EEVblog Electronics Community Forum
Electronics => RF, Microwave, Ham Radio => Topic started by: peter-h on August 05, 2017, 08:29:45 am
-
(http://peter-ftp.co.uk/screenshots/2017-08-05_092838.jpg)
The BF996 is unprotected.
From time to time, the input is found to be blown up and it is the BF996. The cause is probably static.
It would need to be clamped to something like plus or minus 5V and just putting some zeners back to back, or just two diodes back to back (for a 0.5V clamp) is probably going to have way too much capacitance.
The reason for C3 - a trivial 3.3pF - is unclear, given the signal passes through say 10m of 50 ohm coax... The operating frequency is about 300MHz.
I have done this sort of thing before but for low leakage (precision ADC input amp) rather than low capacitance. That was done with low leakage reverse biased diodes.
-
Looking at L19 and C1 I think it is unlikely that it's static. Are you sure that it's not getting too much input signal?
-
Given the usage (connected to a rod antenna) I can't see how this could happen normally, but obviously it is possible with some high power RF source passing by... I can't control that.
-
This is really interesting. At work we use varistors for ESD protection on RS232/422/485 signals, 14V clamping voltage, but they are quite "leaky".
The ESD130-B1 seems perfect. I didn't know this stuff existed.
Does anyone make a 50 ohm BNC inline product with one of these inside? I have a couple of places I could insert such a thing. I know one can get flash protectors like that, but they have a high clamp voltage - of the order of 100V - and are mainly for lightning protection.
-
Hi,
The BF996 is unprotected.
Actually, the datasheet says it's got zener diodes internally. Curiously, they don't give an ESD rating, so it's not clear if they're good for anything at all... |O
But yeah, even if they gave a rating, it would probably be a paltry 2kV (HBM), if that. An ESD strike to the antenna will easily deliver that, and it will be at a lower impedance (because of the coax and filtering components), so, more destructive than an HBM condition.
It would need to be clamped to something like plus or minus 5V and just putting some zeners back to back, or just two diodes back to back (for a 0.5V clamp) is probably going to have way too much capacitance.
The reason for C3 - a trivial 3.3pF - is unclear, given the signal passes through say 10m of 50 ohm coax... The operating frequency is about 300MHz.
Well, coax is a transmission line, it's not a capacitance. If it's properly matched, then it's 50 ohms, period -- no capacitance at any frequency!
3.3pF at 300MHz is hardly trivial, it's 160 ohms. :)
The response of the input network is approximately:
(https://www.seventransistorlabs.com/Images/UHF%20Amp%20Front%20End.png)
As you can see, the output voltage (the voltage at the gate) peaks around, well, low UHF, let's say -- it has a bandpass characteristic. The above is plotted with a load capacitance representative of the gate capacitance plus "select" capacitor. It's not obvious if L5 is providing resistance or inductive reactance. Probably, it's a bit of both, which has a damping effect, and the gain and bandwidth at 300MHz will be more useful.
I have done this sort of thing before but for low leakage (precision ADC input amp) rather than low capacitance. That was done with low leakage reverse biased diodes.
The tricky thing here is not just the magnitude of capacitance: as you note, zeners (typically 100pF+) are right out, here. But if you used ~3pF diodes (say, replacing C3 with junction capacitance), you have a more subtle problem: the capacitance varies with voltage, so large signals will experience a different frequency response than small signals. It's a knock-on effect, so signals will get mixed together under this condition, and now you have weird spurs all over: IMD. This is especially tricky in an RF front end, where you want maximal dynamic range and linearity.
The suggestions others have made are about as good as you can get. Haven't looked, but I'd expect the Mini-Circuits limiters have IMD data, and show good (linear) operation up to the limit. There are commercially available ESD diodes for radio and high speed digital applications, which have the downside that, although they have little capacitance (under 1pF, and the change in capacitance versus voltage is a fraction of that), they're simply small to begin with, and can't withstand too many ESD strikes before failing. (A replaceable inline module would indeed be handy. :) )
Tim
-
Thank you for the brilliant post! And the one above it.
I am going to sort out an inline solution.
I can't see any off the shelf BNC to BNC module but there are lots of modules in suitable packages and I can just buy one from Ebay and remove the internals :) The challenge would be to maintain 50 ohms connector to connector... Presumably a PCB with a ground plane and a controlled impedance trace is the way to go. I will see if I can google something. Infineon sell an eval kit for one of their limiters which has an SMA at each end of a little PCB, but I need BNC.
