Thanks everybody. A prototype batch of 20 29mm Blue Ravens is submitted and is scheduled to ship to me around July 5. I'm getting married on the 9th, so I may not get time to test it until the week of the 17th.
Blue Raven in 29mm bulkhead form factor
- Thread starter Adrian A
- Start date
Help Support The Rocketry Forum:
Hey, congrats on your wedding plans! Honeymoon at LDRS?
Congratulations!
Ha!
Neutronium95
Well-Known Member
- Joined
- Jan 17, 2010
- Messages
- 1,082
- Reaction score
- 1,580
I'll buy one as soon as it's available.
Congratulations on the impending wedding. Wishing you all the best for the future.
Congrats Adrian!!!
Congratulations Adrian and best wishes to your bride to be !
( And thanks for the update )
--kjh
I decided to do a little kick-starter program for the first 14 units that will be available for sale. If you want to pre-order one or two, this is the link to the hidden store page. As it says on that page, if this first round of units works as-is, or just requires small mods I can do by hand, they will be shipping in mid-late July. If there is a major problem with this design that needs a new production round to fix, these pre-orders will be first in line for the next production round, likely shipping in late August.
Last edited:
It fits! Next is to figure out if it works...(That’s a Blue Raven inside a 29mm coupler inside a 29mm airframe)
Adrian --
The 29mm Blue Raven is beautiful !
So ... I imagine you'll be selling a "Featherweight 29mm av-bay harness attachment upgrade kit" to seal the ends of my AV-Bay and to connect my Shock Cords
Like this: Featherweight 38mm av-bay harness attachment upgrade kit but to fit the ends of a 29mm coupler with three threaded rods, instead of four ...
I will want one when they're available.
Or ...
Would it be OK to use the Active and Passive Ends of the existing 29mm AV-Bay with the new 29mm Blue Raven ?
Would that not place 2 mag switches in the circuit ?
Won't I at least need a second passive bulkhead ?
Is it too soon to ask what you might have in mind when I am ready to order a new, 29mm Blue Raven ?
Thanks again Adrian !
The 29mm Blue Raven is a really a beautiful thing !
-- kjh
p.s. the 38mm av-bay harness attachment upgrade kit is wonderful !
I am building a 38mm Min Diameter Model Rocket around it now !!
The 29mm Blue Raven is beautiful !
So ... I imagine you'll be selling a "Featherweight 29mm av-bay harness attachment upgrade kit" to seal the ends of my AV-Bay and to connect my Shock Cords
Like this: Featherweight 38mm av-bay harness attachment upgrade kit but to fit the ends of a 29mm coupler with three threaded rods, instead of four ...
I will want one when they're available.
Or ...
Would it be OK to use the Active and Passive Ends of the existing 29mm AV-Bay with the new 29mm Blue Raven ?
Would that not place 2 mag switches in the circuit ?
Won't I at least need a second passive bulkhead ?
Is it too soon to ask what you might have in mind when I am ready to order a new, 29mm Blue Raven ?
Thanks again Adrian !
The 29mm Blue Raven is a really a beautiful thing !
-- kjh
p.s. the 38mm av-bay harness attachment upgrade kit is wonderful !
I am building a 38mm Min Diameter Model Rocket around it now !!
Adrian --
The 29mm Blue Raven is beautiful !
So ... I imagine you'll be selling a "Featherweight 29mm av-bay harness attachment upgrade kit" to seal the ends of my AV-Bay and to connect my Shock Cords
Like this: Featherweight 38mm av-bay harness attachment upgrade kit but to fit the ends of a 29mm coupler with three threaded rods, instead of four ...
I will want one when they're available.
Yes, I will do that. In the meantime, you can make your own by just drilling a 29mm diameter bulkhead for the three rods and whatever structural attachments you want.
The active bulkhead is not what you would want. I'd have to think about whether the passive bulkhead could work.
Yes, because the 29mm Blue Raven needs to live inside an av-bay that's sealed from ejection charges, and it doesn't provide that sealing bulkhead itself (unlike the 29mm av-bay for the Blue Raven)
I'm glad you like it!
The first two units I have tested have a bad gyro/accel sensor chip or chip installation. I can solder down my a sensor chip from my inventory on there and it works. I'll try re-installing the chip on some of the assembled units to see if it was just an assembly issue or if it's a parts problem.
