Pongbot table tennis robot - review and comparison

[plunder]

where did you get the retrofit mechanism? I'd like to try it out, give it a fair chance against the Pongbot.

I got all my Black Friday stuffz from PowerpongEuropeDOTcom but the order page for the retrofit kit isn't up because they're in the process of stocking up all replacement parts (spare parts, out of stock).

From what i understood they're sourcing the special material from GleitGutDOTde (who does not ship to the US), and the stuffz can also be found on amazonDOTde (who doesn't ship this item to the US either).

The material is very similar to HD-PE, PE-HD, and UHMW-PE, but has a lower sliding friction coefficient (Gleitreibungskoeffizient) and is absolutely resistant to wear/abrasion/consumption/scratching in our application. It's adhesive foam-backed "teflon sheet for gliding purposes" which is put in place at the sides and top of the metal deflector.

Ingenious simple effective solution. You'll be amazed and pleased.

(They tested various international manufacturers/suppliers of this kind of product and determined that this is the one and only solving all issues, see my 8 bullet points.)

 

[dingyibvs]

@dingyibvs quid pro quo

Oh right, it was about 9 (~230mm) inches from the table to the bottom of the ball, similar to the Amicus Prime when set to the middle red line (as the ball deflects from the top deflector). That was a quick measurement, I'll do a more exact one tomorrow.

 

[plunder]

I use 200+ balls, there are def frequent "jams". It's not a matter of pressure, I don't think, maybe you just haven't used it long enough? I don't recall many instance during my first few months using it, but now it's very frequent.

btw I did notice that not all balls "work well" in the robot. No problems with my new Powerpong balls or my new Joola Training 40+, but my super old worn smooth Andro 40 balls weren't working so well. But that has nothing to do with jamming. As mentioned, with the retrofit kit the new balls stay grippy, doht get "polished" by the deflector (with gorilla glass, foil, o.a.) anymore.

You didn't really explain what you mean with "jams", in detail. So i caht understand, caht imagine, sorry:
Do you mean that the rotary conveyor keeps trying to rotate and feed but somehow there's no further balls being fed into the conveyor tray, because the bunch of balls (in the black nylon tray) kinda stopped moving, so the jam be basically within the bunch of balls?

In my nylon tray, there's a strong flow of balls into the conveyor tray. I wouldn't understand how/why there could be a jam anywhere! But i hear ya.

That's what we're here (forum/community) for. Trying to help each other, figuring out together idiosyncrasies, tweaks, mods, limitations, etc. of our robot and getting the most (and longest service life) out of it.

 

[jk1980]

i also have frequent ball jams on that conveyor mechanism. I do have more than 130 balls in there for sure. If i just stir the balls with my hand, then it will work again, but it seems like there is a specific combination of balls in that round countainer in which the rotating fins are located in will stop the balls from getting fed in.

I did find a possible solution somewhere in the internet, whereby somebody blocked the entry of balls from one side (i think the left third of the opening), so that balls would only fall in from one side. Unforuntately, when i finally was eager enough to try it out, i did not find it anymore.

 

[dingyibvs]

btw I did notice that not all balls "work well" in the robot. No problems with my new Powerpong balls or my new Joola Training 40+, but my super old worn smooth Andro 40 balls weren't working so well. But that has nothing to do with jamming. As mentioned, with the retrofit kit the new balls stay grippy, doht get "polished" by the deflector (with gorilla glass, foil, o.a.) anymore.

You didn't really explain what you mean with "jams", in detail. So i caht understand, caht imagine, sorry:
Do you mean that the rotary conveyor keeps trying to rotate and feed but somehow there's no further balls being fed into the conveyor tray, because the bunch of balls (in the black nylon tray) kinda stopped moving, so the jam be basically within the bunch of balls?

In my nylon tray, there's a strong flow of balls into the conveyor tray. I wouldn't understand how/why there could be a jam anywhere! But i hear ya.

