EMRFD Message Archive 12476
Message Date From Subject 12476 2016-03-05 05:26:04 k1rf_digital_stev... 4 pole bandpass filter with FSA 4157 switches Hi all. I just uploaded a schematic for a 4-pole bandpass filter with FSA4157 switches by K1RF_digital_steve in the Files section. This might give you some ideas for making your own filters. It includes provisions for isolating each pole for ease of tuning, either with 0.1 inch spaced through hole jumpers or surface mount 0805 zero ohm resistors. The FSA4157s and associated circuitry can also be depopulated by bypassing them with 0805 sized zero ohm jumpers. Note that the pullup/pulldown resistors on the FSA 4157s are rather high values. This is because for multiple filters they are all effectively in parallel at RF and you don't want to mess up the 50 ohm impedance match. I had already designed on paper a set of 7 bandpass filters for a general coverage receiver based on using T44 size toroids as a compromise between size and good large signal performance and use of very low cost 1 percent 0805 cog/npo ceramic capacitors. 0805 size was selected based on reasonable ease of hand placing/soldering and very low cost with high accuracy (between 10-30 cents max per cap.) I have all the toroids would and am laying out the PC board now for them. I held off building these because I was not happy with using relays (cost and unreliable contact after years unless you rin a current through the contacts) ) or using the run-of the-mill FSA type devices. I'm hoping the FSA4147 will give very good performance. I'm about to find out...
Refards, "Digital Steve", K1RF
12478 2016-03-05 13:01:24 Will Kimber Re: 4 pole bandpass filter with FSA 4157 switches
What would be wrong with using an old fashioned rotary wafer switch with the filters mounted around the outside of the switch? Somewhat similar to the old TV turret tuners of yore but with filters outside and stationary rather than inside and moving..
If you want/need to power operate it a Ledex switch motor would turn it - if they are still available.
I realise the ultra miniaturists will have heart attack at the suggestion but it avoids non linear silicon and poor contacts from relays in the signal path
12479 2016-03-05 13:49:33 Steve Dick Re: 4 pole bandpass filter with FSA 4157 switches Hi Will. I have had the same thought from time to time. In fact, several LPF kits out there do use a rotary switch on a muilti-filter PC board or plug-in filters on a motherboard containing a rotary switch. Their advantage is low switch loss. Their disadvantage is reliability of the contacts, especially over time, unless it is a really high quality switch. Remember the T.V. tuners of yore before they came out with synthesized solid state tuners? The tuners were fairly well built with ball rolling contacts on the giant rotary tuner to make them easier to turn. One would sometimes have to use the proverbial tuner cleaner on them to keep them happy. I also thought about using a single pole by N rotary switch to select the FSA4147 or other solid state switches. Its advantage would be that you just need a single wafer on the switch and the switch could be remotely located from the filters for an easier mechanical configuration. The switch contacts closing the control signal to the solid state switches would have a more solid switching action than directly switching between filters with switch contacts. You also have maximum flexibility with electronic switches as you can enable them from processor digital outputs or decoded digital outputs for software controlled switching.Regards,“Digital Steve” K1RF 12480 2016-03-05 16:13:06 sohosources Re: 4 pole bandpass filter with FSA 4157 switches Steve and the gang,I have been daydreaming from time to time as of late about experimenting with some type of "universal" rotary switch driver. Something like using an Arduino-controlled stepper and a gear-reduction drive to "step" big ceramic military surplus switches back and forth through their ranges.In my case this was for remote, high-power antenna matching hardware, but mini versions should be possible, perhaps eliminating the need for a reduction drive?I'm sure these exist on the high end of the affordability scale...but they should be possible with typical ham stuff.I'm also imagining using a second switch wafer to provide "confirmation of switch closure and position," if desired. It's always a good idea to be reasonably certain that a switch is in the position you think it is... :) Especially when dealing with heavy iron.Regards,--Kirk, NT0ZMy book, "Stealth Amateur Radio," is now available from
www.stealthamateur.com and on the Amazon Kindle (soon)
12482 2016-03-06 01:37:19 dx11 Re: 4 pole bandpass filter with FSA 4157 switches
Hello Kirk and others,
Several ideas came up to me after your daydreaming questions.
In the ham world I see a lot of attention for stepper motors, understandable for several reasons, but I myself am a fan of low voltage DC motors. These are available in a large number of sizes and power levels, often connected to a planetary gearbox to reduce speed and increase torque output. On ebay they can be found for decent prices. Arduino’s can be used with motor shields with an H-bridge on them. Together with PWM they can be controlled very precisely. Some form of position sensing will be required. Discrete switches can be used or a potentiometer on the output shaft using only one of the analog input pins of the Arduino.
A good source for a DC motor with reduction drive might be an old (or new) cordless screwdriver. The adjustable drill chuck comes in handy to connect it to the input shaft of a mechanical rotary switch. Output torque is considerable and should be able to drive even large switches. Voltage requirement is in the order of 12 Volt. They can draw a large current, but both speed and torque (is linearly related to current) can be controlled with PWM of the Arduino.
Switches can be operated with a drive like this, but also rotary variable capacitors or rolling inductors. A remote controlled antenna tuner would be possible, one with decent switches instead of relays!
To protect it all, end switches on every shaft would be needed with a diode over every end switch to return the shaft from its end position.
Steve and the gang,
I have been daydreaming from time to time as of late about experimenting with some type of "universal" rotary switch driver. Something like using an Arduino-controlled stepper and a gear-reduction drive to "step" big ceramic military surplus switches back and forth through their ranges.