The power switch that is supposed to be mag-controlled is just puts a diode drop in series when it's off. So I probably designed the P-channel FET in backwards. Otherwise, I'm pretty psyched that so much is working on the first try.
The power switch that is supposed to be mag-controlled is just puts a diode drop in series when it's off. So I probably designed the P-channel FET in backwards. Otherwise, I'm pretty psyched that so much is working on the first try.
- Joined
- Feb 3, 2012
- Messages
- 6,354
- Reaction score
- 5,565
Those fine-pitch parts are a pain, huh? It's got me looking at pick-and-place machines... I'm not sure I can mount an SSOP reliably (0.65mm pitch), although I think I've got the knack of the ADXL375's (0.8mm)... but I gave up on the 0.5mm ADXL372's.
Yeah, I'm not sure what the problem was. The third unit was fine, so that points the finger toward the assembly rather than the part. Not sure why these are having issues when this part that hasn't been a problem for the hundreds of regular Blue Ravens so far. Replacing one or two of these parts by hand is o.k. with a microscope and a hot-air rework station. I tin the pads on both sides with a fine-tipped soldering iron and then use the hot air tool to float it into place.
On the magnetic switch, the schematic symbol for the FET is oriented correctly, so maybe this particular part has a non-standard pinout? I'll look into this further later.
The Bluetooth range on the 29mm Blue Ravens is good, at least as good as on the original Blue Raven.
Last edited:
There is a technique where you set the iron to the correct temperature, flux all the pins, and with a tip that has a hollow to hold some solder you just start at one end and move he iron along at about a 45deg angle. As you move along the solder flows to the joint, and as you move past it the excess solder goes back to the tip. You do need to "bump" solder a couple of corner pins to hold the package in place before soldering a side like this. This works a treat for the 0.5mm-pitch parts.
You can see the technique in this video by the IPC.
Having a good iron really helps (I use and recommend JBC). Also, remember that solder follows three things; heat, flux, and other solder. By controlling those you can control where the solder goes.
I confirmed that the high-side MOSFET I selected for the built-in magnetic switch has the source and drain backwards from the usual convention for 3mm x 3mm FETs. With a simple part substitution from my inventory the magnetic switch is working properly. I'll need to do a few more tests but it's looking like the board layout is sound.
I fit-tested two redundant Blue Raven 29mms, a tracker, and 2 batteries (the tracker would share with one of the Blue Ravens) inside a 29mm coupler tube, and the length was 2.5."
I did notice that the small pattern nuts for the threaded rods had some tight clearance to other components on the board, and can even touch components they're not supposed to if the threaded rod is not centered in the hole. For the next production run I'm tightening up the clearance hole for the rod and looking to see if I can squeak out a little more clearance around the nuts.
I did notice that the small pattern nuts for the threaded rods had some tight clearance to other components on the board, and can even touch components they're not supposed to if the threaded rod is not centered in the hole. For the next production run I'm tightening up the clearance hole for the rod and looking to see if I can squeak out a little more clearance around the nuts.
Thanks for the reminder Adrian ...
I been meaning to ask what material those 4-40 Small Pattern Nuts are made of.
I found 8-18 and 316 stainless steel that 'may be mildly magnetic' and they also have brass and plain old steel Narrow Hex Nuts at McMaster-Carr:
McMaster-Carr Narrow Hex Nuts
Anyhow, no rush, I've got a few of them that you provided with the 38 mm ( -and-or- the 29 mm ) AV-Bay.
But I was thinking that the Small Pattern Hex Nuts would be a 'better fit' all-around for the inside nuts on the Featherweight AV-Bay Coupler Bulkheads.
Thanks again Adrian.
-- kjh
I been meaning to ask what material those 4-40 Small Pattern Nuts are made of.
I found 8-18 and 316 stainless steel that 'may be mildly magnetic' and they also have brass and plain old steel Narrow Hex Nuts at McMaster-Carr:
McMaster-Carr Narrow Hex Nuts
Anyhow, no rush, I've got a few of them that you provided with the 38 mm ( -and-or- the 29 mm ) AV-Bay.
But I was thinking that the Small Pattern Hex Nuts would be a 'better fit' all-around for the inside nuts on the Featherweight AV-Bay Coupler Bulkheads.
Thanks again Adrian.
-- kjh
I'm getting ready for a test flight of the 29mm system tomorrow at Tripoli Hartsel.
My test rocket has a spot for my tracker in the nosecone, and I kinda need that mass up there because after losing my last aft section from a broken shock cord, I'm using a 29mm aft section that has fins a little smaller than I would prefer.
So I decided to put 2 29mm Blue Ravens into one av-bay with 2 small batteries in between. I have a couple of different forward sections for my test rocket. One has an extra-long coupler so that I could put 2 rectangular Blue Ravens end-to-end. My other one has a 2" long coupler, and I wanted to see if I could get two 29mm Blue Ravens with 2 batteries into that length. The normal 160 mAhr batteries that we ship with the av-bay kits are a little long for that, but I had previously purchased some 110 mAhr batteries, and I'm giving those a shot. I switched out the JST 2.5 mm connectors for JST 1.25 mm connectors.
Here's the whole av-bay, including structural sealing bulkheads and harness/charge attachment bulkheads, and charge holders. It all comes in at 36 grams.
The top Blue Raven in this photo has the battery connector facing outward, but fortunately, there are 2 holes I could use to thread the wires through to the other side. I'm also using some aluminum spacers underneath the small-pattern nuts so that I have room to use a 3/16 nut driver. This made assembly much, much easier than trying to tighten the nuts in close quarters with needle-nose pliers.
On the bottom side, the electrical components are low enough profile so that I didn't need more than the small-pattern nuts as a spacer before the sealing bulkhead. On the green bulkheads that go on the outside of the coupler, there is room for standard 4-40 nuts, which are easier to use for attaching the deployment charge wires.
The small battery cells I'm using have some cell protection circuitry, so I fully assembled the av-bay and ran a simulated flight test with a 1.1 Ohm load and a fully-charged battery to see if the cell protection circuit would trip off.
The cell protection circuitry did not get in the way, despite a current briefly exceeding 2.5 Amps.
My test rocket has a spot for my tracker in the nosecone, and I kinda need that mass up there because after losing my last aft section from a broken shock cord, I'm using a 29mm aft section that has fins a little smaller than I would prefer.
So I decided to put 2 29mm Blue Ravens into one av-bay with 2 small batteries in between. I have a couple of different forward sections for my test rocket. One has an extra-long coupler so that I could put 2 rectangular Blue Ravens end-to-end. My other one has a 2" long coupler, and I wanted to see if I could get two 29mm Blue Ravens with 2 batteries into that length. The normal 160 mAhr batteries that we ship with the av-bay kits are a little long for that, but I had previously purchased some 110 mAhr batteries, and I'm giving those a shot. I switched out the JST 2.5 mm connectors for JST 1.25 mm connectors.
Here's the whole av-bay, including structural sealing bulkheads and harness/charge attachment bulkheads, and charge holders. It all comes in at 36 grams.
The top Blue Raven in this photo has the battery connector facing outward, but fortunately, there are 2 holes I could use to thread the wires through to the other side. I'm also using some aluminum spacers underneath the small-pattern nuts so that I have room to use a 3/16 nut driver. This made assembly much, much easier than trying to tighten the nuts in close quarters with needle-nose pliers.
On the bottom side, the electrical components are low enough profile so that I didn't need more than the small-pattern nuts as a spacer before the sealing bulkhead. On the green bulkheads that go on the outside of the coupler, there is room for standard 4-40 nuts, which are easier to use for attaching the deployment charge wires.
The small battery cells I'm using have some cell protection circuitry, so I fully assembled the av-bay and ran a simulated flight test with a 1.1 Ohm load and a fully-charged battery to see if the cell protection circuit would trip off.
The cell protection circuitry did not get in the way, despite a current briefly exceeding 2.5 Amps.
Attachments
There was a little odd behavior, though, when I did the test.. When the main charge fired, I heard one of the Blue Ravens start up. At first I thought it was the one that I had turned on resetting, but it kept going through the whole flight. Instead, it was the other Blue Raven, which was powered off but with outputs wired in parallel with the other Blue Raven.
