what did you wrench on today??

[alke46]

Not m/c related but today I tore into my zero turn lawn mower. It has been leaving oil droppings on the garage floor so I had to put cardboard under it to catch the oil.
My 1st time for removing the hydrostatic transmission. Need to find out what is causing the leak. I am suspecting some sort of seal failure inside the two halves. Will talk to the maintenance dept at the shop where I bought it and discuss the issue soon.

IMG_20230619_121302136 by Larry Alkire, on Flickr

IMG_20230619_170235996 by Larry Alkire, on Flickr

IMG_20230619_170242576 by Larry Alkire, on Flickr

IMG_20230619_172734153 by Larry Alkire, on Flickr

[Buffalo Bill]

Yikes Bill. How did you fare?

Not so bad, luggage rack bruised my left calf as the bike slid away. That swelled up for a week.
Bike is ok, just the stator cover had a tiny hole. Brazed that shut.
So lucky a good friend drove down with a trailer. took me home.

[Suzukian]

Glad to hear you're O.K. What did you Braze that with?

[dorkburger]

Glad that you're ok Bill.

[Buffalo Bill]

Glad to hear you're O.K. What did you Braze that with?

I bought a MagTorch kit and some US Forge self fluxing 1/8" aluminum rods. Did the job but no regulators, so flame changed as tank pressure dropped.
I like brazing and think it's worth getting a light duty gas welding kit with regulators.
Lot of things could be gas brazed easier, than electric welded.

[Rich82GS750TZ]

I know next to nothing about welding or brazing of any kind, but this video popped up a few weeks ago. I didn’t know this was possible.

[Buffalo Bill]

Excellent video, amazing build.
Did not like his method of putting a saw blade on a grinder, no thank you.

[Carter Turk]

Hey Bill, unfortunate about the low side. Off camber so hard parts were closer to the pavement .
Did anything touch to cause the low side? Just curious

[Suzukian]

Excellent video, amazing build.
Did not like his method of putting a saw blade on a grinder, no thank you.

I use a produce called "Alumaloy". I've used it for decades. I've fixed my aluminum ladders, lawn mower engine cases, etc. The stuff is amazing. The way that guy built it up is quite amazing. I usually put a piece of steel behind the surface I want to build up, as it will not stick to steel. I've milled this stuff, when I had to fix an aluminum cylinder head on a car, and you could not see the area I had filled with it. There's an actual exchange of electrons, so this is more than a braze, it's actually binding to the surface on an molecular level. You will be able to grind that as smooth as you wish to without worrying about affecting it.

https://www.alumaloy.com/

[joequesmith]

Washed out my diaper. Onto what appears to be an instrument cluster problem. The coolant temp bounces between cold and over temp. Sensor and wiring to ECM all appear good.

diaper.jpg

[joequesmith]

Inserted a fixed resistor in place of the sensor and the temp indicator is still jumping. Dug out my tablet, plugged it in and the temperature is stable as the cluster's indicator continues to jump.
The cluster uses a basic serial interface. Nothing near as advanced as CAN. The reset of the cluster appears fine so time to pull it. This is the second cluster I have had fail. The tach died on one which turned out to be a bad IC which I can't seem to source. Time to bust out the soldering iron.

This GS is a pretty old but compared with my GS1100E's, it's a rocket ship. ECM, what's that? lol.

cluster.jpg

[joequesmith]

Testing the cluster on my home made simulator.
HIL.jpg

[joequesmith]

I was able to replicate it with the simulator. It's really an odd problem. The simulator drives a fixed signal into the ECM. The logger shows the quantization noise but that's it. The cluster on the other hand just wanders. The other bad cluster had the stepper driver fail. Maybe this is a similar problem. I found a few damaged clusters that I hope to scavenge to repair the two problem clusters. One looks like it was in a fire crash, the other had hardly anything left. As long as the ICs are still alright, I should be able to use the. In the mean time, the last working one was donated to the bike for now.

I wasn't aware TI bought Pierburg, but now that the electronics, fuel pump and motor are sorted, time to start putting it back together.

[bobgroger]

I was able to replicate it with the simulator. It's really an odd problem. The simulator drives a fixed signal into the ECM. The logger shows the quantization noise but that's it. The cluster on the other hand just wanders. The other bad cluster had the stepper driver fail. Maybe this is a similar problem. I found a few damaged clusters that I hope to scavenge to repair the two problem clusters. One looks like it was in a fire crash, the other had hardly anything left. As long as the ICs are still alright, I should be able to use the. In the mean time, the last working one was donated to the bike for now.

I wasn't aware TI bought Pierburg, but now that the electronics, fuel pump and motor are sorted, time to start putting it back together.

Ferrite beads on the signal wires? Suppression ignition wires?

[joequesmith]

Ferrite beads on the signal wires? Suppression ignition wires?

Ferrite would normally be used for higher frequencies. The simulator doesn't really have ignition wires. It uses a custom wound pot core to simulate the coils. The ECM reads the analog signals and then sends that data to the cluster over a digital serial bus. The cluster then decodes that information and drives the various gauges. These are mostly stepper motors. So basically its all digital.

Because it is digital, I can sniff the serial bus with a PC and some software, basically making my own cluster. When I look at the data the ECM sends out, it's fine. Everything on the bad cluster is stable except for the coolant temp.

When the gauge is first powered up, each gauge's needle will go to their maximum position, then to their minimum as a sort of self test. There are no problems during this self test. This particular GS (gen 1, K6) uses the temperature gauge to warn the driver of a problem. Don't ask me why they did it this way. If the check error is displayed, the coolant temperature gauge will read max but the over temp LED will remain off. I guess it was done because the driver may not notice the CHEK on the LCD but the over temp may get your attention. This all seems to work correctly.

In this case, there are no faults and the coolant temperature acts like an analog gauge, moving around.

The tachometer on the other bad cluster is the problem. I can heat the stepper controller and it will start to work but it's not a soldering problem. It appears the IC itself has gone bad. Once I get the junker cluster, I will try and sort it out.