday, November 04, 2018 11:29 AM
To: [email protected]
Subject: Re: [HP-Agilent-Keysight-equipment] 8642B displays A6A1 transient failure F3 when audio peaks at FM mod input ? 700 MHz -140dB Mod out 1.00VL at 10Hz in FM mod in AC coupled?
I am playing around with FM deviation, with test setup above 33khz is the threshold which it is about to throw an alarm, I can reduce the FM deviation to 1 kHz and below no problem, seems any fm deviation above 33 kHz deviation causes an alarm even to 1.5 MHz. Interesting
|
I connected the 8642 to a spectrum analyzer and found when I ran the generator into a transient failure condition the deviation was off and at times way to much deviation but upon bringing the generator out of the transient failure conditions it would restore to normal, this at 100 MHz
|
Well let's see. ?Assuming your 10Hz mod rate and 33kHz deviation, the peak phase deviation at the VCO would be 33k/10=3300 radians. The loop divide number is 1080 so the peak phase deviation at the phase detector is 3300/1080=3.056 radians. ?Looks close enough to pi that the phase detector range is probably +/- pi radians. ?Hey Brian, do you remember what you used for the phase detector? ?I never kept any of my '42 schematics.
Jim
toggle quoted message
Show quoted text
On Nov 4, 2018, at 11:28 AM, wilson2115@... wrote: 700 MHz -140dB Mod out 1.00VL at 10Hz in FM mod in AC coupled?
I am playing around with FM deviation, with test setup above 33khz is the threshold which it is about to throw an alarm, I can reduce the FM deviation to 1 kHz and below no problem, seems any fm deviation above 33 kHz deviation causes an alarm even to 1.5 MHz. Interesting
|
The peak phase deviations are: 20k: 20 radians 10k: 40? 5k: ? 80 2k: ? 200 1k: ? 400 500: 800 100: 4000 60: ? 6667
If your earlier results showed the loop starting to lose lock at 3300 radians the 100Hz and 60Hz cases here would be exceeding the phase detector range in the FM Loop. ?This is normal operation and not a problem with the unit I believe. ?
Switching to DCFM should provide proper deviation at these lower rates, although in DCFM the FM VCO free runs so there can be a small carrier frequency error (which I believe the firmware corrects for initially when switching to DC mode.)
Jim
toggle quoted message
Show quoted text
On Nov 4, 2018, at 1:42 PM, wilson2115@... wrote: I performed some additional tests with a spectrum analyzer. The generator is 100mhz -30dB Mod out 1.00VL to FM mod in AC coupled mode From the images shown below from top to bottom with 400 kHz total FM deviation the top image starts with a 20 kHz sine 10khz 5khz 2khz 1khz 500 hz 100 hz 60 hz, each horizontal division is 100 khz <7EA33570_A22E_4332_8490_B7E1A021394C.jpeg><61295152_B701_47D9_AC57_EFA49C7FC6AC.jpeg><FD9F38B7_22BC_4C1F_B585_247F62A3FF3D.jpeg><BCDD84F8_AF6D_4823_A288_5A2F1E8329CB.jpeg><F45D6052_1CE8_4F79_AB2D_DD2992726B21.jpeg><E552ED72_D16C_4C0A_B674_6C59F6B3B132.jpeg><DE68E5DE_A982_4F81_A5EB_B090C8DC7092.jpeg><A2A3E22C_754B_473F_B2ED_048A93CF07CA.jpeg>
|
By the way, the analyzer is sweeping 2MHz so each horizontal division is 200kHz.