When the main output fires, you can see about 0.3 Amps of current which quickly ramps down, which is what happens when a board is first powered up and charges its hold-up capacitor. I had to go back to a simplified diagram I had previously made for a shared-rod dual altimeter setup to convince myself that nothing was amiss:
Microcontroller 1 is off and microcontroller 2 is on. Not shown (due to laziness) are the body diodes for each of the MOSFETs represented by the switch symbols. But the body diode of the Blue raven 1 main output switch is key to this sneak path when Blue Raven 2 turns on its main output and the main charge is disconnected:
If you want to follow this, I recommend starting at Batt2 + side. When its power switch is on , it energizes the Arm Out rod, which normally wouldn't turn on Blue Raven 1. But when Blue Raven 2 connects Main output to its own ground, it allows the return current for Blue Raven 1 to flow back through the Main 1 body diode and through its own main switch in the normal direction to complete the circuit. This is also why you see the Main output with some voltage on it when the main switch isn't being fired and there is no Main charge connected. I don't think any of this is harmful for this application. Normally both altimeters will be on, and I want both of them to fire each of the charges.
When the main output fires, you can see about 0.3 Amps of current which quickly ramps down, which is what happens when a board is first powered up and charges its hold-up capacitor. I had to go back to a simplified diagram I had previously made for a shared-rod dual altimeter setup to convince myself that nothing was amiss:
Microcontroller 1 is off and microcontroller 2 is on. Not shown (due to laziness) are the body diodes for each of the MOSFETs represented by the switch symbols. But the body diode of the Blue raven 1 main output switch is key to this sneak path when Blue Raven 2 turns on its main output and the main charge is disconnected:
If you want to follow this, I recommend starting at Batt2 + side. When its power switch is on , it energizes the Arm Out rod, which normally wouldn't turn on Blue Raven 1. But when Blue Raven 2 connects Main output to its own ground, it allows the return current for Blue Raven 1 to flow back through the Main 1 body diode and through its own main switch in the normal direction to complete the circuit. This is also why you see the Main output with some voltage on it when the main switch isn't being fired and there is no Main charge connected. I don't think any of this is harmful for this application. Normally both altimeters will be on, and I want both of them to fire each of the charges.
Attachments
I have revised the board design to provide a little more room for the threaded rod pads, and to specify a part with the correct pinout for the magnetic switch pass FET. The holdup now is that 7 out of the first 10 boards have had a problem with the gyro/accel part. Not sure if it was a bad batch of parts or (more likely) an assembly issue. I'm in contact with my assembler to investigate the issue. In the meantime I'm working on my production tester that I'll use to evaluate the other 10 boards that are still on the panel.
Adrian --
How is testing going on this little beauty ?
I've been thinking about the threaded rod electrical connections and my existing AltAcc AV-Bays a bit.
Say I've got an existing AV-Bay with a sled at least 1-inch wide.
Q1: Would I not be able to lay the 29mm Blue Raven flat on the sled, connect a battery fastened under the sled via a 1.25mm JST 'extension cord' and then connect Ground, Apogee and Main to the sled via brass screws with short stand-offs to the three existing threaded rod connections ?
If so, the three brass screws could be wired to a pair of screw terminal blocks with a shared ground.
But I believe I would like to try to align one axis of the accelerometer and gyro with the vertical axis of the sled.
Q2: Which way should I try align the mounting holes for the 29mm Blue Raven on the sled ?
Thanks !
-- kjh
How is testing going on this little beauty ?
I've been thinking about the threaded rod electrical connections and my existing AltAcc AV-Bays a bit.
Say I've got an existing AV-Bay with a sled at least 1-inch wide.
Q1: Would I not be able to lay the 29mm Blue Raven flat on the sled, connect a battery fastened under the sled via a 1.25mm JST 'extension cord' and then connect Ground, Apogee and Main to the sled via brass screws with short stand-offs to the three existing threaded rod connections ?
If so, the three brass screws could be wired to a pair of screw terminal blocks with a shared ground.
But I believe I would like to try to align one axis of the accelerometer and gyro with the vertical axis of the sled.
Q2: Which way should I try align the mounting holes for the 29mm Blue Raven on the sled ?
Thanks !
-- kjh
that is really really fricken cool, I'm planning to get some of these for Christmas. I spoke with @bad_idea and some others at the TNT Shootout about the Blue Raven and it's such an impressive unit.
Thanks Adrian (and all the vendors out there) for pushing the hobby forward and perpetually raising the bar.
