HP 8350B sweeper behavior
|
Martin,
toggle quoted message
Show quoted text
if the trace is still noisy, whether or not you activate the CW filter on your 83522A, the circuit involved in this filter is very simple: on the A6 module, it is a U22 driver, a K1 relay and a capacitor C14, 330uF 40V. Regards Yves -----Message d'origine-----
De?: yves_tardif@... <yves_tardif@...> ·¡²Ô±¹´Ç²â¨¦?: 19 novembre 2023 20:18 ??: '[email protected]' <[email protected]> Objet?: RE: [HP-Agilent-Keysight-equipment] HP 8350B sweeper behavior Martin, I'm back... My 8350B, on which I was testing, suddenly broke down, when it was turned back on, it was working 30 seconds earlier...what a coincidence! But, it wasn't too complicated: a capacitor shorted on the +5V power supply on the processor board. By removing the PCBs one by one, I was able to quickly locate the source of the problem. As for the noise on the 1GHz signal, my Spectrum probably has more flexibility, I adjusted the parameters to have exactly yours, based on the image you sent earlier today. Apart from the selectivity of the 10kHz resBW filter which is a little different, I would say that the noise floor is approximately the same as well as the amplitude, the signal shown on your Spectrum shows relatively a lot of noise. However, by default the CW filter should be activated on your plug-in, if this is not the case, do it and compare the result with the image (with/without CW filter). If the image looks like the one on the right with the CW filter ON, then there is no problem. If the CW filter was enabled when you created an image of your Spectrum, then there is a problem...but if so I think it will be quite simple to fix. Do the test and tell me the result, I will tell you which circuit is involved in this process. -----Message d'origine----- De?: [email protected] <[email protected]> De la part de Martin via groups.io ·¡²Ô±¹´Ç²â¨¦?: 19 novembre 2023 13:39 ??: [email protected] Objet?: Re: [HP-Agilent-Keysight-equipment] HP 8350B sweeper behavior On 19 Nov 2023, at 18:50, Jeff Kruth via groups.io <kmec@...> wrote: What is interesting to me is the expectation of greater accuracy than waspossible... I'm probably spoiled by these synthesised generators. When I was at school, many years ago, I build a generator myself. Certainly horrible performance, but I did not have any tools to measure that. Now I have the tools, so... But what about the noisy SA trace of the 83522A, compared to Yves's trace or the one from the older module? cheers Martin |
|
On 20. Nov 2023, at 06:35, Yves Tardif <yves_tardif@...> wrote:
if the trace is still noisy, whether or not you activate the CW filter...Bonjour Yves, The trace is as noisy with or without CW-filter. I already wondered what this CW button is doing at all :-) When its on, the LED behind is not as bright as the others, perhaps there's a reason. Thanks for the details, will check circuit diagram and inside! cheers Martin |
|
Hi all,
I can hardly believe what I saw when I took out the A6-board. A capacitor with only one leg... what the heck is so special about this one that its so different to all the others - and leaking, too! Well, as a replacement for the odd value of 300uF I found a 470uF 40V that fit into the support. Now the CW-Filter switch is working and the spectrum looks clean until down to 100kHz span with 1kHz RBW.? Merci Yves! Attached some pics of the cap. Enjoy! cheers Martin |
¿ªÔÆÌåÓýOn 20. Nov 2023, at 21:42, paulswed <paulswedb@...> wrote:
:-)
Apparently there is no damage to the traces. I used a glassfibre-brush to get the dirt off. These traces only lead to the test points, so no functional impact.? cheers Martin |
|
On 21. Nov 2023, at 02:32, Ed Breya via groups.io <edbreya@...> wrote:
If finding reasonable wet Ta caps is difficult or outrageous, you can do some experiments with Al types, with proper selection and testing, and acceptance of good but not necessarily perfect original performance...I did a very thorough selection before choosing the cap with the best dimensions ;-) Do you know the magnitude of currents going into the YTO's coil? Since the capacitor is in parallel to the coil, the voltage-to-current converter is unable to see and correct for its leakage. As a consequence: higher CW frequency -> higher current -> higher voltage -> higher leakage -> higher shift in frequency when I switch CW-filter on and off. I'll check that today once the unit is stable (takes an hour or two). cheers Martin |
¿ªÔÆÌåÓýHi, ? I have 3 plug-ins, and for the 83522A, I know that it is the original capacitor, since it was purchased new from HP at the time, and I was the only user of this device. I bought it when the company ceased operations. ? So for the 83522A and 83540A, the capacitor is on A6. For the 83595A, the capacitor is on A8. I removed the 300uF capacitor on the 83595A and noticed right away that it was much heavier than a 330UF 40V capacitor. Having no performance problem, I put the original capacitor back in place after inspection. ? De?: [email protected] <[email protected]> De la part de Jim Adney ? On Mon, Nov 20, 2023 at 02:09 PM, Martin wrote:
The wet Ta caps that I'm familiar with, and like to use, have a glass seal around the pos. lead. This one appears to have a rubber seal, but the odd thing is that the leak appears to have been at the neg, case, end. How does this happen? Is this a typical failure mode? |
|
Some more pics of the broken CW-Filter capacitor.
I made some experiments to measure the effect of CW filtering on frequency: CW 1000 MHz (after 2 hours with CW filter on), frequency CAL'd to 1GHz: CW Filter to off: +5kHz. CW 10 MHz (after 30 min later): frequency -75kHz, CW Filter to off: +2 kHz CW 2400 MHz (immediately after): frequency -2MHz, CW Filter to off: +15kHz CW 2400 MHz (after 90 minutes): frequency -48kHz, CW Filter to off: +10kHz Its interesting to see the charging and forming effect on the cap as the delta between CW filter on and off gets smaller. But these variations are much smaller than the variation of nearly 2MHz I get when the current through the YTO changes and probably heats it up a little. All in all I cannot see why this capacitor is critical, except if you do not want to see the slightes change in frequency when turning CW filter on or off. Why do that in CW mode? In sweep-mode CW filter is turned off anyway. BTW, does the older 86222B has a CW filter, too? There is nothing on the front panel, its output in CW mode is comparable to the 83522A with CW filter on. cheers Martin |
Martin
Stefan