But you were aware of the existence of Schaeff Inc?
Actually they developed all the electronics for their products before the Great Downsizing. Which consisted of all the different models of the W-Series and ECHO forklift product lines. The other products they have on their web page now were acquired from other companies. They also wrote all the control software for the forklifts, which I occasionally had a part in, between my hardware design projects.
During my years there I put much work into improving the design of the system controller and operator display PCBs in the W-Series, and I designed the system controller, operator display PCB, and power supply / hydraulic coil driver PCB for the ECHO model, among other things.
I was. The only separtely excited development work I was aware of at the time was at Raymond though.
My emphasis started off in the other direction. I did the control SW and as resources grew tighter (fewer people) I started picking up the electronics. A physics background seems to produce a bit of a JOAT.
I worked for SRE Controls. At one point we were down to 3 people. One person in production, one management etc and me. I was R&D, customer support, and 1/2 dozen other things for a while. They've since gone through receivership and re-emerged.
Resources were almost always quite constrained.
One of my experiences with customer service early on involved, for lack of a better term, a Samson test. I had gone down with another employee to replace a controller at a customer site. We replaced it, got some feedback on the behaviour of the diagnostics and sent the truck back into service. The driver took the forklift, placed it against a load bearing pillar and pressed the accelerator to the floor! The wheels slipped an
1/8 of a turn or so and the truck stalled. He held it there for a moment, sort of nodded said "it seems to work" and headed off. At which point I sort of picked my jaw up off the factory floor and we left.
When I started at Schaeff in 1994, several people there were ex-Raymond employees, including the General Manager and the guy I mentioned previously that spent an hour trying to figure out why the clock on a controller board was running at 700 Hz.
Schaeff's forklifts have "stall protection". We just programmed them to use the encoders on the traction motors to detect when there was no movement for a couple seconds even though significant power was being dumped into the motors. It would just stop and display an error message. If you returned the control handle to center it would automatically reset and let you drive again.
I learned real fast to never underestimate the ability of a customer to abuse a forklift.
Normally service issues were handled by the dealers and Schaeff's service department. But occasionally when they were stumped one of us engineers would go out to the customer site with Schaeff's service people.
One of the sites I went to was a customer that managed to scrape all the paint off the sides of the forklift by bumping the sides into things, and then dragging the forklift against whatever it was, such that all the paint was gone on the sides. And the steel was always polished nice and shiny where the paint was missing, meaning that they were doing it constantly, since otherwise the bare steel would get some sort of thin layer of rust rather quickly. Oh, and the quarter inch steel plate that the side of the forklift was made of would be bent inward a bit. That sort of made me wonder about the structural integrity of the building they were working in.
One of the common things that customers would do is to plug the battery charger into the forklift instead of the battery. If you've worked on forklifts you probably know this, but for others, someone way back when decided to design the standard forklift battery connector in a symmetrical shell shape, so that there is no male or female style connector. Two identical connectors can be plugged into each other by turning one 180 degrees. This allows the customer to plug the charger cord into the forklift battery connector instead of plugging the charger into the battery connector.
The problem with this is that a lot of older chargers are just a giant transformer and full wave diode bridge. No filtering on the output even. And a charger for a 36V forklift can go as high as 70V at the peaks of the full wave rectified sine waves. Schaeff's controllers didn't appreciate that, and it tended to let the magic smoke out.
So I had an idea. We put a textbook over voltage crowbar circuit into the controller. The idea was that if someone plugged the charger into the forklift and turned on the key, the crowbar would short out b+ and b- and blow the 5 amp fuse for the electronics. The electronics would be saved and it would also send up a red flag to the user that they did something wrong and they should figure out what they did, and then replace the fuse.
Implementation was a bit harder, since one of these chargers can put out 350 amps. It turned out that in the few milliseconds that it takes a 5A fuse to disappear in a blinding flash you would still get a current pulse of about 1000 amps through the crowbar circuit. I ended up finding an SCR from Motorola that was rated for something around 800A for one half cycle of 60Hz. And in a TO-220 package at that. A little underrated for the task, but it was the highest current part I could find that could be mounted on the PCB, and it worked.
The problem that goes back to not underestimating the ability of the customer to abuse the forklift was that instead of replacing the fuse properly, they would wrap the blown fuse with some of the foil paper from their pack of cigarettes, or replace it with a 30A fuse. At that point, the crowbar would try to do it's job, but in the process the trace on the PCB between the power connector and crowbar circuit would disappear.
I ended up redesigning it with a different circuit that used a
100V p-channel MOSFET to just open power to all the electronics when the voltage went over about 50V.
That wouldn't have worked for us for several reasons. First there was no separate speed feedback and there is no good way of determining speed of a series wound motor w/o one. Probably more important was hill-hold, you had to be able to hold the tractor at stop on an upward incline. That's really a superset of starting on an incline.
We did have three levels of current limit. A HW limit that protected the controller againstunreasonable demands. It would normally only activate under degenerate conditions. A short term limit with a 10-100mS response meant to offer some controller protection and to protect the mechanicals from over-torque. And a third long term limit operating over minutes to hours that was meant to protect the motor from overheating. Those last two would have some of the same effect as your stall protection.
