CSM Carbsmart setup

[Finless]

They are using a MINI servo (hs81) and I used a micro servo (HS65).. The Align servo bracket is for a micro servo and there is no problem getting to the glow plug.

Bob

[snappa]

I use an hs55 and it works fine

[ZephanRC]

Just got a chance to test it yesterday. I had to shorten the link to the servo so it would go a 2 more clicks lean as it was running full lean the entire time. Once I did that I could really feel the performance. I guess I was running a little rich manually the entire time!

Did some test with hard punch out and trying to bog the motor and quickly set the heli down and found the carb smart had richened the needle perfectly!

I left the gain at 50% and temp set to 110F. All the guys at the flying site noticed I was getting more power in flight now. They will be calling GrandRC on Monday and ordering some :)

It really does work as advertised!

Bob

Bob,
This is the last update that I saw as far as settings. I just got my carbsmart and have everything mounted/setup except I have been holding off on mounting the sensor and had a few questions.

- Has your mounting of the sensor still been holding up? Would you still use the same approach? Trim heatshrink, epoxy end, quicksilver (i was going to use the alumina epoxy) the sensor in place, RTV around whole thing?

- You said that you ended up 2-clicks leaner then your original setup video. So your mid-position is now about 12 O'clock (about 1.25 turns)? Your full lean about 1:30 (about 1.125 turns) and full rich about 10:30 (about 1.375 turns)?

- You mentioned that it did indeed richen the mixture with hard punch outs. Does it run up around the full lean point most of the time during "typical" flight? What do you get with the "finger on the backplate" test after a "typical" flight?

I hope to get the sensor mounted this week and do some testing with my eagletree, both bench testing (with heat gun) and flight testing. I hope to log temp (separate sensor, but I wonder if I could "Y" the CSM sensor) and servo position.

[Finless]

Yep both my sensors (T700) are still working fine the way I mounted them.

I have increased my throw a little more so full lean is not about 2:00 and full rich about 10:00. When you start out it will be in full lean so hover for about 30 seconds or so and let the engine warm up and it will pretty quickly start working. Depending on ambient temps and fly style, mine runs about 1:00 most the time but I notice as the tank gets below 1/3 full it starts to richen a little which shows fuel supply without a header tank does change!

Now I don't suggest you set up for 2:00 lean until you know your "hover point" and every engine may be different. What I did on my 90 setup was hover side in and have someone with a telephoto lens look at the needle arm. In hover if I saw the arm come off full lean a little bit (2-3 clicks) I knew I was pretty close. Then I flew around as hard as I could with hard climb outs and then brought it back into the hover and had the guy look again. Yep it richened up a little more. Now I knew I had it pretty much right.

Just be careful about full lean as it will always start up at that position until the motor warms up. If your too lean you can damage the engine. So it's a little bit of a game to find the full lean position. Bottom line in the hover after the engine has warmed up your full lean should not be more than about 2-3 clicks from your hover running position.

Bob

[ZephanRC]

Thanks for the update. Sounds good. It actually sounds like I have been running a little leaner than you for my "normal/hover" to start with about 1:00 (which puts my full lean a click or two above the 3:00 position), but like you say every engine, location, etc. is going to be a bit different. I'll post some eagletree results if I get them.

[gorgin4325]

Hi,

Could this be used instead of RTV to mount the sensor on the cooling head as shown in the setup videos by Bob? http://www.grandrc.com/inc/sdetail/59094

I already ordered the arctic silver stuff from Grand RC.

Thanks
G

[TheBum]

Hi,

Could this be used instead of RTV to mount the sensor on the cooling head as shown in the setup videos by Bob? http://www.grandrc.com/inc/sdetail/59094

I already ordered the arctic silver stuff from Grand RC.

Thanks
G

JB Weld is not flexible and, in my opinion, would effectively be no different than using the Arctic Alumina exclusively.

[gorgin4325]

Thanks. But does it really have to be flexible?

[TheBum]

Thanks. But does it really have to be flexible?

