Anyone using AGF A06CLS servos?

[Landru]

Hey there,

Recently I heard tell of AGF servos. Then I discovered that Aloft sells the AGF A06CLS. It seems like a very good fit for DLG use. And if appearances are anything to go by (yeah, not much, I know), they seem to be a quality product.

Has anyone got any flying hours on these servos?

In general, I go for the "best" servos I can get my hands on, especially if they're to be installed in an expensive model like a molded DLG. That said, I've run into failures and quality issues with some of the big name, big dollar servos so I'm on the lookout for alternatives.

TIA

 

[Konrad]

I've held them in my hands and played with them (no flight report yet). They center well and have phenomenal speed. Those are the two primary performance criteria I look for in a servo (I like to use as small a servo arm as possible to get the most torque and resolution out of my set ups). To be honest they aren't quite at the same performance level as the KST X0-8, but they are in the same class. (KST X0-8 uses a custom chip and has what appears to be a narrower dead band and lower idling current. If you can fit them in, I think they are a bit better than the KST X0-6). Basically they fall in between the KST X0-8 and X0-6 both in performance and size.

The only "down side" I see with them is that the pig tails are too long. I'd like to see servo manufactures offer servo leads just long enough to allow you to pull the servo out of the servo pocket. As I have to make up my own wiring harness this would save on space and weight! Or better yet just have a connector at the servo body like we had with the Volz servos of old.

Now I'd like you to name names. I see where it does nobody any good to protect products that don't live up to your expectation. What big name servos have let you down, and why?

I'll start, the KST X0-8 is on my do NOT use list unless reworked! As you can see there is no retention feature for the motor. I had the motor literally fall out of the housing. I was shocked to see that there were no screws holding the motor in (by design) nor did I see any adhesive used. (I assume the motor housing is just a press fit into the servo housing (red). The new series of servos from KST now have the motor held captive by the bottom part of the servo housing.

This is the kind of objective quality data that I find helpful! I should state that the time between failure for the KST brand is very very high, like hundreds of hours.

P.S. This was an inflight failure on my elevator servo! Because of the skill of the pilot I was able to control pitch with the fully articulated trailing edge. OK, I was real lucky in that I was on the slope and had the distance to the valley floor to figure out a fix on the way down.

 

[Wayne]

We have sold a lot of the KST servos Konrad is showing and this is the first time I have seen a report of one doing that. Not an excuse, just wanted to add one more data point.

We like the AGF A06CLS servos a lot. These were originally designed for some little rock crawler, not DLGs. Yeah, kind of funny. I think we are one of the few shops that order them with a normal servo plug. AGF is after us to raise the prices, but I am avoiding that as I would like to see more folks give these fine servos a try.

I will be very clear, KST has spent a fortune to get the centering performance of their X08 and larger servos. They have a custom IC chip that probably cost about as much as your house to develop. Not many companies have the pockets deep enough to do this. And still KST does have some weaknesses with this chip. I do think it is hard to beat that servo regardless of price. I am amazed when I see inferior servos from some other brands asking for the same money or in some cases even more money than KST. IMHO KST still makes the best servo regardless of price when you are looking at 8mm or thicker servos.

With this said, there are some very good servos that can save you some money. Not everyone has the budget for KST, and not every model needs to have a KST budget for servos.

I wish KST would treat me better, but such is life.

This AGF servo is as close to KST performance as we have tested, and talking with the company is a joy. They understand what they are making and have real answers when questioned. That is rather rare in the servo game these days.

 

[Konrad]

I agree with all you said.

I tried to indicate that my issue was rare in that I've had close to hundreds of hours of flight times (run times) on my KST fleet and this was for me the first of this type of failure. I do point to some basic mechanical issue that I think contributed to this failure. As there appears to be a design issue with the older X0-8. I hope others will learn from my experience and add an industrial adhesive to the motor and housing.

Again the later KST replacement for the X0-8 does have a captive housing.

I'd like to learn what brand and type of failure that @Landru is experiencing. He has said that his requirements had nothing to do with price point. He wanted the best reliability/performance. I've used the entry level MKS 6100 family servo and don't think they hold up against the X0-8 family.

 

[Landru]

Hey Konrad and Wayne,

Thank you for the outstanding responses!! Gosh, I love the Aloft forums.

Since you asked, Konrad, I'll add a few comments on some of the servo failures I've run into.

Most recently, a compound gear in an X06 failed in flight causing the servo motor to spin endlessly. The affected servo was on one aileron in a 1m DLG. Happily, I got the model on the ground with no damage, albeit mostly by good luck. Close inspection of the compound gear in question revealed what I took to be a manufacturing error. There are photos and details in this RCG thread.

