Page 2 of 2

Re: VR board

Posted: Mon May 27, 2019 11:09 am
by AndreyB
puff wrote:
Mon May 27, 2019 10:57 am
just healthier trigger disk with more teeth...
I am not so sure about that. What is the exact math? Let's assume 150 RPM for cranking, we know 85ms watchdog. How many tooth is the threshold for the issue?

Re: VR board

Posted: Mon May 27, 2019 6:33 pm
by puff
okay, let's calculate:
150 rpm means 150/60 = 2,5 rotations per second, right?
one rotation per 0.4 seconds?
with two equal teeth you'd probably have two pulses per 0.4 seconds. loosely that means 0.2 second per tooth?

with subaru trigger the pattern was somewhat unusual - with a large gap (1/3 of the whole wheel ~ 1/3 of 0.4s = 133 ms with 150 cranking rpm?)
so probably with 150RPM the threshold is somewhat about two equally spaced teeth? not sure though if VR sensor would easily detect them.

Re: VR board

Posted: Tue May 28, 2019 12:45 am
by kb1gtt
Also Hall does not care about gain, as hall is always a huge signal. The gain might cause some false triggers as it will be more sensitive to the noise when the watch dog has timed out. However if you do not get erroneous pulses, then your sensor side is not picking up noise, so you are good.

Hmmmm, should I encourage you to make a simulator with a discovery's DAC?

Re: VR board

Posted: Tue May 28, 2019 12:56 am
by AndreyB
kb1gtt wrote:
Tue May 28, 2019 12:45 am
Hmmmm, should I encourage you to make a simulator with a discovery's DAC?
You know the probability of this is zero. @abricos I know you have a VR sensor and wheel outside of the car, can you post a picture? How do you control RPM on that bench setup? Is there a chance to change trigger wheel to something Subaru-like?

Re: VR board

Posted: Tue May 28, 2019 1:27 am
by Abricos
russian wrote:
Tue May 28, 2019 12:56 am
kb1gtt wrote:
Tue May 28, 2019 12:45 am
Hmmmm, should I encourage you to make a simulator with a discovery's DAC?
You know the probability of this is zero. @abricos I know you have a VR sensor and wheel outside of the car, can you post a picture? How do you control RPM on that bench setup? Is there a chance to change trigger wheel to something Subaru-like?
I don't have Subaru trigger wheel ... I have million electric motors to attach trigger wheel ...

super unit free energy )))

Re: VR board

Posted: Tue May 28, 2019 8:43 am
by kb1gtt
Someone sent me 2X Subaru wheels, should I ship them to someone who can get around to trying them out?

Re: VR board

Posted: Wed May 29, 2019 9:36 am
by Simon@FutureProof
No need for a physical wheel, I have been using this to generate crank triggers for a while.
http://www.daqarta.com/

It has a specific menu for generating crank triggers in different shapes. Hook it up via an isolating transformer and you get really good low voltage representations of a VR crank trigger.
With a good sound card the square wave is not too bad either. (read that as a 10 year old ebay cheap but good quality sound card ;) )

I have a MAX9926 on a breakout board I could test pretty easily but I'm jammed up with work for likely about a week.

Re: VR board

Posted: Wed May 29, 2019 10:15 am
by kb1gtt
I've used these radio-shack transformers for things similar to this. Any suggestions on a particular part number? I'm thinking the transformer should be 600 ohms, 1:1 and reasonably high frequency response.

https://www.ebay.com/sch/i.html?LH_CAds=&_ex_kw=&_fpos=&_fspt=1&_mPrRngCbx=1&_nkw=Radio+Shack+273+1374+Audio+Isolation+Transformer+1+1+Turns+Ratio+Lot+of+7&_sacat=&_sadis=&_sop=12&_udhi=&_udlo=&_fosrp=1

Re: VR board

Posted: Wed May 29, 2019 11:00 am
by puff
By the way, speaking about the possible tests of VR interface in general (just to check if I get square pulses out of a sine wave) - is it a good idea to use 18V transformer (220AC to 18 AC) as a source of that sine signal?
Thing is, when I used a multimeter in AC voltage measuring mode, I got only less than 1V from a sine wave on my smartphone, and just a bit more than that - when I passed the signal through the audio amplifier.