Testing is going pretty well. I had a successful 120G flight in a 29mm test rocket, and tomorrow I'm going to shoot for the J record on a new 38mm rocket, with two of these and a Featherweight GPS tracker for the electronics, inside a 29mm airframe tube that's underneath my chute cannon. That flight is expected to pull 100 Gs, sending the rocket 0 to Mach 2.8 in 1.0 seconds.
I updated the design to provide a little more room for the mounting nuts, and re-oriented the accelerometers so they are aligned with each other, as they are on the standard Blue Raven, which will allow them to be calibrated without any special software. I also needed to change the FET that's used for the built-in magnetically-activated power switch. If I don't get any bad surprises tomorrow I'll place an order for another short production run with the updated design, probably on Monday.
A1: Yes.
A2: You will need to ID the accelerometers on the board and align them. The accels are rotated to have their axes aligned and fit on the board, but they're not aligned with anything else on the board. I would actually shoot for a 45 degree off-angle so that the primary rocket axis is picked up by 2 axes of accelerometer, which will extend your more-accurate accelerometer range to 45 Gs, up from the single-axis 32 Gs. The 400G accelerometer works, but it's not as accurate as the lower-range one.
Last edited:
Thanks for the reply Adrian,
As always, there are lots of ideas to ponder.
Good luck with your launch today -- It sounds like another whiplash-inducing flight
Looking forward to your launch report as well as the performance report on the new 29mm Blue Raven !
-- kjh
EDIT: spelling - lanch -> launch
As always, there are lots of ideas to ponder.
Good luck with your launch today -- It sounds like another whiplash-inducing flight
Looking forward to your launch report as well as the performance report on the new 29mm Blue Raven !
-- kjh
EDIT: spelling - lanch -> launch
Last edited:
The flight result was posted over in the High Power forum. So far what I have posted about the post-flight analysis was focused on the fin flutter from that flight, and I noted that the main chute was tangled when I recovered the rocket, leading to a hard landing and two broken-off fins. But I did have another surprise when I looked more closely at the electrical data.
Background: For this flight I had two Blue Ravens sharing the same threaded rods that are connected to the Apo and Main channels. The top Blue Raven had a standard 160 mAhr battery, and the bottom Blue Raven used a 400 mAhr cylindrical cell that it shared with the GPS Tracker. The tracker's ground is connected to the bottom Blue Raven's ground. The threaded rod with the white sleeve is the apogee output.
First the expected part, at apogee.
The bottom Blue Raven detected Apogee first, and it fired the apogee charge. With the large cylindrical battery, the initial current was pretty high, approaching 16 Amps. (if it had continued for more than 1 msec, the Blue Raven would have cut back the current). The apogee voltage, which is recorded with the low rate data at 50 Hz, shows the switch turning on.
The accelerometer recorded a backward impulse from the nosecone getting ejected, some harness reefing, and then a big forward impulse when the nosecone reached the end of the cord. So far, so good.
But then, the main charge continuity voltage dropped to zero, and kept dropping to zero at 1 second intervals, with some exceptions:
The accelerometer data also shows that the main chute charge fired at this time:
But the software says "hey it wasn't me, I waited until it got down to the main chute deployment altitude"
Here is a closeup of the times when the top and bottom Blue Ravens fired the main channel intentionally for the usual 1 second duration, at 262 and 267 seconds in to the flight, respectively:
Then it occurred to me that the GPS transmitter is transmitting once per second, and so I took some more measurements and confirmed that the duration of the repeated "on" events matched the transmission duration. Later I was able to repeat the unintentional firing at my workbench by putting the Apo charge wires right next to the transmitter antenna when all three units were powered up in a flight-like configuration, confirming that the main chute output switch was getting turned on by the RF output from the transmitter.
Next question is how could the main channel output switch could be susceptible to the RF transmissions? One part of this is that the transmitter shares the ground with the bottom Blue Raven, which connects its high side with the top Blue Raven through the Arm+ threaded rod. I don't think this susceptibility would be possible except in this tight 3-device configuration. Another part is that the Apogee charge wire was taped down parallel to, and very close to, the tracker antenna. It did not cause the interference until the Apogee charge had fired, probably because firing the charge breaks open the connection, turning the apogee wire into another single-ended antenna. But the biggest contributor is one that is now fixed in the newest firmware build, available in phone build 206. This build is available through TestFlight and FireBase, and it should soon be available through Apple's App store and Google Play.
