I was playing about with Jitsi Meet and I've got an install of it running. Who fancies picking an evening to all jump on for a chat? If you're on a PC don't need to install anything, just make sure your camera and microphone works (headsets work just fine). There's an app for mobile users too, which seems to work okay. Unlike competing products that sound like an 80s ice lolly, they don't send all your info to China ;-)

I feel like the impact wrench is more likely to apply a "straight" torque to the bolt but a breaker bar is pulling it sideways. I've always had good results getting things like rusted-in bleed nipples free using a T-bar and a 6-face socket, and using the T-bar like you would with a tap and die - equal pressure on both sides so you're only turning it and not applying any sideways force.

If I leave mine parked on any suspension height with the timer relay pulled, it will eventually end up straining against the end stops of the shocks, way higher that "oh shit I'm beached" mode.

I think the non-return valves are leaking allowing tank pressure to blow past the "up" valve into the bag valve gallery, where it then blows past the bag valves.

Bear in mind that the valves are just seated by the spring and held closed by air pressure on the back of them - if there is sufficient pressure on the other side of the valve it will lift the rubber disc off its seat and the valve will leak. This is what the NRVs on the "up" and "down" valves are supposed to prevent.

I suspect this is why you get the "dancing front" problem when you're stopped at lights - with your foot off the brake the front suspension valves both open to equalise the pressure between left and right, and if the "up" NRV is leaking then air will just flow straight into the bags.

I'm pretty sure there's more than that because in the (sadly a little outdated) docker-compose example on github it also mentions a plugin for subscribing to topics and messages, and one that stops the cacheing bug that means that sections stay flagged as having unread messages.

It used to work but then it broke and I couldn't get it working again without breaking signups. I'm not 100% sure that the Track Topic button works properly either.

I keep trying to make time to update the forum software and you'd think with the whole lockdown thing and all I'd have that, but we're busier than ever.

That triangular hub there looks like an old diesel flexplate would be about right... :-)

It did indeed slot back into place with a particularly satisfying "thunk", with a bit of turning and jiggling. When it's not fully on the splines it's quite wobbly on the shaft, and once it's on properly it's quite clearly "happier".

While Miah and I had the engine out to do the core plugs, it looks like the torque converter's fallen off the gearbox a bit. Any hints on getting the splines lined up properly so I don't wreck the pump when I refit it?

Gilbertd wrote:

When RRTH sent me a PM suggesting that I should become a mod

The hell...? :-D You need to get him on here! Someone signed up with RRToadHall years ago on here and I messaged him at the time but he swore up and down it wasn't him. I've always maintained that he would be a welcome and valued member of the rrpub community if he did.

I fitted "proper" Boge shocks about two years ago and they lasted about six months. The rubber bushes fell apart, and when I took the shocks off to replace the bushes I found that two were partially seized and two were doing sod all.

So far the Shitpart replacements at a fraction of the price seem to be holding up.

That is such an awesome story, and great photos! I might nick a bit of one for the next banner if you don't mind ;-)

The servos have a pot in them that develops dead bits, and the motors wear out. They use the same servos and more or less the same heater box as the Citroën XM! On those I found that I could clean up the motors a bit and get a bit more life out of them particularly if I avoided running the blower full chat when moving the flaps, because the motor controller senses an increase in current and uses that to decide when the flap is at the end of its travel. It'll run the motor each way until it stalls to get an idea where the end points are. If it can't move the flap easily because it's sticky or because the airflow is pushing back on it and the motor is a bit worn and drawing more current anyway, it'll decide that the end point is actually right in the middle of travel, or all its movement is up one end. You can try resetting it with nanocom or whatever and see if that gets it sorted.

kultur wrote:

Your confusing mud tracks for trails mate. See video posted by Morat.

I don't watch videos. The stuff I drive on is more kind of jaggy bits of mountain rather than mud trails.

If you mean the idiots that deliberately pick the worst route up something, I don't see the point in that. I drive offroad to get 500kg of tools and equipment and three or four people - including me - safely up a hill, get a day's work done, and get them all back down safely at the end of it.

kultur wrote:

Winches, big tires, and lifts all serve a purpose when you actually offroad on trails like the Rubicon and Moab.

I've seen folk driving on those "offroad" trails, and they look like the farm tracks round here. I drive a bone stock P38 on far messier stuff than that on a daily basis.

Why would you panic-buy bog roll when you can shoplift the Daily Mail? Of course you're risking catching its arse cancer...

Well if you're in touch wish him all the best from me!

I didn't get far with replacing the NRVs. My cunning plan was to find off-the-shelf "push fit" ones that could be machined down to fit the holes in the block, but I haven't really found anything that doesn't fall apart or have too small a hole up the middle to let enough air through.

I looked at using an Arduino to drive six big MOSFETs. The normal controller pack provides about a 100ms pulse of 12V to the coil then PWMs it at 30% duty cycle at around 24kHz, using a bunch of opamps, transistors and passive discretes. I figured that since microcontrollers are as cheap as opamps and actually designed for PWM I mway as well just use that to drive everything.

I got a prototype pretty much working, drew a board, and then never really got a chance to finish it off, same old story. I'll put the files up on github if anyone fancies a shot at it.

Finally after 18 months or so Google Adsense has reached the payment threshold of 60 quid, so that's the domain name paid for the next couple of years.

Not bad.