|
|
Post by jmastropole on May 1, 2007 18:35:02 GMT -5
Hi Bob,
Nice looking product! Not wanting to lug my $40K Agilent VNA around for simple measurements in the field ... and considering that I could have bought 3 of your units for the repair cost when I last zapped the VNA on an antenna mission ... I'm very interested.
However, I was having trouble finding info regarding measurement speed and any limitations regarding step size and sweep width limitations.
As I have frequent need to develop automated test software (must choreograph the unattended control of several different instruments), I would like to know if the serial I/O protocol is documented for public consumption. Not to worry... I have the knowledge required to process raw data, perform calibrations and apply vector corrections as you do in your impressive Windows software.
Could be that I got some bad or outdated info ... but I am puzzled by the 30 ppm frequency accuracy/stability in view of the wide availability of low-cost high-performance TCXO's (e.g., by CMAC, RAKON et. al.). I'm guessing that customers would gladly pay the extra $5 for 1ppm level performance as required for quartz crystal measurements.
Best Regards,
Jim Mastropole
RF/Microwave Eng. Consultant
Great Falls, VA USA
|
|
|
|
Post by Bob on May 2, 2007 9:12:20 GMT -5
I was having trouble finding info regarding measurement speed and any limitations regarding step size and sweep width limitations. As I have frequent need to develop automated test software (must choreograph the unattended control of several different instruments), I would like to know if the serial I/O protocol is documented for public consumption. Not to worry... I have the knowledge required to process raw data, perform calibrations and apply vector corrections as you do in your impressive Windows software. Could be that I got some bad or outdated info ... but I am puzzled by the 30 ppm frequency accuracy/stability in view of the wide availability of low-cost high-performance TCXO's (e.g., by CMAC, RAKON et. al.). I'm guessing that customers would gladly pay the extra $5 for 1ppm level performance as required for quartz crystal measurements. Hi Jim, The measurement speed is around 25-30 points per second. It depends somewhat on the speed of the PC since the program is very computer intensive. There is no limit on the step size and sweep width. There can be up to 1000 sample points in a sweep. The rs232 i/o protocol is very simple if someone wants to try their own program. The processing in the PC is quite involved though. The xtal oscillator spec of 30ppm is the initial spec from the factory. It is not temperature compensated but for indoor use, it seems to be very stable. I've checked it a few times with my radio and it's been within 50Hz at 10MHz. It can be calibrated to WWV if you want to. If someone wants to replace it with a more stable oscillator, it's very easy to do. There are only four pins to unsolder with a solder sucker. I'll send them the pinout information. The oscillator runs on 3.3V. RAKON does have a good line of xtal oscillators but they don't list a distributor in the US. It looks like their min order is about $100. 73/ Bob |
|
|
|
Post by jmastropole on May 2, 2007 15:56:10 GMT -5
Thanks Bob,
RAKON just merged with C-MAC and is now setting up sales reps here in the states. I buy 5X7 and 3.5X5mm SMT TCXO and VCTCXO parts from them regularly. You might want to consider going with a dual-footprint for the oscillator on your next board spin ... or just ditch the big part all together as most of the new technology "action" is SMT based.
Say Bob ... what you really need to do is get that board down to the size of a business card and replace the RS232 port with USB (check out the Silicon Labs CP2102 USB bridge chip). Get your DC power from the USB port via a switching regulator.
It is my guess that within a week of purchasing one of your systems, I will probably do all that repackaging myself ... just for fun. As icing on the cake, I would likely fabricate a nice close-fitting CNC milled housing and go with an press-fit SMA connector. Without any heroics, I know I could get the overall size down to 2.000 X 4.000 X .300 inch. Your manufacturing costs for this would be about the same (or less) relative to your current product. You may be interested in seeing some of my design handiwork ... there are a number of websites out there featuring RF communications products which I have developed.
Best Regards, Jim
|
|
|
|
Post by wb4eni on May 2, 2007 17:36:19 GMT -5
PMFJI, but I'd rather keep the option of using the RS232 port. I'm using 2 older laptops around the shack running Windows 98SE and DOS (Yes, I still run antennaware in DOS!). One has a single USB, the other has two USB ports. I need those ports for the mouse, a scanner and an outboard hard drive where I carry all my files. I suspect most hams use older (i.e., cheap and expandable) computers for antenna work rather than state-of-the-art machines running the latest bloatware. I have two of those for professional endeavours and I'm not taking them out below my antennas. :-).
73, Marius
WB4ENI
|
|
|
|
Post by wb4eni on May 2, 2007 17:37:26 GMT -5
Correction -- Read "expendable"
|
|
|
|
Post by wa3off on May 9, 2007 14:49:07 GMT -5
This is a self imposed limitation of the PC app, not a limitation in the AIM4170 unit, correct?
Dave, WA3OFF
|
|
|
|
Post by Bob on May 9, 2007 19:54:57 GMT -5
This is a self imposed limitation of the PC app, not a limitation in the AIM4170 unit, correct? Dave, WA3OFF Yes, it's just a software limit in the PC program. The AIM hardware doesn't save the sample data. 1000 samples may be useful if you use Zplots and want to zoom in on some sections of the scan. You can specify a smaller max number in the config file if you want to. 73/ Bob |
|
