Home  |  Forums  |  914 Info  |  Blogs
 
914World.com - The fastest growing online 914 community!
 
Porsche, and the Porsche crest are registered trademarks of Dr. Ing. h.c. F. Porsche AG. This site is not affiliated with Porsche in any way.
Its only purpose is to provide an online forum for car enthusiasts. All other trademarks are property of their respective owners.
 

Welcome Guest ( Log In | Register )

2 Pages V  1 2 >  
Reply to this topicStart new topic
> D-Jetronic: Using My New CHT Rich / Lean Test Box, And it shows pretty much what I though I knew was wrong...
pbanders
post Nov 4 2016, 06:58 PM
Post #1


Senior Member
***

Group: Members
Posts: 939
Joined: 11-June 03
From: Phoenix, AZ
Member No.: 805



Some of you have seen the recent posting about the CHT spacer that helps warm-up, and the discussion that followed. I proposed making a test box that would use the CHT to adjust the mixture to either rich or lean while running to see what was really happening. The idea of inserting a pot into the CHT circuit is as old as the hills when it comes to D-Jet, but this is the first time I've enabled both rich and lean adjustment, plus put in a test point so I can monitor the resistance of the CHT as I drive around.

I just did my first run, there will be more, but I have a few interesting results and observations already.

First, there has been a lot of conjecture and theory about what's going on with the 914 D-Jet during warm-up and hot starting (where the car has sat for 5+ minutes after stopping). The conjecture for crappy warm-up was that the CHT heated up to quickly, leaning out the mixture and causing crummy idle. Hence, the spacer, a factory (VW) solution. Space the CHT above the head, slow the thermal conduction, and fool the ECU into thinking the engine is cooler and needs more fuel. Turns out it works pretty well, but doesn't fix the hot start problem.

I conjectured that the spacer made the hot start problem worse, because the sensor cooled off faster than the head, and made the mixture too rich. I came up with a way to start the car (key off , 1/4 throttle open, key on, start) that seemed to verify this, as more air with more fuel worked.

But, as we learn in science, experiment trumps theory. So, I went out with the box today, here's what I found.

Warm-up with the spacer proceeds well. Resistance of the CHT starts at about 2200 ohms and drops to around 400 ohms in the first 5-10 minutes of driving (sort of a cool day here in Phoenix). The fun begins when you turn the key off. According to my theory, the CHT sitting away from the head cools off, and the resistance increases, causing a lean hot start condition.

Not so, not at all. Key off the resistance of the CHT plummets, dropping from 400 ohms down to 100 ohms in just a few minutes. Why? Duh, there's no cooling air, the fact that it's sitting away from the head is irrelevant. When you try to restart, the CHT thinks the car is way hotter, and wants a leaner mixture, and it won't start for crap. The core of the engine is at nearly the same temperature it was before you turned it off after 5 minutes, it wants a much richer mixture.

Enter the "rich" potentiometer knob. I can dial in exactly the same resistance I saw when I went key off. When I do that, the car starts right up. So, no, it's not too rich, it's too lean.

I also played with the knob in a number of other idling situations. It was very easy to make a quick adjustment when the idle bogged from any loads (e.g. fresh air fan, heater blower, lights) to richen the mixture slightly, and the idle stabilized very well.

I have to stop the car and turn the motor off to measure the CHT resistance, but I can monitor the voltage on the CHT continuously, and compare that to the SPICE model I did for the circuit years ago. I'll do that in the future.

Oh, and I played with using the lean adjustment, but it was never needed. The 914 always wanted more fuel, more, more, more!

Right now, I'd have to say that adding a 500 ohm series pot to the CHT that you can adjust from the driver's seat is pretty awesome. I know others have done this before and said the same thing, I'm saying, "me, too".

I'll be experimenting with this setup through the week, will report more on what I find.
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
BeatNavy
post Nov 4 2016, 07:17 PM
Post #2


Certified Professional Scapegoat
****

Group: Members
Posts: 2,922
Joined: 26-February 14
From: Easton, MD
Member No.: 17,042
Region Association: MidAtlantic Region



Good info, Brad! And it makes perfect sense that the mixture is too lean. My AFR readings after warm-startup were way above 14, and I was thinking "is the low idle really caused from a too-rich condition?"

