Automotive 6 Volt Generator Transistor Voltage Regulator

[rjenkinsgb]

Why not do a circuit board layout with the basic "core" parts you need for the control circuit on it, to build around?

Use EasyEDA / JLCPCB and select parts they have in stock and you can have a few boards with surface mount fitted, in a week or so. The cost of small board from them is peanuts.

They list the LT6700 as well as many 311 style comparators,

 

[KeepItSimpleStupid]

Because of the wierdness of using a hall effect module, I think it makes sense to use a commercial module used in GM charging systems.
e.g. http://www.techsmartparts.com/en/products/product-categories/battery-current-sensor/

Would be a nice and rugged module.

http://www.underhoodservice.com/state-of-charge-charging-systems/

The battery current sensor is a serviceable component that is connected to the battery cables. The battery current sensor is a three-wire hall effect current sensor. The battery current sensor monitors the battery current. It directly inputs to the BCM. It creates a 5V pulse width modulation (PWM) signal of 128 Hz with a duty cycle of 0-100%. Normal duty cycle is between 5 to 95%. Between 0-5% and 95-100% are for diagnostic purposes.
[unquote]

Another: **broken link removed**

I can't find any decent information about the sensor pinout. It's "probably" +5 from the BCM, ground and an open collector output.

Something like
PWM to DC (design): http://www.egr.msu.edu/classes/ece480/capstone/spring15/group10/Application Notes/Kyle.pdf

can hopefully convert the output to something useable for your design.

You need something really rugged.

Then there is these http://www.pololu.com/category/118/current-sensors guys, but they are likely not going to be reliab;e/

You can look at either the output of the alternator or the input/output of the battery.

Again, ideas.

Maybe a buck-boost converter, an ACS sensor and a undervoltage lockout looking at just alternator output. You can also create a charging indicator too or a real ammeter feed.

I'd hate to have to add a 2.5V to 15V buck boost regulator for %V out to your design.
http://www.analog.com/en/products/ltc3111.html#product-overview
Shutdown current is 1uA, so you could be OK there too.

Sorry, been pre-occupied with stuff like GPS stolen when car was in the shop and my credit card was used in Nigeria.

 

[KeepItSimpleStupid]

Found it:

https://www.brakepartsinc.com/dam/jcr:dd879eb0-b5a0-4eb3-862b-2cb49f40e18d/13

It's called
Remy Techical Service Bulletin November 2017

The BCM sends 5 volts and supplies a ground to the battery current sensor. The battery current
sensor is an amp clamp that measures current flow into and out of the battery. The current sensor
then translates flow into a variable duty cycle signal. The sensor then sends this variable duty
cycle to the BCM to indicate current flow. The BCM monitors this current flow and sets a B1516
08 code if the variable duty cycle is outside the range of 5-95%. The BCM will set a B1516 66
code if polarity is reversed. This code can also be set if using a battery charger without placing
the chargers ground onto the engine block.
[unquote]

It does give the wire colors, but not the pin #'s. It doesn;t have to worry about a 0-5V output that is ratiometric.

 

[KeepItSimpleStupid]

the LT-6700-1 is only available in the TSOT-23 IC which is a Surface Mount part, very difficult to solder in. So I want to come up with a Break Out board that already has a OP amp on it. If a design can be done that uses a ready made breakout board,

Check out www.proto-advantage.com

This https://www.proto-advantage.com/store/product_info.php?products_id=2200230 is likely the adapter. They will procure and solder if it' digikey part for an additional cost. Back to hurry up and wait again.

I drew up a few things in EASYEDA, but haven;t gotten to the point of getting a board made. I do have a project I'd like to do though. Only a few parts, but a tricky layout and small.

 

[KeepItSimpleStupid]

Battery Control Modules Theory and operation: https://ryangottlieb.com/?page_id=73

 

[KeepItSimpleStupid]

Another blurb:

Thanks Eddie39. I put in my VIN and the site shows me the part number that Autozone gave when I went there which is the 2nd Design; part number 13505369. I am still confused. Will this actually work? I ask because when I look up the different designs, one is "(ACDELCO 10306471) 1st Design; FOR VEH WITH 10V SUPPLY FROM BCM)" and the other is () 2nd Design; FOR VEH WITH 5V SUPPLY FROM BCM).

from here:

Which Battery Current Sensor

www.cadillacforums.com

 

[Danwvw]

Had a blank post on the previous page and responded there: Thanks
The Allegro ACS773 50 Amp Hall Effect precision current sensor looks very promising to current limit the LT-6700-1 Design as it could solder right into the main D+ buss in place of the shunt and the spare channel of the LT-6700-1 used to set a threshold.

 

[KeepItSimpleStupid]

I guess you can look at >some value (overcharging) or < some value (not charging enough) and light up the idiot? A red/yellow LED would be cool.