The frequency range here is 329-335MHz so quite narrow. That also is a long way away from the ~205MHz peak in the graph here, which is odd... maybe that is intentional.
Somebody will realise the frequency range is the ILS glideslope. I reckon what blows it up is airport radar. If you park at any major airport, you get zapped with a directional beam of few kW every few seconds, and while the frequency is a lot higher, it is rather wideband due to nonlinearity of other things which pick it up (and possibly re-radiate it) as one gets a loud noise in the headsets.
-
The challenge would be to maintain 50 ohms connector to connector... Presumably a PCB with a ground plane and a controlled impedance trace is the way to go.
Not even that much is necessary (though it's easily done) -- at 300MHz, 1/4 wave is 25cm, so trace lengths much less than that (a few inches) have very little effect, as long as they are still ballpark (it's hard to do lower than, say, 20 ohms, or higher than 150 ohms, on a PCB).
A strip of copper clad, with the device in the middle, and connectors on each end (especially edge-launch types), makes a fine inline device, and can be made very close to 50 ohms, even at high frequencies.
The frequency range here is 329-335MHz so quite narrow. That also is a long way away from the ~205MHz peak in the graph here, which is odd... maybe that is intentional.
Well, like I said, it depends on the load capacitance. The peak can be slid up or down by adjusting that parameter. Or by canceling out its capacitive reactance with some inductive reactance. It's within 40%, which isn't horrible at a frequency where device properties and PCB parasitics are paramount.
Somebody will realise the frequency range is the ILS glideslope. I reckon what blows it up is airport radar. If you park at any major airport, you get zapped with a directional beam of few kW every few seconds, and while the frequency is a lot higher, it is rather wideband due to nonlinearity of other things which pick it up (and possibly re-radiate it) as one gets a loud noise in the headsets.
So, a couple concerns...
1. This goes in aircraft? A landing instrument, at that? What kind of paper trail is required here? Not a dev kit, I'm guessing...
2. Front ends can get overloaded by strong enough out-of-band signals. Your description isn't quite right, but the general idea is, yes. Fortunately, a simple solution is: filter those front ends better! ATC radar is 1GHz, so a lowpass in the 300-500MHz range should help here. And one at 150-500MHz would help with the radio.
Tim
-
It's a homebuilt aircraft. Limited to VFR, so no, not used for real. It's a function in the radio however and it ought to work.
If it can be done as an inline device, no problem.
There have been big re-radiation issues with 121.50 MHz ELTs, with their PI-tanks getting excited by VHF on nearby frequencies and blanking out the GPS carrier.
Curiously I had big problems on USB2 (400mbps) where breaking the cable shield for even 2cm was enough to break the comms totally. One could use it if connected via an old USB1 hub...
-
It's a homebuilt aircraft. Limited to VFR, so no, not used for real. It's a function in the radio however and it ought to work.
Ah, okay.
Curiously I had big problems on USB2 (400mbps) where breaking the cable shield for even 2cm was enough to break the comms totally. One could use it if connected via an old USB1 hub...
Well, yeah: breaking the shield (any gap length) is the same as adding a 1:1 transformer in series with the wires inside, where the primary winding has the shield's ground loop voltage applied to it. More than about 1.5V of common mode noise and the input receiver threshold is violated, and you get gibberish data.
The only reason USB (Full Speed or higher) can exist, is because it must be 100% shielded!
Bizarrely, you see a lot of appnotes suggesting to add ferrite beads in series with the power/ground wires, or even the shield itself. This goes to show you how appnotes should be taken with ample grains of salt...
(RS-485 has the same considerations as well, but the input CM range is a whopping 12V, so it can tolerate quite a lot of noise, shielded or not. It's also much slower, so, much easier to filter RFI away without affecting BER. And if that's still not enough, galvanic isolation can be used to extend the CM range massively. :) )
Tim
-
You could experiment by putting a suitable PIN diode across the input inductor L19. ... or if you want a quick inline solution for a 50R transmission line you can make a fairly effective limiter (up to 500MHz) using 4 cheapo/jellybean 1N4148 diodes arranged as two back to back pairs spaced slightly apart. This should give low loss at 300MHz and good VSWR and will begin to limit at a few milliwatts. But the diodes won't limit very well at frequencies above about 500MHz. So no good for radar signals. However you could include a lowpass filter section ahead of the limiter to stop this issue?