toggle quoted message
Show quoted text
On Nov 4, 2018, at 1:42 PM, wilson2115@... wrote: I performed some additional tests with a spectrum analyzer. The generator is 100mhz -30dB Mod out 1.00VL to FM mod in AC coupled mode From the images shown below from top to bottom with 400 kHz total FM deviation the top image starts with a 20 kHz sine 10khz 5khz 2khz 1khz 500 hz 100 hz 60 hz, each horizontal division is 100 khz <7EA33570_A22E_4332_8490_B7E1A021394C.jpeg><61295152_B701_47D9_AC57_EFA49C7FC6AC.jpeg><FD9F38B7_22BC_4C1F_B585_247F62A3FF3D.jpeg><BCDD84F8_AF6D_4823_A288_5A2F1E8329CB.jpeg><F45D6052_1CE8_4F79_AB2D_DD2992726B21.jpeg><E552ED72_D16C_4C0A_B674_6C59F6B3B132.jpeg><DE68E5DE_A982_4F81_A5EB_B090C8DC7092.jpeg><A2A3E22C_754B_473F_B2ED_048A93CF07CA.jpeg>
|
Now that I think about it during the test I performed, I did notice at 100 MHz at just slightly below 100 Hz I was hitting the absolute lower limit. Thanks again for all your help with this Jim! Also, if anyone has a 8642 sitting around to see if similar results can be reproduced.
I should now discuss how I arrived at this problem to begin with. I was feeding audio from the iPad and the output was going into a radio. I was testing the various modulation and filter bandwidths of the radio and occasionally the display would flash and I would get a A6A1 transient failure F3. Since I had the generator on for a while I assumed there was some sort of thermal instability. But I found from recent experimentation applying sine waves have I finally came to this conclusion here.
|
12.9 MB of crap? Can't you reduce the photo sizes and crop the unneeded areas into something reasonable?My Email inbox is only 300 MB. 23 messages this size would fill it.
toggle quoted message
Show quoted text
-----Original Message----- From: wilson2115@... Sent: Nov 4, 2018 4:42 PM To: [email protected]Subject: Re: [HP-Agilent-Keysight-equipment] 8642B displays A6A1 transient failure F3 when audio peaks at FM mod input I performed some additional tests with a spectrum analyzer. The generator is 100mhz -30dB Mod out 1.00VL to FM mod in AC coupled mode From the images shown below from top to bottom with 400 kHz total FM deviation the top image starts with a 20 kHz sine 10khz 5khz 2khz 1khz 500 hz 100 hz 60 hz, each horizontal division is 100 khz
|
People don't think these days.? All the cellphones with cameras have multi megapixel cameras these days generate huge image files.? Great if you really need the resolution, but a huge waste of resources (i.e. bandwidth).? Folks just send them without thinking about the consequences.? When I need to send pictures in these circumstances, I use a freeware program "Irfanview", which makes it easy to resample the images to a much more manageable size (pretty much to your choosing).? I recommend it!
Daun
Daun E. Yeagley II, N8ASB
toggle quoted message
Show quoted text
On 11/4/2018 6:38 PM, Michael A. Terrell wrote: 12.9 MB of crap? Can't you reduce the photo sizes and crop the unneeded areas into something reasonable?My Email inbox is only 300 MB. 23 messages this size would fill it.
-----Original Message-----
From: wilson2115@...
Sent: Nov 4, 2018 4:42 PM
To: [email protected]
Subject: Re: [HP-Agilent-Keysight-equipment] 8642B displays A6A1 transient failure F3 when audio peaks at FM mod input
I performed some additional tests with a spectrum analyzer. The generator is 100mhz -30dB Mod out 1.00VL to FM mod in AC coupled mode
From the images shown below from top to bottom with 400 kHz total FM deviation the top image starts with a 20 kHz sine 10khz 5khz 2khz 1khz 500 hz 100 hz 60 hz, each horizontal division is 100 khz
|
I admit I have messed up on my part, I will try not to let this happen again and apologize for the inconveniences I have caused, take care.
|
I've used it, and 'Paint' for many years to resize images for websites and to create manageable image sizes for PDFs. I agree that it is a great program.
Michael A. Terrell
toggle quoted message
Show quoted text
-----Original Message----- From: Daun Yeagley <daun@...>
Sent: Nov 4, 2018 6:57 PM
To: [email protected]
Subject: Re: [HP-Agilent-Keysight-equipment] 8642B displays A6A1 transient failure F3 when audio peaks at FM mod input
People don't think these days.? All the cellphones with cameras have
multi megapixel cameras these days generate huge image files.? Great if
you really need the resolution, but a huge waste of resources (i.e.
bandwidth).? Folks just send them without thinking about the
consequences.? When I need to send pictures in these circumstances, I
use a freeware program "Irfanview", which makes it easy to resample the
images to a much more manageable size (pretty much to your choosing).? I
recommend it!
