work in progress rusEfi own wide band controller board add-on using CJ125

Hardware inside and outside of the ECU
User avatar
kb1gtt
contributor
contributor
Posts: 3711
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA

Re: CJ125 board

Post by kb1gtt »

Can you connect a meter in series with the heater and measure the current?
Welcome to the friendlier side of internet crazy :)
User avatar
AndreyB
Site Admin
Posts: 11173
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: CJ125 board

Post by AndreyB »

So like
3) DIAHD silkscreen on pad #6 should be DIAHG
4) add a note describing UR (R8) and UA (R15) low pass filter / slew rate limiter
5) add a note explaining different C2 and C9 values
Poking @ about cutting the heater wire and adding A-meter into the cut.
https://rusefi.com/s/howtocontribute
very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions
my skype is arro239
User avatar
kb1gtt
contributor
contributor
Posts: 3711
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA

Re: CJ125 board

Post by kb1gtt »

The GND pad is a large heat sink. How confident are we that we got a good solder flow on the leads? If we have 10A of current running through a 0.1ohm resistance, that would generate (10^2)0.1ohm = 10 watts.

What is the voltage on the gate? Measured referenced to the GND side of the MOSFET. Perhaps there is a voltage drop on the GND side of the MOSFET which is causing the gate voltage to be not fully driven. Perhaps it's in the transient instead of saturated state.
Welcome to the friendlier side of internet crazy :)
User avatar
kb1gtt
contributor
contributor
Posts: 3711
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA

Re: CJ125 board

Post by kb1gtt »

Looking at the above picture, you need more solder on the MOSFET leads. The other leads probably need more solder as well, but it's less noticeable as you have much less current flow.
Welcome to the friendlier side of internet crazy :)
puff
contributor
contributor
Posts: 2842
Joined: Mon Nov 11, 2013 11:28 am
Location: Moskau

Re: CJ125 board

Post by puff »

I'd say, for it to be a filter, the resistor should be before the capacitor? Unless the STM chip is an output, rather than an input.

Sent from my XT1058 using Tapatalk
puff
contributor
contributor
Posts: 2842
Joined: Mon Nov 11, 2013 11:28 am
Location: Moskau

Re: CJ125 board

Post by puff »

blind me, but where do I see your CJ125 code?
Atomic-dm
contributor
contributor
Posts: 80
Joined: Thu Nov 21, 2013 2:43 pm
Location: Saint-Petersburg, Russia

Re: CJ125 board

Post by Atomic-dm »

kb1gtt wrote:I think the question is for UR (R8) and UA (R15). I forget why those had the 1k resistors. I'm reasonably sure it's a low pass filter / slew rate limiter. I forget where I got those values from. I'm sure it was off a datasheet or example some where, but I forget where.
RC filter usually converts PWM signal to analog. But out from cj125 already analog (not pwm)
User avatar
kb1gtt
contributor
contributor
Posts: 3711
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA

Re: CJ125 board

Post by kb1gtt »

We can change those to 0R's and see how it works. I think it was a generic anti aliasing filter.
Welcome to the friendlier side of internet crazy :)
puff
contributor
contributor
Posts: 2842
Joined: Mon Nov 11, 2013 11:28 am
Location: Moskau

Re: CJ125 board

Post by puff »

Btw, don't we need a voltage divider (don't remember the voltage level from cj125, but as far as I remember it works from 5v?)
Rhinoman
contributor
contributor
Posts: 256
Joined: Thu Sep 24, 2015 2:14 pm
Location: Wiltshire, UK

Re: CJ125 board

Post by Rhinoman »

R12 is connected incorrectly.
Atomic-dm
contributor
contributor
Posts: 80
Joined: Thu Nov 21, 2013 2:43 pm
Location: Saint-Petersburg, Russia

Re: CJ125 board

Post by Atomic-dm »

Rhinoman wrote:R12 is connected incorrectly.
correct observation
User avatar
AndreyB
Site Admin
Posts: 11173
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: CJ125 board

Post by AndreyB »

3) DIAHD silkscreen on pad #6 should be DIAHG
4) add a note describing UR (R8) and UA (R15) low pass filter / slew rate limiter
5) add a note explaining different C2 and C9 values
6) R12 is not routed right
Attachments
screenshot2.png
screenshot2.png (76.5 KiB) Viewed 6914 times
screenshot.png
screenshot.png (33.71 KiB) Viewed 6914 times
https://rusefi.com/s/howtocontribute
very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions
my skype is arro239
User avatar
AndreyB
Site Admin
Posts: 11173
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: CJ125 board

Post by AndreyB »

R12 moved, recalibrated, solder added

Same issues: hot mosfet. Dmm shows 0a on heater wire. Ideas?
https://rusefi.com/s/howtocontribute
very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions
my skype is arro239
Abricos
contributor
contributor
Posts: 798
Joined: Mon Aug 18, 2014 12:32 am
Location: Carteret, NJ 07008

Re: CJ125 board

Post by Abricos »

I did as told ...
the result is zero ...
It gets very hot mosfet ...
maybe mosfet burned out ???

Image
Image
Image
Image
User avatar
kb1gtt
contributor
contributor
Posts: 3711
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA

Re: CJ125 board

Post by kb1gtt »

What is the voltage on MOSFET pin 1?, I guess same for the other pins.

There are many protections on this MOSFET, burt out is not likely. More likely it's oscillating or not driven fully, or something like that. Seems odd it's getting hot but 0A. Where are the watts coming from? Could the MOSFET be switching on and off really fast, at such a high frequency that the DMM doesn't see it? Can you put a scope on the gate, or source?
Welcome to the friendlier side of internet crazy :)
puff
contributor
contributor
Posts: 2842
Joined: Mon Nov 11, 2013 11:28 am
Location: Moskau

Re: CJ125 board

Post by puff »

could it be that the DMM is broken?
User avatar
kb1gtt
contributor
contributor
Posts: 3711
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA

Re: CJ125 board

Post by kb1gtt »

Did you check the DMM with another DMM to verify the internal fuse was OK? Typically the 10A doesn't have a fuse, so it's probably OK. Also can you connect it to another DMM in the diode setting. That should measure like 10mA. It would verify the meter display is OK.

However if that was open circuit, then it still doesn't explain why it's getting hot. Could you replace the MOSFET? I would not do that yet, but I wonder if it is a possibility.
Welcome to the friendlier side of internet crazy :)
User avatar
AndreyB
Site Admin
Posts: 11173
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: CJ125 board

Post by AndreyB »

kb1gtt wrote:Could you replace the MOSFET? I would not do that yet
I have more of these MOSFETs obviously. Source code says we are running PWM at 500Hz.
https://rusefi.com/s/howtocontribute
very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions
my skype is arro239
puff
contributor
contributor
Posts: 2842
Joined: Mon Nov 11, 2013 11:28 am
Location: Moskau

Re: CJ125 board

Post by puff »

what does the scope shows on the gate of Q1? may be the 3.3V is insufficient to drive it? Is Atomic-dm using the same 3.3v level chips to drive it?
User avatar
AndreyB
Site Admin
Posts: 11173
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: CJ125 board

Post by AndreyB »

@ do you have any of http://www.ebay.com/itm/112035273840 boards by any chance? I could ask @ if he would share his STM32F103 firmware prototype with you if you are interested.
https://rusefi.com/s/howtocontribute
very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions
my skype is arro239
User avatar
AndreyB
Site Admin
Posts: 11173
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: CJ125 board

Post by AndreyB »

puff wrote:Is Atomic-dm using the same 3.3v level chips to drive it?
Yes, @ is using same STM32F103 minimal board as we do. We are also using his firmware. The only difference is that we are using our board and our mosfet.

