From what I can deduce the DSA-815 with a tracking generator is probably one of the better price/performance alternatives when it comes to tuning filters for 2-way radios. There are probably better performing instruments out there, but this will do the job at a price that at least some of us can afford.

contact me if you still having troubles with 8920

If you had the correct Pass-Reject loop for the can, it is fairly simple. However, it also requires a special can top to allow for the required anti-resonant tuning cap. It might be possible to get this done by Telewave but I haven't worked there for a number of years thus don't know their current policies.

@@jimeagle1 Thank you for the quick reply James. Unfortunately I know little about Cavities. For the Pass-Reject loop does it have to go inside the can or attach to the connector at the top? Same for the anti-resonating tuning cap? Finally could please point me to where I might get the construction details.

As I indicated, this isn't really something that can be "homebrewed". The factory might do it for a price but it takes significant changes to the top of the unit and the loop inside. Best done at the factory.

Chris Scott You bring up a good point. I am not a cavity filter designer and I was only repeating what I was told about the reason for use of the Invar. However the use of +1 material with +10 material would still cut any temperature related detuning in half. By the way, generally there is also a small value capacitor used in many cavity combiners and/or duplexers and these can certainly have negative temperature coefficients. Are there any cavity designers out there who like to enlighten us on this subject?

Sergio Mackay There will be a notch both above and below the main peak response for most duplexers. Use a wide and sweep (10 MHz or more) to find the lower (or higher) notch. Readjust Main peak tuning and Notch tuning in order to notch below or above the peak response. If you cannot get the desired Peak / Notch relationship you might need to adjust the loop size. You might want to check out videos at Telewave.com for further instructions about duplexer tuning. There are also several videos on KZfaq about tuning up duplexers.

Please explain what you mean by a 'very high mistake in the amplitude'. I have not found this to be true when using the Normalize mode with a SPAN that is as narrow as practical to show both the peak and null of a cavity. One does need to use the Accy sweep mode to ensure frequency accuracy and appropriate RBW, VBW, and Sweep speeds to achieve proper results just like any other Spectrum Analyzer.

James Eagleson In the Hamburg University of Applied Sciences is a lab for communication technology. My group did a simple Amplitude Modulation. On the scope screen we saw modulation factor about 50%. The Rigol DSA 815 showed 39,99%. I tried to find the mistake in my messuring, however, the mistake is to find in the technical data. Amplitude is shown inaccurate by bigger than 1,5 dB / dBm. The difference explained why the scope had shown verything right and the loss of 10% apmlitude at the Rigol. So 10% is many if inaccurate amplitude shoulbd be below 5%.

ElmshornBoy Were you using the Rigol in the Linear mode or Log Linear mode?

ElmshornBoy Yes, there does appear to be a discrepancy in the Zero Span, Linear Scale representation on the screen of the Rigol DSA815. It shows the bottom of the modulation at about 40% when peak modulation is shown at 100%. Thus I am not certain what Rigol's Percentage scale is trying to show us. If I use the Peak Table with a dBm Log reference, the carrier falls at -0.03dBm and each sideband falls at -12.03dBm just as I would expect for 50% modulation. At the same time the DSA815 Percent scale shows each sideband at 25% when the carrier is set to 100% which appears to be measuring percent of power in each sideband (i.e. 25% power in each sideband for a total of 50% modulation, which is the correct answer). It is unclear what Rigol is trying to represent by their 100% reference when the unit is placed in wide bandwidth (example 20kHz RBW), Zero Span, and Linear modes. It does not appear to be modulation depth, however.

Ma!ny thanks! I was a pleasure to tlak with such an competent and experienced Engineer. :-)