Who would like to beta test one of my new voltage regulators?

-
Right now I haven't worked out how to make an adapter to stock wiring. I don't have the resources to do an injection molded connector, but if I can find similar contacts I can probably do something with a silicone mold and two part plastics. However, you can get the opposite side of the Amphenol connector to adapt wiring configured for the motorsport version back to stock wiring.

Need to solve a minor overheating problem with my test car before I'm ready to say these are holding up well. I'm guessing the radiator shroud I used to hold the fans is creating more problems than it solves.
 
Right now I haven't worked out how to make an adapter to stock wiring. I don't have the resources to do an injection molded connector, but if I can find similar contacts I can probably do something with a silicone mold and two part plastics. However, you can get the opposite side of the Amphenol connector to adapt wiring configured for the motorsport version back to stock wiring.

Need to solve a minor overheating problem with my test car before I'm ready to say these are holding up well. I'm guessing the radiator shroud I used to hold the fans is creating more problems than it solves.
I rewired my 71 Duster myself, without a factory loom. Thus I do not have any factory wiring molded connectors. So I can adapt your setup to my car without the molded connectors, you are concerned about
 
Will this work with the 69 stock round alternator?
 
It is the Roundback through model year 1969.
1969 seems to have been something of a changeover year. If you have the "isolated field" style here, it will work: Identifying Chrysler Alternators (1960-1976) The grounded-field won't work with this design, but once I have this one headed for production I'm going to look at those too.

1700444379107.png
 
Most of the round back alternators are grounded-field, but there's a few isolated field versions. I can't see the terminals from that angle.
 
Latest batch works on the bench but runs into issues on the car. At least the regulator still passes a bench test after testing on the car, which is an improvement over what happened the last time. The plan is once I have one working on the car to run it in for a while before making the others available.

I wonder if I can find one of those alternator test benches you see at parts stores, secondhand. That would avoid the risks from testing on a car, while bench tests clearly don't cover all the bases.
 
Hey Matt, I would like to try one on the A100 I am building... it will be a month or two before I am to that point though.. If you want someone to bounce ideas off send me the schematics and we can discuss what the issues are you are seeing. Did you account for the huge L you see looking into the field winding in your control loop? (retired electrical engineer in power systems, switching converters, motor controllers, etc).....
 
Hey Matt, I would like to try one on the A100 I am building... it will be a month or two before I am to that point though.. If you want someone to bounce ideas off send me the schematics and we can discuss what the issues are you are seeing. Did you account for the huge L you see looking into the field winding in your control loop? (retired electrical engineer in power systems, switching converters, motor controllers, etc).....
Thank you for the offer. I think what happened was that I slowed down the switching too much, overcompensating for an earlier design that switched field current at way too high a frequency and ran the field coil like it was a boost converter. I basically copied the Mopar design for my latest revision with an adjustment circuit added but had to guess at capacitor values, and I think I guessed an order of magnitude or so too high.

And I may have a line on a way to make an adapter to stock Mopar wiring as well - I ordered some contacts that may fit along with new capacitors yestereday.
 
Thank you for the offer. I think what happened was that I slowed down the switching too much, overcompensating for an earlier design that switched field current at way too high a frequency and ran the field coil like it was a boost converter. I basically copied the Mopar design for my latest revision with an adjustment circuit added but had to guess at capacitor values, and I think I guessed an order of magnitude or so too high.

And I may have a line on a way to make an adapter to stock Mopar wiring as well - I ordered some contacts that may fit along with new capacitors yestereday.
So this is a switching type design vs a linear to try to save power and keep the FET cooler? I should stick my scope on these Chineese ones I have and see if they are linear or not. My guess they are not because FET is just pop riveted to the housing they are not potted or anything. A linear design would burn some serious power for sure.
 
It has been awhile.. It is about 7 ohms looking into the field winding correct? So the max current is in the 2.5A range if you are CW on the alternator?
 
Been researching a playing around with a charging system model I found on a forum with spice a bit.. I had never really spent anytime understanding how these really worked... I get it now. They are just bang bang type controllers so they are inherently stable... I did read that modern regulators run constant PWM and vary the duty cycle. There are also other considerations for off nominal conditions like startup and recharging the battery running full alternator output for a significant amount of time putting undo stress on alternator diodes, bulk head connectors, etc.

The frequency really ends up being variable depending on the capacitance of the battery, load and peak charging current. You mentioned running the frequency too high, are you running fixed freq and modulating duty, it does not seem like the PWB photo had enough components to do that?
 