In my case this was for remote, high-power antenna matching hardware, but mini versions should be possible, perhaps eliminating the need for a reduction drive?
I'm sure these exist on the high end of the affordability scale...but they should be possible with typical ham stuff.
I'm also imagining using a second switch wafer to provide "confirmation of switch closure and position," if desired. It's always a good idea to be reasonably certain that a switch is in the position you think it is... :) Especially when dealing with heavy iron.
12484 2016-03-06 06:58:02 Bill Carver motor driven tuners, etc Six wire motors are easy to drive with four MOSFET with series
resistors. I had 24 VDC: subtracted motor voltage, divided by motor
current (both on stepper motor nameplates) to get resistor values. Four
wire motors require switching wires between ground and the supply
voltage, specialized chips are available for those, and have their own
PWM to eliminate the resistors in series with winding. Any micro,
Arduino or otherwise, will operate the motors.
Counting the steps is one way to control rotation. I operated for a day,
back and forth, to make sure the motor did not skip steps (too much
mechanical load and that's possible). You really should have some way
to ensure that the switch or capacitor shaft is where you think it is.
Especially if there is non-interrupt Arduino software involved. None
skipped, but when vacuum caps are involved I did not want to take
chances. I put a finger of aluminum on the shafts which passed through
the gap of a "photo interrupter" (LED source + photo transistor in a
single assembly) as a backup. For a 1.8 degree stepper the interrupter
signal should be detected every 200 steps: if not, a step has been skipped.
Alternatively one could put a potentiometer on the shaft and read the
voltage as feedback. Not quite as precise,Motors with shafts out of BOTH
ends are nice, but uncommon surplus items. I've driven pots off toothed
belts and pulleys which are available from people like "Stock Drive
Products" or "Sterling". Other small parts guys also have them, but
SDP/Sterling had much lower prices. But.....those little mechanical
parts are surprisingly expensive!
For switch positions you can operate a wafer with a string of resistors
connected to the Vcc of the Arduino/microcontroller, and read a
different voltage at each switch position.
I operate an 8877 amplifier with four motors (fourth motor for the
cathode tank), nice to "remote" the blower noise! Also the capacitor and
roller coil of a remote tuner that sits at the base of the 160m/80m
vertical. Would be neat to send DC power and control signals down the
coax, but I sent power on separate wires since the one coax feeds
multiple antennas, and used 2.4 GHz to send commands to the tuner.
Automation is kinda fun. But there's an old saying that applies there:
"If anything can go wrong, it will". Make sure your software is
bulletproof, then add feedback to ensure the position is what you think
12488 2016-03-06 14:34:21 Kirk Kleinschmidt Re: motor driven tuners, etc Bill and the gang,Yes, determining the position of a rotating switch shaft is an interesting and somewhat neglected study in amateur circles. I, for one, wouldn't mind seeing a "cookbook"-style article in a magazine or two.In the late '90s I was working as a tech writer for a company that was developing a computerized valve-position/predictive diagnostic system for process-control valves. The engineers used string pots (EXPENSIVE) and magneto-resistive sensors (probably 3 cents apiece nowadays) to precisely determine a valve's shaft position, speed, stiction, torque, overshoot, etc. These figures can determine whether a valve needs servicing during scheduled downtime instead of relying on a fixed maintenance schedule. In big paper mills and oil refineries, downtime might be measured in terms of "millions of dollars an hour," so NOT fixing a big valve if it doesn't need fixing could save huge bucks.It was interesting: They were seeing valve shafts and I was imagining how I could use similar shaft-position sensors to make a massive-duty antenna rotator from "farm parts" such as the worm gear drive from a big manure spreader! I envisioned mounting it at the tower base. Back then microprocessor-controlled goodies were a lot less accessible.My problem is, I understand the systems from the big picture perspective...but my coding skills stopped with BASIC in 1980... :)Regards,--Kirk, NT0ZMy book, "Stealth Amateur Radio," is now available from
www.stealthamateur.com and on the Amazon Kindle (soon)
12489 2016-03-07 17:20:56 k1rf_digital_stev... Re: 4 pole bandpass filter with FSA 4157 switches Folks, the FSA4157 is problematic due to relatively high capacitances pointed out to me on the hermes lite yahoo group by Graeme, ZL2APV. He stated: "The bit that concerns me with the FSA4157 switches is the off port's 12pF capacitance. If you are switching say 6 filters these ports will be commoned at 72pF and will connect to the filter when switched. Port A is 40pF so you could have a total of 100 odd pF presented to the filter. Of course this value could be included in the filter itself and a series parallel conversion done to absorb it. Bus switches could be worth a look and there are 1 to 8 fanouts which would be ideal but speed, through resistance and capacitance were the limiting factors when I last looked at them a couple of years ago. There could be advances now so I should look again.
73, Graeme ZL2APV"
I liked the '4157 switch because of its low loss and high linearity, but this is definitely problematic and should be considered in any designs. Transmission gate analog switches have a trade-off. Lower on-resistance generally means higher capacitances and vice versa. My apologies for missing this important point. Relays anyone?
"Digital Steve", K1RF
12490 2016-03-08 09:32:59 Ashhar Farhan Re: 4 pole bandpass filter with FSA 4157 switches there is a way around it though.first, instead of just an on-off switch, consider using two switches. the first disconnects from the band pass filter and the second grounds the disconnected bandpass filter. in effect, the capacitance of the disconnected bandpass fiter's switch is now grounded. this can be absorbed in the filter.- f