That's what we're here (forum/community) for. Trying to help each other, figuring out together idiosyncrasies, tweaks, mods, limitations, etc. of our robot and getting the most (and longest service life) out of it.

Yes, that's exactly what happens. It's like the balls form an "arch" over the hole where it feeds into the machine. Personally I think if the feeder uses something more substantial than thin metal rods to stir the balls it'd solve the issue. Right now the rods are not able to collapse the "arch" most of the time. I'm not terribly bothered by it, not nearly as much as the sidespin anyway so I haven't bothered to fix it.

 

[plunder]

@dingyibvs @jk1980 Okay then i'm relieved that it's just that. I understand that there's some potential that balls near the rotary tray could block each other out (inhibition phenomenon), some photos of the situation were posted in the other thread. Fortunately, with my setup this blocking situation becomes highly unlikely because the cardboard ramp makes the conveyor be fed from the sides, away from the pole. And it is also said that there should be only 1 layer of balls on the black nylon.

Some users experiment with duct tape to maintain the feeding into the rotary tray, i built a ramp out of cardboard and reduced the number of balls, all are good workarounds! I herewith regard the potential issue of "jams" fixed.

 

[Jslick89]

I just received my Halo S Pro and did some basic setup stuff. I have to work tonight so not much time to test it out extensively. After more thorough testing I'll write a thorough review, with the Butterfly Amicus Prime as the primary comparison. Here are some first impressions thus far.

1) The box is gigantic, it's very well packaged, but it's 1.2m x 0.6m x 0.6m. Only weighs like 23kgs but it was hard to get my short arms around it to move it from my door to the garage by myself after delivery. On the plus side, it's very well packaged, so very little chance of damage during shipping.

2) I thought the main and obvious advantages it would have over the Amicus Prime would be the software, the ability to move to anywhere and shoot balls instead of only from the center of the table, and the lack of weird sidespin. But OH MY GOSH is it powerful! It is so much more powerful than the Amicus Prime, I actually feel the weight of the ball when it hits my racket. With the Amicus Prime, while it can generate a superhuman level of spin (watch Ponfinity's video on the Amicus), the speed is very subpar. The resulting ball when you turn up the spin is a very unnatural ball, too much spin for the speed. When I turn the Halo's spin and speed all the way up, it actually feels like a real loop, like I'm blocking my training partner's loop.

3) Going through the settings, I can see a few obvious disadvantages compared to the Amicus Prime, including 2 major ones. First is that each combo only has a maximum of 8 balls. The Amicus Prime has 10, and you can combine up to I think 10 combos together into a single "sequence", creating a maximum of 100 different balls in one single drill. In addition, the Amicus allows variability in placement for each ball (e.g. you can program a ball to go to the middle +/- one position to the left or right), further adding variability in your drill. You can literally just have 1 ball in your programming and that could be sufficient for a FH drill as it can place it anywhere on the FH half table.

@pongbotstore Is there any way to recommend some new software features to the developers? I feel like a few tweaks can really add to this already excellent robot, and the features can be applied to all other robots as well as all future robots so I think it's a very worthwhile endeavor.

1) Agree- very well packaged and protected within a semi-hard foam capsule. The box for the epad is very nice too lol.

2) I, too, feel the Pongbot delivers a significantly higher quality ball than the powerpong 5000 i was previously using. I honestly haven't even set the spin and speed past 6 for topspin because its already that powerful. but you do make me want to try to create a blocking drill at full power. might hurt if i miss the ball or get hit on my finger though LOL.

3) I did want to say, for each ball within a drill, you can increase the number of those balls. for example, ball #1 can be programmed to be repeated twice, three times, four times and so on. I do agree it would be cool if we could combine multiple drills together! I also wish there was a longer "reset time" after each drill. so far the only work around is to set the ball delivery frequency on the first ball to 0%...or rest while the robot cycles through a whole drill sequence.