-
Microwave pin diodes and schottky diodes are commonly used for input protection. When a pin diode is used, it turns on shorting out the transmission line during its storage time. Avago has several application notes on this subject.
What I would do is add a shorted 1/2 wave line to the input using a coaxial t-adapter. This will pass a broad range of frequencies (and their harmonics) but short out everything else.
-
Thank you all for great ideas.
There isn't really any space inside the radio and I have no desire to hack it around too much. I have to build a BNC-BNC solution. Unfortunately that means that there won't be any bias voltage available for anything.
Those Infineon limiters look great.
-
Ideally you need to find out what is causing the damage. A limiter PIN diode across L19 would provide a limiting action for big local RF signals from HF through to maybe 2GHz. But it isn't an ESD device as such.
Note that the earlier simulation for the input network looks wrong to me. The second resonator is missing and so is the 27pF grounding cap. The equivalent circuit/models and the simulation given earlier looks bizarre to me. I think there is an obvious BPF response there and the centre frequency will be close to 330MHz because 64nH and 3.3pF resonate here. It looks like two resonant tanks coupled by a series 180nH inductor. However the resonant tank near the FET is grounded via 27pF so the network will also provide (unwanted) voltage gain at a lower frequency. Maybe 56MHz?
But this looks like a 330MHz front end to me. Ideally I'd need to see the real hardware to model it properly but I think it would look something like the plot below:
Note that there isn't anything there that will combat stopband re entry modes up at several GHz. The shunt capacitors will only need a few nH package and layout inductance to permit higher frequencies to pass through this circuit easily with minimal loss. So maybe you need to find a decent LPF that maintains a good stopband up to the frequency of the local radar. You could try the ESD diodes as well but I'd be tempted to start with a decent LPF and a diode limiter.
Is the 9V power supply clean without any large spiky switching transients on it? It's unlikely this would be the cause of the damage but worth ruling out.
-
I used to loose my SDR front end about once a year when using it with a whip antenna or with a roof antenna.
I used a Motorola BAV99 at the input connector for the antenna bridged to the coax (and case) ground right at the connector. The BAV99 has two RF diodes in an SMD package. Pin 3 to the antenna and pins 1 and 2 to ground. In effect the two diodes provide voltage clamping for anything over the bandgap of either polarity. No noticeable deterioration in the 200-900MHz range that I could notice.
The SMD package makes it possible to do a very compact connection right at the case connector. If you do not want to go inside the box a simple bnc to bnc adapter would be easy to make. It would also be simple to remove if you do not like it. The BAV99 are pretty cheap too.
-
I am happy to try anything which is inline-BNC because I can get a colleague with an IFR4000 to do a full ground test, including checking the signal level at which the receiver flags-out, so we can check if the solution is creating any attenuation whatever.
BAV99:
(http://peter-ftp.co.uk/screenshots/2017-08-06_202224.jpg)
That will obviously clamp anything to +/- 0.6V or so.
-
I have to build a BNC-BNC solution. Unfortunately that means that there won't be any bias voltage available for anything.
This is why I suggested a 1/4 wave shorted transmission line. (1) It can be added externally between the antenna and receiver and attaches with whatever connectors are currently being used. The 1/4 wave transmission line is just a length of coaxial cable, about 6 inches long at 300 MHz, which can be bent or coiled up to fit in the available space.
(1) I said 1/2 wave earlier but it should have been 1/4 wave.
-
The shorted coax stub will provide strong attenuation at low frequencies and at multiples of 600MHz but that's about it. Unless the interfering signal lies in one of these zones then the coax stub won't help much? The shunt 64nH inductor in the receiver front end should provide loads of attenuation at low frequency already?
Here's a couple of images of my cheapo homemade diode limiter made from 4 cheapo switching diodes placed in a coaxial housing. It uses decent N type connectors and has very low loss up to 500MHz and very low VSWR. eg >0.05dB loss at 300MHz and better than 1.025:1 VSWR. This limiter is used to protect my spectrum analyser and my power meter when there is risk of damage at LF through VHF. It should withstand 5W continuous power. It was carefully constructed to maintain a very low VSWR and low insertion loss to minimise its contribution to overall measurement uncertainty.
