Raven V3 E-Match Test, Battery Voltage Dipping concerns

[dvdsnyd]

Hi All,
Happy New Year!

I just finished wiring up my avionics bay for my current build. It's a pretty stock Madcow Super DX3 using a Raven V3 for the flight controller.
I wanted to give everything a full test with e-matches to build confidence in my system.
I am using the Raven in it's stock configuration, where Apogee is controlled by the Apogee and 3rd outputs, and the main is controlled by the Main and 4th outputs on the Raven.

I powered the Raven up and connected it to my laptop via a USB cable, started the FIP. I ran a short test using the Test Flight Simulation capabilities of the FIP. The flight went to about 1600 feet. All of the charges fired as expected.

After downloading the simulation data from the altimeter, and looking it over, I have a few questions/concerns:

-I was surprised to see that the voltage dipped very low on all 4 outputs as they fired their respective charge.
-The overall Battery voltage parameter doesn't show the same dips as each of the 4 outputs.
-The current draw remained constant, showing a draw of zero amps throughout the entire simulation. I was expecting to see a nominal draw for the altimeter, and then some peaks as the charges fired.

My setup consisted of a new, off-brand 9 V alkaline battery. The short-circuit amp draw from the battery was 5-6 amps before I put it on the sled for testing.
I was using run-of-the-mill "orange" leaded e-matches for this test. The altimeter is turned on using a Featherweight screw switch. I have it wired as shown in the start-up guide.

I have attached the FIP file if anyone would like to look into it further. I have also posted a quick image file that contains each output parameter (battery voltage)graphed along with the simulated altitude along with current draw and overall battery voltage for anyone else who'd like to play along.

My main question is should I be concerned with any of these things? A few other questions: What is causing the battery voltage to dip? Why was the current draw 0 throughout the simulation? Am I missing something with how the FIP/test simulations work with the Raven?

Please let me know if there is anything that I could explain more clearly or add to in anyway to help get my questions answered. Thank you for taking the time to help me look into this.

Dave

 

[cvanc]

(for anyone who saw this earlier, I had to edit it. I realized the battery was actually OK and removed comments accordingly)


It all looks perfectly normal and healthy. Here are some specific comments:

1- It's normal for each pyro outputs' data to dip near zero during the fire command - that is the way Raven pyro data always looks. The pyro voltage data is high (at the level of the battery, whatever that may be) before fire and drops near zero during fire. After fire, if the e-match retains some continuity, the data will come back up to at/near battery voltage. However, if the e-match becomes an open circuit as a result of firing, the pyro voltage for that channel will stay at zero after fire. You can see this clearly in your data as all four e-matches retained some continuity and the voltage of all 4 channels went back up after fire to battery level.

2- It's also normal to read no current spikes at all from the pyro events if you have fast e-matches. The current draw channel does not sample frequently enough to record the current surge of some popular e-matches. They burn too quickly for the spike to get recorded. I have three Ravens and have exclusively used Quest Q2G2 and Crazy Jim's pyro products and I have never seen a current spike from a pyro event on any flight. Higher current matches, or a short circuit on an output, would be captured and recorded in the current data.

Hope this helps. Lots of folks in the forum are well experienced with Ravens so feel free to ask any questions that may come up, and good luck flying.

 

[dvdsnyd]

Hi Carl,
Thanks for your response and explanation of the pyro outputs and the insight about the current channel sampling rate. I was hoping the issues were more related to my lack of understanding than an error in my setup or Raven.

I still have one question however. If you look at the current channel of the original FIPa file I posted, it reads zero amps for the entire simulation. Why is this? I would expect that the Raven would have a nominal current draw of around 100 milliamps or so.

I went back and did a few bench tests with some Christmas Tree bubs connected to the charge outputs of the Raven, and am getting current draw spikes as expected, just not a nominal current draw. The current channel reads 0 amps in between the spikes. You can see this in the attached FIPa file.

I've been spending a bit of time looking at this thread to compare my files too. (Maybe this is where I should have posted my original questions?)

Anyways, you posted a file in post #4 of that thread - L910 Early main, which shows a current draw of 125 milliamps throughout the entire flight. However, if you look at some of the subsequent posts, #5,#6. The current draw channel yields pretty erratic data that even appears to go negative at times. Post #13 is a really good candidate for this as well. All of the current channel data is negative in the file in post #13.