So based on your findings, it is pretty straightforward to add resistance during this condition.... (IMG:style_emoticons/default/beer.gif)
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
914four
post Nov 4 2016, 07:29 PM
Post #3


Senior Member
***

Group: Members
Posts: 510
Joined: 4-March 07
From: Rainbow City, Alabama
Member No.: 7,582
Region Association: South East States



This helps me to understand that even a perfectly tuned D-Jet system does not behave the same at all times. Maybe I should not be so critical of the small variances I see.

Thanks for the insight into how the CHT affects hot start condition.

Kelvin
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
JamesM
post Nov 4 2016, 08:45 PM
Post #4


Senior Member
***

Group: Members
Posts: 1,875
Joined: 6-April 06
From: Kearns, UT
Member No.: 5,834
Region Association: Intermountain Region



QUOTE(pbanders @ Nov 4 2016, 04:58 PM) *


Not so, not at all. Key off the resistance of the CHT plummets, dropping from 400 ohms down to 100 ohms in just a few minutes. Why? Duh, there's no cooling air, the fact that it's sitting away from the head is irrelevant. When you try to restart, the CHT thinks the car is way hotter, and wants a leaner mixture, and it won't start for crap. The core of the engine is at nearly the same temperature it was before you turned it off after 5 minutes, it wants a much richer mixture.




I made this same discovery while tuning a Megasquirt setup running stock d-jet sensors. Key off and both sensors (CHT and IAT) will badly heatsoak. The CHT causes less of an issue with Megasquirt as is only affects the cold start map and being a digital system has a fixed endpoint to its effect however the IAT heat soaking causes a very noticeable leaning of the running mixture on hot restart with Megasquirt and continues to do so until enough air has passed the sensor to cool it down to the actual air temp. Given I was seeing hot restart IAT readings above 160 deg on a 70 deg day and the IAT is a closed element sensor in direct contact with a metal manifold, this cooling of the sensor can take quite a while.

If I recall you (or someone) determined that the IAT has a smaller effect on mixture in d-jet systems but it might be interesting to see if the IAT is also contributing to the hot restart issues there.
User is online!Profile CardPM
Go to the top of the page
+Quote Post
pbanders
post Nov 4 2016, 08:52 PM
Post #5


Senior Member
***

Group: Members
Posts: 939
Joined: 11-June 03
From: Phoenix, AZ
Member No.: 805



QUOTE(JamesM @ Nov 4 2016, 07:45 PM) *

QUOTE(pbanders @ Nov 4 2016, 04:58 PM) *


Not so, not at all. Key off the resistance of the CHT plummets, dropping from 400 ohms down to 100 ohms in just a few minutes. Why? Duh, there's no cooling air, the fact that it's sitting away from the head is irrelevant. When you try to restart, the CHT thinks the car is way hotter, and wants a leaner mixture, and it won't start for crap. The core of the engine is at nearly the same temperature it was before you turned it off after 5 minutes, it wants a much richer mixture.




I made this same discovery while tuning a Megasquirt setup running stock d-jet sensors. Key off and both sensors (CHT and IAT) will badly heatsoak. The CHT causes less of an issue with Megasquirt as is only affects the cold start map and being a digital system has a fixed endpoint to its effect however the IAT heat soaking causes a very noticeable leaning of the running mixture on hot restart with Megasquirt and continues to do so until enough air has passed the sensor to cool it down to the actual air temp. Given I was seeing hot restart IAT readings above 160 deg on a 70 deg day and the IAT is a closed element sensor in direct contact with a metal manifold, this cooling of the sensor can take quite a while.

If I recall you (or someone) determined that the IAT has a smaller effect on mixture in d-jet systems but it might be interesting to see if the IAT is also contributing to the hot restart issues there.


It's second-order but that doesn't mean it doesn't have a significant effect. Good info from you on the heatsoaking of both sensors, all I can say in my defense is, "duh" (IMG:style_emoticons/default/smile.gif).
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
pbanders
post Nov 4 2016, 08:53 PM
Post #6


Senior Member
***

Group: Members
Posts: 939
Joined: 11-June 03
From: Phoenix, AZ
Member No.: 805



QUOTE(914four @ Nov 4 2016, 06:29 PM) *

This helps me to understand that even a perfectly tuned D-Jet system does not behave the same at all times. Maybe I should not be so critical of the small variances I see.

