rgreen24

My Feedback: (6)

Any pics of your set up Keith?

smcharg

My Feedback: (1)

As an update, I decided to go with two PLR5 failsafe switch w/ regulator.  The reason is I just don't like the idea of unplugging my RX connection to the lipo every time I take my canopy off.  I do not draw power from the motor's 10S pack and keep them totally separate.  Yes, it added 50g to my airplane but I'm still at 4850g RTF so the weight wasn't a concern.  I do like the ability to check each regulator with the failsafe switch just by turning each one on individually and then both on for the flight.  For me, this made the most sense.  This check is very similar to checking the magnetos in a Cessna...switch to each one to make sure they are functioning and back to both for the flight.

Thanks for everyone's input,
Scott

OhD

I just tested the set up and it seems to work fine. I did measure about 0.4 volts peak to peak at the input to the Jaccio regulator but the output was clean as I expected it would be. A few years ago I worked with a company that was flying a big helicopter with a TV camera/transmitter on board. The RC system was using the same battery pack as the TV system and every time a servo drew current they could see it on the video. We put a regulator between the battery and the TV system and it filtered out the ripple on the input and solved the problem.

I had been thinking about a redundant power system and this looks like an easy way to do it. While I'm at it I will put a voltmeter on the connection to the bottom two cells that will tell me when I plug it in that the pack is charged (or not). You never want to take off with a pack that isn't charged.

Dean Pappas

Hello Jim,
Is that ripple at commutation rate or at the chopping rate? Sorry I couldn't read the timebase.
400 mV on two cells! I guess it's not that surprising. So that works out to maybe 2V peak to peak across the entire pack.
Sounds like more filtering at the ESC would lessen the current peaks at the battery. I wonder if the batteries will respond well to that.

Regards,
Dean Pappas

OhD

The timebase was .2 msec per division so it looks like it is the 8KHz PWM frequency. I never went to full throttle but I would assume the ripple would disappear at 100% duty cycle as you would get a fairly pure DC at that point. I don't know what the effective series resistance is across the two cells but the 400 millivolts seems reasonable if the R was 10 milliohms and the I was 40 Amps.

We need the regulator to get down to 6 volts and it does the filtering so I don't see a need to do anything else at the battery/ ESC interface.

What have you been up to? What ever happened to electric control line? I can't keep up with all the stuff that is going on but I find a lot of folks getting ahead of the pattern guys when it comes to new technology.

Regards, Jim

Dean Pappas

Hi Jim,
So it's the chopping rate that we see. Okay, that's high enough to easily filter out of the regulator output, and low enough not to cause any RF issues, seeing how the RX ground is going to end up riding on top of a common mode noise signal at the same frequency.
I wonder if there is any real high frequency crap as well? I can't imagine that the big electrolytic caps we see on any firm's ESCs are any use at RF. I would imagine that 2.4G is completely unaffected, but 72 Meg just might get unhappy.

E-Stunt? Let me see if I can sum it up in just a few words. It's GREAT! There are a fair number of holy-war, "don't change my event" naysayers, but it works better than even the best piped two-strokes and four-stroke setups.

1)We are running with the ESC set in helicopter RPM governor mode, and if you have the gain adjusted high enough and the RPM control loop response speed set fast enough (just a safe margin of stability, but noticeably underdamped) then ...
2)The vertical uphill RPM is only 100 RPM or so below that of level flight and,
3)vertical dive RPM is maybe 100 or so above level flight.
( for comparison, excellently set up four-strokes and piped two-strokes exhibit about 1000 RPM differences from climb to dive and the classic muffled 2-4 Stunt run shows close to 2000 RPM difference)
4) There are experimenters looking to make the system recognize where the plane is on the circle so as to compensate for the few hundred RPM variations we still see, but this is still far from ready for prime time.
5) Podium places at Worlds have been taken with electric, and this year's European Continental Championship has been won with Electric.

Now for the cross-pollination part of our discussion ... The constant RPM setup surely will work for Pattern: it will turn the throttle into an "airspeed setting lever".
The only problem is that this only works if you have enough overhead to achieve the governed RPM at the slowest speed needed while in a vertical climb.
As a data point, static current at the set RPM is about 135% of the cruise flight current at the same RPM.
I strongly suspect that the very long vertical downlines we fly, and the lack of control-line-drag will require an RPM governor that engages the regenerative brakes.