Is there any explanation to this? Is this normal? What is the use for the current draw data? In my mind, data like this is supposed to be useful for diagnosing potential problems. However, I am having a really hard time understanding what it means to have a current draw of zero amps throughout the flight...I am pretty new to the Raven, so any additional insight on these questions/concerns is appreciated.
Again,
Thanks for taking the time to look into this and offer your help!

Dave

View attachment test2_Xmas Bulbs Bench, 010120015.FIPa

 

[kjs]

Dave,

I went and checked a bunch of the FIPa files from when we test them (with Christmas lights) and some number of them show zero current even with the Christmas lights for pyro, some have 62.5ma and some few spikes to 125ma (resolution is obviously 62.5 ma per bit). We'll have to maybe review if there are any issues in how the measurement is made - or in timing relative to other things, but I can say that your unit is not unlike other Raven3s in it's behavior.

One other thing regarding the graph. If you right click on the left axis that says 'Volts Battery (V)' and select 'Edit' then you can set the minimum value for that curve to zero. It then gives a better representation of how little the voltage drops relative to the full battery voltage (else the auto scale makes the drop look alarming even though it might only be a 100-200 mv).

Kevin
FIP developer

 

[cvanc]

Interesting.

I hadn't really gone looking for the Raven current draw before (never needed it). All I wanted to see - just for curiosity's sake - was the match draw and once I found "my" matches did not register I kind of stopped looking at the current data. In reviewing my files I see some of my Ravens read the altimeter draw as zero and some about one-tenth of an amp.

Attached is a file from one of my bench tests long ago, where all 4 outputs are driving 2 ohm resistors. It shows that the current recording process does work, and that the idea some matches are just too fast for the recording to capture seems valid.

View attachment 2ohm1402a.FIPa

Hope this helps.

 

[CarVac]

-I was surprised to see that the voltage dipped very low on all 4 outputs as they fired their respective charge.

The Raven is common high, so based on the exhibited behavior the circuit goes:


[Battery negative]---[Raven FET switch]---[pyro voltage measurement point]---[ematch]---[battery voltage measurement point]---[battery positive]


When the FET switch is open, most of the resistance is from the FET and so the measurement point sees the full battery voltage. When the FET switch is closed (firing), most of the resistance is from the e-match and the voltage dips to near zero.

As cvanc says, after the FET switch reopens, any residual conductivity of the ematch will increase the measured voltage back to the battery level. But the residual conductivity does drop over time, so the corresponding voltage also drops over time.

-The overall Battery voltage parameter doesn't show the same dips as each of the 4 outputs.

The battery voltage doesn't drop that much because it's being measured on the other side of the pyro load.

-The current draw remained constant, showing a draw of zero amps throughout the entire simulation. I was expecting to see a nominal draw for the altimeter, and then some peaks as the charges fired.

Same as what cvanc says: the e-matches are very fast, so they're not sampled.

 

[kjs]

Carl - Thanks for that data! I'm glad to see the 2 ohm resistor load behaves as expected.

 

[kjs]

CarVac - Thanks for your summary analysis.

Dave (dvdsynd) - I also just now caught your comment about battery going very low - actually be careful of the scaling of the battery data as it is scaling for range. I should maybe always set battery graph min to be 0 volts. Your actual voltage drop if you compare the graph to the legend is only ~8.88v down to 8.60 v before coming back up to 8.88. If you right click on the battery voltage axis, you can select 'Edit' and set the minimum to be zero volts and then the graph looks much more reasonable.

 

[dvdsnyd]

Dave,

I went and checked a bunch of the FIPa files from when we test them (with Christmas lights) and some number of them show zero current even with the Christmas lights for pyro, some have 62.5ma and some few spikes to 125ma (resolution is obviously 62.5 ma per bit). We'll have to maybe review if there are any issues in how the measurement is made - or in timing relative to other things, but I can say that your unit is not unlike other Raven3s in it's behavior.