|
OK. Some programs won't even open files that large.
toggle quoted message
Show quoted text
-----Original Message----- From: wilson2115@... Sent: Nov 4, 2018 7:04 PM To: [email protected]Subject: Re: [HP-Agilent-Keysight-equipment] 8642B displays A6A1 transient failure F3 when audio peaks at FM mod input I admit I have messed up on my part, I will try not to let this happen again and apologize for the inconveniences I have caused, take care.
|
|
|
There's a plugin for Thunderbird called "Auto Resize Image", which
does, very configurably, exactly what its name suggests. When you hit
"send" on an outbound email with image attachments, it can bring up a
dialog box with a preview of each image, information about its size
(both byte count and image dimensions), and ask you whether to resize it
to a preconfigured (by you) size, some other size, or leave it as-is.
I like it a lot.
Of course you're pretty much screwed if you're using a web browser for
email, but then that was already the case at the beginning, so.. ;)
-Dave
toggle quoted message
Show quoted text
On 11/4/18 6:57 PM, Daun Yeagley wrote: People don't think these days.? All the cellphones with cameras have
multi megapixel cameras these days generate huge image files.? Great if
you really need the resolution, but a huge waste of resources (i.e.
bandwidth).? Folks just send them without thinking about the
consequences.? When I need to send pictures in these circumstances, I
use a freeware program "Irfanview", which makes it easy to resample the
images to a much more manageable size (pretty much to your choosing).? I
recommend it!
Daun
Daun E. Yeagley II, N8ASB
On 11/4/2018 6:38 PM, Michael A. Terrell wrote:
12.9 MB of crap? Can't you reduce the photo sizes and crop the
unneeded areas into something reasonable?My Email inbox is only 300
MB. 23 messages this size would fill it.
-----Original Message-----
From: wilson2115@...
Sent: Nov 4, 2018 4:42 PM
To: [email protected]
Subject: Re: [HP-Agilent-Keysight-equipment] 8642B displays A6A1
transient failure F3 when audio peaks at FM mod input
I performed some additional tests with a spectrum analyzer. The
generator is 100mhz -30dB Mod out 1.00VL to FM mod in AC coupled mode
?From the images shown below from top to bottom with 400 kHz total FM
deviation the top image starts with a 20 kHz sine 10khz 5khz 2khz 1khz
500 hz 100 hz 60 hz, each horizontal division is 100 khz
--
Dave McGuire, AK4HZ
New Kensington, PA
|
At least on the default iOS mail application, it asks right before sending anything with image attachments, and gives 4 choices for total attachment size: S, M, L and original. S is usually < 1/50th of original. It’d be weird if Android mail app or gmail app didn’t do the same…
Cheers, Kuba
toggle quoted message
Show quoted text
4 nov. 2018 kl. 18:57 skrev Daun Yeagley <daun@...>:
People don't think these days. All the cellphones with cameras have multi megapixel cameras these days generate huge image files. Great if you really need the resolution, but a huge waste of resources (i.e. bandwidth). Folks just send them without thinking about the consequences. When I need to send pictures in these circumstances, I use a freeware program "Irfanview", which makes it easy to resample the images to a much more manageable size (pretty much to your choosing). I recommend it!
Daun
Daun E. Yeagley II, N8ASB
On 11/4/2018 6:38 PM, Michael A. Terrell wrote:
12.9 MB of crap? Can't you reduce the photo sizes and crop the unneeded areas into something reasonable?My Email inbox is only 300 MB. 23 messages this size would fill it.
-----Original Message-----
From: wilson2115@...
Sent: Nov 4, 2018 4:42 PM
To: [email protected]
Subject: Re: [HP-Agilent-Keysight-equipment] 8642B displays A6A1 transient failure F3 when audio peaks at FM mod input
I performed some additional tests with a spectrum analyzer. The generator is 100mhz -30dB Mod out 1.00VL to FM mod in AC coupled mode
From the images shown below from top to bottom with 400 kHz total FM deviation the top image starts with a 20 kHz sine 10khz 5khz 2khz 1khz 500 hz 100 hz 60 hz, each horizontal division is 100 khz
|
Michael A. Terrell
Dave McGuire