1995 Ford E-150

[kb1gtt]

In the picture of the setup, I see you are using input 5, 6, 7, 8 and 11, but I only see notes about TPS on input 5, IAT on input 6 and CLT on input 7. What is on inputs 8 and 11? Do I recall 11 was hall?

Can you set TPS to about 3/4 scale, then set IAT to about 1/2 scale then tell me the voltages of the R263/C260 node as well as the TPS and IAT input voltages at the screw terminals and the 5V voltage measured at C251. I suspect that the power voltage may be dropping due to loading. What is you power supply?

[E4ODnut]

It's been an interesting morning.

I started by seeing if I could find out what happened to my CLT signal. Turns out there was an open between U202 pins 1 and 2 and R274. I either damaged the chip or pulled a pad off the board, most probably the latter. So then I thought I'd better check for continuity on all connections.

Inputs 1, 2, 3, 4 and 7 are not available. Having said that, I may be able to restore 3 and 4 with a bit of luck.

These are the connections I have now and outputs respond to the pots.
TPS - Input5 - Output 5 - PA4 (R250 pulled)
IAT - Input 6 - Output 6 - PA3 (R261 pulled)
CLT - Input 8 - Output 8 - PA1 (R281 pulled)
MAP - Input 9 - Output 9 - PA0 (not connected, no resistors pulled)
PIP - Input 11 - W211 - PC6 (R310 pulled)

Still available for EGO or whatever (no resistors pulled yet):
Input 10 - Output 10 - PC3
Input 12 - W212 or W212 to Out 12 - PC1

5 volts supply comes from my power supply and is typically between 5.04 and 5.06 volts. Depending on the position of the pots the most it will change is 0.01 volt.

Testing so far is with Discovery unplugged from the analog board.

Perhaps what we should do now Andrey is to make the firmware changes to reflect the channels I am now using and see what the console has to say with Discovery plugged in?

[AndreyB]

I've updated the code, please program current binary

Code: Select all

	engineConfiguration->tpsAdcChannel = 4; // input channel 5 is PA4, that's ADC4
	engineConfiguration->iatAdcChannel = 3; // input channel 6 is PA3, that's ADC3
	engineConfiguration->cltAdcChannel = 1; // input channel 8 is PA1, that's ADC1

[E4ODnut]

I was able to repair channels 1 and 3 (I think) so they are now available to us as well. I'll get right on the firmware update.

[kb1gtt]

What are you using for MAP? You will most likely need to remove the pull up and down, probably replace the pull down with a cap. Is this the typical MPX4100 sensor?

[AndreyB]

I suggest we focus on TPS, IAT & CLT and only get to MAP once these three seem to be working

[E4ODnut]

MAP is a MPX4250 but it's not connected to the analog board at this time. It's a 2 bar sensor with full output ~ 4.5 volts if I recall. At atmospheric it is generating about 2.2 volts.

More interesting stuff.

RPM still responds to changes in 555 timer settings. CLT gauge responds. TPS and IAT gauges do not.

TPS - PA4 range 0.156 volts to 0.289 volts

With TPS at full output:
IAT - PA3 range 0.175 volts to 0.242 volts
CLT - PA1 range 0.205 volts to 0.306 volts

With TPS at minimum output:
IAT - PA3 range 0.060 volts to 0.123 volts
CLT - PA1 range 0.075 volts to 0.146 volts

[AndreyB]

Wait a second. According to the schematic http://i.imgur.com/cnEZuAX.png TPS should output between 0 and 5v. Does it? If you disconnect TPS from the analog board?

For 0 to 5v input range on input 5, I am expecting to see 0 to 2.5v on PA4.

[E4ODnut]

Voltages to ground:
TPS Input 5 = 5.00v
R252 in = 5.00v
R252 out 1.678v
R253/C250 = 1.68v
R254 in = 0.583v
R254/255 = 0.290v
PA 4 = 0.290v

I see similar voltage drops across R262 andR282

[AndreyB]

This is not right. Please disconnect the analog board from everything else to simplify matters and just leave +5 VCC, GND and +5 TPS input

Expected behavior:
TPS Input 5 = 5.00v
R252 in = 5.00v
R252 out 5v
R253/C250 = 5v
R254 in = 5v
R254/255 = 2.5v
PA 4 = 2.5v

R251 should either not be there or it should be 500K. R250 should not be there.

[E4ODnut]

I'm looking at those Shottky diodes and wondering if the polarity is correct. If it's wrong then that might explain the voltage drop because it would be connecting to ground.

[E4ODnut]

R250 has been pulled. R251 is 500K.

[AndreyB]

Can you verify the diode bag says 497-2521-1-ND? That's if you still have the little bag.

Do you feel comfortable removing one of the diodes just to verify if they play a role? They are there just for protection.

