VR sensor board mods for uS vs AC Zener voltage clamp

[Bruce Bowling]

To compute the voltage/current output of the VR sensor is complicated. You can measure the voltage at different RPMs (across the VR) to get a trend, if desired.

- Bruce

[Bruce Bowling]

No need to perform the mod as listed in this thread http://www.microsquirt.com/viewtopic.php?f=89&t=23063 ?

I have a gen 1 board, just like the one pictured.

If you have the red-colored board, then yes you should do the diode mod. The original diode had leakage issues at higher temperatures and could cause issues.

Also, no need to an AC Zener clamp regardless of this mod?

No mods for the AC voltage, unless you are feeding several hundred volts into it, and then you would just increase the series resistance.

What wattage resistor would you recommend? What is the main purpose of the resistor in this case as you have recommended? To lower voltage, lower current, or both?

THANKS!

I would use a 1/2 watt resistor, or parallel two 1/4 watt resistors at 10K resistance.

The resistor is for current limiting purposes mainly.

[Aaron Silidker]

Are you still able to assist people with the mods to the boards as you had listed in that thread? I am not comfortable soldering a small board like this one.

So the voltage is not the issue, the current is the main issue? I thought the voltage is the issue because it causes the diode to reverse avalanche.....

Thanks again everyone

[Aaron Silidker]

To compute the voltage/current output of the VR sensor is complicated. You can measure the voltage at different RPMs (across the VR) to get a trend, if desired.

- Bruce

I measured at 80, 240, and 400 on the lathe to get a trend. It was basically linear as expected and projected to a 600 volt p-p at 10k rpm redline on the bike.

I was just more curious how I could calculate current in this circuit without knowing the resistance. I guess the easiest thing to do here is measure current, then I can calculate how the resistor will affect it exactly.

If I parallel two 10k resistors, I would have a total resistance of 5k correct? I will just use a half watt 10k, they are easy to get from anywhere it seems.

[Bruce Bowling]

I measured at 80, 240, and 400 on the lathe to get a trend. It was basically linear as expected and projected to a 600 volt p-p at 10k rpm redline on the bike.

Is this test with no VR load? If it is unloaded (i.e. just wires in the air) then it can generate a lot of voltage.

Try putting a 10K resistor across the VR wires (as a load) and scope across the 10K resistor - does the amplitude change? Repeat with a 1K resistor - does it decrease?

- Bruce

[Aaron Silidker]

Yes, this test was with no load. The "misinformed person" mechanical engineer in me did not realize the load makes that much of a difference because I didn't think the load from the uS would be significant.

I will test again tonight and get some more voltages for you guys to analyze. Are 1/4 watt resistors okay for this bench test? It is all I have easily available.

I take it when I test this with different resistor loads at a given speed, we can figure out what type of voltage the uS will see knowing what type of resistance is internal to the appropriate circuit for the VR in on the uS. Correct?

Also, are you still available to help with the board modifications to update my Gen 1 uS?

[Peter Florance]

Yes, this test was with no load. The "misinformed person" mechanical engineer in me did not realize the load makes that much of a difference because I didn't think the load from the uS would be significant.

I will test again tonight and get some more voltages for you guys to analyze. Are 1/4 watt resistors okay for this bench test? It is all I have easily available.

I take it when I test this with different resistor loads at a given speed, we can figure out what type of voltage the uS will see knowing what type of resistance is internal to the appropriate circuit for the VR in on the uS. Correct?

Also, are you still available to help with the board modifications to update my Gen 1 uS?

I can do it if you can get it to me
I'm surprised you don't have some EEgrad student you can task for that. They have to be useful for something..



The diode is the only thing you need to change internally and I'm not sure it's an issue with as much output as your sensor has.

[Aaron Silidker]

Yes, this test was with no load. The "misinformed person" mechanical engineer in me did not realize the load makes that much of a difference because I didn't think the load from the uS would be significant.

I will test again tonight and get some more voltages for you guys to analyze. Are 1/4 watt resistors okay for this bench test? It is all I have easily available.

I take it when I test this with different resistor loads at a given speed, we can figure out what type of voltage the uS will see knowing what type of resistance is internal to the appropriate circuit for the VR in on the uS. Correct?

Also, are you still available to help with the board modifications to update my Gen 1 uS?

I can do it if you can get it to me
I'm surprised you don't have some EEgrad student you can task for that. They have to be useful for something..



The diode is the only thing you need to change internally and I'm not sure it's an issue with as much output as your sensor has.

I wish I trusted them with any sort of fabrication tasks incuding this! I will PM you for your information to send the uS to you. They have been somewhat helpful with this project, but you guys are way more helpful being familiar with the uS and familiar with using your hands, unlike most students.

As far as the diode is concerned, I thought it was an issue with VR input over a certain voltage which I am way over. From what I understood from Bruce's post about it, the reverse voltage that makes it avalanche backward is low, then it shuts some transistor off when this happens.

Correct me if I am wrong!

I know later on he also posted about using the updated firmware.

When I get to that point, I will tackle it. I am confident in my hardware with respect to injector, sensors, etc...I am now also quite confident in my wheel and VR sensor providing a very clean signal. I am sure it will be no problem to get the uS to run everything. Then I just have the fun of calibrating it all!

[Aaron Silidker]

Just did some tests with a resistor mounted in series.

With 0 ohms I was just under 22 volts p-p at 400 RPM
With 1m ohm I was just over 12 volts p-p at 400 RPM
With 10m ohms I was just over 2 volts p-p at 400 RPM

The only other thing connected to the VR besides one series resister is the oscilloscope.

Thoughts on the above numbers?

[Bruce Bowling]

As far as the diode is concerned, I thought it was an issue with VR input over a certain voltage which I am way over. From what I understood from Bruce's post about it, the reverse voltage that makes it avalanche backward is low, then it shuts some transistor off when this happens.

Correct me if I am wrong!

Yes, the diode swap is due to temperature leakage....

To answer an earlier question - will the high voltage affect the diode or transistor? In short, it will not because in one direction there is the diode 0.7 volt drop protecting the transistor, and in the other direction the transistor emitter-base is forward conducting and it also drops 0.7 volts.

To illustrate, I ran a pspice simulation of the VR circuit, and input a +/600 volt (thats 1200V p-p), low impedance source, measuring different points in the circuit. Here is the schematic and a simulation of the VR input voltage - you can see the +/- 600 volt swing at 2000 Hz (way fast for even high tooth count wheels at high RPMs):



Next, here is the same run, looking at the base voltage (blue) and the emitter voltage (green) of Q4:



Here is a zoom of one half cycle of the above plot to see the voltage amplitudes:



The voltage swings stay within 0.7 volts of the 5-volt supply, so neither the diode or transistor is in danger of excess reverse voltage.

- Bruce