[hpsdr] Toroid evaluation for filters, transformers, IIP3

[Glenn]

At 10:57 +0800 27/01/10, Steve Hubbard wrote:
>***** High Performance Software Defined Radio Discussion List *****
>
>Hi Glenn,
>
>Are the results at the top of 
>http://www.xs4all.nl/~martein/pa3ake/hmode/bpf_intro.html with the 
>NP0 capacitors or polystyrene? I too have found that the Philips NP0 
>monolithic caps can degrade IP performance quite severely. However, 
>multilayer COG ceramics gave results as good as, or better than, 
>silver mica. I was looking for good out of band IP3, where the 
>unwanted signals were 5/10% away from the wanted.
>
>Is the schematic of the 40 m filter available? I'll have a play if I get time.
>
>I agree with Phil that it's difficult to measure IP3 at this level 
>of performance. This is particularly true when the test tones get 
>reflected back into the generators by the filter under test. What 
>test frequencies were you using?
>
>73 de Steve VK6ABZ......now living in VK5

Hi Steve

It is pa3ake that made the filters and measurements:

IP3 measurement
Signal Generators and Hybrid Combiner:
http://www.xs4all.nl/~martein/pa3ake/hmode/ip3measurements.html
Quote: "...
Signal Generators and Hybrid Combiner

I am using 2 completely separate and double shielded AD9951 DDS's 
(I0CG's design with 500MHz external clock). The DDS's produce about 
+8dBm output level with the ERA-1 MMIC. To reduce the 3rd order IMD 
products of the 2 DDS's together when connected to the hybrid 
combiner I have build 2-stage J310 grounded gate amplifiers that have 
a reverse isolation of better then -75dB up to 6M and around 4dB 
gain. This allows for IMD tests up to the +6dBm output level at the 
output of the combiner.
...
I am using a homemade hybrid combiner that isolates the input ports 
by more then 50dB on all bands except 160M where it gives only -48dB 
when the output is precisely terminated with 50 ohms. On 10M the 
isolation is -56dB peaking even to almost -70dB on 40MHz. The 
isolation provided by the combiner is much less unfortunately, as 
termination is not exactly 50 ohms resistive during real IP3 
measurements. When connected to my homemade step attenuators the 
isolation is only -30dB decreasing to around -20dB on the higher 
bands.
...
So this system is suitable to measure up to +50dBm on 160M-20M. On 
17M-10M +45dBm is the limit. I have not determined the cause of the 
increased IMD at higher frequencies. It could be the combiner or the 
output amplifier or both.
...
This picture (Wandel Goltermann SNA-62) confirms at least -90dBc IMD 
products with 1dBm input levels on 40M (OIP3>=+46dBm). Funny enough a 
few -80dBc or less AD9951 spurs are visible too! This is the best I 
can squeeze out of this instrument.
..."

Frontend - Filters - BPF's:
http://www.xs4all.nl/~martein/pa3ake/hmode/bpf_all.html

Links from page - the filters contains lots of reactive components - 
and he has on purpose lovered the loaded Q to increase the IP3 
(higher Q higher currents and voltages). Measurement graphs are also 
present:
http://www.xs4all.nl/~martein/pa3ake/hmode/bpf_160m.html
http://www.xs4all.nl/~martein/pa3ake/hmode/bpf_80m.html
http://www.xs4all.nl/~martein/pa3ake/hmode/bpf_40m.html
http://www.xs4all.nl/~martein/pa3ake/hmode/bpf_30m.html
http://www.xs4all.nl/~martein/pa3ake/hmode/bpf_20m.html
http://www.xs4all.nl/~martein/pa3ake/hmode/bpf_17m.html
http://www.xs4all.nl/~martein/pa3ake/hmode/bpf_15m.html
http://www.xs4all.nl/~martein/pa3ake/hmode/bpf_12m.html
http://www.xs4all.nl/~martein/pa3ake/hmode/bpf_10m.html
http://www.xs4all.nl/~martein/pa3ake/hmode/bpf_6m.html

regards,

Glenn/OZ1HFT

 1264577204.0