DIY PSU Build for Flashing

[aboulfad]

Originally Posted by terahertz

...Pin 9 for the 12v sense is usually used so the PSU can adjust to the 12v dropping from variable load and it will increase the voltage automatically. We are hijacking that to make it raise the voltage to 13.5, so it doesn't have that same automatic voltage adjustment anymore. Either way, as long as your voltage stays steady in the 13-14v range, you should be good to go.

Ok flashed three ECUs this morning, I don't find the 600PB operation very reliable: the voltage was hovering around 13.35V during flash with low beam/DRL on, fan and radio on. The fact that the voltage drops and stabilize for every new additional load factor makes this PSU behaviour unpredictable under higher load.

If the sense pin 9 is defeated then the PSU has no way to adjust automatically the voltage, then it will be a bit risky to use as a reliable power source and we are winking it... My bro's clamp meter was unreliable, so we couldn't verify the current draw from the PSU.

Also, I had one or both PSUs shutdown in their own, the green lights went off, don't know why and if there's an issue w my wiring or if they encountered an over voltage. My setting was 13.7V with no load.

[terahertz]

Originally Posted by aboulfad

Ok flashed three ECUs this morning, I don't find the 600PB operation very reliable: the voltage was hovering around 13.35V during flash with low beam/DRL on, fan and radio on. The fact that the voltage drops and stabilize for every new additional load factor makes this PSU behaviour unpredictable under higher load.

If the sense pin 9 is defeated then the PSU has no way to adjust automatically the voltage, then it will be a bit risky to use as a reliable power source and we are winking it... My bro's clamp meter was unreliable, so we couldn't verify the current draw from the PSU.

Also, I had one or both PSUs shutdown in their own, the green lights went off, don't know why and if there's an issue w my wiring or if they encountered an over voltage. My setting was 13.7V with no load.

Voltage changing under load is pretty normal. Your voltages were still very safe. Higher than mine.

Shutting off is really bad though. Mine never shut off, and mine were steady at 13.30v once it reached equilibrium (the battery started off at 13.40v and discharged until 13.30v).

I did not keep changing the load like you did, but it was rock solid the whole time. It was left running for over an hour after the flash ended because I was helping someone do an oil change on another car. I'd peek at the multimeter and it was 13.30v every time.

Maybe 13.7v is a little high? Mine was 13.5v.

[aboulfad]

Yeah that's what I was thinking too, I read somewhere that the OVP is set to 13.65 and conflicting with other sites mentioning 13.8, so I dropped the unloaded voltage to 13.6 for my next flashing session (GHAS, VDC, EGS). Now shopping for a good clamp multimeter.

[HDDMark]

Originally Posted by aboulfad

Yeah that's what I was thinking too, I read somewhere that the OVP is set to 13.65 and conflicting with other sites mentioning 13.8, so I dropped the unloaded voltage to 13.6 for my next flashing session (GHAS, VDC, EGS). Now shopping for a good clamp multimeter.

I got my PSU today and the remainder of the supplies should be here Wednesday. Based on your idea of using a digital multimeter to monitor the voltage and amp draw, I ordered one too.

I'm interested to hear of your driving impressions after flashing the DME, DSC and ICM. Also interested if the updates to GHAS and VDC provide much benefit.

Your progress in doing this is making me feel more comfortable about doing it too.

Mark

[aboulfad]

Originally Posted by HDDMark

...
I'm interested to hear of your driving impressions after flashing the DME, DSC and ICM. Also interested if the updates to GHAS and VDC provide much benefit.

Your progress in doing this is making me feel more comfortable about doing it too.

Mark

Feel free to ask or PM me if you need help. If you do your homework you will be fine. Driving impressions in few days. My update strategy is to update important ECUs only, mostly related to engine, drivetrain, suspension...

[HDDMark]

So I finally got my act together and finished building my PSU thanks to the input from that other thread and from you guys.

I ended up getting a digital multimeter to monitor the voltage and amp draw similar to what aboulfad linked to earlier. What took me the longest was figuring out how I was going to wire up the multimeter and conceal all of the wires. A buddy at work said he could 3D print a custom enclosure for me but he's out of town for a few weeks so I had to come up with another solution. I ended up going to Lowes to see what I could cobble together.

I ended up using a 4x4x4 plastic electrical junction box with a blank lid. I actually attached the plastic junction box to the end of the PSU with the terminals. Right now I'm just using zip ties which is lame but it works for now. I just needed to cut holes in the side of the junction box for the terminals on the PSU and fan. I then cut vent holes in the other side of the junction box to allow air flow through it. I cut holes in the blank lid and mounted the multimeter and switch in it. I also drilled a small hole in the top of the lid and expoxied the potentiometer on the bottom side of the lid so I can adjust the voltage without taking the lid off. Fortunately, the shunt for the amp meter barely fits in the bottom of the junction box diagonally.