Thanks for the insight into how the CHT affects hot start condition.

Kelvin


I have to say, I was amazed at how easily I could trim the idle with my rich adjustment, and for just about any condition. When I get done with the experimental side of this, I'm definitely adding a knob I can adjust from the drivers seat. Like you say, it's the trimmer we never had.
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
pbanders
post Nov 4 2016, 09:21 PM
Post #7


Senior Member
***

Group: Members
Posts: 939
Joined: 11-June 03
From: Phoenix, AZ
Member No.: 805



OK, a couple of pics of the setup. Right now, the cable from the engine compartment is just going through the side gap in the engine cover and coming over the roll bar.

Attached Image

This is the box, you can see it has two pots, one for rich and one for lean. The center switch chooses which function is active, and the switch on the lower left isolates the combination of the box and the CHT so that you can read the resultant resistance. When it's in "Run" mode, you can measure the voltage at the node. Connections from the engine compartment are the CHT (white), the wiring harness connection to the ECU for the CHT (red), and ground (brown, taken off the mount for the vacuum limiter and MPS). The red banana lead on the bottom test point, which is connected to the DMM shown in the next pic (black banana lead is DMM ground).

Attached Image
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
cary
post Nov 5 2016, 07:28 AM
Post #8


Advanced Member
****

Group: Members
Posts: 3,900
Joined: 26-January 04
From: Sherwood Oregon
Member No.: 1,608
Region Association: Pacific Northwest



QUOTE(pbanders @ Nov 4 2016, 07:53 PM) *

QUOTE(914four @ Nov 4 2016, 06:29 PM) *

This helps me to understand that even a perfectly tuned D-Jet system does not behave the same at all times. Maybe I should not be so critical of the small variances I see.

Thanks for the insight into how the CHT affects hot start condition.

Kelvin


I have to say, I was amazed at how easily I could trim the idle with my rich adjustment, and for just about any condition. When I get done with the experimental side of this, I'm definitely adding a knob I can adjust from the drivers seat. Like you say, it's the trimmer we never had.


Thanks for all your research.

I think I'll run the pot wires to the front when I finally install an AFR gauge to replace the clock. I'm thinking I'll put the pot right beside the AFR gauge.
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
Java2570
post Nov 5 2016, 07:41 AM
Post #9


Senior Member
***

Group: Members
Posts: 649
Joined: 7-May 11
From: Fishers, IN
Member No.: 13,035
Region Association: Upper MidWest



Great info, as always, Brad! Glad to see you posting here....your research has really helped out a lot. Perhaps you can provide a parts listing and layout of your final box....once you get all the testing done? I've played around with a pot on my CHT but it's pain to make adjustments when it's in the engine compartment. (IMG:style_emoticons/default/beerchug.gif)
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
jasons
post Nov 5 2016, 08:19 AM
Post #10


Jackstand Extraordinaire
****

Group: Members
Posts: 2,002
Joined: 19-August 04
From: Scottsdale, AZ
Member No.: 2,573
Region Association: None



Let me know if you need another data point in town. I have a well functioning 2.0 djet (thanks to your sites help) and I have a variable resistor on the CHT sensor. I'm not sure how it's mixed but my car idles rock solid at 1k RPM after warm up. During warm up it idles around 1.5k and sometimes goes through an RPM hunting cycle. But after it's warmed up, it idles and runs fantastic. Without the resistor, it hunted between low idle and 1k.
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
TX914
post Nov 5 2016, 08:36 AM
Post #11


Alan-B
**

Group: Members
Posts: 167
Joined: 27-July 14
From: USA
Member No.: 17,689
Region Association: None



That is totally excellent. Thanks for sharing your results. Couple of questions (if I had electrical skills marginally better than a chimp I could find out myself):

• You surmised that when engine is stopped the core of the engine remains about the same (temp) while the CHT sensor becomes hotter. Is this because heads are aluminum and soak in heat build-up?
• Any idea how long it takes on restart for the CHT sensor to rid itself of heat build-up (i.e. return to normal)? Since you can measure voltage while running, is it possible to infer from your CTC voltage vs. resistance graph how long it takes?
• I suppose ideally compensation would be removed gradually as CHT sensor returns to normal, although I don’t suppose a rich condition for too long would be harmful.
• Given compensation will be manual, perhaps a light on the box would useful as a reminder to revert to normal operation.
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
GregAmy
post Nov 5 2016, 09:58 AM
Post #12