Before that, I ran into an adventure with a Dymond D47 in another 1m DLG. Granted, D47s aren't in the same league as KST for quality or price. Still, they do tend to be highly regarded. The problem I encountered turned out to be a dry solder joint on the PCB. One of the leads was very poorly soldered but still managed to maintain contact -- most of the time. In use, the symptom was very intermittent erratic servo response, often during launch or immediately after. I figure that's when G forces caused the dry joint to open momentarily. The servo always performed fine in ground testing so it took a while for me to twig to the problem. For a while, I even thought the problem must have been with the pilot. IIRC I posted photos and details somewhere on RCG. For contrast, I logged over 250 flying hours on another pair of D47 on the tails of a 2m glider without any problems.

As for MKS, I had a couple of DS6100 fail on power up of a 2m glider. Totally dead. The reason remains a complete mystery. Another pilot suggested there might have been a voltage spike from the model's ESC/BEC. That seems plausible. According to the fellow, there's a little known voltage spike issue with Castle Creations ESC. Still, I have to wonder why only two of six DS6100 in the model were killed. The BEC output voltage was set for 5V.

Years ago, I bought a good supply of Hitec HS-55 servos for parkflyers and 1.5-2m glider tails. I don't think they were over taxed in those applications. I always ran them on 5V from a BEC. Still, every single one has now failed or was starting to act erratically before being pulled.

I have a few more servo war stories but I'll leave it at that for now.

Am I a magnet for servo failures? Sometimes I think so. If I'm abusing gear, I'm at a loss to know how. I'm purposefully gentle with all my models. Plus, they're all gliders these days. With the exception of DLG launches or the odd high speed descent, they've all lead a happy and sedate life in the thermals.

Back on the topic of the AGF servos, based on your info Wayne, I'll definitely consider them for some builds. For the new Falcon DLG, I'm starting to think KST A08 are the way to go. Any idea when you'll be getting new stock on the horizontal flange version?

Thanks again, you guys!

 

[Wayne]

I ordered KST long ago, still waiting for it to show. I'm told it should be soon. (My statement about them treating me better..)

The D47, never have figured out exactly who makes those, they seem to be the same as many similar servos on the market, yet people single out the Dymond as being better. In my testing they are maybe a middle grade servo with so-so centering and seem to get worse the longer you use them.

 

[Landru]

Thanks, Wayne. I'll keep my eyes peeled for the arrival of new KST stock. I'm sorry to hear you're not getting great treatment.

As you said, the D47 centering definitely isn't anything to write home about, especially by current standards. I bought a dozen of them years ago when they were perhaps the best option available at the size and weight. Apparently, they're now discontinued.

I've seen the newish CHA DS06 suggested as a higher quality substitute, albeit at a higher price.

As an aside, there's a brand new requirement here in Canada to register every model above 250g with the government. And that's on top of the 400 foot restriction. Sigh. I guess I'll be spending more time with smaller gliders and smaller servos.

 

[Konrad]

The D-47 mystique may be as a result of history.

I recall using the D-47 in a lot of my 7cell Speed 400 racers and LMR glider 20 to 25 years ago. I recall how much better these servos performed compared to the other 5 gram class servos of the time such as the Cirrus and GWS servos. It was very noticeable, and as a result the D-47 still has a soft spot in my heart. I still use them as I bought 2 dozen of them before Dymond got into trouble with their law suits. In modern times I still use them in many of my less demanding models, such as my F3-RES ships.

FYI; I've had the MKS 6100 fail on 2 cell 18650's, no BEC spikes.

 

[Doc J]

Hey there,

Recently I heard tell of AGF servos. Then I discovered that Aloft sells the AGF A06CLS. It seems like a very good fit for DLG use. And if appearances are anything to go by (yeah, not much, I know), they seem to be a quality product.

Has anyone got any flying hours on these servos?

In general, I go for the "best" servos I can get my hands on, especially if they're to be installed in an expensive model like a molded DLG. That said, I've run into failures and quality issues with some of the big name, big dollar servos so I'm on the lookout for alternatives.

TIA

I tested some for AGF about 6 months ago. I like the programmable widget dongle that comes. Easy to use and really versatile. The servos passed all the tests well and didnt fault on the 24 hour continuous test - that one normally sorts ot the good from the bad - so I can only say that they seem pretty good.

Cheers,

Doc.

 

[Landru]

Excellent. Thank you.

I'm curious: Is there any load on the servo in your 24-hour test? Or is the servo simply cycling continuously without load?

I found info on the programmable features here:

AGF-SPV3 USB Servo Programmer for AGFRC Programmable Servo

servo card AGF-SPV3 programmer card can be only compatible with AGFRC SERVOS with ASS function,could change servo speed,servo torque etc servo parameter.

www.agf-rc.com

 

[Doc J]

Excellent. Thank you.

I'm curious: Is there any load on the servo in your 24-hour test? Or is the servo simply cycling continuously without load?

I found info on the programmable features here:

AGF-SPV3 USB Servo Programmer for AGFRC Programmable Servo

servo card AGF-SPV3 programmer card can be only compatible with AGFRC SERVOS with ASS function,could change servo speed,servo torque etc servo parameter.

www.agf-rc.com

Hi Landru - I use a rubber band to tension the 100mm (4") test arm in both directions. The load is about 200g max both ways.

I test the backlash (measured) before and after a successful no glitch test. The mechanical construction (unmeasured) is tested in a similar way but looking at fit, finish, and play.

Some, and I repeat Some (but not all) of the servos coming from China are improving very rapidly recently.

Cheers,

Doc

 

[Landru]

Thanks, Doc.

That seems like a very useful test. By any chance are your results published online?

Best,
Andrew

 

[Doc J]

Thanks, Doc.

That seems like a very useful test. By any chance are your results published online?

Best,
Andrew

Unfortunately, no; Andrew I did a number of tests for several companies. I don't suppose they'd want each other to see the results.

Cheers.

Doc.

 

[Landru]

Understood. I figured that might be the case.

 

[Landru]

<snip>FYI; I've had the MKS 6100 fail on 2 cell 18650's, no BEC spikes.</snip>

Thanks for that. What were the circumstances? As noted, mine failed on the ground.

 

[Landru]

Here's a positive little mini review of the AGF A06CLS by a well known DLG pilot:

Like you, Doc, Pierre praised the like the ease of use of the programming device and app. Having struggled with other programmable servos, I find that a plus.

 

[Konrad]

For my failed 6100's it was on initial start up prior to flight and with batteries less than fully charged. This was for 2 units. I think I've only owned 6 servo 6100. And of those 4 were used but low time, from a F3F racer purchase. I don't recall which population the failed servos came from. BTW; the failure was on the board not the motors.

BTW; The test parameters and data capture methodologies is very critical in validation testing.
For example the test describe earlier would most likely be considered an accelerated wear test (severe stress testing). As the continuous running for 24 hours would be unrealistic and violate any duty cycle limitations. This kind of test can generate valuable wear data in a short period of time for components such as the gear train and feedback pots (if used). This is assuming the heat buildup is controlled and none of the components go into thermal runaway.

I use a simple acceptance test to weed out infant mortality failure in all my servos prior to being presses into service. I run them in a free state at full speed for 5 minutes. Then I check their centering and general condition.

As was mentioned the servo quality we get is improving. Two notable exceptions are the Corona brand. Many of these have poor PCB in that the leads (motor, pot, cables) are bridge soldered to the IC chip. This runs the risk of shorts and thermal damage to the chip. If the Corona servo passes my acceptance test I get ok performance out of them. But on some Corona servo models I've had a 50% failure rate! Another servo that has shown horrible quality are the ones Freewing uses in their EPO foamies.


That KST X0-8+ that failed was a low time servo, less than 2 hours run time, and most of that was set up time! This is the first servo in a long time that literally fell apart. The last time I had a servo fall apart (not counting fractured leads) was the Futaba S128 & S132 from the early 80's. If this happens again I'm thinking I might have to put my servos on a vibration table as part of my acceptance testing.

 

[Landru]

<snip>I don't recall which population the failed servos came from. BTW; the failure was on the board not the motors.</snip>

<snip>I use a simple acceptance test to weed out infant mortality failure in all my servos prior to being presses into service. I run then in a free state at full speed for 5 minutes. Then I check their centering and general condition.</snip>

Thanks, Konrad. My suspicion is the failure was also on the PCB in my DS6100. Mind you, I haven't tried to confirm that.

Like you, I now do an acceptance test on every new servo. In my case, it's a 10-minute stint of cycling continuously with no load on the servo. I use a servo tester programmed for something like a 5-second sweep. (I don't recall the exact time at the moment.) Perhaps testing at full speed would be more effective?

Recently, I started using a longish 'wand' attached securely to the servo output, the purpose being to reveal any irregularities in the motion.

 

[Konrad]

I like to use a X10 glass and my nose to look for failed components. You know what burned electrical components smell like. But often failed components will also swell fracturing the packaging (electrical housing). It is just a quick and often dirty way to analyze the failure.

 

[Wayne]

CHA sent me some servos to test. I'd have to double check our findings, but as I recall they were decent, but the price killed my interest. I think they were more expensive than KST?? As I recall their 6mm servo was the best 6mm servo I have tested for centering.. I may be remembering wrong..