It's not pretty but it seems to work. I hooked it up to the car and turned on the ignition with the headlights, stereo and AC fan and it was drawing about 7.5 amps at around 13.4 volts.

Time to flash. One question though. Is there a way to monitor the battery voltage in E-Sys? I don't think I've ever seen it but that would be cool.

Thanks,

Mark

[terahertz]

Wow. Great set up!

I haven't heard of a way to monitor voltage from esys. I believe there's a secret menu on the cluster to view battery voltage but I'm guessing it would go away if the kombi was flashed and subsequently rebooted.

[aboulfad]

HDDMark , Cool so you set up one PSU, which is fine too. BTW, I hope those thin green/red wires aren't going to be used for flashing... (Oh they are for measuring the load). I am still waiting for my display to arrive how come you got it so fast? no voltage in Esys but ISTA/D/P via ICOM does.

And BTW, when you flash, turn on your low beams, from user manual, the ignition shouldn't go off, just additional precautions! Please post back in my thread after you flash with any comments, it'd be nice so I can update main post.

[HDDMark]

Originally Posted by aboulfad

HDDMark , Cool so you set up one PSU, which is fine too. BTW, I hope those thin green/red wires aren't going to be used for flashing... (Oh they are for measuring the load). I am still waiting for my display to arrive how come you got it so fast? no voltage in Esys but ISTA/D/P via ICOM does.

And BTW, when you flash, turn on your low beams, from user manual, the ignition shouldn't go off, just additional precautions! Please post back in my thread after you flash with any comments, it'd be nice so I can update main post.

Thanks guys. What gauge wire are you using from the PSU to the car? The red and green wires are the two wires for attaching to the car. I'm just using some 10 gauge stranded copper that I had. It should be good for ~30 amps which should be enough.

I got the multimeter from amazon.

http://www.amazon.com/DROK-Multimete...ilpage_o00_s00

[aboulfad]

I used 6ft of AWG10 from a battery booster cable, it's good enough for 25-50Amps for that length.

[aboulfad]

Originally Posted by HDDMark

Well I finally flashed the DME, DSC and ICM to 58.3...

My single homebrew PSU worked albeit during the initial part of the flash the water pump ran for about 4 minutes and the battery voltage dropped from about 13.28 volts down to about 13.12 volts which had me worried. After that stage the voltage rose back to about 13.16 and stayed there. The multimeter integrated with my PSU was showing a solid 13.6 volts and around 10 amps. I was under the impression that this thing should be putting out 20+ amps during a flash. I'm wondering if the voltage drop through my leads is causing more drain on the battery and the PSU to output less current that it should. Maybe time for some shorter and larger leads...

I am starting to wonder about his PSU, both i & terahertz have two wired up in parallel, so in theory they should put out close to 90Amps, let's say it's 50% efficient with our wiring and it's design, so that should leave close to 40Amps with AWG10 for less than 6 ft.

In theory the voltage should only drop slightly, now given that the modded PSU doesn't have the voltage sense pin, so what is really going on?!? As the demand for current increases due to any electrical operation (flash, ignition, ...) , I wonder if it's the battery that is supplying the current!!!

In theory, current will flow over the least resistant path, I.e. with the thickest cable (and battery cables are very thick), so what proves that this mod'ed PSU is actually supplying the current... And why is the voltage dropping, unless the cable quality is really crap...

HDDMark , did you measure the battery voltage before and after the flash?

[HDDMark]

If I remember correctly, the battery before flashing was about 13.28 volts at the terminals. It dropped down to a low of about 13.12 during the flash but started increasing about half way through (after the water pump stopped) to about 13.16 volts at the end. I left the PSU on the terminals for a while after finishing the flash and it was able to get the battery voltage back up to 13.32 volts so it is capable of charging the battery just not seemingly at a sufficient current during the flash. Meanwhile the voltage on my integrated multimeter was rock solid at 13.6 volts.

I think that maybe with this set up you may be correct about the battery being the path of least resistance for the current draw. I'm no electrical engineer but what about using our potentiometer to regulate the PSU output based on the voltage at the battery terminals instead of at the PSU ?

We'll get this figured out.

Originally Posted by aboulfad

In theory the voltage should only drop slightly, now given that the modded PSU doesn't have the voltage sense pin, so what is really going on?!? As the demand for current increases due to any electrical operation (flash, ignition, ...) , I wonder if it's the battery that is supplying the current!!!

In theory, current will flow over the least resistant path, I.e. with the thickest cable (and battery cables are very thick), so what proves that this mod'ed PSU is actually supplying the current... And why is the voltage dropping, unless the cable quality is really crap...

HDDMark , did you measure the battery voltage before and after the flash?

[aboulfad]

Hahaha, I am ashamed to say I am an EE but I used to hate circuits, my passion is for DSP and telecom I need to buy an DC amp meter. Btw your battery voltage was too low before flash hence why I put a tip that to flash w a full battery ~13.3-13.45V for our LifePO4.