A "random" feature on a specific ball, like you mentioned would also be great! I miss that function from the NEWGY 2055 and PowerPong 5000. So far, my work around for that was to program a a 4 or 5 ball drill with slightly different variation on each ball, and then run the drill on the "random sequence" format. it works well enough but it did take a little time to program the drill.

 

[dingyibvs]

1) Agree- very well packaged and protected within a semi-hard foam capsule. The box for the epad is very nice too lol.

2) I, too, feel the Pongbot delivers a significantly higher quality ball than the powerpong 5000 i was previously using. I honestly haven't even set the spin and speed past 6 for topspin because its already that powerful. but you do make me want to try to create a blocking drill at full power. might hurt if i miss the ball or get hit on my finger though LOL.

3) I did want to say, for each ball within a drill, you can increase the number of those balls. for example, ball #1 can be programmed to be repeated twice, three times, four times and so on. I do agree it would be cool if we could combine multiple drills together! I also wish there was a longer "reset time" after each drill. so far the only work around is to set the ball delivery frequency on the first ball to 0%...or rest while the robot cycles through a whole drill sequence.

A "random" feature on a specific ball, like you mentioned would also be great! I miss that function from the NEWGY 2055 and PowerPong 5000. So far, my work around for that was to program a a 4 or 5 ball drill with slightly different variation on each ball, and then run the drill on the "random sequence" format. it works well enough but it did take a little time to program the drill.

The Amicus Prime and the Powerpong have a small keyfob sized Bluetooth controller. It's super useful and I hope pongbot adds it too. I could hold that in my left hand and start/stop the drill whenever I want. With the pongbot I have to run to the Epad or my tablet and hit stop while the robot continues to spit balls out.

 

[Jslick89]

The Amicus Prime and the Powerpong have a small keyfob sized Bluetooth controller. It's super useful and I hope pongbot adds it too. I could hold that in my left hand and start/stop the drill whenever I want. With the pongbot I have to run to the Epad or my tablet and hit stop while the robot continues to spit balls out.

That would be helpful! I know the nova s has a little remote control. For now. I mostly control my Omni s with my cell phone app. I put my
Phone in ball basket next to the table.

 

[plunder]

That was a quick measurement, I'll do a more exact one tomorrow.

Thanks, yes please do so thanks!

What i'd be concerned about the Pongbot (any model) is the bouncing height: how high (i.e. low) would the ball bounce off from my table? Can the robot produce really low bounce heights, really low flight trajectories, really low serves/services? From all Pongbot vids i've seen so far, similar to most returnboard vids, the ball bounces off the table pretty high. 🤷‍♀️

Training with high flying balls is, to me, rather witzlos (pointless).
Training with low flying fast balls would be much more realistic, challenging, and interesting. [It's possible to do so with Amicus]

On the forum, unlike in sponsored youtube vids, let's try to be clear, detailed, open, and honest about the quirks and limitations of a product.

 

[dingyibvs]

Thanks, yes please do so thanks!

What i'd be concerned about the Pongbot (any model) is the bouncing height: how high (i.e. low) would the ball bounce off from my table? Can the robot produce really low bounce heights, really low flight trajectories, really low serves/services? From all Pongbot vids i've seen so far, similar to most returnboard vids, the ball bounces off the table pretty high. 🤷‍♀️

Training with high flying balls is, to me, rather witzlos (pointless).
Training with low flying fast balls would be much more realistic, challenging, and interesting. [It's possible to do so with Amicus]

On the forum, unlike in sponsored youtube vids, let's try to be clear, detailed, open, and honest about the quirks and limitations of a product.

Yes, that was one of my concerns as well. You can of course make the launch point higher by putting something beneath the robot, but not lower. With that said, I've found that to be less of a concern over time.

