Reassessing the principles behind the heave add-on

Readers who have followed the project from the start know we’ve followed the maxim that heave (or up and down motion) is nice to have, but not essential. Lifting the whole platform, and user, is very expensive in terms of power and engineering strength which is largely why motion simulation is so expensive to get into. Rotating the user is cheap, so that’s what we’ve concentrated on figuring out first.

We realised early on though that even though heave wouldn’t be a major focus it would be awesome to add sooner or later. If our average customer is happy to pay X for a sphere they might be happy to pay X again to get that sphere to move up and down. Ideally the mechanism for heave would fit inside the base and be backwards compatible. It seemed like a good principle, but perhaps not the best idea.

We want to get the sphere cockpit perfected, but having to design around a future mechanism makes the task rather more complicated. We would need to actually design and build the heave add on at the same time. This falls into the realm of ‘feature creep’ which is pretty dangerous when trying to make something on a budget, spending a good chunk of your prototyping budget on a future feature isn’t the best plan.

Instead I’ve realised that there is a great opportunity for simplifying the idea behind the heave add on. We always believed that the ‘end game’ for the platform was a sphere mounted on a six degree of freedom platform which would give the most options for users so why not have that as a guiding principle instead. A simple lifting mechanism inside the base doesn’t move us towards that final goal.

We also recognise that while people react well to the sphere lots of people are still in the market for a ‘traditional’ simulator. They still want an affordable 2 or 3 degree of freedom system (pitch/roll + heave) which the sphere design doesn’t fulfil.

So, it makes sense to actually design TWO simulators. The spherical cockpit version and an affordable base with pitch, roll and heave that can also be updated to add sway, surge and yaw. We treat both designs as separate unit that work individually but can also be combined. So users could buy one, or the other, or both. They might decide to upgrade either unit. You also have the option (if you had both) to separate the modules if you had a friend over to play 🙂

As usual the philosophy is to ‘keep it cheap’ so I threw together a quick frame version.

simple frame60  Clearly this is no good, what is the point of making something affordable if it’s too ugly to want?

Instead I think it would be better to have something with more elegant lines. This could be made by bending three sheets of aluminium and reinforcing them with an external frame.
simple frame2

The sphere is already quite large so this would add more bulk to the setup, so instead of three large parts we create smaller ones with the center missing, as demonstrated by the right hand picture below.


This means the heave add on is much smaller when used separately, perfect for your own chair or with a ‘playseat’.


If you want the whole setup you could use the same parts but add three panels to fill in the center of the platform. It’s possible these could be dispensed with since the base for the sphere would cover any gaps depending on how it’s designed.


This idea is more like a 3D sketch to get the ball rolling and creative juices flowing, it’s very early in the whole process but I think this could be the way to move forward.

Two platforms that could be used alone or combined would be a great way to give users more choice, as well as increase functionality.

Comments, as always, are welcome.

My MailChimp Experiment

Normally I write a newsletter and send it out once it’s done right away. I’m hardly an expert, we’ve only sent six.

This time I though we’d try something new. Apparently the best time to send mails is around 10am, I guess more people are at their desk and more likely to open them but this is tricky to time when you have a massive list of mails.

MailChimp does have a “Send with Timewarp” Feature and also “Let MailChimp optimize time for maximum engagement” but these are both paid for options….

Instead what I did was export my list to excel and reorder the list by the recipients time zones. This idea isn’t perfect and about 2/5 of my list didn’t have that as an option, but it was a start. I suspect a lot of these were due to the fact that I imported my list from another program.

I wrote my campaign as a template and then was able to schedule each major time zone list to land at approximately 10am. The unknown timezone addresses I sent at 6pm GMT, which was 10am on the US west coast. Most of the list are Americans and Europeans so this is a reasonable guess.

The last three campaigns have had a 50% open rate over all, so I will report back in a week or so if this has improved. It took about 10 minutes to paste in all the values but if it drives up the engagement it might be worth the extra effort.

How many pre-signups before launching my crowdfunding campaign?

I posed this question over on reddit and got an interesting response.

“A crowdfunding campaign, in itself and providing you manage it properly, will attract sign ups. Based on my own experience, I now use this rule: 1/3 of the funding goal, based on the expected average pledge.

If your goal is 25k for a $25 average, then (25’000 / 25) / 3. So roughly 300 sign ups.”


If this is true, and you need $25k then you could perhaps expect to be 100% funded on day 1 and end up with at least 300% at the end…. right?

Read the full thread here

I blocked China again….

Seems only one person actually looked at the site, but 211 hack attempts were registered by Cloudflare…

Next worst offender is the US but they’re also the best customers… :p

Another UK reply

One went into my spam folder but at least they sent a decent reply

“Unfortunately your requirements are outside of our fabrication capabilities and would have to decline on this occasion.”

They advise me to look for a ‘pressings’ company, which I thought they were…. time to hit the google again!

The decline of British industry

After I uploaded my RFQ (request for quote) to Alibaba I went to bed and woke up to 10 companies in China asking for more information and giving me quotes.

I email 11 British companies and 24 hours later one couldn’t ready my dwg files (the standard file type), and after receiving my pdf say ‘this isn’t a job for us’.
Another replied that ‘Tooling is going to astronomical.’ but quickly followed by ‘Were going to decline to quote, sorry’ after I email them back saying that I realise that.

The other nine companies are presumably so swamped by business the person who checks the email is out hammering panels :p

I haven’t given up on making them locally just yet, but the enthusiasm of the Chinese companies is quite a big contrast to the UK…

Incidentally if you know somewhere that can make the panels please get in touch.