The output channels are set up as timers with push-pull outputs, in order to implement the current limiting feature. I thought that the output switch gates were pulled down by the microcontroller before the the outputs are turned on by activating the hardware timer. But when I measured it, it turns out that before the timer output is activated, the line going to the output FET gate only pulled down by the external 10kOhm gate pull-down resistors. I have modified the firmware now so that the gate is pulled down to ground by the microcontroller before it is activated, which increases the measured pull-down strength from 10kOhms resistance to about 13 Ohms, a factor of more than 700x. This change will improve the RF susceptibility for both the 29mm Blue Ravens and the original Blue Ravens. I confirmed that this prevented the susceptibility from happening in the same test configuration where I could reproduce the problem in the lab, and even when the transmitter antenna was right on top of the apogee charge wire, before or after firing.
Background: For this flight I had two Blue Ravens sharing the same threaded rods that are connected to the Apo and Main channels. The top Blue Raven had a standard 160 mAhr battery, and the bottom Blue Raven used a 400 mAhr cylindrical cell that it shared with the GPS Tracker. The tracker's ground is connected to the bottom Blue Raven's ground. The threaded rod with the white sleeve is the apogee output.
First the expected part, at apogee.
The bottom Blue Raven detected Apogee first, and it fired the apogee charge. With the large cylindrical battery, the initial current was pretty high, approaching 16 Amps. (if it had continued for more than 1 msec, the Blue Raven would have cut back the current). The apogee voltage, which is recorded with the low rate data at 50 Hz, shows the switch turning on.
The accelerometer recorded a backward impulse from the nosecone getting ejected, some harness reefing, and then a big forward impulse when the nosecone reached the end of the cord. So far, so good.
But then, the main charge continuity voltage dropped to zero, and kept dropping to zero at 1 second intervals, with some exceptions:
The accelerometer data also shows that the main chute charge fired at this time:
But the software says "hey it wasn't me, I waited until it got down to the main chute deployment altitude"
Here is a closeup of the times when the top and bottom Blue Ravens fired the main channel intentionally for the usual 1 second duration, at 262 and 267 seconds in to the flight, respectively:
Then it occurred to me that the GPS transmitter is transmitting once per second, and so I took some more measurements and confirmed that the duration of the repeated "on" events matched the transmission duration. Later I was able to repeat the unintentional firing at my workbench by putting the Apo charge wires right next to the transmitter antenna when all three units were powered up in a flight-like configuration, confirming that the main chute output switch was getting turned on by the RF output from the transmitter.
Next question is how could the main channel output switch could be susceptible to the RF transmissions? One part of this is that the transmitter shares the ground with the bottom Blue Raven, which connects its high side with the top Blue Raven through the Arm+ threaded rod. I don't think this susceptibility would be possible except in this tight 3-device configuration. Another part is that the Apogee charge wire was taped down parallel to, and very close to, the tracker antenna. It did not cause the interference until the Apogee charge had fired, probably because firing the charge breaks open the connection, turning the apogee wire into another single-ended antenna. But the biggest contributor is one that is now fixed in the newest firmware build, available in phone build 206. This build is available through TestFlight and FireBase, and it should soon be available through Apple's App store and Google Play.
The output channels are set up as timers with push-pull outputs, in order to implement the current limiting feature. I thought that the output switch gates were pulled down by the microcontroller before the the outputs are turned on by activating the hardware timer. But when I measured it, it turns out that before the timer output is activated, the line going to the output FET gate only pulled down by the external 10kOhm gate pull-down resistors. I have modified the firmware now so that the gate is pulled down to ground by the microcontroller before it is activated, which increases the measured pull-down strength from 10kOhms resistance to about 13 Ohms, a factor of more than 700x. This change will improve the RF susceptibility for both the 29mm Blue Ravens and the original Blue Ravens. I confirmed that this prevented the susceptibility from happening in the same test configuration where I could reproduce the problem in the lab, and even when the transmitter antenna was right on top of the apogee charge wire, before or after firing.
- Joined
- Aug 6, 2022
- Messages
- 3,269
- Reaction score
- 2,968
Gotta love a vendor who crashes his own rockets before customers have a chance, and then fixes the problem and tells us all about it.
A sniff of capacitance to ground on those long lines that can be come antennas, to ground, is a good way of reducing RF-induced voltages. Doesn't need much to give a low impedance at high frequencies.
Similar threads
