AndyMark EVO Dog Gear Replacement Issue


We are sending this email out today to all of the teams who have purchased EVO Shifters from AndyMark:

AndyMark is reporting an issue with the EVO Dog Gear (am-3492). The EVO Dog Gears in the field may break under high loads.

This past weekend at a pre-ship scrimmage, a FIRST Robotics Competition team experienced a break and sent us a report. This report cited mistakes in our design. We are committed to making products you can trust, especially on something as critical as your drivetrain, and will fix this mistake by sending out stronger dog gears to all of the early adopters of this product.

Summary and Action

Our current EVO Dog Gear has a counter bore for the connecting shoulder screw which is too deep.

We are committed to keeping your trust, and to ensure your EVO Shifter performs to your expectation we will send out updated EVO Dog Gears with much shallower counter bores and longer shoulder screws to all EVO Shifter customers. We know this will require precious time to replace at competitions or Robot Access Periods, so we also will provide a credit of $25 for each EVO Shifter purchased. Replacement EVO Dog Gears and longer shoulder screws will begin shipping to you by next Monday, Feb. 27th.

Issue Details

The current 4-tooth Dog Gear with Deep Counter Bore has a 0.188 deep counter bore for the head of a carbon steel shoulder screw, as shown in Picture A. The shoulder screw is similar to McMaster-Carr part 92012A505, but is made of carbon steel, not stainless steel. The breakage in the field this past weekend was made in the area of the counter bore. We realize this area is a weak point of the dog gear.

During our summer 2016 testing, this counter bore was shallower and did not fail. When we chose a production shoulder bolt, we picked one with a taller head than our test version. After choosing this taller head, we erroneously made the counter bore deeper in order for the top of the screw to be flush with the outside of the dog gear. Our solution is to make a new EVO Dog Gear with a very shallow counter bore, as shown in the drawings below.

am-3492, the original 4-tooth Dog Gear with Deep Counter Bore 4 Tooth Big Counterbore Dog Area.PNG

am-3492a, the New 4-tooth Dog Gear with Shallow Counter Bore 4 Tooth Dog Tooth Area.PNG

This shallower counter bore allows more material in the cross section of the dog gear, shown in the top portion of the am-3492a picture. We are confident this fix provides a robust dog gear for FIRST Robotics Competition purposes, exceeding the amount of material we tested during our prototype tests.

Here is a drawing (PDF link) of the New 4-tooth Dog Gear with Shallow Counter Bore.

EVO Dog Gear Replacement Process

The total time needed for this change will vary depending on your robot’s design, but it will only take about five minutes once removed from the robot. We are editing a video depicting the process, which will be posted to the EVO Shifter product page later today (edit: it’s finished). Total filming time was ten minutes, so we do believe this is a quick process.

We apologize that this creates another task to complete either before the Stop Build Day deadline, during a Robot Access Period, or during the practice day of a regional competition. If you have any questions, please email customerservice at or call us toll-free at (877) 868-4770.

Andy B.

High quality customer service at its finest right here. Nothing but the best from Andymark for sure.

I can’t think of another company that handles issues this well. Truly one of the best out there!

Here’s the “report”, for all curious. Kudos to Andymark for the quick turnaround. I’ll send you the bill for the consulting hours (jk).

For any team who’s first regional happens to be Pittsburgh and you need to make this dog gear swap there- if you want help come find me with Team 4027. Myself and several of our students know this gearbox inside and out now.

I personally think, teams who use AndyMark EVO shifters, should be given a chance to unbag the robot and replace the dog gear before the competition.
If the dog gear fails during the competition, you might as well be out. If you do it on the first day in the pit, you are taking away the precious practice time from your team. A 5 minute job ,according to AM, may take hours depend on your robot design. I’m really surprised that FIRST didn’t address the issue. To me it is. But that is my opinion. I wonder what CD community thinks about it?

Making rule changes based on potential COTS failures is a can of worms you don’t want to open. Specifically, FIRST endorsing one Vendor over another is something that should not, and will not, happen.

Why make an exception for an AM product? There are loads of other arguments that could be made for defects that should be allowed unbag time to swap out.

It’s a can of worms best left shut.

Yes, there were other COTS defects in the past, and in my 15 years i can name a few, however none were as time consumable as this one

If a gearbox component replacement takes you out for an entire tournament, you did not properly design for maintenance. I say this from the perspective of a team that had to replace multiple gears in our drivetrain last year. The first two each cost us a match, but the third (and final) time was finished during the span of a timeout in the playoffs.

If teams with this issue are given unbag time, I would hope that the following also get some:

–Teams who lost time due to snow days.
–Teams who lose time due to school test days (mandatory time off).
–Teams who don’t get their kits until after Kickoff Day.

If it made sense to do so, I’d suggest that this be retroactive over the last 10 years.

You see where I’m going with this, of course. Once the “you get unbag time because X” starts, it’s more and more difficult to justify not giving unbag time to teams who have reason Y instead.

What this announcement does is give teams with these shifters time to plan their approach for the events so that the gearboxes can be coming out of the robot right about the time the robot is out of the bag. 5-10 minutes later, the gearbox is going back in.

In 2007 there was a design problem with the carrier plate of the kit of parts transmission from Banebots that caused the transmission to fail in high load situations. Many teams contributed time to test and gather data so that a proper fix could be found and Banebots and FIRST worked non-stop to find a fix and get it out to teams as quick as possible. The fix arrived very late in build season (week 6 if I remember correctly) such that teams basically had to choose between sacrificing testing time and/or finishing their robot depending on where they were at in order to get the transmissions beefed up, or running the weaker plates and praying a failure didn’t occur in competition. Even when the KOP included transmissions had a design problem FIRST didn’t allow anyone extra time for it; I don’t see why a gear for a COTS item only a small percentage of teams are using would justify an extension.

Read up on the double D carrier plate issue here. The response time on that issue coupled with other transmission failures is why I have stayed away from Banebots ever since.

I do remember pretty well the problem with the Banebots, I think P80, even though we didn’t use them. The problem with the shifter this time it’s quite different, it’s designed to mount in a certain way on AM14u3 frame. I’m sure teams that used them will understand me.

I haven’t used it, but I understand quite that it mounts to the AM14U3. I really like EricH’s argument. Students have lost days to snow, testing, etc. Why is this any different? I would argue that using a brand-spanking new gearbox is a choice of a team, and one that invites failures in your robot. The risk should be known by teams during design. However, no team can design around snow.

Why should snow days not grant you extra time, but this should?

As much as I disagree with some of the arguments, I totally agree with you that I took some risk in using a brand new design box. Thank you

We have the AM14U with the EVOs installed. We where able to pull the gearboxes with the motors attached from the bottom of the chassis before we bagged the robot. We did pull the front wheels to loosen tension on the belts to provide the angle needed to remove them, and we where able to rotate and move them enough to get them out. We have the long configuration so the gap was even narrower between the motors.

It took about 30 minutes from the time we started unloosening the first bolt until they where out, but it is no different in what we all deal with sometime in our career or life. Things just do not go as planned. I applaud AndyMark for recognizing the issue with only one failure reported so far, that I am aware of, and providing an updated Dog gear before competition. They could of sat on it and waited to see if more failed after the first week before recognizing it publicly. Definitely an inconvenience, but possible to overcome. Being in a District state does mean we have the opportunity to replace the gearboxes in our shop versus on Friday at the competition. That does relieve some of the pressure in getting this done, but even at that we would of dealt with doing it as soon as we could at a Regional event.

And I do applaud AndyMark for recognizing the problem. I am using AM shifters from the day they were introduced, i think 2006, and never had a problem, even when other teams were reporting a problem with shearing the dog gear roll pin. Our robot was going after Regionals to five off-season beatings with no super shifters problem.
On the EVO issue, it will definitely be a disruption on a day when you have to worry about inspection, working out programming issues,organizing the pit. When you have a small team that becomes a little problem. For now during our practice before the bag, we lowered the shifting to 25 PSI, i think higher PSI contributed to the dog breaking as well.

Are these being sent automatically to all EVO Shifter customers, or do we have to request a replacement?

When can we expect them to arrive? Teams (like ours) going to week one Regionals are headed out Wednesday afternoon, Feb. 29.

We could not get our EVO shifters to shift using any less than full pressure (60 psi). And sometimes the spring-return cylinder didn’t have enough force to shift back when pressure was removed. The only fix we could think of was to turn the single acting cylinder into double acting (adding an airline to the unused port). Have any other teams had to do this?

So far no failures during testing of our competition 'bot & test 'bot.


I can confirm that your parts shipped yesterday, and South Carolina is typically a 2 day ship point. Your team should receive your replacement gears tomorrow.

No customer will have to request replacements on these gears, and we are shipping them to all of our customers who ordered EVOs or the Dog Gear itself. We have gone through and determined which teams ordered EVO Shifters and what week they are competing, and shipped the first batch out to week 1 customers.

We are looking to ship the remainder to our customers today and/or tomorrow.

I believe AndyMark estimates the extend and retract at 40PSI, we did it at 25-30 PSI, it was probably a bit slower, not sure