Put it this way: If it isn't flexible, it will have to be quite hard not to crack under the vibration. My experience with JB Weld is that it's a good adhesive for tightly pressed metal surfaces but not a strong filler. It's definitely more brittle than regular epoxy.

[gorgin4325]

Put it this way: If it isn't flexible, it will have to be quite hard not to crack under the vibration. My experience with JB Weld is that it's a good adhesive for tightly pressed metal surfaces but not a strong filler. It's definitely more brittle than regular epoxy.

Thanks

[dingfangle]

I just put mine in and am looking forward to trying it. I just wanted to say thanks to everyone in here for the tips and especially to Bob for the great videos, I don't know where I would be with out them.

[Taffage]

Put mine in with regular 30 min epoxy last summer and so far so good.

[ZephanRC]

Well I got my CarbSmart installed and my EagleTree logger hooked up and I thought that I would share my results. I did the install pretty much per Finless’s videos. I epoxied the sensor in between the third set of fins up against the top of the head with the Arctic Alumina epoxy and followed up with RTV silicone for strain relief. I used a HS81 servo mounted with a couple of spare servo plates and this seemed to work just fine (see pic). I purposely did not show my mixture needle setting in the picture as I want to get some opinions on the results without anyone knowing the actual needle setting for full-rich, full-lean and mid-point. I will tell you that the mid-position is what I was using for a mixture setting before installing the CarbSmart. The logging was done with a separate EagleTree temp sensor mounted between the fourth set of fins and a “Y” cable to log the servo position.

I got some interesting results and I’d like to hear comments from the helifreak community (See Chart). As you can see, ambient temp was about 16°C (61°F). After starting the engine (OS50Hyper), I pushed/held the button on the CarbSmart a few times to show full throw on the servo and temporarily disconnected the temp senor to show the servo mid-point just before setting the CarbSmart to 100°C. Full-lean on the servo is 100 on the chart and full-rich is zero with mid-position at 50. The left scale also shows °C for the temperature graph.

You can see that the temp was held somewhere around 90-95°C. This is below the 100°C target, probably because the EagleTree sensor is not seeing the same temp as the CarbSmart sensor. Whatever the actual temp was, the CarbSmart appeared to hold the temp within +/- 5°C. (+/-9°F). The first part of my flying was basic circuits, loops and rolls with one landing to take a look at where the mixture servo was (it looked pretty much full-lean). In the second half of flying I did multiple full pitch climbouts and one tic-toc (although not held very long, because I can’t do them that well). BTW, at the end of the session, with the engine still idling, with the “finger on the backplate test”, I could hold it there indefinitely, but it was hotter than it felt when I didn’t use the CarbSmart.

Something that I found interesting is the peaks/valleys of the sensor position and the peaks/valleys of the temp almost exactly line up (see blue lines). If anything, the changes in direction of the servo movements appear to “lead” the temperature peaks/valleys. This suggests to me that the CarbSmart is actually looking at rate-of-change in temp or possibly even the “acceleration” in temp to predict its compensation (good thing).

What do you HeliFreaks think? CarbSmart working well? Without knowing what my actual needle range was, do you think that this looks like a good range (I will post my actual settings later once I get some opinions)? Any other comments or interpretations?

IMO ,the CarbSmart appears to do a really good job and I am happy with the setup and results.

[forjer]

That looks good. No flat lines out of your servo movement. How many degrees are you getting out of the needle? I'm near 90 deg worth of travel and mine will reach and stay full rich sometimes. Yours looks better than mine.

[ZephanRC]

I am getting about 90 degrees of mixture control as well. Your temp curve appears a little flatter than mine (good thing). What gain are you set at on your CarbSmart? I forgot to mention it in my post. I have the gain about 50%. I'm not really sure why yours ended up at full-rich with the temp curve you are showing, since it didn't react that much with the temp increase just before that. Could be differences in govenor as well. I didn't mention this either in my post, but I'm running a GV-1. Could be pitch setup and/or flying styles as well. I'm running +/-12 deg in pitch and +/-7 deg cyclic and I was doing full pitch climbouts and one tic-toc in the test.

