Paul & Sabrina's cheap DIY 144v motor controller

[jedsmd]

Originally Posted by Astro

My first guess would be a low supply voltage.
Is the supply voltage still around 5v when it is in the truck?
Maybe try using an independent supply voltage to test it when it is bolted in position.

Or try taking it off the firewall but still connected to vehicle supply to see if the mounting position is the cause.

The cable allows the throttle arm to move all the way to the physical stop?

Could the wiring changes to the controller still have a series resistor in the supply circuit or something like that that allows the supply to test as 5v unconnected but when connected to the pot box it doesn't deliver full 5v?

Okay did some experimenting

Being mounted to the firewall has nothing to do with it. bad guess

The 5V supply from the controller is fine it remains at 4.94V.

Drum roll ......

The controller is loading the signal line. If you disconnect the signal line from the controller the signal line rises to 4.8v at max throttle. Connect to the controller it can only rise to 3.85v.

I think they can only supply 2ma of current on the output, but the pull down resistor is sucking 1mA. Maybe the pulldown should be more like 47k instead of 4.7k.

yep, but I'm not sure it is a problem I need to fix as it all works fine with the program variables set for the voltage drop.

[Astro]

With the signal wire loaded with 4.7k to ground the current flowing in the signal circuit is 0.008mA at zero throttle and 0.844mA at full throttle.

i didn't have a 47k resistor handy to test the effect of increasing this value, actually i was using a 4.7k SMD resistor held by alligator clips as all my through hole components have all hidden themselves somewhere in the study. Obviously in fear of these new SMD interlopers.

[mrbigh]

Originally Posted by Astro

With the signal wire loaded with 4.7k to ground the current flowing in the signal circuit is 0.008mA at zero throttle and 0.844mA at full throttle.

i didn't have a 47k resistor handy to test the effect of increasing this value, actually i was using a 4.7k SMD resistor held by alligator clips as all my through hole components have all hidden themselves somewhere in the study. Obviously in fear of these new SMD interlopers.

Always there is a Radio Shack store with some spare parts on hand....and "if" the store clerk have an idea or knowledge of what U R looking for....

[jedsmd]

A new problem...

For monitoring battery charge I use an 'EV Display' from Cleanpowerauto

EV Display - CleanPowerAuto LLC

The display goes bonkers when the controller is making power. The display is stable when the controller is not making power, but the displayed numbers for remaining charge are now rubbish.

I think this started when I reinstalled the spread spectrum software for the hall effect throttle. I had been using the 8khz fixed frequency until recently and hadn't had any problems.

I had tested the spread spectrum earlier and had not noticed any problem with the gauge so that leaves me confused. Paul did you make any changes recently other than the settings for the hall effect throttle?

Unfortunately It is hard to go back to earlier software versions to test because of wiring changes to the controller.

Paul please set me up a version of the 8Khz fixed freq with the hall effect values so I can test it with the display.

Thanks
Jerald

[MPaulHolmes]

I just love mysteries like this. I'll do that now... new software sent!

[jedsmd]

Originally Posted by jedsmd

A new problem...

For monitoring battery charge I use an 'EV Display' from Cleanpowerauto

EV Display - CleanPowerAuto LLC

The display goes bonkers when the controller is making power. The display is stable when the controller is not making power, but the displayed numbers for remaining charge are now rubbish.

.....

I traded Emails with Dimitri of CleanPowerAuto. They have seen this type of EMI from newer controllers and they recommended a simple filter to add to their device:



I have ordered the parts from digikey and will implement.

Meantime I will try the fixed freq code Paul just sent to see if it makes a difference

[MPaulHolmes]

Do you have the isolated dc-dc feature? Or is the EV Display ground the same as the pack ground?

[jedsmd]

Originally Posted by MPaulHolmes

Do you have the isolated dc-dc feature? Or is the EV Display ground the same as the pack ground?

The 12volt sending unit ground is connected to vehicle ground. The pack positive and negative are connected directly to the sending unit (and are isolated from vehicle ground). There are no dc-dc's outside of what may be internal to the gauge.

Which raises a related question. Does the ground for the throttle hall sensor need to be isolated from vehicle ground?

[MPaulHolmes]

The throttle sensor ground is the control board ground, which is a filtered version of the output ground of the isolated dc-dc that's bolted to the roof of the controller.

[Astro]

Originally Posted by jedsmd

The 12volt sending unit ground is connected to vehicle ground. The pack positive and negative are connected directly to the sending unit (and are isolated from vehicle ground). There are no dc-dc's outside of what may be internal to the gauge

Sounds like the pack is being grounded to vehicle ground through the remote sending unit. In the PDF they say that "If you need galvanic isolation between pack negative and your 12V supply ground, then you must use optional isolated DC/DC converter as shown on the wiring diagram."
If you have a spare 12v battery of some sort to power the remote sender independently of the vehicle 12v system it would prove if the isolation or lack of is causing the issue.

Maybe also put a meter between the pot box ground wire and its casing to see if they are internally connected. I would have expected there to be no connection between the hall sensor and the casing but you never know till you check.