Products > Test Equipment
EIP 545B Frenquency counter
douardda:
--- Quote from: HighPrecision on January 24, 2018, 01:33:41 pm ---If you have the possibility to measure the input RL (VSWR) setting the YIG coil current for a specific tune frequency, you can determine if the coupling between input RF link (that is critical) and the YIG sphere is compromised or not.
--- End quote ---
I do have a spectrum analyzer (Advantest R3465, 8GHz) but no TG (well I do have an R3561L that can be used as TG for the R3465 up to 2GHz, but it also needs repair, the output level jumping all over the place). So I may be able to do some measurement "by hand" (sweeping my RF gen), but not completely sure how to do such a measurement properly (I am mostly a newbie in RF).
What is the best way of setting the YIG current to match a given frequency (bypassing the automatic search mechanism)? I don't see any TEST mode for that. Could the min/max frequency parameters be used to achieve this?
--- Quote from: HighPrecision on January 24, 2018, 01:33:41 pm ---I have some in paper form, unfortunately NOT in pdf, another member has already asked me for a manual, at the moment don't have much time to do long and tedious scanning sessions...
--- End quote ---
oh please! please! please! :-) eg. a few pages at a time. Another solution: send it to me, I do the scanning then I send it back.
David
douardda:
Humm this is gone to be pretty tricky I'm afraid. According to these docs:
http://www.microlambdawireless.com/uploads/files/pdfs/ytfdefinitions2.pdf
http://www.microlambdawireless.com/uploads/files/pdfs/ytodefinitions2.pdf
--- Quote ---The YIG sphere is typically mounted on the end of a thermally conductive rod (normally beryllium).
This is done for two reasons: 1) the rod acts as a “tuning stick” for orienting the YIG sphere in the resonant circuit and
2) YIG has best performance when it’s temperature is kept constant.
--- End quote ---
So (obviously) the YIG sphere must be oriented correctly...
When I look at it under the microscope, I can see the rod's mark on the YIG sphere, so in theory, it may be possible to put it back roughly at its original position/orientation.
This will really be tricky... especially since my cheap USB microscope needs to be really close to the observed device to have high enough magnification, so I won't have a clear path to the YIG sphere... I'll try to move the sphere under microscope with a tiny plastic tip ASAP, just to see if it's actually practicable.
Note that the document describing the principles of a YIG filter do describe a single stage bandpass filter as well as a two-stages band-reject filter, using 2 YIG spheres 'in series'. The topology in the EIP input stage seems to be a 2-stages bandpass filter: each YIG resonator is mounted as a bandpass, with the input and output coupling loops aligned at 90°; both filters being in series (the output of the first one, the one that took off the rod in my device, is the input of the second one, via the wire used as coupling loop over both of them).
douardda:
--- Quote from: douardda on January 24, 2018, 07:21:49 pm ---What is the best way of setting the YIG current to match a given frequency (bypassing the automatic search mechanism)? I don't see any TEST mode for that. Could the min/max frequency parameters be used to achieve this?
--- End quote ---
I cannot test this right now since the YIG is disassembled on my bench, but I think this is achievable but using TEST10 (memory read/write). According to the SM, it seems possible to directly write into the YIG frequency controller (at addresses 0x1840 and 0x1842); for a desired frequency F (in MHz), just have to write F/2 as hex value in these registers.
For example, if I want to tune the YIG filter to 1200MHz, F/2 is 600, ie. 0x0258 in hex.
So I would have to write:
* 0x02 at address 0x1842 and
* 0x58 at 0x1840.I'll give this a try ASAP...
HighPrecision:
--- Quote from: douardda on January 24, 2018, 10:28:18 pm ---but using TEST10 (memory read/write). According to the SM, it seems possible to directly write into the YIG frequency controller (at addresses 0x1840 and 0x1842); for a desired frequency F (in MHz), just have to write F/2 as hex value in these registers.
--- End quote ---
Yes, the addresses are correct for your new CPU board (old version is 0x984x).
These TEST10 commands are to be inserted just after power up of unit (switch OFF and then ON).
douardda:
A quick follow up:
- I've been able to fix the -160MHz offset default setting
- I've been able to partially fix the 5 digits resolution default setting
- I've not yet been able to fix the sensitivity problem
For the default settings, I just replaced a few instructions in the initialization routine by NOPs, as described in https://whatever.sdfa3.org/eip-545b-rf-frequency-counter-part-4.html
The problem with the resolution is that I do have all the digits displayed, but the actual resolution remains on 5 digits; I guess there is also some initialization code that configure the gate generator (A107) to modify accordingly...
For the sensitivity problem, I've not been able to put the YIG sphere in place: it is very hard to manipulate by hand. However, I a wondering if this is the only problem; I also suspect the 15dB boost in A201 to be faulty.
I've picked up the schematic of the A201 board (the IF amplifier/detector part):
I've been suspecting the Q1 transistor, but I've been fooled by the fact it's a JFET, so it does measure low resistance between the source and the drain when the gate is not high.
I've desoldered it from the board, but it looks to work just fine (cheked with 20$ LCR/transistor meter).
On this schematics, I'm not sure what kind of signal is expected on input E9 (from the YIG): when measuring with a multimeter, I have very low resistance to the ground (maybe 0.2 or 0.3 ohm).
Looking at the output of the YIG, this E9 being the green wire, it looks to be, indeed, connected to the ground on the ceramic board:
So I guess it should be considered as ground-level (once again, being a noob in RF, I've been fooled by the presence of R1/C1 there (especially R1), as both side of the wire are grounded, but as usual with RF, it's a "transmission line", not an equipotential, so I guess they make sense).
[edit]in fact, when Q1 is blocked, the gain of this amplifier is something like 32 (1+61.9/2), and when the Q1 is on, the gain drops to around 1, so the ~15dB boost controlled by the E7 input.[/edit]
So I need to dig more...
For those playing at home (as Dave would say), here is a picture of the whole ceramic board in the YIG enclosure, including the mixer (the IF signal generated by the VCO arrives on the pin at the top of the picture) :
David
Navigation
[0] Message Index
[#] Next page
[*] Previous page
Go to full version