One other thing regarding the graph. If you right click on the left axis that says 'Volts Battery (V)' and select 'Edit' then you can set the minimum value for that curve to zero. It then gives a better representation of how little the voltage drops relative to the full battery voltage (else the auto scale makes the drop look alarming even though it might only be a 100-200 mv).

Kevin
FIP developer

CarVac - Thanks for your summary analysis.

Dave (dvdsynd) - I also just now caught your comment about battery going very low - actually be careful of the scaling of the battery data as it is scaling for range. I should maybe always set battery graph min to be 0 volts. Your actual voltage drop if you compare the graph to the legend is only ~8.88v down to 8.60 v before coming back up to 8.88. If you right click on the battery voltage axis, you can select 'Edit' and set the minimum to be zero volts and then the graph looks much more reasonable.

Kevin -
Thanks for your work with the FIP. It's a great piece of software and adds tremendous value!
Thanks for the graphing tips as well, I'll have to give them a shot!

With regards to the current draw data showing zero amps in my tests. I'd just like to make sure I am understanding correctly, so sorry if this is redundant. As others are pointing out, some electric fire too quickly to be able to measure any sort of peak amp draw with how the Raven saves data. I am understanding this, however, I still don't understand why my current draw is zero amps for the ENTIRE simulation...
But, as I understand, there isn't a problem with the Raven. It's obviously firing the matches, so the current is being delivered. You don't seem to be alluding to a problem with how the Raven is storing the current data, but to more of a problem with the FIP in downloading the data from the Raven? Hopefully that makes sense, please feel free to set me straight, if I am confused.

I'm going to try to do some bench tests with some low value resistors. Would you suspect that I will still see a nominal amp draw of zero amps throughout the simulation, before and after match firing?

The Raven is common high, so based on the exhibited behavior the circuit goes:


[Battery negative]---[Raven FET switch]---[pyro voltage measurement point]---[ematch]---[battery voltage measurement point]---[battery positive]


When the FET switch is open, most of the resistance is from the FET and so the measurement point sees the full battery voltage. When the FET switch is closed (firing), most of the resistance is from the e-match and the voltage dips to near zero.

As cvanc says, after the FET switch reopens, any residual conductivity of the ematch will increase the measured voltage back to the battery level. But the residual conductivity does drop over time, so the corresponding voltage also drops over time.



The battery voltage doesn't drop that much because it's being measured on the other side of the pyro load.



Same as what cvanc says: the e-matches are very fast, so they're not sampled.

CarVac -
Thanks for going through the sensing schematic with me. It makes sense the way you explain it.

Interesting.

I hadn't really gone looking for the Raven current draw before (never needed it). All I wanted to see - just for curiosity's sake - was the match draw and once I found "my" matches did not register I kind of stopped looking at the current data. In reviewing my files I see some of my Ravens read the altimeter draw as zero and some about one-tenth of an amp.

Attached is a file from one of my bench tests long ago, where all 4 outputs are driving 2 ohm resistors. It shows that the current recording process does work, and that the idea some matches are just too fast for the recording to capture seems valid.

View attachment 250566

Hope this helps.

Carl,
Thanks. I'm not sure if I have 2 ohm resistors in my stash. Would any other resistor value work as well? what about something common like 1K? I'd like to do a test with resistors to see how the data responds.
With regarding to "needing" the current draw data. I'm not totally sure I do either, but being pretty new to using the Raven, I wanted to see what the data looked like. So, naturally, seeing a flat line reading 0 amps made me question if everything was working properly.




All -
Thanks for all of your efforts in helping me understand how the Raven works and answering my questions! This is a great community!

Dave

 

[kjs]

Dave,

I'm not sure why I would see close to zero current for me with my Christmas bulb test setup so you may still see zero. I'll need to get with Adrian to understand the test circuit first. I can only say that what you are seeing matches what I see - with Christmas bulbs for me so certainly not like an ematch - but my setup has a 4.5v 1A power supply behind it (instead of a 3.3v battery). I may set it up to test with batteries. The only odd variable I can think of is the big capacitor that is there may be providing the current and not part of the current sense (but I would think we would get a current surge from the battery to re-charge the cap if that was the case.

[So although I may work on FIP, Adrian is hardware expert so accept my inputs only as "pondering speculation" until I can get with Adrian - or until I can spend some significant time looking at the circuit myself.]