Last edited:
The board I'm using is similar to the one from this thread where I tore down an off the shelf regulator:

Alternator voltage regulator teardown and analysis

However, I've added back in the temperature compensation and added a trim pot to adjust voltage. I'm using a Nippondenso alternator with around 3-5 ohms field coil resistance. A previous design I tried used a microcontroller but it was running a bang-bang output with no deadband other than its ADC resolution, which was a big mistake.

Running a FET in the linear region might give you a very accurate voltage, but I don't think it's worth the elevated temperatures. Especially since the temperature sensor is inside the regulator since installing an external battery temperature sensor is way more complexity than most installers would want to deal with, for fairly little gain. If I were going to add an external sensor, it would be some sort of battery current meter. But I'd rather get a solid basic one working first.
 
I offered to test your regulator on my 71 Duster 318, Am I waiting for a reason or should I just give up on ever seeing this regulator
 
Progress, finally. Thanks to Jim Kueneman, who checked with an alternator simulation and realized the circuit I had copied off the stock type regulator could be very finicky depending on capacitor size and some quirks of the alternator. So I made some changes so that the regulator acts more like a Schmidt trigger and instead of having a minimum time between switching the field off and on, it targets holding the voltage in between a minimum and maximum value. Here's a data log from the onboard EFI when testing my new regulator and a stock-replacement regulator - the sampling is at around 20 Hz, so it won't catch abrupt spikes, but it shows the new regulator (left) is holding voltage as well at that sampling rate as the old one (right).

So what comes next? In about this order:
  • Modify the remaining batch of regulators in the same way as well as upgrading one of the components for improved reliability
  • Test a set of contacts I've found to see if they'll plug into a stock regulator connector (I'll need to make a housing, probably using a silicone mold and 2 part epoxy)
  • Install one of the regulators more permanently on my Dart
  • Drive it a lot to make sure it's working well
  • Send out the remaining four regulators for testing

regulator graph annotated.png
 
The weather hasn't exactly been good for installing the regulator and racking up a lot of hours of operation, so I've been working on the problem of how to deal with plugging into the stock harness when the plug isn't available. I don't have the budget for something like Inventor or Solidworks, or even Alibre, and don't want to deal with Fusion 360's cloud-only storage, so I decided to give Freecad a try. It's nowhere near the same league as commercial CAD packages, but it was enough to get started:
connector CAD model.png


I've spent the past weekend getting a newly arrived cheap 3D printer up and debugged, and now:

printed first attempt.jpg


The hard part is going to be figuring out what to use for contacts; that's a work in progress...
 
Then withdraw without the whining. There's probably a dozen people that have offered, and you're squawking.
You think it's just as simple as cutting it out of cardboard and mailing it to you? He is providing a service to us and yer bitching...
If something goes/went wrong, i can only imagine... :rolleyes:
I offered to test your regulator on my 71 Duster 318, Am I waiting for a reason or should I just give up on ever seeing this regulator

I am going to withdraw my offer to test your device.
 
I understand if someone needs a regulator right now and doesn't want to wait. Right now I could have the devices completed and shipped out in a few days with just a bench test and a few minutes of running an example alligator-clipped to my alternator, but I prefer to do the first long term test on my own car in case there are any more surprises.
 
I think I was the first to offer and it still stands. No sweat if someone else better fits your needs. Mine charges and that's all that matters.
 
Minor setback: I bought a tube of something advertised as silicone potting compound for the first batch...

IMG_20240211_145803757.jpg

... turned out to be more of a silicone sealant, and not appropriate for use as a normal potting compound. I don't trust the first batch to keep out water at all, and it may even trap water.

For my Dart, this will not be a problem - the car uses a trunk mounted voltage regulator to match the temperature of a trunk mounted battery. But for a unit under the hood, this is likely to be an issue. If anyone reallt wants to try one of these regulators, I can send one at no charge and with no guarantees. There is one more available with the Amphenol AT connector and three with spade terminals. But I'd recommend waiting for the preproduction prototypes; I'm going to be working on redesigning these based on what I've learned from testing them on my alternator so far.

Meanwhile, I have made some progress on figuring out the connector pins needed. So the plan is to offer the production run with stock type and Amphenol connectors, unless there's real interest in a spade terminal type option as well.
 
I decided to use the spade terminal version rather than the motorsport connector version - I was feeling lazy and didn't want to run an extra wire. The 3D printed adapter means I can easily switch back if anything breaks on the test run.

IMG_20240216_150609406.jpg


Ran for 20 minutes without any trouble. Next step will be to lay out the production board with a stock type connector. Is there any interest in the motorsport connector version as well? Because this is going to take two circuit boards with the way I plan to set up the stock connector.
 
Hi Matt, I’m happy to beta one for you and can do the connector however which way.
Cheers
Tim
 
-
Back
Top