Advanced Member
****

Group: Members
Posts: 2,262
Joined: 22-February 13
From: Middletown CT
Member No.: 15,565
Region Association: North East States



Install a timed latching relay that automatically puts the resistor in series when powering the starter solenoid? Additional resistance shouldn't be a problem when the car is stone cold.
User is online!Profile CardPM
Go to the top of the page
+Quote Post
pbanders
post Nov 5 2016, 03:22 PM
Post #13


Senior Member
***

Group: Members
Posts: 939
Joined: 11-June 03
From: Phoenix, AZ
Member No.: 805



QUOTE(Java2570 @ Nov 5 2016, 06:41 AM) *

Great info, as always, Brad! Glad to see you posting here....your research has really helped out a lot. Perhaps you can provide a parts listing and layout of your final box....once you get all the testing done? I've played around with a pot on my CHT but it's pain to make adjustments when it's in the engine compartment. (IMG:style_emoticons/default/beerchug.gif)


That schematic and parts listing is in the other topic about the CHT spacer.
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
pbanders
post Nov 5 2016, 03:29 PM
Post #14


Senior Member
***

Group: Members
Posts: 939
Joined: 11-June 03
From: Phoenix, AZ
Member No.: 805



QUOTE(TX914 @ Nov 5 2016, 07:36 AM) *

That is totally excellent. Thanks for sharing your results. Couple of questions (if I had electrical skills marginally better than a chimp I could find out myself):

• You surmised that when engine is stopped the core of the engine remains about the same (temp) while the CHT sensor becomes hotter. Is this because heads are aluminum and soak in heat build-up?
• Any idea how long it takes on restart for the CHT sensor to rid itself of heat build-up (i.e. return to normal)? Since you can measure voltage while running, is it possible to infer from your CTC voltage vs. resistance graph how long it takes?
• I suppose ideally compensation would be removed gradually as CHT sensor returns to normal, although I don’t suppose a rich condition for too long would be harmful.
• Given compensation will be manual, perhaps a light on the box would useful as a reminder to revert to normal operation.

HTH,
Alan


The heat soak happens because the cooling air supply is cut off to the sensor. When you're driving at a constant load for a long enough time, the sensor equilibrates at some temperature determined by the thermal conductivity of the head, the radiation from the head, and the temperature and flow rate of the engine cooling air. When you shut the car off, the sensor does NOT, as I stupidly thought, suddenly cool off more quickly, but instead it continues to heat up because of the lack of cooling air.

I'm hoping to characterize from cold to hot to cold how long this process takes. It's a lot slower than you'd think. If I can infer it from the CTC voltage I will.

The additional compensation provided by using a pot only is required briefly, under a minute from what I could tell. Could an automatic system be designed? Most likely, but it won't be me who is doing it, I'm not good enough at circuit design.
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
stugray
post Nov 5 2016, 04:14 PM
Post #15


Advanced Member
****

Group: Members
Posts: 3,824
Joined: 17-September 09
From: Longmont, CO
Member No.: 10,819
Region Association: None



QUOTE(pbanders @ Nov 5 2016, 03:29 PM) *

The heat soak happens because the cooling air supply is cut off to the sensor. When you're driving at a constant load for a long enough time, the sensor equilibrates at some temperature determined by the thermal conductivity of the head, the radiation from the head, and the temperature and flow rate of the engine cooling air. When you shut the car off, the sensor does NOT, as I stupidly thought, suddenly cool off more quickly, but instead it continues to heat up because of the lack of cooling air.


(IMG:style_emoticons/default/agree.gif)
In fact I can see in my datalogger exactly when I shut the engine off because the CHTs spike while the EGTs drop.
I could even look to tell you how quickly it begins to drop again, but I rarely left the DL on for long after shutting down.

QUOTE(pbanders @ Nov 5 2016, 03:29 PM) *

I'm hoping to characterize from cold to hot to cold how long this process takes. It's a lot slower than you'd think. If I can infer it from the CTC voltage I will.

