Coil pack testing
-
- Former Member
- Posts: 38
- Joined: Tue Jan 23, 2018 12:04 pm
- RX-8: 231
- Colour: Velocity Red
- Has thanked: 0
- Been thanked: 7 times
Coil pack testing
Hi all
I have created a coil pack setup in my garage and decided to do some testing of packs that came with my car. I have no idea if any packs are genuine, however 3 of them are marked as N3H1-18-100C. I had a NGK spark plug and a decent spark lead.
The setup is very simple, I have a current monitor (ACS712 20 AMPS) on the 12V input side to the coil pack to monitor current pulled by the primary current during dwell. I used a FPGA development kit to drive the trigger signal, the board was lying around and is complete over kill, but had some buttons so I can change the dwell period easily. I bought a capacitive pickup probe from hantek and a simple 100Mhz ossiliscope, then after a few hours writing some VHDL code and soldering wires I had a tester.
So onto the important bit, the results.
The coils marked as N3H1-18-100C produce a much longer burn time and seem to saturate with a much larger dwell than the non-marked packs. When using the exact same dwell period, the N3H1-18-100C packs have a much longer burn. I suspect this shows that these unmarked packs are worse for the engine leaving less fuel burned, but I am no expert, maybe someone else can share that knowledge. Anyway I am going to carry on testing.
I have created a coil pack setup in my garage and decided to do some testing of packs that came with my car. I have no idea if any packs are genuine, however 3 of them are marked as N3H1-18-100C. I had a NGK spark plug and a decent spark lead.
The setup is very simple, I have a current monitor (ACS712 20 AMPS) on the 12V input side to the coil pack to monitor current pulled by the primary current during dwell. I used a FPGA development kit to drive the trigger signal, the board was lying around and is complete over kill, but had some buttons so I can change the dwell period easily. I bought a capacitive pickup probe from hantek and a simple 100Mhz ossiliscope, then after a few hours writing some VHDL code and soldering wires I had a tester.
So onto the important bit, the results.
The coils marked as N3H1-18-100C produce a much longer burn time and seem to saturate with a much larger dwell than the non-marked packs. When using the exact same dwell period, the N3H1-18-100C packs have a much longer burn. I suspect this shows that these unmarked packs are worse for the engine leaving less fuel burned, but I am no expert, maybe someone else can share that knowledge. Anyway I am going to carry on testing.
You do not have the required permissions to view the files attached to this post.
- These users thanked the author rb_rx_gtr for the post (total 3):
- warpc0il (Sun Aug 04, 2019 5:28 pm) • 13Black (Sun Aug 04, 2019 6:34 pm) • Cousin63 (Thu Sep 19, 2019 11:24 am)
-
- Spin Doctor
- Posts: 32537
- Joined: Sat Sep 06, 2008 10:56 pm
- RX-8: 231
- Colour: Lightning Yellow
- Location: Groomsport, Co Down, NI
- Has thanked: 444 times
- Been thanked: 2080 times
Re: Coil pack testing
This could prove very interesting.
Are you up for people sending you their (suspect ) coil packs for testing and comparison?
Sent from my SM-G800F using Tapatalk
Are you up for people sending you their (suspect ) coil packs for testing and comparison?
Sent from my SM-G800F using Tapatalk
-
- Forum Moderator
- Posts: 9053
- Joined: Fri Aug 04, 2017 4:56 pm
- RX-8: Turbo
- Colour: Nordic Green
- Location: cornwall
- Has thanked: 97 times
- Been thanked: 539 times
Re: Coil pack testing
id be interested to see you stress test a d585 with that rig, against a genuine, fake, and rrp coil..
i suspect the harder you push a d585 the weaker the spark gets...
after a fair amount of reserch on my part, along with first hand and second hand experiece, i dont think they stack up against even mazda gen coils.., the maths just doesnt work out....
i suspect the harder you push a d585 the weaker the spark gets...
after a fair amount of reserch on my part, along with first hand and second hand experiece, i dont think they stack up against even mazda gen coils.., the maths just doesnt work out....
-
- Former Member
- Posts: 1968
- Joined: Thu Mar 01, 2018 12:47 pm
- Location: south east London
- Has thanked: 38 times
- Been thanked: 116 times
Re: Coil pack testing
Mazda c coil is not a bad coil Clive run his race car on them.
LEXUS LS430 4.3 v8 family car
Done exhaust and work wheels
Mx5 nd soul red Mike daily and work horse
Jessy’s Jaguar xkr 4.2 supercharged for fun family days
Done exhaust and work wheels
Mx5 nd soul red Mike daily and work horse
Jessy’s Jaguar xkr 4.2 supercharged for fun family days
-
- Forum Moderator
- Posts: 9053
- Joined: Fri Aug 04, 2017 4:56 pm
- RX-8: Turbo
- Colour: Nordic Green
- Location: cornwall
- Has thanked: 97 times
- Been thanked: 539 times
Re: Coil pack testing
thats kinda my point lol. d585s should perform at least as well, but on paper, they cant.
-
- Former Member
- Posts: 4726
- Joined: Sun Jan 10, 2016 10:14 pm
- RX-8: 231
- Colour: Velocity Red
- Location: Leicestershire
- Has thanked: 265 times
- Been thanked: 359 times
Re: Coil pack testing
Why would a D585 perform WORSE when you charge it for longer (I.e to the dwell they were designed to run at) ?