I was just surprised at the forklift operators confidence that the tall skinny pole he was pushing wouldn't give way under the sidways push he was giving it. Stall tests were part of the production tests for the controller itself.
Isn't that the truth. I ran into another one today. I got a call from one of my clients to deal with a service call. Apparently a customer had used the EV's 36V battery to boost the battery on an ICE powered Forklift!
We had a client that essentially ended up building there own forklift. Sideloaders in their case. First the mast wasn't strong enough so they milled their own out of a block of metal. Then the bumpers were being bent when they hit the end of the, narrow, end of aisles so they beefed those up, finally the end of aisle bumpers were being bent so they replaced those :)
They can be keyed to prevent that, but I think I've only seen one facility that actually did that.
For control circuitry, I think I'd use a polyswitch in and a zener for that.
One of the app notes I did for a line contactor dealt with the power section for that case. Although I think that the original complaint was that they didn't like the spark they got from charging up the caps. In any case if the key was left on and the battery unplugged the contactor would drop out and plugging it back in would not cause it to pick up, you had to move the key to the start position to do that and it had a spring return away from that position.
Oddly enough we didn't run into much problem with that. For a couple of reasons I suspect. One was that the power section could withstand that kind of voltage as long as it wasn't switching with the possible exception of the coil drivers. The control was probably protected with a polyswitch. The other reason though would be if you did plug the charger in with the key on you would hear that startup diagnostics and that might be enough to startle you into disconnecting again. If the key was off the only item to be affected would be the power section. We did get a few that might be traced down to that though.
A bigger but related issue we had was with hot batteries. A battery just off charge can sometimes have a quite high voltage on it. We ended up adding an overvoltage shutdown where the controller refused to start if the voltage was more than 55V.
One of the last projects I did was a separately excited controller for series wound motors. It did this, although we did dispense with the feedback sensors. You can't get to quite the degree of control but it physically more robust and it's not an autonomous vehicle after all. It did hold pretty well though.
As near as I can tell pretty much the same keyswitch type is used for them all. In fact the aftermarket ones I've seen use the same 'key' for all trucks. Most controllers seem to ignore the strt position though. That's where the line contactor app note came in. Even though all it involved was an extra diode I'm not sure more than one outfit ever used it.
Ours worked over a nominal 24-48 and had to be setup as well. I was asked occasionally to auto-recognize the battery voltage but always resisted to avoid confusing a higher voltage battery with a bad cell and a hot low voltage battery. Doing that on every startup seemed a bit risky to me. Also that was only one of the items that had to be setup so we ended up doing an assisted setup where we would ask the installer to confirm the nominal voltage was what we thought it was.
It gets even worse when you move to higher voltages (as we did). I don't think there is much hope of certainty when the nominal voltages you are trying to distinguish are 60V, 72V and 80V, all of which are reasonably common once you move above 48V.
Now that you mention hill-hold, I realize my explanation above wasn't right. The stall protection cutout required the control handle to be pushed forward significantly with no movement from the forklift for a few seconds, precisely so that it wouldn't cut out when sitting on a ramp or going up a ramp extremely slowly.
Schaeff's W-Series has a closed loop control system. If you have a 7 mph top speed, and you push the handle half way forward, it will go 3.5 mph, no matter what the load was, or even if you were going down a ramp. Yes, you had to push the handle forward to go down a ramp. I thought it felt weird at first, but later when I got used to it I decided that it was the coolest feature for a forklift.
I was always impressed when I stepped on the brake on a ramp and you could hear the traction motors kick on and hold position.
Schaeff's current limits were all in hardware. But we also had a fancy software routine that monitored the current going into the motor and did an estimation of motor temperature. I was skeptical when they first started developing that code, but in the end it proved to be surprisingly accurate and more than sufficient to protect the motors from overheating.
Schaeff's forklifts had a key it was kind of like a car ignition. It had an off, on, and start position. The start position was spring loaded to return to on. So if they unplugged the battery and then plugged in the charger, it wouldn't go into run mode anyway, so you wouldn't hear any of the startup diagnostics that tested the contactor and motors and stuff.
Schaeff's forklifts were mostly 36V but they occasionally sold
48V ones as well. We used the same controllers on each, but they had to be software configured for the right battery voltage for the control systems to work right. So even though the electronics could take a hot 36V or 48V battery, we still had a software check that the user had connected the proper voltage battery. It was a bit picky to get that working and not have a hot 36V battery be detected as a 48V, or a low 48V battery detected as a 36V battery, and cause a wrong battery type error.
We tried to come up with a way to calculate motor speed in software and dispense with the encoders on the traction motors, but we could never get the ramp hold to work as well as we wanted.
The encoders were kind of overkill. They were 256 count encoders, but the ASIC that Schaeff designed for motor control stuff counted all 4 edges, so you got 1024 counts per motor revolution. There was a 20:1 gear reduction to the wheel, so you got 20480 counts per wheel revolution. I don't remember exactly, but the wheel diameter was in the ballpark of a foot, so that means we could measure something like 1/6500 of an inch of movement. Way overkill, but once we found an encoder that was capable of surviving the environment, the high resolution sort of came along with the package.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.