@ also has a PCB for our board so there is a chance he will solder it one day.
https://rusefi.com/s/howtocontribute
very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions
my skype is arro239
Rhinoman
contributor
contributor
Posts: 256
Joined: Thu Sep 24, 2015 2:14 pm
Location: Wiltshire, UK

Re: CJ125 board

Post by Rhinoman »

russian wrote:@ do you have any of http://www.ebay.com/itm/112035273840 boards by any chance? I could ask @ if he would share his STM32F103 firmware prototype with you if you are interested.
I don't but I could source one. My Oshmon PCBs have been despatched but I haven't received any notice that my CJ125s are on the way.
User avatar
kb1gtt
contributor
contributor
Posts: 3711
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA

Re: CJ125 board

Post by kb1gtt »

Does it still get hot if the PWM is changed to 0% AKA is this a copper bridge or some kind of physical problem. Then does it got hot if PWM is 100%. This will tell is if it's an switching issue, or a DC issue.
Welcome to the friendlier side of internet crazy :)
puff
contributor
contributor
Posts: 2842
Joined: Mon Nov 11, 2013 11:28 am
Location: Moskau

Re: CJ125 board

Post by puff »

Or Q1 was chosen incorrectly? Or a faulty board?

I wonder why the guys who have a scope don't use it?

I wonder, whats' the role of DIAHG pin on CJ125 - am I right that it is pulling it high (closing Q1) when something goes wrong, overriding PWM from STM chip? Could it be that CJ is driving it up, but doesn't have enough power, so that Q1 is semi-closed (but then again, it doesn't explain 0A?)

What happens when you apply PWM to power up an inductive load? (don't we need a diode?)

Don't seen any resistance on the Q1's gate (which is present in other similar schemes)
Atomic-dm
contributor
contributor
Posts: 80
Joined: Thu Nov 21, 2013 2:43 pm
Location: Saint-Petersburg, Russia

Re: CJ125 board

Post by Atomic-dm »

maybe pwm signal very fast for this mosfet.

if you have more one mosfet, i recommend change mosfet. if it not help - do pwm more slower
puff
contributor
contributor
Posts: 2842
Joined: Mon Nov 11, 2013 11:28 am
Location: Moskau

Re: CJ125 board

Post by puff »

Have you used a clamping diode across the heater contacts?
Do you have a resistor between the mosfet gate and STM?
User avatar
AndreyB
Site Admin
Posts: 11173
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: CJ125 board

Post by AndreyB »

We are wasting time. We need one of the EE guys to have the setup in front of him.

Abricos to ship stuff to kb1gtt, or atomic-tm to assembled his board, or Rhinoman to assrmbled his board.
https://rusefi.com/s/howtocontribute
very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions
my skype is arro239
User avatar
kb1gtt
contributor
contributor
Posts: 3711
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA

Re: CJ125 board

Post by kb1gtt »

Hmmm, looks like 1kHz is the absolute maximum for this chip, and if you are at the maximum delays for some reason, then 333Hz is the absolute maximum. Even at 500Hz, your spending allot of time in the transition part of the pulse. So it might be to fast of a switching frequency. Also what is the duty? Is this 50% duty? You probably can't be very far from 50% at 500Hz. Might be worth trying something like 100Hz or below.

Also also, is the port configured correctly? AKA not drive high, then float. Is it driving both high and low? If it's driving hi and low, is it doing it with the weak pull ups, or the strong transistors?
Welcome to the friendlier side of internet crazy :)
User avatar
AndreyB
Site Admin
Posts: 11173
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: CJ125 board

Post by AndreyB »

I've just compiled hopefully a 125Hz version of that firmware, @ will probably try tomorrow.

Code: Select all

pwm.c
	TIM_Time_user.TIM_Period=255;  // russian: 1023 replaced with 255               
main.c 
        float maxU = 255;// russian: 1023 replaced with 255
https://rusefi.com/s/howtocontribute
very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions
my skype is arro239
Abricos
contributor
contributor
Posts: 798
Joined: Mon Aug 18, 2014 12:32 am
Location: Carteret, NJ 07008

Re: CJ125 board

Post by Abricos »

New information:

@ has a resistor divider on UR analog line with two 100K resistors. On our board we can remove C2 and install 1K resistor.

Also apparently PWM is much slower than we though: original firmware is just 11Hz, modified firmware is 44Hz. Could this be causing any problems? Should we play with 11Hz or 44Hz or 88Hz or 172Hz?
Attachments
2zamer_255.png
2zamer_255.png (4.74 KiB) Viewed 6821 times
zamer_original_hi_res.png
zamer_original_hi_res.png (4.52 KiB) Viewed 6821 times
Post Reply