For any add-on boards such as VR conditioners, optos and OEM interface boards
By Old Grey
#11113
Cool, will look at that.

From memory the resistance goes down for higher temp in my car.

I wanted to differentiate the 12V from the 5V so I labeled them different, but I'll have to change that - I've seen EEVblog cock-on about neg rails so much I should've tweaked -.

Just noticed I don't have the neg from the reg tied to the circuit.
By Old Grey
#11148
The more fancy $1.20 board with duty cycle seams good and very stable on my DSO138 - the others weren't -, except for the annoyance of having to swap jumpers for under 3000rpm. I tried to parallel the 2/3 caps/jumpers but can't get any better range.
http://www.ebay.com.au/itm/351550562512
With 5V input it puts out 3.5Vmax - I wonder if it's enough to trigger a Speedy
With 6V input it puts out 4.5Vmax
When duty goes below 20% or over 98% it gets flaky, but at least it has a big hit range so should work just by plugging it in without knowing the duty.
Image
User avatar
By cx500tc
#11153
3.5V out should be enough to register a state change. I think the lower limit for a logic high is around 2.7V and upper limit logic low is 0.7V. There's a bit of a gap in there for hysteresis and any voltage in that range may or may not be interpreted correctly- one reasons most MCU's like the 'duinos have "INPUT_PULLDOWN" and "INPUT_PULLUP" options to mostly ensure one state or the other in the absence of a legitimate input signal. If you're in doubt, a small signal transistor could switch the 5V rail in and out for certainty. The current is negligible so a mosfet wouldn't be needed. Something like a 2N2222:

Image

A pull down resistor to ground on the gate might be needed though to ensure "off" state.
By Old Grey
#11171
I was trying to simplify the circuit and mod-ed that osc, but when I added the 5V reg it's now down to 2.5V.
The eBay osc's output voltage changes with the input - about 1.5V less - but when I add the 5V reg to my simplified version it now seams too low.
So it's either your trans, or run a 7V reg and voltage divider to the rest of the board.
Image
The circuit can now do idle to 18,000rpm fixed 60% duty cycle.
User avatar
By cx500tc
#11173
If you're only measuring the voltage output it could appear to be lower than it really is. Think along the lines of AC voltage where the peak voltage differential is more than the average, a.k.a. RMS - root mean square. While the voltage output is a switched DC voltage, it is technically an alternating voltage with a positive bias equal to 1/2 the peak voltage so oscillates around that instead of 0V like a 'normal' AC source would do.

An oscillator running 50% duty between 5.0 and 0V will read as 2.5V.

Or at least that's my theory. ;)

As for the regulator, you need sufficiently large electrolytic caps on the input and output stages so as to absorb transient voltage draws, but also need non-polarized to filter out harmonics; 1.0uF and 0.1uF are common values and we see those often used as "decoupling" capacitors placed near IC's.

I'm pulling things out of my retentive region here but with a 7805 regulator, you should have capacitors like electrolytic around at least 4.7uF and 0.1uF non-polarized rated to at least 2x the input voltage on its input, in that order, and the same in reverse order on the output rated accordingly.
By Old Grey
#11180
It's showing 2.4Vmax, -0.9Vmin, 3.3Vpp and 1.77Vrms on the scope - on a multi it's 1.7V on AC -. I read that it has to be a 5V sq wave, so is that just the positive component, Vmax or the full wave, Vpp.

The caps came off the LM7805 datasheet, should I change them.
User avatar
By cx500tc
#11197
Old Grey wrote:It's showing 2.4Vmax, -0.9Vmin, 3.3Vpp and 1.77Vrms on the scope - on a multi it's 1.7V on AC -. I read that it has to be a 5V sq wave, so is that just the positive component, Vmax or the full wave, Vpp.

The caps came off the LM7805 datasheet, should I change them.
Well, I feel somewhat vindicated since you are getting a voltage reading on the AC scale. 1.7VAC sounds about right considering the readings you're getting on the scope. Having a negative voltage does seem suspicious to me though. That might imply some back-EMF is occurring, and that could be a result of how you have the 7805 set up. Non-pol caps are a must.

2.4V will probably not register as a logic high to Speeduino. Like I mentioned before, 2.7V is about the bottom end of what would be considered high with 5V CMOS logic signals.

As for the 7805, I'm looking at this data sheet- https://www.fairchildsemi.com/datasheets/LM/LM7805.pdf, and it's not calling out the capacitors as electrolytic type so that may be part of your problem since you are using polarized. I've nothing against that but there should be non-polarized caps there too. They will help block any negative voltage signal from affecting the regulator and since you're seeing that, I'd suspect that may be part of your issue.


So I suggest you do, as a modified version of the datasheet I referenced above:
V+ -> 10uF/33V electrolytic cap -> 0.33uF/33V non-pol cap -> 7805 -> 0.1uF/16V non-pol cap -> 10uF/16V electrolytic cap -> load
I'd be cautious about the caps on the input side and their voltage rating: you may want to go up a bit just in case the voltage may exceed their rating. Heck, maybe up-rate all the caps just to be safe. ;)
By Old Grey
#11207
It doesn't seam to matter if it's powered from the PS, PS to 5V reg, more caps, no caps, all these osc circuits seam to have a neg component.

I frizzled my 555, and now the new one runs at 3.0Vmax, -0.7Vmin. This one is the best so far with the least neg of the 5 circuits.
With my fixed 330r resistor replacing the duty pot, at idle duty is 50% and at 24,000rpm it's 98% duty.

I didn't have a 2N2222 so I used a BC547 and the circuit inverted the wave, so that didn't pan out.

Seams the best way is use a 7V reg - 4.75Vmax, -0.9Vmin, 5.7Vpp and 4.35Vrms - and voltage divide down to 5V for the rest of the circuit, as it cuts down on parts.
User avatar
By PSIG
#11211
It seems something is amiss here, as a 555 should output a clean 5V square wave, and be of almost no load to the system. Is there possibly a different failed component, or a mis-connection to account for it?

David
User avatar
By cx500tc
#11216
Yeah, the transistor will invert the signal if you're pulling from the high side of the transistor, which is the diagram I posted. You should be able to invert the diagram, and by that I mean pull the signal off the other side of the transistor with the resistor below it to get the opposite effect. Consider the difference between "high" side driver and "low" side driver, or maybe, rather, one that switches the voltage source as opposed to one switching the ground.

What's going on is that with the 2N2222, or BC547 you're using, the "out" to the ECU is high when the transistor is off, and low when it's turned on. Moving that "out" and resistor to the other side will result in the opposite- "out" will be grounded by default with the transistor off, through the resistor, but get voltage when the transistor turns on.

If you do invert the drawing, you should also be able to bump the resistor up to 4.7K to 10K without any problem other than maybe a slightly delayed transition, but the increased resistance will keep current consumption down.

I'm concerned though that you're getting a negative voltage component- that shouldn't happen. As @PSIG mentioned, you should be having a fairly clean square wave and very little load. Back-EMF is almost certainly the reason for the negative voltage but I can't see any particular reason why you are seeing it. And if you have non-pol caps of around 0.1uF on the 7805 output, that should block any AC related component if you are measuring the voltage right at the regulator. If you're measuring it at the 555, maybe you need a cap there?

What does your latest, greatest schematic look like? (I don't mean that to sound as cheeky as you may think. I am interested.)
  • 1
  • 10
  • 11
  • 12
  • 13
  • 14

Hi, I am trying to assign Signed values to the x-a[…]

Vr Conditioner Noise when cranking

New version 202305 don't fix the issue. Now after[…]

blitzbox

I've finally figured out why MAP and Lambda weren'[…]

Hello, I bring news!! Let me tell you that after […]

Still can't find what you're looking for?