Dip-stick heater test -- no go

<div class='quotetop'>QUOTE(bwilson4web @ Jan 19 2007, 11:29 AM) [snapback]377726[/snapback]</div>

So, you couldn't get your rod in the hole? I almost had the same problem with the block heater. But seriously, is there a reason for picking this over the block heater, or do you have both?

 

Oil tends to stick and form a glop at the end of the oil dip stick heater. Not recommended!

 

<div class='quotetop'>QUOTE(bwilson4web @ Jan 19 2007, 11:29 AM) [snapback]377726[/snapback]</div>

Those dip-stick oil heaters have been around for decades. You'd think that by now, someone would have been able to come up with a version that is able to accomodate the bends in the dip-stick tube.

 

<div class='quotetop'>QUOTE(bwilson4web @ Jan 19 2007, 01:15 PM) [snapback]377794[/snapback]</div>

Thanks, want to hear more about your thermistor hack. I've installed the block heater, figure I'm getting mabye 2MPG better overall during cold weather given my commute.

 

forget the thermistor hack as your wasting your time. Get another oil pan for the PSD and get a tube machined up and tig weld it into the pan and put another block heater in the PSD unit and get it warm. Block heater in the ICE left on for 3 1/2 hours at an OAT of 5C get's the ICE to 27-30C.
Leave it on over night and it get's the ICE to 42C maybe a bit higher. Put one in the PSD as it has the same mass and the same fluid capacity as the ICE and get it up to the high 30's low 40's Celcius. That will significatly improve your mileage by reducing the fluid viscosity drag on the drive unit. IMHO

 

if your down around freezing I doubt that it's the ICE temp that's forcing the ICE to run, I get the same thing and if you look at your Battery temps you'll see that they are probably in the 2-7C range and it's the battery temps that are forcing the ICE to run to build heat to warm the battery. Just bump your MiniScanner to read battery temps and you'll notice that when they get to about 12C your ICE "problem" is gone.

 

I was doing some timing on my first start of the morning but it came to a halt with a bicyclist hit by a car. I spent several minutes there so I'll try again tomorrow. In the afternoon I have a two mile run to a traffic light at 90KMH and if it's backed up no amount of ev will help. I normally shut it off or take the alternate route which is 7 miles to the first traffic light. I'll do some timing to see if the addition of the circuit will save me much.

 

Bob were into a warm spell again and today I clocked about 20 seconds where I could have forced the "ev" with a temp sender hack. First stop today I showed 72C on the miniscanner. I have some short climbs when leaving the house and then a flat run in traffic that should be going 30 mph but travels at 45 mph. Then I start an undulating climb of about 1000" in 47 blocks, the undulations is where I get HV battery heating with the draw down and recharge and by then I have warm air circulating thru the battery compartment and the temp by then is in the 12-16C degree range. Even after a weekend sitting the distance to that temp is pretty consistant. I'm currently doing 46MPG US on my commute, how does that compare to your milage?

 

yes the battery "boost" on first startup is significant. What I try and do is to give it a bit of quick acceleration to road speed and if no tailgaters I back off before the speed limit until I see less than 20 amps draw. I'm starting to think that the boost is on till the computer senses that the HC absorber cat is open to full flow and not recircing the exhaust gas thru the absorber. That signal I can only assume comes from the downstream O2 sensor or so it seems to me. I've tried watching the voltage but it's hard to decipher what is the point of the "boost cut off" but it seems to me to be about the time the HC absorber open. One day I'll tie in a vacuum gauge to the valve and watch the vacuum switch, and O2 voltage and the "boost".

 

Bob-

I assume that you're spoofing the coolant temp sensor reading when you're talking "thermistor hack"?? I've been doing this for more than a year now on my Classic.. I started with a manual 25K 15-turn pot in parallel across the coolant temp sensor leads, and manual monitoring with my Miniscanner, but now I've got a little PIC microcontroller and a bunch of programmable digital potentiometer chips (in parallel to minimize the wiper resistance) that automatically optimizes the temp reading for me.. In warm weather, it will ramp the temperature reading up to 70C rapidly, if the car has been run recently, it will move to 70C almost immediately, and in really cold weather (ie below 0C), it delays operation for a bit and allows the engine to warm up a bit to avoid MIL and codes.... Subject to the aforementioned conditions, the algorithm is rather simple- if the temperature is below 70C, it adjusts the potentiometer value so that the temperature reading to 72C (which I picked rather arbitrarily); once it rises above 72C, the potentiometer switches out of circuit and it goes into "monitoring mode", where it watches the temperature reading and switches in again at 70C and maintains the temperature reading at 72C until the engine warms up again and the cycle repeats..

