I’m thinking about potting my ESCs and putting them outside the enclosure on my sub. I see that the BlueESC used to do this, but it was discontinued in 2017ish (When will BlueESC be available? - #19 by donoldham). Is ESC potting a reliable solution / Do we know why the BlueESC was discontinued?
Some background: I’m trying to fit a Zed stereo camera and Jetson nano in a 4 inch enclosure, space is very tight. To save space in my enclosure, I’m wondering if I can move components (like ESCs) out of the enclosure. ESCs seem like an ideal candidate for this, but now I’m wondering why the BlueESC was discontinued. Was the core concept flawed (were thermals unmanageable, did repeated thermal expansion/contraction damange the potting compound, was the PWM signal too noisy over any reasonable wire length?) or did BlueRobotics just decide to focus resources somewhere else?
That’s unfortunately from before my time with Blue Robotics, so I’ve brought up your post internally to see if I can get a response for you. Should hopefully have an answer within a day or two
Ah, its been a while since I’ve thought about this. We discontinued the original Blue ESC in April 2016, and spent about a year and a half afterwards working on a new version, without much success.
The short answer is that potting is not a reliable solution to keeping electronics waterproof, at least the way we had been doing it back then. Even now, we are moving away from potting in general for manufacturability and reliability reasons, which is why we recently released the new WetLink Penetrator.
It’s possible that we may have been able to overcome the issues with enough time and effort, but it was determined not to be worthwhile, given a lack of strong demand and and already solid reliable solution with the Basic ESC. We prioritized other projects.
I completely understand the manufacturability reasons to move away from potting, but can you elaborate on the reliability issues of potting? (should I replace my potted cable penetrators with Wetlink penetrators?)
Until someone else in the know answers you- I can throw in my two cents.
I think there reliability issues of potting electronics come with large volume production. It’s too difficult to pot a large volume of products with a low failure rate. In other words, without testing each part on the production run, it is too hard to ensure the entire run of parts will perform reliably.
If you were to pot and test your own ESCs which perform well, you could be sure of their reliability. If you were to pot 5000 ESCs (like Blue Robotics), then you couldn’t guarantee they would work unless you tested an unreasonable amount of them.
In my (inexperienced) opinion, I think that if you had a set of potted penetrators that have worked reliably in the past, and you’ve used them for some time already, they have proved their resilience in their current use case.
I’d say this depends quite a bit on your operating environment(s) and the context you’re operating your ROV in. I believe the biggest contributor to issues is large temperature fluctuations causing delamination between the potting compound and the smooth anodised penetrator surface. That’s something that you can regularly check for and keep an eye on, but ideally the less regular checking that needs doing the more pleasant it is to use a product, which is a big part of why we’re transitioning to the new WetLink epoxyless design - it’s consistently more reliable because it doesn’t rely on a bond that can break down over time/through exposure to adverse conditions.
Of course the deeper you go the higher the pressure differential is from outside to in, so the effects of delamination would then be exacerbated by water trying harder to force its way into any gap it can find.
This is also definitely a factor - while scale can come with efficient processes and concentration of relevant knowledge, it’s much easier to miss something in batches of thousands than it is to do in a set of ~10 that you’re personally going to use. That’s effectively a design problem - consistency of an epoxy join is much harder to guarantee than that of a machined or fixed-shape moulded part.
The best suggestion I can give is to consider the conditions your penetrators get used and stored in (with respect to temperature fluctuations and pressure cycling), as well as the importance of your equipment and the work you’re doing (to you and possibly clients). As with all such things there are questions of what’s your level of acceptable risk, and how can you bring down the operating risks (storage consistency, equipment inspections, different products), along with a tradeoff between the cost and installation time of changing to a new thing vs the maintenance/inspection time of the old thing and potential cost of component failure.
Thank you Eliot, Adam, and Tyler, that perfectly answers my question. I’ll have to think harder about potting my own ESCs since they’re liable to get hot.
I’d definitely recommend upgrading potting penetrators - while storage temperatures, water temperatures, and operating depth all contribute to failure, the ease of switching vs. the cost of damaged electronics is an obvious choice, at least for me. I’ve ben using BR penetrators since before the BlueESC was discontinued, and have always felt betrayed when a proven device failed unexpectedly, from one day to the next.
WetLink for the win!