-
I am happy to try anything which is inline-BNC because I can get a colleague with an IFR4000 to do a full ground test, including checking the signal level at which the receiver flags-out, so we can check if the solution is creating any attenuation whatever.
BAV99:
(http://peter-ftp.co.uk/screenshots/2017-08-06_202224.jpg)
That will obviously clamp anything to +/- 0.6V or so.
By the way I do assume E1 in the schematic is the antenna input with no extras. If so, L19 and R12 at the input seem to preclude the use of powered antenna preamp with a DC bias or phantom power injector (L19 would be a dead short). That is why a wonder about your comment:"" Unfortunately that means that there won't be any bias voltage available for anything.""
For other applications with a receiver that is capacitor coupled one has to be careful where the clamp diodes are installed, not to short the DC in the antenna coax but still protect the input receiver FET...
Same would apply for a TxRx switching setup with a shared antenna.
-
The shorted coax stub will provide strong attenuation at low frequencies and at multiples of 600MHz but that's about it. Unless the interfering signal lies in one of these zones then the coax stub won't help much? The shunt 64nH inductor in the receiver front end should provide loads of attenuation at low frequency already?
And it will short out static and DC. That shunt 64nH inductor should be doing it but maybe the overload or whatever is going on is destroying it also.
-
Correct - the antenna is feeding that input directly, via some metres of coax. It is made up of two rods, each 23cm long, and IIRC one feeds the centre conductor and the other goes to the shield.
I reckon there are only two possibilities: lightning (not a direct strike; one would notice that ;) ) or powerful RF of unknown frequency.
-
At a guess, the FET amplifier might (typically) get damaged at an input power level of +20dBm at 300MHz if the local signal is on the same band as you.
So the local 300MHz transmitter would have to be in the ballpark of 100W transmit power located maybe 5 metres away assuming you both have similar antennas. So you would typically have to be sat alongside another vehicle that has a high power transmitter in it.
The input filter looks to have an extra response at about 56MHz so there is the risk that a local 56MHz transmitter could damage the front end. But unless the layout of the front end filter is very poor then you would typically need to be transmitting up at several GHz to get through the front end filter efficiently with test frequencies above 330MHz.
If the layout and components are poor then there is the possibility of having other narrow BPF responses up at a GHz or so but you would still need to have a very powerful transmitter very close by.
Have you ruled out any common mode effects via your coaxial cable screen? Is there a balun or a clamp filter on the coax? How well grounded is the coax run?
-
The coax is grounded at various bulkheads before it gets to the radio.
The power requirement is easily met by being anywhere near any airport that has a radar on it - basically any large airport. You only have to drive (or fly) past the airport. I am sure it is radiating a few kW.
There are more esoteric explanations. For example one pilot reported avionics damage on each occassion when flying at some 10,000ft above a military installation. It looks likely there is a SAM site there and the operators are just messing about, tracking planes flying overhead.
Light aircraft don't get tested for immunity in the way the big ones do...
-
That shunt 64nH inductor should be doing it but maybe the overload whatever is going on is destroying it also.
It would definitely be worth checking if the 64nH inductor is still intact. I'd expect it to be made from reasonably thick TCW but it might be a little SMD part. The other way to damage this inductor would be via DC connected at the coaxial input but I'd expect this to give visible clues.
-
From the schematics, considering C1, a simple DMM resistance check should show near dead short. If L19 is open then you would read R12.
Radio off, of course....
-
This problem has occurred before, and on that occassion, as on this one, both of the two radios had a blown up MOSFET in the GS receiver. I have just put a meter on the input of one of these and it indeed reads a dead short as one would expect due to that inductor.
I agree the inductor should provide adequate static protection, but the height of the antenna makes it extremely unlikely anybody will be touching it, and any static picked up during flight ought to get rapidly dissipated via this route.
However an interesting angle is that any antenna sticking out of the airframe is also acting as a static wick. The spectrum of the discharge is going to be rather broadband. But if this was a problem, one would expect it to be widespread.
-
Is there enough room to fit something like a https://www.digikey.com/product-detail/en/littelfuse-inc/SESD0402X1BN-0010-098/SESD0402X1BN-0010-098CT-ND/5233522 (https://www.digikey.com/product-detail/en/littelfuse-inc/SESD0402X1BN-0010-098/SESD0402X1BN-0010-098CT-ND/5233522) SESD0402X1BN in the radio? The loading is quite low, and the 10v clamping voltage should be enough to save the gate. The 0402 size package with pads on the bottom only is a bit of a pain to solder, but if you can get it to straddle from a TX line to ground somewhere in the circuit, it could be fit in.