[E4ODnut]

497-2521-1-ND was the part ordered and the part received. I had a look at the data sheet and the pin out is a bit confusing. The mark on the diode is "D95" which would indicate a single diode with the right pin no connection and the left pin reverse biased to ground. If that's the case then it's not our problem, but it's also not the right diode. I'll pull one and see.

[AndreyB]

Yep, that's what we have just realized - the BOM has a wrong part number, it's a single diode not a dual. But it should still work right, and I know I have tested Frankenstein with the same exact schematic same exact diode part number.

[E4ODnut]

I have to go for a while, will report back in a few hours.

[E4ODnut]

I pulled all the diodes except for D310 and D320, being a single diode they only would do half the job anyway.

No change. I'm still getting a large voltage drop across R253, R263 and R283. That would indicate there must be some load to cause the current to cause the voltage drop. I don't know how those Op Amps are supposed to work so I can't offer any suggestions.

[AndreyB]

Can you try making a close-up picture? If there is continuity everywhere where it should be, I can only think about the op-amp being upside down.

[kb1gtt]

Sorry about the diodes, I blundered that in the BOM. I didn't expect the removal to work, as the 10k resistor will limit the current at 5V to .5mA. The diodes at like 500mA were around 1V. So I was reasonably sure they OK. I would wager a guess that the caps have shorted for some reason. Can you remove C250 and see if that make is start playing nicer.

[kb1gtt]

It would appear we are chasing a 5kohm from the op-amp input to GND. I calculate that like this (5v-1.678v)/10kohm = .3mA passed by the 10k into that node. The op-amp is 1Mohm or greater, so we don't expect any significant current to flow there. So we have 1.678/.3mA = 5kohm.

I could see this happening with mild ESD events, and perhaps rapid temperature changes. I've seen this happen with a normal iron instead of a SMT iron tip. Basically the normal iron tip is fat and allows heat to flow rapidly. A SMT tip is long and narrow which slows down how fast heat can flow to the chip. Is you iron temperature controlled? Do you know how hot it's getting? If you can turn down the temperature, that helps prevent from heating the components to fast.

[E4ODnut]

None of the caps read a short.

[kb1gtt]

5k to GND probably won't register as a beep test or short test. It might also not register on the ohm reading for you meter, as the meter will typically only generate around 1V for the ohms tests. You probably get more leakage as you increase voltage across the cap. Only way to really test it is to remove the cap.

[E4ODnut]

C250 gone. No change

[kb1gtt]

Hmmm, I still don't suspect the op-amp, those are usually fairly good about thermal cycling and ESD. Can you put one meter lead into the input screw terminal, and see if you measure 10k at the op-amp input? I wonder if there might be a poor connection across either the 10k or the 100ohm.

Also can I get some details about the iron?

[E4ODnut]

I checked the circuit resistance before as part of my troubleshooting. With the TPS input disconnected I show 10.14k from Input 5 terminal to U202 pin 12. The schematic shows I should have 10.1K so that is well within tolerance. I show 0.509 Meg from pin 12 to ground.

I'm using a Circuit Test SX-500 station with a fine tip that measures ~ 1/32" at the end. Temperature is set to ~ 400C.

[kb1gtt]

Looks like a good station. You might have it set a bit high, according to this fellow they claim 315C for SMT and 370C for thru hole http://electronics.stackexchange.com/questions/1980/whats-the-proper-soldering-iron-temperature-for-standard-031-60-40-solder

I seem to recall circuits 1-4 are toast. Do you see the same issue there? It's getting hard to claim it's not the op-amp. Can you preform some tests on that area of the board?

[E4ODnut]

I did a major test. The results are not good. Only Inputs 3, 10, 11 and 12 survived the transplant. It's enough to play around with a bit but that's all.

Andrey, where would you like me to distribute things so you can change the firm ware, again?

Jared, if the BOM is up to date?

If so I'll put together another Digikey order to build another board, but this time I'll only populate what I need, and hopefully in the right orientation.

[AndreyB]

Fresh firmware has

Code: Select all

	engineConfiguration->tpsAdcChannel = 7; // input channel 3 is PA7, that's ADC7
	engineConfiguration->cltAdcChannel = 13; // input channel 10 is PC3, that's ADC13

there is also a new command 'analoginfo' which might help a bit.

[E4ODnut]

Oops, I screwed up. Input 10 did not survive, It should have read Input 9.

So, TPS will now be on Input 3.
CLT will be on Input 9???.
I'm assuming we'll leave PIP on Input 11.
That leaves us IAT for Input 12 connected to what pin on Discovery????

If this works for you then I assume I'll pull R230, R201 and R321. Is that correct?

One more question. I had a hellova time trying to clean off that tack flux for the SMDs. Has anyone got a recommendation for anything better than alcohol (not the drinking kind)?

[AndreyB]

I use http://www.amazon.com/MG-Chemicals-Heavy-Remover-Aerosol/dp/B005DNQWV0

You would need to jump two pads of W212 - this case inp 12 is PC1.

TPS does not need pull-up and both thermistors need only a pull-up.