When I first got the Amicus Prime, I actually had it on the lowest possible setting, even lower than the 1st red line. I wanted the most difficult balls to practice against, so I figured a lower ball would be useful. After a while, I realized that it's actually not realistic for most balls. With topspin practice, the ball is usually actually a few inches higher than the net when it crosses your opponent end of the table. So now I only use the lower launch points on the Amicus for 2 purposes. 1) Slower, very spinny balls, to simulate extremely spinny, low trajectory opening loops for counter-looping practice (+6 spin, 16 speed). 2) Low, long pushes (-2 to -3 spin, 12-13 speed). When I do regular topspin practices, I have the robot raised to the middle red line, which is where the Pongbot robot is at, so that's perfectly fine.

Another issue with the Amicus/Powerpong is that really bad with services. It can't make a long, fast services because the first bounce cannot be near the end line. It can't do sidespins well because of the sidespin it naturally introduces. Also, I think it's because of the deflector mechanism, but it can't do strong short topspin services. Even at max +7 topspin, which is superhumanly spinny for regular loop practice, it for some reason still doesn't feel like a very spinny service. It really only does strong backspins well, but even then it can't produce quite the sidespin trajectory I'd like for it to produce.

 

[dingyibvs]

Anyhow, I made some more precise measurements. From the table to the middle of the ball would be ~235mm. Which is slightly below (~10mm) where the middle of the ball would be when it hits the top deflector on my Amicus Prime when set to the middle red line.

 

[plunder]

the middle of the ball would be when it hits the top deflector on my Amicus Prime when set to the middle red line.

the Amipong has 4 painted rings, and my pdf manual talks of "number of rings visible", which is clear what exact pole setting this description means. Saying "middle red line" makes me scratch my head 🗣️ 🤔
probably you mean "set to the standard position with the upper 2 rings still visible", so the aluminum pole is secured right at the 2nd ring, yes? Okay, i got the same measurement then. [since the 4th ring is the highest allowed position and the 1st ring the lowest, it's correct to count in this direction ⬇️; so, "secured to the 2nd ring" be defined as the "standard position" -- this way we also manage to avoid naming colors because the Powerpong uses grey painted rings, with only the 4th ring being red as a clear warning to not secure the pole higher than that!]

I could argue that either ejection height (Pongbot vs Amipong) be a bit too high, i.e. producing rather easy balls (bouncing off quite high) but that would be a draw then, in the standard position.

The Halo ejection height is fixed, so i caht imagine if it can produce fast long low serv(ic)es, i.e. long (=fast) services which are also low.

The Halo app accounts for 3 distinct room positions at the table. Not bad. Still a lot room for coverage improvement, around the table, far off the table, and at other height levels.

 

[dingyibvs]

the Amipong has 4 painted rings, and my pdf manual talks of "number of rings visible", which is clear what exact pole setting this description means. Saying "middle red line" makes me scratch my head 🗣️ 🤔
probably you mean "set to the standard position with the upper 2 rings still visible", so the aluminum pole is secured right at the 2nd ring, yes? Okay, i got the same measurement then. [since the 4th ring is the highest allowed position and the 1st ring the lowest, it's correct to count in this direction ⬇️; so, "secured to the 2nd ring" be defined as the "standard position" -- this way we also manage to avoid naming colors because the Powerpong uses grey painted rings, with only the 4th ring being red as a clear warning to not secure the pole higher than that!]

I could argue that either ejection height (Pongbot vs Amipong) be a bit too high, i.e. producing rather easy balls (bouncing off quite high) but that would be a draw then, in the standard position.

The Halo ejection height is fixed, so i caht imagine if it can produce fast long low serv(ic)es, i.e. long (=fast) services which are also low.

The Halo app accounts for 3 distinct room positions at the table. Not bad. Still a lot room for coverage improvement, around the table, far off the table, and at other height levels.

Yes that's right, the 2nd line from the top, so the 2nd lowest position. You can actually lower the robot even lower than the top line, so it does have the advantage there. So how did you get your retrofit? I think that'd be an interesting addition to the Powerpong/Amicus.

As for the position at the table, the Halo can actually be placed anywhere. It suggests three locations, but you can place them anywhere you want. Place it 1-2 meters back from the table and you've got a fast long service. The problem with this solution, however, is that if you want a different service then you have to move the robot, so you can't have a random long/short service training.