-
There probably would be but it is too late since both radios are reassembled.
I will build a BNC inline adaptor with that device. It looks amazing. 0.1pF... I need to find out how to do an inline adaptor which maintains 50 ohms all the way.
-
OK... here goes a dumb question :)
If I just solder two 50 ohm BNCs back to back, there will be something like 25mm of PTFE-covered centre conductor which will have no shield. There will also be about 5mm of centre conductor which will be completely bare - where I will solder the voltage limiter.
How bad will this be at ~300MHz?
I could create a shield for the PTFE section, either with a machined brass tube or by painting it with silver paint, if necessary.
I will probably use a little metal box for the whole thing, just to make it neat.
-
BNC...what? Cables cut open? Connectors? ???
Impedance is about distance between the conductor and the shield. Maintain that, and you're set.
Discontinuities on the order of 1/20 wavelength long (i.e., ~5cm) are not much impact on the signal.
You still want shielding, otherwise you might be letting interference right back in.
Tim
-
There is already a BNC-BNC joint in the cable, so I would just be inserting one connector's worth of loss.
-
OK... here goes a dumb question :)
If I just solder two 50 ohm BNCs back to back, there will be something like 25mm of PTFE-covered centre conductor which will have no shield. There will also be about 5mm of centre conductor which will be completely bare - where I will solder the voltage limiter.
How bad will this be at ~300MHz?
I could create a shield for the PTFE section, either with a machined brass tube or by painting it with silver paint, if necessary.
I will probably use a little metal box for the whole thing, just to make it neat.
Typically I use a small blue Pomona box with the two BNCs of choice already installed at each end. Inside the box you can use a short shielded cable to reach the component. Typically I'll use a small length of rigid coax. For smaller stuff a 1" to 2" length of 3/8" copper pipe to which I insert the threaded connector part into and then solder (some light machining required). Done properly both are waterproof too. If critical then check for VSWR or by TDR to make sure all is OK.
-
Excellent - thanks. I started looking for little BNC boxes on Ebay but you reminded me of the "ITT Pomona" boxes which used to be everywhere in the 1970s... Digikey sell a 2931. I have some semi-rigid RG coax too.
-
I have finally built it. Any comments would be appreciated. I wonder if the way the semi rigid coax shield is grounded is ok. Unfortunately there is no easy way to connect to the BNC connectors directly. I could have cut the BNC pins down in length to reduce the unshielded portions. But this is for about 300-350MHz.
(http://peter-ftp.co.uk/screenshots/2017-10-21_162950.jpg)
-
That looks like a nice job.
If I remember correctly you mentioned a 330MHz working frequency so, once with the cover installed (and grounded to the case, assuming the cover is metal. Shave some of the paint at each corner.), the losses should be tolerable.
If easy to do, you might double up on entry and exit grounds but I doubt if it will make an appreciable difference.
You might want to do a few receiver tests with the box in and out to see if you can measure any loss in sensitivity.
-
Actually, the datasheet says it's got zener diodes internally. Curiously, they don't give an ESD rating, so it's not clear if they're good for anything at all...
Irrelevant now, but Infineon BF998 is AEC Q101 qualified, which includes ESD rating AFAIR.
-
I don't think this is ESD. The inductor at the radio input should kill that - as posted above. I think it is RF, at just the "right" frequency.
-
Well, ESD is a wideband pulse, a direct strike will still excite a tuned circuit. An RF pulse is an RF pulse. >:D
Tim
-
A late update: the insertion loss of this little device has been tested and it is unmeasurable, at some fraction of a db.
Thank you all for your input :)
-
and the receiver front end is still OK or only time will tell?
Thanks for the feedback :)
Cheers.
-
Only time will tell... however I would be amazed if it gets blown up again.
I also tested the box with a 0-30V power supply, via a ~10k resistor to limit the current, to make sure it actually worked as a clamp. That 0402 chip was so small that checking the soldering needed a x40 microscope.
-
A further update: It worked perfectly in flight on a real ILS - as expected after being tested with an IFR4000 tester and found to have an insertion loss below what is measurable.