LCD Instrumentation for Cougar Controller

[ianbartie]

I'll do something like your battery suggestion or stick a voltage divider across the variable output of my supply. Have to take the kids to netball shortly so the car will have to wait a while.
The new pump is a VBS pump kit from YT Stabletech in Taiwan. A few places in the US sell them. I have not given it a test yet 'cause I still doing the new rubber mounts, new wiring & new hoses. Was a whole redesign job but I think it will be worth it.
I'll put some pics up on my web site this afternoon.
I found the Thomas pump to be very, very noisy & fairly slow in sucking down the vacuum. Would take 7.5 seconds initially & then run for 2.5 seconds every brake press.
These new pumps are 40dB and are really fast. They are a rotary vane type instead of a reciprocating diaphram type. They run for less than a second on a brake press & take just a couple of seconds for initial suck down.
They do draw 35A but when I worked out the watt/seconds that both the Thomas (8A) & the VBS drew, the VBS only draws 1.5 times the power of the Thomas because it is so fast.

[Freyguy]

Originally Posted by mora

Oh noes. New PICs give the same black boxes on the upper row of the screen. Current draw is within the limits this time though. Maybe something else fried during first testing. I don't have switches, leds or U8 installed though. I suppose switches should be momentary on, so would it matter if they are not installed anyway? How about leds?

Mora,
Install gren LED D7 and also momentary switch SW2. Pressing SW2 should toggle the LED on and off. If you see no LED activity, then your PIC may not be programmed or the oscillator may not be working. Do you have an o-scope?

-Chris

[ianbartie]

I made a voltage divider using a 1k resistor, a 220R resistor & a 470uF cap across the 220R & a 68 ohm resistor on the divider output.
Put it onto my power supply variable output & I was able to get down to .18V & get up to 3V reasonably accurately.
Hooked it up onto the sender wires, connected the DMM, turned the voltage right down, crossed my fingers & switched on the ignition.
Lo & behold, the low fuel warning light came on. Turning the voltage up made the gauge come up to full, turning it down & down it went -hehehe. Stupid car doesn't know there is no fuel tank any more.
Anyway, the voltages were lower than predicted with 0.18V at empty & 1.75V at full. This was measured with my DMM between the sender wire & ground.
It follows to reason that the temperature gauge works similarly. Shouldn't be that hard to get it to think it is a battery amp meter instead.

Here is a pic of the new vacuum pump.
Should have it fully installed tomorrow sometime.

[Freyguy]

Hey thats good news! What was the response like from when you adjusted the power supply until movement of the needle? I'm convinced there's a computer (micro) interface here that may ignore sudden changes. Not that big a deal for this fuel sender interface, but you'll want quick response on the temp gauge. Since temperatures change slowly, the micro may used a slow sample rate on this or large sample averaging. While you have the setup, you should connect to the temp wire and see what the gauge does.

I've seen that pump kit before. You'll notice on the left side of the pump a beige colored block. I believe this is the muffler which is the main reason for the 40dB level. The Thomas has none. I read a post the other day where it was recommended to make a muffler from 2ft of heater hose filled with pantyhose. Pantyhose woven into a metal coil made from coathanger inserted into said hose. Then attach to exhaust port. I might give that a go.

[ianbartie]

I'll do a couple of quick voltage changes in the morning (shuteye time now) & see what happens. I'll try both the fuel gauge & the temp gauge but I'll put the scope on the temp gauge first just to be sure.

Re pump. The block is the muffler which has 13cc of oil in it. I believe the oil is the lubricant for the vanes as well. Don't quite know how it all works just yet to keep it so quiet. Being a vane pump, they are inherently quieter than any reciprocating type pump.
The pantyhose in the rubber hose trick will certainly work. Just don't stuff it in too hard otherwise there will be too much back pressure on the pump & you won't get full vacuum. I did find some pneumatic mufflers that would reduce noise output by about 30dB. In either case the muffler or the stocking will clog up with crap & reduce the flow to the point it will need replacing.

[mora]

Originally Posted by Freyguy

Install gren LED D7 and also momentary switch SW2. Pressing SW2 should toggle the LED on and off. If you see no LED activity, then your PIC may not be programmed or the oscillator may not be working. Do you have an o-scope?

I installed both leds and nothing happens when pressing SW2. I can rig up crude scope using soundcard input.

Could somebody send me a precompiled .hex file? I'm not sure if I'm doing conversion from .asm to .hex right.

[mora]

Oh yes, now some text on the screen. "Resetting SOC" and there it stays. I exported raw 1.3a.asm file as .hex and it began "working". I think I have exported 1.3b before, but it was for 2480.

[mora]

Now it must be only software issue as I got things working once with version 1.2a software. All the info displayed etc. But: if I alter the file in any way, or simply save it without modifying anything, export it as .hex and burn I get only row of blocks on display.

I found quickbuild -button from MPLAB software. Versions 1.2a and 1.3a both fail on quickbuild but 1.3b and 1.2b go fine. If I burn 1.2b or 1.3b on chip I get stuck on "Resetting SOC" message. Guess this is because b-versions are for 2480 chip.

[ianbartie]

Hi Mora
Make sure you are using MPLAB 8.60 & no later. I found later versions will not do a successful build on a 2420 chip.
You also need to choose the right asm file for your chip. The 'a' version is for the 2420. The 'b' version will not work on a 2420 chip.
The procedure is fairly simple doing it all via MPLAB.
Start MPLAB, browse to & open the correct asm file for your chip.
Probably a good idea not to change anything initially until you are sure it is programming & working right then start changing constants.
Select 'Project' then 'Quickbuild'
It will build the file & tell you if it is successfull.
Plug your Pickit2 programmer onto the chipset.
I got a clone from eBay for $20. Works fine.
The programmer powers the chip during programming so turn the 12v display supply off.
Next, choose 'Programmer' then 'Pickit2'
It will show that it has found the programmer, found the chip & powered up the chip.
Select 'Programmer' then 'Program'
It will program the chip & tell you when it's finished.
Select 'Programmer' then 'Set Vdd Off'
Unplug the programming cable & turn on the 12v supply.
You should be good to go.
I found my desktop works fine for programming but my old laptop does not do successful builds using the same version MPLAB.

[Freyguy]

So after driving my EV to the market, I glance at the SOC value and see 84% before turning off ignition. 20 minutes later I’m back in the car, and when I turn it on, the SOC shows 151%! I cycle power again, but it comes back to 151%. Hmmmm. I drive home, and the SOC percentage drops like normal from the 151% value. This is the first time I encountered this, but I believe I know what is happening.
The updated capacity is written to EEPROM every 60 seconds. I think it was in the midst of an EEPROM write while I powered it off, thereby corrupting the stored capacity value. I thought about various ways to handle this, but the most bulletproof (I think) is to check battery amps before writing capacity to EEPROM. If battery amp draw is less than 2 amps (vehicle is stopped) skip the EEPROM write. I left a 1 amp buffer since my DC/DC converter draws about that if my vacuum pump is running after I stop.

I’ll add this to version 1.4 along with the EEPROM write immediately after Resetting SOC . Still want to hold off release until 1.3 is beat up a bit more to uncover additional bugs. Required code changes below:

Insert these 5 lines after line 400-
tstfsz bAmpH ; test for battery amps
bra Tempmov ; less than 2 amps
movlw .1 ; if so, then
cpfsgt bAmpL ; skip SOC adjust
goto Alarm ; and writing to EEPROM

Add label “Tempmov” to line 406 (was line 401)
Tempmov movff bAmpL, bAmptempL

Add label “Alarm” to line 448 (was line 443)
Alarm movlw ALRM_lim