With the temperature sensor spoofing in place, where I used to have nothing better than 9L/100km bars for the first 5 minutes (in most cases it was pegged at the top), but with the mod in place, I'm seldom ever worse than about 7.5L/100km, and on most days, get much better; even now in the winter, on most days, I'm still better than 9L/100km on the first 5 minutes.. The savings are actually even more apparent in the winter when combined with radiator blocking because you get engine stop on all but the very coldest days, and stealth under virtually all conditions..

 

I ended up using a digital potentiometer because I started out with a manual 15-turn pot. when experimenting; I had considered various approaches such as switching out the temp sensor completely and spoofing the voltage with either a DAC or some sort of switched capacitor PWM, but it always came down to complexity and ease of programming, especially as I've got all the stuff necessary for PIC microcontroller development, and these chips have built in I2C for communicating with the digital pots and ADCs for monitoring the sensor voltage.. Using a parallel connection also appealed because the default state of the pots is max. resistance, so if anything zapped, the actual temp sensor reading dominates and the vehicle functions normally. I never thought about the diode/resistor combo, but that's my fault for being so digital oriented..

The only technical hitch came in because most of these digital pots are meant to be in the gain stage of op amps and they have fairly significant "wiper resistance" such that just a few mA of current will result in a pretty large voltage drop.. It ended up not being a problem because the parts I got from Analog Devices (free samples even) were dual and quad devices that I could wire and control in parallel to minimize the wiper resistance.. The only variables then were the number of steps available in the pot and the max. resistance.. I'd have to double check, but I believe I ended up using a quad part with 200K max resistance and 1024 steps (ends up being 50K with 1024 steps when wired all in parallel).. Again, the free samples available from Analog Devices and a few other companies meant that I had a variety of components to experiment with at zero cost to me..

In terms of construction, the only real problem was that most of the parts came only in surface mount varieties- I ordered the largest packages I could (eg, the SOT type, which could still be manually soldered to, and which Digikey sells "adapter boards" for), but a number of the parts only came in the new tiny packages with the microscopic lead pitches- and these ones I convinced our PCB guy at work to add a few "break out adapters" on the scrap corners of some of the stuff he was working on (these are the "frames" of the PCB that are snapped off after everything is complete, typically we lay out test lines here to ensure that the etching is of the correct width but these portions are basically scrap) and mounted the chips using the hot-air reflow equipment at work..

I did a lot of experimenting to figure out how to optimize the insertion of the modifying resistance and how to vary it.. As you probably know from your own experimenting, the Prius programmers thought they'd be clever and do all sorts of consistency checks so inserting the resistance too early can result in MILs and codes, which wouldn't be nearly so bad if they didn't require you to switch off the vehicle and restart to get out of the "limp home mode".. I ended up writing code for all sorts of initial conditions based on thermistor readings- if the coolant is > 40C, the engine is still hot from a recent operation and it's virtually always safe to ramp up to 70C immediately; in between 25C and 40C, I wait 20 seconds, ramp up to 50C to get rid of the low temp light (it also seems to have some positive effects on available engine power), and then slowly ramp up to 70C. It's lower temperatures that cause some consternation because this is the situation where you have to watch how quickly you begin spoofing the readings.. If you're unlucky, you get an MIL right away; if you're really unlucky, the MIL doesn't come on until a bit later when you're already moving- I found one set of start up conditions that I thought was perfect- the car would consistently allow engine stop within 2 minutes startup, but the problem was that if you had to stop again before the engine had really had chance to warm up, the minute you hit the accelerator to go, it would pop the MIL and start moving ahead really slowly on battery only, which is rather unsafe.. I never managed to tweak this setting to fix this, so unfortunately, in most cases you have to delay the insertion of the pot and slow the ramp rates to gradually get the reading up..

There seems to be some sort of interaction between the hybrid system and the climate control system as well- obviously if you use the A/C or turn the heat way up, the engine will start, but I've discovered that you can fiddle with the climate control to force the engine to stop as well.. I can use the A/C and Max A/C buttons as an "engine stop" command on my car- as long as the temperature knob is set to something reasonable, and the fan isn't going at max, if I turn on the A/C and the Max A/C button, and then immediately switch the A/C off, if the outside temperature is warm, the engine will stop immediately; if it's colder, it may take a couple of A/C on/off cycles to get the message across..

I think the next project will be to figure out exactly what the climate control and hybrid control interactions are and potentially spoof these..