The additional compensation provided by using a pot only is required briefly, under a minute from what I could tell. Could an automatic system be designed? Most likely, but it won't be me who is doing it, I'm not good enough at circuit design.


There are commercial systems to "fool your ecu" even on cars as modern as my BRZ.
One that comes to mind is "unichip".
They just give the ECU altered data from MAP, MAF, IAT, AFR, etc to fool the ECU into making more power (operating outside predefined parameters).

My datalogger in my race car is an Arduino based system monitor that measures multiple temperature and voltage inputs (EGTs, CHTs, AFR, RPM, Throt Pos, Oil Press, Oil Temp).

To measure just 1X Cylinder Head temp (CHT) & 1X Intake Air temp (IAT), I could program a chip the size of your thumbnail to control a variable potentiometer based on a table lookup.
(IMG:http://www.914world.com/bbs2/uploads_offsite/www.jameco.com-10819-1478384056.1.jpg)
You could update the tables over usb to tune it.
To do it right would require 2X miniature thermocouple amplifier boards @$15ea

It could read the two temps CHT & IAT and decide what temp to present to the ECU as a variable resistance output.
If it could also read MAP, AFR & RPM, it could adjust the rich/lean bias in the ECU by moving that CHT input to the stock circuit.

I can plug my cellphone into imy DL and read the temps all in real time:
(IMG:http://www.914world.com/bbs2/uploads_offsite/i366.photobucket.com-10819-1478384056.2.jpg)
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
pbanders
post Nov 5 2016, 06:24 PM
Post #16


Senior Member
***

Group: Members
Posts: 939
Joined: 11-June 03
From: Phoenix, AZ
Member No.: 805



Stu, very nice, thanks for posting that, I'll look into it.
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
alexkirkham
post Jun 14 2018, 10:09 AM
Post #17


Newbie
*

Group: Members
Posts: 17
Joined: 23-April 18
From: London, UK
Member No.: 22,066
Region Association: None



wouldn't it just make sense to connect a time delay relay like:

https://www.12voltplanet.co.uk/adjustable-d...ff-12v-10a.html

across say a 300 ohm in series resistor with a 2min (or so) delay, so that:

on cold start: extra 300 ohms for 2 mins: makes no difference to the already high resistance

on hot start: significantly enriches the mixture for the start

and then when the relay switches after 2 mins it just bypasses the resistance and the mixture goes back to normal

I'm going to try that!

Alex
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
Chris914n6
post Jun 14 2018, 01:08 PM
Post #18


Jackstands are my life.
****

Group: Members
Posts: 3,287
Joined: 14-March 03
From: Las Vegas, NV
Member No.: 431
Region Association: Southwest Region



Would it make sense to remove the sensor from the air stream, then adjust the resistance to function correctly as it reads actual engine temp?

Seems this is only a problem because VW tried to adapt Bosch Djet to a non-water cooled engine.
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
troth
post Jun 14 2018, 02:17 PM
Post #19


Member
**

Group: Members
Posts: 150
Joined: 17-August 16
From: CT/WA
Member No.: 20,305
Region Association: None



QUOTE(Chris914n6 @ Jun 14 2018, 03:08 PM) *


Seems this is only a problem because VW tried to adapt Bosch Djet to a non-water cooled engine.


For what it's worth:
Something very similar happened on my water cooled go kart engine. After a race you would pull into the pits and shut down. The water temp read out would immediately start climbing as much as 20 degrees. Did not have an effect on performance since it was carbed, but would certainly give your heart rate a bump when the warning light came on.

So I don't think it's an issue of water vs. air cooled.
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
Dave_Darling
post Jun 14 2018, 03:34 PM
Post #20


914 Idiot
**********

Group: Members
Posts: 14,974
Joined: 9-January 03
From: Silicon Valley / Kailua-Kona
Member No.: 121
Region Association: Northern California



It's a well-known phenomenon called "heat soak".

--DD
User is offlineProfile CardPM
Go to the top of the page
+Quote Post

2 Pages V  1 2 >
Reply to this topicStart new topic
1 User(s) are reading this topic (1 Guests and 0 Anonymous Users)
0 Members:

 



- Lo-Fi Version Time is now: 29th March 2024 - 09:25 AM