Mine was 13.45 before flash then dropped to 13.3V during flash for the same trio ECUs. That tells me that this PSU is not useful if it's not able to maintain a stable voltage.

[terahertz]

I used 8 gauge jumper cables.

[aboulfad]

Ok I have read the basics of PSUs and the one we are using, and as terahertz eluded to earlier, here is what is going on:

1. As we are using the sense pins to increase the voltage, this PSU has no way of regulating voltage and becomes victim to the laws of physics.
2. Current and Voltage are inversely proportional, (ohms law), so as the energy demand increases during flashing, the voltage will inevitably drop with this PSU using our configuration. (Increase load from car and watch Voltage drop and then stabilize)
3. Due to 1 and 2, the rated energy output of this PSU will drop by a factor of (1-12/13.5) ~11%, not counting any losses.
4. Other sophisticated PSU (Schumacher Inc-700A) must use one of linear/switched regulated modes to maintain stable voltage without sense pins.

Recommendations:
- use short (< 3-4ft) thick (AWG8) battery cables
- set output voltage to 13.5 (accounting for a 0.1-0.2V drop) to match our LifePO4 working battery voltage, don't go higher.
- make sure starting battery voltage without load (ignition off) is above 13.3V (a PSU will not charge our battery, or very very slow) (funny, even the flash programming entry of the Inc-700A manual says the same)

Question: not knowing yet how much current could be drawn during flashing, how can we guarantee voltage to remain above 13.0V given "2" above?

Links:
http://www.teamwavelength.com/info/powersupply.php
https://en.m.wikipedia.org/wiki/Swit...e_power_supply
http://www.rcgroups.com/forums/showp...postcount=1720
http://www.rcgroups.com/forums/showp...&postcount=603

LifePO4 battery voltage state as per my Optimate lithium charger TM291:

[jwzimm]

Would it be possible to add in a proper voltage sense circuit at the end of the power output cables somehow? Since this PSU is able to adjust output voltage actively we should be able to use that to properly regulate it while in use.

[aboulfad]

Originally Posted by jwzimm

Would it be possible to add in a proper voltage sense circuit at the end of the power output cables somehow? Since this PSU is able to adjust output voltage actively we should be able to use that to properly regulate it while in use.

That's the whole purpose of the sense pins... and since we are using them to increase voltage (connection S- to +5V via a pot), the internal voltage regulator is blind. Read above links.

[jwzimm]

Originally Posted by aboulfad

That's the whole purpose of the sense pins... and since we are using them to increase voltage (connection S- to +5V via a pot), the internal voltage regulator is blind. Read above links.

I understand that. What I am asking is, is there a way to do both. If an active feedback circuit could be constructed that would take the output voltage at the jumper ends and feed that back into the sense circuit (with a Pot on it to increase the base setting) it would be able to actively regulate the voltage at the required level (assuming it doesn't have to go too high that it trips on overvoltage protection).

[aboulfad]

Not sure if that is possible, further modding will require knowledge on the internal design of that PSU. From a blackbox connection, either use the sense pins to provide feedback to internal voltage regulator or adjust voltage. If you find a circuit or a method, please do share !

[jwzimm]

Originally Posted by aboulfad

Not sure if that is possible, further modding will require knowledge on the internal design of that PSU. From a blackbox connection, either use the sense pins to provide feedback to internal voltage regulator or adjust voltage. If you find a circuit or a method, please do share !

Well it would seem to be pretty simple I would think.

All we are doing when we are connecting the sense pin to the +5V reference through a Pot is using that +5V reference to feed back a control signal to the sense pin. The Pot is added to allow for us to modify this reference to increase the output voltage of the supply by reducing the input voltage seen at the sense pin (using the Pot as a controllable voltage adjustment).

If we constructed a tiny 12V to 5V power supply to provide that +5V reference instead of using the power supply +5V pin we could feed that new supply with the output voltage taken at the jumper cable ends (or even at the output terminals assuming minimal voltage drop along the cables). Since the reference voltage would now vary linearly with the power supply output voltage the active feedback system would be restored. The converter should be unregulated such that the output voltage does vary though (most 12VDC to 5VCD converters are regulated). Perhaps something like this: http://www.rapidonline.com/mornsun-g...verters-560167 (P/N H0503S-2W or G1205S-2W).

[te37]

Have not read the whole thread but would having a running car hooked up via jumper cables accomplish the same thing? Or would it not be enough power?

[aboulfad]

I am not sure I follow many points with your idea. maybe a diagram could help explain it better ? For the new reference to change linearly with the PSU output voltage, that would mean this mini PSU should be unregulated ! and we still need to set the desired output voltage, how is that accomplished ?

Anyways, further readings at rcgroup also mention that the PSU capacity will drop significantly as the voltage is increased which may have been mitigated by the fact we are hooking two in parallel.