In my opinion, this approach will get close to the point where sensored motors will be needed for the frequent changes from driving to braking at high energy levels.
We already see that ESCs have issues with the transition from braking to running hard at low RPM. Sensors will fix that.

take care,
Dean P.

OhD

Thanks for the report Dean,

Not sure I understand "...static current at the set RPM is about 135%..". Does that mean the current at zero velocity?

I believe we have the overhead required as I have done the Masters pattern never exceeding 2500 Watts and our batteries are getting better. However, I'm sure there will be conditions where we might need more power but we don't need perfect regulation. I agree the sensors might be the answer to the control problem. It looks like Futaba uses them in their brushless servos. I've also heard rumors they might have an RPM controlled power system coming.

Stay tuned, things just keep getting better, Jim

jgg215

Jim and Dean,
Seems like the right participants are here for me to ask a question.

It appears that we all would be be better off if the throttle stick operated as an RPM request.  This would eliminate or greatly reduce any problems with reduced power availability later in the flight as the batteries discharge. We could all get used to a given stick position being a given RPM(power for a given set of conditions). The batteries these days typically have excessive power available at the beginning of the flight.

I don't see why any additional hardware or sensors would be required to do this. The RPM is currently reported in the CC ICE as an output plot so it is clearly available in the controller software. All that would be required in the software is a simple control loop based on throttle stick input and RPM feedback. Possibly more processing power would be required but the code is simple enough.   It would be best if you could set min RPM and max RPM as parameters input through the data link. Any deviation from a straight line between the two RPM end-points could be handled in the transmitter.

What am I missing here? Why isn't this available to me and at a reasonable premium? I am fairly new to electric,at least for pattern and this appears to me to be the only substantial disadvantage of electrics compared to the YS.

John Gayer

OhD

Hi John,

I'm no control system or servo designer but I do know the dynamics are quite complicated. Controlling the RPM at a given value is not the problem, it is the requirement to change the desired RPM quickly as we move the stick quickly. Measuring the RPM cannot be done quickly. What you read out is an average of some number of cycles over time. To have the desired frequency response you'd need to measure angular velocity many times per revolution. Then with the high inertia you'd need lots of change in torque in both directions to control the angular velocity.

Having said all this, I'm sure it will be done eventually. Right now I still feel the advantages of the electric system far outweigh the change in "throttle feel" from the beginning until the end of a flight. The guys that practice can control the speed as well as the path of the aircraft. I'm still working on the path as well as he speed.

Jim

Jetdesign

My Feedback: (8)

Wouldn't a 35V regulator on the power batteries give constant throttle feel? Isn't that why we use regulators on the Rx battery in the first place (consistent control feel)?

I'm learning this as I've been working on understanding my power setup and looking at the graphs. Thought I was dropping down in power, really I was running a conservative setup and staying in the 'fresh zone' for a longer period of time. Throttle feel is pretty consistent once the battery voltage drops down and levels off - the whole point of using lipos is for that 'voltage plane' which levels out for quite some time.

Not that I think I would bother using a regulator, at least at this point - really just need to learn to slow down my trim pass and fly at constant speed from there.

OhD

The ESC IS a regulator that is regulating voltage to the motor rather than regulating RPM, as a function of stick position. You would never want to put two regulators in series.

Yes we do use a regulator to keep the servo voltage and speed consistent as the battery voltage drops during flight and from flight to flight. It shouldn't be required in Futaba's new programmable brushless servos where you can program speed. I will try to run some tests and let you know if this is true.

Jim

nonstoprc

My Feedback: (90)

Sounds like the ESC could do a better regulating job by taking into consideration the voltage: higher voltage - shorter power pulses; lower voltage - longer power pulses.

I can definitely feel more power when the pack is fresh (close to 42v) with CC HV 85. With CC ICE HV 80, the extra "power surge" is less evident.

jgg215

Jim,
I see your point about how long it would take to determine the RPM to a reasonable accuracy. It would depend on the frame rate used in the processor but it is certain to be a substantial number of frames.
Having this capability cannot be too far off if Dean can control to a static RPM within +/- 100 RPM for controlline. Sounds like everything is in place except the response time of the RPM delta,
John

Dean Pappas

Hi All,
Jim, yes zero-airspeed is what I meant by static conditions, and you're right about needing to evaluate the available overhead.
Oh yes, and you're too darned modest.

As far as RPM governed operation goes, I can easily say it will lessen the pilot's workload, but I already think airspeed control is easier with E-Power.
When well optimized, the exponential response time to a step change in load is about 1/4 second. This is the case with the Castle Phoenixes, at least.
The CL setups use no braking except for a hard brake at motor shutdown. Our planes are cleaner and heavier and lacking control-line drag we need the brakes going downhill.

Like I said, the high energy transition from pulling to regenerative braking and back to pulling, while the RPM is commanded to maybe 5500 is more than any of the ESCs out there are capable of.

I think there is a way around the problem, so that E throttle management is less of a handful.

later,
Dean

smcharg

My Feedback: (1)

HI guys,
   I went out and tested my new setup with the two Tech-Aero regulators and separate packs yesterday.  I flew several flights and when I got back home, I pulled the packs out to check voltage.  The pack that has always been there with a single regulator read the same as it usually does after 3 flights and the backup pack seemed to almost be at a full charge.  I check to make sure each pack is powering the receiver as described before by turning on each switch separately and then both on for flight.
  Maybe I had the wrong idea about this so I'm just trying to confirm.  How does this actually work?  Does the rx draw off the main pack that's plugged into the "Battery" slot on the RX and discard the pack that's plugged into a spare channel or does it draw off of both at the same time equally?  I would think if it was equal and assuming both packs started out at the same voltage, they would at least be close to each other.  The "primary pack" seems like it got used a lot more than the other.  How does this actually work for the RX?

Thanks,

OhD

The receiver will draw current from the regulator that is set to the highest voltage. You should be able to put a meter on the system when you do your one at a time test to detect the difference between the two regulators.

Jim

rcpattern

My Feedback: (45)

Scott,

How do you have the voltages setup? The regulators have to be matched for them to draw equally. The programmer Ed has is very handy for this. If the regulators are close, it will draw from each, but as Jim said, if they aren't, it will always draw from the higher one until the voltage drops below the other.

Arch

smcharg

My Feedback: (1)

Hey Arch,
   I used the programmer box and they are both set at 6.0V.  Maybe I should check again but unless the regulators have arbitrarily changed their setting, they'd both be set to the same.

jgg215

Scott,
If you have an ESV with a built-in load, use that to confirm you are  getting 6.0 volts from each regulator separately under load, disconnected from the receiver.  If they both read 6 volts, lower the one that is taking the load to 5.9 and try again. I'm pretty sure you won't need t do that as they won't both be at 6 volts.
The receiver doesn't care which slot you plug the battery into.
John

rcpattern

My Feedback: (45)

Yea, that was going to be my suggestion. Slightly adjust one of them and see what it does. Another idea may be to swap batteries. One of your flight packs may be stronger as well. Granted, this should NOT make a difference with the regulators, but I have noticed that in some cases it does.

Arch

smcharg

My Feedback: (1)

I'll do these tests but just to make sure I got it, it doesn't really matter that one draws first and/or second. Since we know that each one works independently we know it is acting as a redundant system yes?

nonstoprc

My Feedback: (90)

It probably is better to allow the two RX batteries to work equally hard as otherwise one of them is always 100% charged. Not very good for the health of LIPOs.

Two tiny 250mah lipo packs should be enough for 5-6 flights with about 50% charge left.

MachLoop

My Feedback: (3)

Scott,

From the Tech-Aero site:

There are two basic schools of thought about how to do a redundant power delivery system and both are valid and effective. These are the Balanced Redundant Supply (AKA Dual Supply) and the Hot Standby Redundant Supply (AKA Backup Supply).

Here is a link to the explanation of how each is configured and works: http://d7kx18lgvr3p4.cloudfront.net/Tech-Tip%200601.pdf

smcharg

My Feedback: (1)

Hey Michael,

That link to the tech-tip answered my question.  Here's the deal:

I put my 2nd battery pack up front under the motor for a multitude of reasons but namely, I switched to a Falcon prop thus losing 67g off the nose of my airplane from the APC prop.  I thought that I should put it there to counter the basically 3 oz of weight I lost off the nose.  When I did this, I had to add an extension from that battery pack to the regulator thus providing a voltage drop across the extension.  This is why my #1 battery was being used more than the backup.  So in a sense, I have created a backup system instead of a redundant dual system.  Once the primary (if you will) pack drops to the voltage received at the regulator of the second pack, they will drain more together.  The voltage drop across the extension is what was causing this. 

Thanks for that article and thanks to all who responded.  Y'all are great.

OhD

Why not charge the back up receiver pack to the storage value and run off the regulator connected to two cells of the power system pack? Then you'd never need to charge the single, small, light RC system pack.

Jim