IMO if you want to practice the service receive or short game, you really would need a very powerful tabletop robot. I hope pongbot makes a tabletop version of the Halo/Omni. Combine that with a stand and move the ball feeder to the top and you'd have basically a Halo that can also be placed on the table. Something like the Tenni-robo actually, which I would've been interested in if they were still available.

 

[plunder]

So how did you get your retrofit? I think that'd be an interesting addition to the Powerpong/Amicus.

plz check again, I detailed about sourcing the retrofit kit in post#24, European Amipong users doht need to wait for the restock but could simply order the material from the amazon link. That solution is heaven-sent since it eliminates all those 8 bullet points.

Do you know the video contribution by Seth Pech? He outlines how he creatively sets up the Amicus, i.e. placing it far off the table, lowering its ejection height, for service-receive practice. I am going to setup and save such drills too:

Do you believe that the Nova S Pro is not powerful enough a tabletop robot for service-receive practice?

 

[Jslick89]

Thanks, yes please do so thanks!

What i'd be concerned about the Pongbot (any model) is the bouncing height: how high (i.e. low) would the ball bounce off from my table? Can the robot produce really low bounce heights, really low flight trajectories, really low serves/services? From all Pongbot vids i've seen so far, similar to most returnboard vids, the ball bounces off the table pretty high. 🤷‍♀️

Training with high flying balls is, to me, rather witzlos (pointless).
Training with low flying fast balls would be much more realistic, challenging, and interesting. [It's possible to do so with Amicus]

On the forum, unlike in sponsored youtube vids, let's try to be clear, detailed, open, and honest about the quirks and limitations of a product.

I see you are very passionate about the Power Pong robot. That is great. I think one key difference to note here is that @dingyibvs has used both products (PongBot AND PowerPong) while it seems you have only used one? In that aspect, @dingyibvs is able to provide a more clear, detailed and honest review and comparison of the products. Likewise, I have used several robots; Newgy 2055, PowerPong 5000 and Butterfly Amicus prime. With most of my practice being on both the Newgy 20255 and PowerPong 5000. Seeing as how he still owns his PowerPong and just recently purchased the Pongbot, i'm not sure what the incentive would be for someone to not be honest?

Anyways, since we are talking about realistic and challenging balls, for me, it's not debatable which robot produces are more realistic side spin shot. As the conversation implies, a big part of table tennis is being able to excel in serve return. Many opponents doing pendulum serves with heavy side spin + heavy underspin. Or tomohawk / reverse pendulum serves with heavy side spin + heavy underspin. When I was using the power pong 5000, it simply could not produce those types of balls, so unfortunately, I had to take those out of my robot practice. Now that I have an OMNI S, I am able to practice returns against those type of serves again. I've programmed perfect half long pendulum & tomohawk serves. and also short and long. and wide.

It seems like you might be saying that the retro-fit solves this side spin issue (ball bouncing off the defelector plate and altering the spin and natural trajectory of the ball) . Would you be able to provide video evidence of this? And even if that were the case, all the retro fit would be doing is solve the problem that the Pongbot doesn't have, but the robot still costs $700 more.

As for the unnaturally high flying balls, I have not had this perception from any of the robots ive used, but then again , i'm only USATT 1700. What is your rating? I know german rating is a little different, but i'm just trying to get an idea of your play level.

Not to say you don't have a valid point with being able to adjust the height of the robot, but even if that was a huge advantage, i'm still not sure is worth $700 in value.

Also, In theory, i believe i could set up my omni-s on a Chair further away from my table if i really wanted to, just like seth pech. I have not had the need to do that yet though

For me, the Pongbot shines in:

1) best price
2) true side spin production
3) high quality ball delivery

The powerpong Omega shines in:

1) Having a more mature app with more user-friendly features
2) being able to easily adjust the height on the robot?

Thats why it was a no brainer for me to get the Pongbot. For me, it was the BEST VALUE for my hard earned dollars. And this is coming from somone who dreamed of having a PowerPong Omega for a long time...until I got my hands on a Power Pong 5000 and realized some of the limitations with that product. It no longer was worth it for me to spend $2000+ on a robot when there was an option in the market that fit all my needs for significantly less of a price.

I think any club level player would be highly satisfied with with the Omni S or the Power Pong Omega. However, one is significantly cheaper and arguably delivers a better ball, and no doubt a better side spin ball. The cheaper one has some room for improvement on the application, but the support team has said they are working on improvements to the application.

Do you believe that the Nova S Pro is not powerful enough a tabletop robot for service-receive practice?

According to the specs on the website, Nova S is "weaker" than both the Omni and Halo. It seems like the Omni and Halo seem to have the same specs: Speed is 2-15m/s and spin is 100 r/s. Nova S is spin up to 60 r/s and speed is 2-15 m/s.

 

[dingyibvs]

I've been using it for a few weeks now, really digging the more realistic speed/spin ratio of the Halo. They also just updated the app a few days ago, adding a couple new functions. I'm now able to clone my custom drills (go to manage, then there's an extra option to duplicate now). It also says there's a function for "serving balls by combo repetition", but I couldn't quite figure it out. Maybe @pongbotstore can help explain it?

It's nice to see the app actually being updated. The Amicus app updated once in the two years I had it, which fixed absolutely nothing and broke the random service ability

 

[pongbotstore]

I've been using it for a few weeks now, really digging the more realistic speed/spin ratio of the Halo. They also just updated the app a few days ago, adding a couple new functions. I'm now able to clone my custom drills (go to manage, then there's an extra option to duplicate now). It also says there's a function for "serving balls by combo repetition", but I couldn't quite figure it out. Maybe @pongbotstore can help explain it?

It's nice to see the app actually being updated. The Amicus app updated once in the two years I had it, which fixed absolutely nothing and broke the random service ability

Regarding the new function for "serving balls by combo repetition," it currently only works in the programmed drills. On the training preparation page, you can choose to start training based on either time or combo repetition. When selecting combo repetition, you can adjust the number of repetitions for each combo. For example, if a training session contains three single ball types (1, 2, 3) and you set the repetition count to 5 combos, the robot will serve a total of 15 balls (each of the three single balls will be served consecutively 5 times).

Hope I explain clearly on this. If not, please let me know. 😊 @dingyibvs

 

[dingyibvs]

Regarding the new function for "serving balls by combo repetition," it currently only works in the programmed drills. On the training preparation page, you can choose to start training based on either time or combo repetition. When selecting combo repetition, you can adjust the number of repetitions for each combo. For example, if a training session contains three single ball types (1, 2, 3) and you set the repetition count to 5 combos, the robot will serve a total of 15 balls (each of the three single balls will be served consecutively 5 times).

Hope I explain clearly on this. If not, please let me know. 😊 @dingyibvs

Thanks! I kinda figured it was for pre-programmed drills. It def would be useful for custom drills too, or decreasing the minimum drill length from 1 min to say 30 seconds. I'm practicing FHs right now, and at 60-70% frequency that's 54-63 FH shots in a minute! Even pros get tired after 40-50 FHs in a row.

 

[Jslick89]

I've been using it for a few weeks now, really digging the more realistic speed/spin ratio of the Halo. They also just updated the app a few days ago, adding a couple new functions. I'm now able to clone my custom drills (go to manage, then there's an extra option to duplicate now). It also says there's a function for "serving balls by combo repetition", but I couldn't quite figure it out. Maybe @pongbotstore can help explain it?

It's nice to see the app actually being updated. The Amicus app updated once in the two years I had it, which fixed absolutely nothing and broke the random service ability

does the app updated automatically? we just had our baby so I haven't used the robot since last week. Excited to clone some of my drills to vary the placements.

edit: just had to relog back into the app and the new duplicate drill feature was there. sweeeeeet!