Re: VR board

Posted: Wed May 29, 2019 12:00 pm
by Simon@FutureProof
kb1gtt wrote:
Wed May 29, 2019 10:15 am
I've used these radio-shack transformers for things similar to this. Any suggestions on a particular part number? I'm thinking the transformer should be 600 ohms, 1:1 and reasonably high frequency response.

https://www.ebay.com/sch/i.html?LH_CAds=&_ex_kw=&_fpos=&_fspt=1&_mPrRngCbx=1&_nkw=Radio+Shack+273+1374+Audio+Isolation+Transformer+1+1+Turns+Ratio+Lot+of+7&_sacat=&_sadis=&_sop=12&_udhi=&_udlo=&_fosrp=1
Not sure on that, 1:1 works fine the one I have is pulled out of an old PC power supply and is 1:1 but I don't see why any other ratio would not work.
It's needed more so that the circuit to the MAX chip is a loop with inductance so gives a proper VR style +- fluctuation instead of a 0-Xv positive wave.

Cheapest source would be pull one out of an old power supply for a laptop or PC, potentially even from a 12v mains plug. Used them because I had them, I pulled a couple of old power supplies apart for the components and threw the bits in a draw ages ago. I have tried all 3 of the smaller ones in this picture with similar results, I believe they are used for isolation in the power supplies anyway.

https://www.techpowerup.com/articles/160/images/seasonic_X560_top.jpg

Re: VR board

Posted: Wed May 29, 2019 12:12 pm
by Simon@FutureProof
puff wrote:
Wed May 29, 2019 11:00 am
By the way, speaking about the possible tests of VR interface in general (just to check if I get square pulses out of a sine wave) - is it a good idea to use 18V transformer (220AC to 18 AC) as a source of that sine signal?
Thing is, when I used a multimeter in AC voltage measuring mode, I got only less than 1V from a sine wave on my smartphone, and just a bit more than that - when I passed the signal through the audio amplifier.
1v output from a sound card has in the past worked for me to trigger both OEM Bosch ECUs and the MAX9926 used on rusEFI and Speeduino.

Re: VR board

Posted: Wed May 29, 2019 4:17 pm
by kb1gtt
My concern with the power supply, is that you are likely to grab a 60 Hz one, instead of a high frequency inductor. It may act like a low pass filter and do odd things. It shouldn't be a huge deal to get a good small one. I'll keep poking around, until I find one that is not stupid priced.

Re: VR board

Posted: Fri May 31, 2019 7:41 pm
by Simon@FutureProof
Sure, I can only really suggest what I have tried and I expect your US power supplies are different to our UK ones.

I simply scoped the signal before and after the transformer to check they matched.

I will do something tonight to confirm if we have this issue.

Re: VR board

Posted: Sun Jun 02, 2019 7:17 pm
by Simon@FutureProof
Confirmed this issue earlier today, it needs a combination of a short trigger duration (like 10 degrees) and a slow RPM (10hz signal frequency causes it).

Seems like the only answer for a lot of these situations is to use a hall sensor for the slow signal.

Re: VR board

Posted: Sun Jun 02, 2019 8:48 pm
by kb1gtt
An alternative solution is to use the MAX chip with a slightly different mode. Then control the gain with an analog signal. Basically a RPM to voltage table.

Re: VR board

Posted: Mon Jul 01, 2019 3:38 pm
by AndreyB
Do I understand that we currently use MAX9926 in A2 operation mode?

Between four MAX9926 operation modes which one are we interested in to avoid the 85ms watchdog? What additional logic would be needed in rusEfi firmware? What additional circuitry would be needed?

Re: VR board

Posted: Mon Jul 01, 2019 4:06 pm
by kb1gtt
Yes A2 has the adaptive peak threshold, which is what causes the 85mS issue. I believe we want the mode B to remove this automatic stuff, and the STM then will control that peak threshold via analog DAC signal. If we did that then the 85mS issue could be removed as the gain is set by the STM and not the MAX chip.

Re: VR board

Posted: Tue Jul 02, 2019 4:38 pm
by puff
populated the max chip onto the frankenstein board.
with a sine gererator I can see the output from the max chip, but the amplitude is 380v, according to my scope. probably, it's not enough for the stm board?

Re: VR board

Posted: Tue Jul 02, 2019 9:28 pm
by kb1gtt
380mV??? 1X probe with a 100X scale?

Re: VR board

Posted: Wed Jul 03, 2019 5:46 pm
by Реконструктор
I remember making such thing based on analog comparator with hysteresis. It was quite universal, worked with hall sensors as well.

Re: VR board

Posted: Wed Jul 03, 2019 6:13 pm
by kb1gtt
Schematic?

Re: VR board

Posted: Wed Jul 03, 2019 6:26 pm
by Реконструктор
kb1gtt wrote:
Wed Jul 03, 2019 6:13 pm
Schematic?
It was long time ago, but luckily, for some reason, my Subaru have VR sensors so I may play with it to reproduce the device once again. :)