 

[Danwvw]

A light would be ok and I can play with that with a different transducer the switch ones but that is not what is happening here.
ACS773
0 Amps=2.5 volts
+50 amps= 5 volts - some value.
-50 amps=near 0 volts
So with a 500 ohm pot it can be adjusted to say current limit anywhere between 0 and -50 amps it would control the gen charging via Pulse Width Modulation like the Voltage Regulator channel of the LT-6700 is doing. All I will do to adjust it is set it just above where it limits with the headlights on. I have concluded that current regulation is needed to protect the cutout diode from overheating when the battery is low. Previous prototypes 2 and 3 with the heat sinks were better at handling overload.

 

[Danwvw]

Experimenting with CMOS Voltage Detectors: ICL7665S

4 PIN Comparator very simple Design and it would charge the same voltage 7 volts out regardless of the 6 volt generator it's on or the headlight load. ADCMP350

 

[KeepItSimpleStupid]

Note: The ACS773 is a 3.3V part AND the output is ratiometric to the power supply voltage.

 

[Danwvw]

Thanks for the comment the schematic I posted should work, I have the supply at 3.5 Volts with a Zener and the ACS773 output across 500 ohms. Anyhow if it outputs say 3 volts at or near 50 amps I just remembered it Zeros at 2.5 volts (3.5*0.5) and goes +or- from there. Still it should be possible to adjust the 500 ohm pot to cross the LT-6700-1 threshold which is 0.4 volts. The ACS773 is on order and more parts ie ICL765S and the ADCMP350 (Which I am really excited about using to simplify the build) for the other voltage only schematics posted too. I will have to install a ACS773 Prototype on the car to know anything about if it will actually regulate current though. The LT-6700-1 Prototype Photo is bench tested and likely one of the best choices too.

From the ACS773 Data Sheet:
"ZERO CURRENT OUTPUT VOLTAGE (VIOUT(Q))
V / 3.3 V CC where: IPOSx and INEGx are positive and negative currents. The output of the sensor when the primary current is zero. It nominally remains at 0.5 × VCC for a bidirectional device and 0.1 × VCC for a unidirectional device. For example, in the case of a bidirectional output device, VCC = 3.3 V translates into VIOUT(Q) = 1.65 V. Variation in VIOUT(Q) can be attributed to the resolution of the Allegro linear IC quiescent voltage trim and thermal drift.
ELECTRICAL OFFSET VOLTAGE (V ) OE
The deviation of the device output from its ideal quiescent value of 0.5 × VCC (bidirectional) or 0.1 × VCC (unidirectional) due to nonmagnetic causes. To convert this voltage to amperes, divide by the device sensitivity, Sens. "

 

[Danwvw]

The SparkFun BreakOut Boards with the THS82 IC OpAmp on them arrived today. I was thinking of the following schematic to use one as a Solid State Dynamo Voltage Control. Does anyone have any experience using the THS82 OpAmps especially as Comparators.

 

[Danwvw]

Today tested the ICL7665S Comparator with its 1.3 volt internal REF and Vin 1.6 to 16 volts. It shows a lot of promise to be used in a similar manner as the LT6700-1 only better because both inputs are inverting and it's an 8 pin Dip.
More testing is needed to see how well its outputs will operate in parallel, Planning to use Set 1 for Voltage and Set 2 Current control. ICL7665S datasheet.

A rather clean square wave 1.5ms/cm ie.. 666 Hz oscillation at threshold without hysteresis configured totally clears up with 50 ohm HYST resistor.

 

[KeepItSimpleStupid]

as the LT6700-1 only better because both inputs are inverting

Don't forget, you can the LT6700-x as both inverting, both non-inverting and one of each depending on the suffix -x.

 

[Danwvw]

Oh, Ok yes that is good Thanks, LT-6700-2 This one LT6700HVHS6-2#TRMPBF but for now just having fun learning about these different chips . The Hall Effect current sensors arrived I will build a new Regulator to fit one and may use the ICL7665s with it but really the Hysteresis network probably won't work because it references the Un-regulated and now I want to reference B+ but yes it could work if I just eliminated the Un-Regulated supply and powered off of B+.

 

[Danwvw]

 

[KeepItSimpleStupid]

Wow! Looks like something to encase in "Silly Putty"?

 

[Danwvw]

Well. not Silly Putty but it got a bit complicated It test out though, Only problem is I failed to really check out the. ICL7655s Logic. Set 2 is effectively Non Inverting Logic as I have things wired. It can work though If I reverse the Hall Effect main current path. The voltage side "E" hysteresis circuit is working but not the Current side "I" hysteresis circuit must be the reversed logic, Got to wrap my brain around that. Thanks "KeepitSimpleS" for following along here.
ICL7655s Logic:

 

[Danwvw]

Well the Hist2 resistor was missing from the prototype but the logic will need to be handled. Here is what I am going to try: