[TheOtherCliff]

Executive summary: Don't buy a Revo clone without reading this and knowing how to test it as soon as you get it.  I would bet that the same problems exist with the Sparky2 clones.

A friend with a Revo vertical stability issue had it turn out to be a bad baro chip.  The board came from an eBay seller, website affiliation unknown at this time.

I also talked with a forum user that had bought two Revos from banggood.  Both had bad MPU6000's (gyro/accel) and low res baros.  Note that the high resolution baro is spec'ed and required for proper Revo operation.

I have access to a Revo just received from thanksbuyer and have done some testing on it as well.  It also has the low resolution baro chip.  This particular board has good gyros, accels, and mags as far as a quick bench test can prove.

So bad clones are known to come from both banggood and thanksbuyer.  These two companies probably get their boards from the same place.  If you have bought a clone (Revo or Sparky2) please test and report back here (what you bought, when you bought it, what you know is broken, what you know works (e.g. it flies in Basic (Complementary) Attitude Estimation Algorithm)) until we have a good feeling for whether any boards from these sellers are OK.  Note that if this thread gets really long, I may delete replies and provide a summary.

I also have heard of clone boards overheating and failing for no apparent reason.

I suggest that you pay with a service that will help you get your money back.  I use PayPal, but had to twist some PayPal arms once, even though I was in the right.

Here is a graph that shows first the bad board plugged in and then a good OP non-clone board.  The difference in the average value isn't important.  The noise is what is important.  You can see that the first half is a much thicker mess than the second half.  In particular, if your graph is about 00,070 thick (top to bottom, about 97,830−97,760 for this graph) or thicker, you have the noisy low res baro.  If your graph is about 00,015 thick or less you have the good baro.

This is in Pressure kPa which is the default baro scope


Note that the good part has a little curve to it.  That is because that board was just plugged in and is warming up.  Ignore the curve and pretend it is straight.

This is in Pressure Altitude which you can enable with Tools->Options->Scopes->Barometer if you know your way around in there.

[ggrif]

Thanks Cliff, will check out mine from BangGood later tonight and post results.

[Mateusz]

I bought from ebay fisher-hobby, revo with a case, it had good arm z mcu, was flyable in complementary mode, so I guess mpu/mcu were fine, but baro was broken. I had huge vertical oscillations in the altitude hold/vario mode, and on scopes baro had 4-6 meters error, while original revo with the same firmware (both temp calibrated and not) had error of 0.5-1m. Also my baro was not changing average readout while I moved FC up and down, original revo worked. I got refund from the seller after him trying to deceive me to close request before he pays back, suggesting that he can send another baro, or to try another software as deadline for opening request was approaching.

[ggrif]

Quick test tonight. The spread is about 12, so good baro?

edit: hmmm, actually more like 6, at least to my non electrical engineers eye! 

[Mateusz]

Quick test tonight. The spread is about 12, so good baro?

edit: hmmm, actually more like 6, at least to my non electrical engineers eye!

I only looked at BaroSensor.Altitude [m] on the left part of scope you can see Bad baro on the right part same scale you can see good baro after switching boards.

Bad baro (lowest 173.5, highest 182.5, difference 9 meters!)
Good baro (lowest 175.5, highest 177, difference 1.5 meter)

[f5soh]

Baro spike can come from a bad powering, an overloaded/bad regulator or bad filtering.
Check if there is ripple on 3.3v regulated. (Revo has 3 regulators)

[Mateusz]

Baro spike can come from a bad powering, an overloaded/bad regulator or bad filtering.
Check if there is ripple on 3.3v regulated. (Revo has 3 regulators)

Mine was not reacting to altitude change, it should I guess despite the noise, and I can't check now, I returned item and got refund.

[ggrif]

Here are a couple of screenshots.  Maybe someone could do a little interpreting for me?
Something funky with BG Accel?

[cato]

I don't see a relevant difference in the two baro scopes, they both are in the same area. Some drift is normal, especially right after you started it. I did not ceck that on my revo, but I know the same behavior from my baro altimeters for biking and hiking.

[Mateusz]

Yes, both plots show the same (yellow) response kPa. I don't know what it should normally show, before I returned my revo clone I only looked at BaroSensor.Altitude in [m] units. That should not jump more than ~1.5m from lowest to highest peak (see attached plot). Mine was jumping 9 meters and not reacting to change of altitude.

The correct baro under magnifying glass should have three lines of symbols, the first line should be MEAS and second MS561101BA03. That is high precision baro http://www.meas-spec.com/downloads/MS5611-01BA03.pdf.

If you see higher than expected noise, or your baro does not react to change of altitude (be careful not to rip off USB connector when testing), but your baro symbols are as above, then it might be cheaper capacitors (leading to ripple on power lines), cheap regulator or damaged baro.

[ggrif]

OK, thanks.
I finally figured out how to add the altitude widget after I posted.

Works fine I think, variation is only 0.6 m and reacts to 1 foot changes in altitude.

Looks like I may have a good one, we'll see how well and how long it flies in a week or two.

[xpantz]

Somebody is pre-selling their Revo clone.  (my OP account was PM'd with this info)

http://www.trianglequads.org/

[TheOtherCliff]

@ggrif It looks like your banggood Revo clone has a good baro (or maybe has good voltage regulators, we don't know the cause for sure yet as @f5soh points out).

It would be nice to find out that (if) it is caused by voltage regulators.  It is possible for a human to replace those.    I jest, but I wouldn't try replacing a baro unless I had a lot of boards that were otherwise junk to me, and I have 4 sizes of kapton tape and a hot air station.

@xpantz And OP/NG is not happy...

[ggrif]

@ggrif It looks like your banggood Revo clone has a good baro (or maybe has good voltage regulators, we don't know the cause for sure yet as @f5soh points out).

It would be nice to find out that (if) it is caused by voltage regulators.  It is possible for a human to replace those.    I jest, but I wouldn't try replacing a baro unless I had a lot of boards that were otherwise junk to me, and I have 4 sizes of kapton tape and a hot air station.

@xpantz And OP/NG is not happy...

I would stay way far away from
Trianglequads.

As for OPNG, they quit making Revos over a year ago. So if someone wants to step in and fill the gap making an "expensive, outdated" controller for $46, more power to 'em.

Sorry, a bit off topic.

[mazevx]

We had a User in german Forum who had the same problems with his revo clone altitude oscilation about 10m while standing on the table inside the house, He fixed the problem by soldering a Filter capacitor direktly to Vccin at the baro did you try something like that?