[forjer]

I too thought that full rich was odd given the first high temp needle position. That graph was actually one of my first flights using the Multigov Pro to control temp. The flight had a couple full pitch climbouts, loops...but mostly FFF. The gain is completely different with the MG...something like 1-500 and I'm at the default 85.

This graph is one of the last flights with my Carbsmart. It was tuned pretty well at 110C.

FYI...Raptor 50, 11.5 collective, 7 cyclic, 2050 HS.

[ZephanRC]

Yup... That looks a lot better. What were you doing (flying) at the end?

[forjer]

Yup... That looks a lot better. What were you doing (flying) at the end?
Probably nothing different that in the beginning, it's just that the engine had been warmed up so its temp will rise easier, as well as the tank is emptying which makes it run leaner hence the compensating richer needle setting.

[ZephanRC]

Well, it does not appear as though I am going to get any more comments on how my SmartCarb is setup from anyone else so, I thought that I would reveal what my actual needle settings were. That may induce some comments.

Full-Rich is a couple clicks above 1 1/4 turns out.
Full-Lean is a couple clics above 1 turn out.
Mid-Position is just above 1 1/8 turn out (which is what I had for a setting before the CarbSmart).

[derelict]

am i correct that you have both... the carb smart and multigov pro on the same machine?

any specific reason you like the carbsmart more ?

Greets

Marcel

[ZephanRC]

No... I have the CarbSmart and a GV-1, but plan to go to an 8V setup and use the rev limiter in the AR7100R.

[vboy55]

my sensor has gone bad anyone got some advice on how to remove it

[lovespicyfood]

Well, it does not appear as though I am going to get any more comments on how my SmartCarb is setup from anyone else so, I thought that I would reveal what my actual needle settings were. That may induce some comments.

Full-Rich is a couple clicks above 1 1/4 turns out.
Full-Lean is a couple clics above 1 turn out.
Mid-Position is just above 1 1/8 turn out (which is what I had for a setting before the CarbSmart).

Well I'm just about done installing my Carbsmart...it's taken forever because life got in the way...

I installed the sensor per Finless' recommendation: Arctic Silver and then RTV for strain relief. It took me awhile to get the linkage setup as I couldn't find an Align 600 nitro pushrod that was long enough. I ended up getting some "all thread" rod and cutting it to length. I'm using 600N balls and an HS65 servo. I did not use a nut with the balls, they are threaded directly into the servo arm and Carbsmart arm. I dremeled off the sharp ends of the balls and CA'd them in place.

I'm powering the Carbsmart via a y-cable from the Aux2 port of my AR7000 receiver. I didn't have any channels left so I figured this was the best channel to draw power from. My Carbsmart shares this with my Multi-Gov. I took out the data wire from the lead powering the Carbsmart.

ZephanRC: Thanks for the needle setting info. I was planning on setting my needle valve at ~ 1.25 turns out in the middle position, but now I might lean that out a little bit.

I'm running an O.S. Hyper .50, Align pipe and Coolmax 30% nitro.

Any comments/advice is welcome. I'm hoping to fire her up this weekend...

Thanks,
Pete

[forjer]

am i correct that you have both... the carb smart and multigov pro on the same machine?

any specific reason you like the carbsmart more ?

Greets

Marcel
I had the Carbsmart when I had no gov, then I got the Multigov Pro and gave away my Carbsmart.

As far as fuel mixture control, I haven't used the MG long enough yet. Thee CS did a terrific job though. Setting up the linkage goes to the MG, the CS is easiers to verify the lean/rich/center positions though. And the CS is easier to change the temp setting, even when you don't want to, while the MG gives you greater temp target options and more temp related options with the multi target temps and the flight's high temp.

They both seem to do well at what they are intended for.

[pvande]

Has anybody done a carbsmart install on an OS 37 - the fins are not as clearly defined as on the larger engines and it's not clear what the optimal position is to ensure direct access to the combustion chamber.

Any advice/pointers would be much appreciated.