Edit: misunderstood
Edit: misunderstood
Last edited by 13Black on Mon Aug 05, 2019 9:47 am, edited 1 time in total.
-= Rotormelon Money Pit Build of Tasteless Questionable Bodgery =-
Experimelon 183whp rebuild | Rotary Revs D585s | PZ Suspension | Hawk & Yellowstuff pads | Extra steering lock | Custom exhaust | Added rearward flames of glory | Unnecessarily repurposed ball gags-= Clicky click for build thread =-
-
- Spin Doctor
- Posts: 32537
- Joined: Sat Sep 06, 2008 10:56 pm
- RX-8: 231
- Colour: Lightning Yellow
- Location: Groomsport, Co Down, NI
- Has thanked: 444 times
- Been thanked: 2080 times
Re: Coil pack testing
Duty cycle starts to catch up with itself as the revs rise
- These users thanked the author warpc0il for the post:
- qwakers (Sun Aug 04, 2019 6:49 pm)
-
- Former Member
- Posts: 1191
- Joined: Sun Mar 11, 2012 11:58 pm
- Colour: Custom
- Has thanked: 30 times
- Been thanked: 92 times
Re: Coil pack testing
As above, you can't sustain full dwell up to the redline because you simply don't have the time available. Having said that you don't necessarily need it as combustion pressure is dropping too.
-
- Forum Moderator
- Posts: 9053
- Joined: Fri Aug 04, 2017 4:56 pm
- RX-8: Turbo
- Colour: Nordic Green
- Location: cornwall
- Has thanked: 97 times
- Been thanked: 539 times
Re: Coil pack testing
this, they get to (and beyond) 100% rated duty before you're at 9k rpm, so if you spend most of your time flat out (ie on track) they self destruct.
-
- Former Member
- Posts: 4726
- Joined: Sun Jan 10, 2016 10:14 pm
- RX-8: 231
- Colour: Velocity Red
- Location: Leicestershire
- Has thanked: 265 times
- Been thanked: 359 times
Re: Coil pack testing
Oh, pushed as in RPM!
Sorry I thought you meant pushed regarding charge times.
See my graph on Phil's Mazdaedit thread, you can safely charge D585s for twice as long as you would a Mazda coil at redline - yes it's less than what you get in the midrange, but like RRB pointed out your VE has dropped also so the demand doesn't require an immense spark, surely.
Also, should we include the discharge time in its duty cycle calculations?
I would have thought that it's the charge time that matters most in terms of stress and longevity. Phil? OP?
But yus, Mazda C coils are still going to be great I expect (as long as they've not been used for 50k).
Sorry I thought you meant pushed regarding charge times.
See my graph on Phil's Mazdaedit thread, you can safely charge D585s for twice as long as you would a Mazda coil at redline - yes it's less than what you get in the midrange, but like RRB pointed out your VE has dropped also so the demand doesn't require an immense spark, surely.
Also, should we include the discharge time in its duty cycle calculations?
I would have thought that it's the charge time that matters most in terms of stress and longevity. Phil? OP?
But yus, Mazda C coils are still going to be great I expect (as long as they've not been used for 50k).
-= Rotormelon Money Pit Build of Tasteless Questionable Bodgery =-
Experimelon 183whp rebuild | Rotary Revs D585s | PZ Suspension | Hawk & Yellowstuff pads | Extra steering lock | Custom exhaust | Added rearward flames of glory | Unnecessarily repurposed ball gags-= Clicky click for build thread =-
-
- Former Trader
- Posts: 1616
- Joined: Fri Jan 27, 2017 8:06 pm
- RX-8: Supercharged
- Colour: Brilliant Black
- Location: Rugby and Stonham Aspal
- Has thanked: 46 times
- Been thanked: 153 times
Re: Coil pack testing
Ref: Duty Cycle.
This thread might interest some: https://www.rx8ownersclub.co.uk/forum/v ... 05&t=79271
This thread might interest some: https://www.rx8ownersclub.co.uk/forum/v ... 05&t=79271
Engine Rebuilds, ECU mapping, OEM & Aftermarket products.
http://Www.ryanrotaryperformance.com
Sales@ryanrotaryperformance.com
Www.facebook.com/ryanrotaryperformance
Carl
http://Www.ryanrotaryperformance.com
Sales@ryanrotaryperformance.com
Www.facebook.com/ryanrotaryperformance
Carl
-
- Former Member
- Posts: 38
- Joined: Tue Jan 23, 2018 12:04 pm
- RX-8: 231
- Colour: Velocity Red
- Has thanked: 0
- Been thanked: 7 times
Re: Coil pack testing
I'd be interested in testing some other packs. First I need to establish exactly what results are important. For example how much does the spark burn time matter? This is the largest difference I have observed so far followed by saturation limit.
From the 2 types of pack I have, the packs marked C (not sure if they are genuine) seem to saturate around 2.7ms of dwell, anymore is just wasted as heat. The clone packs are just about 2ms. I suspect there is fewer winding's on the clone coils.
I am using a car battery to power the system, so take that into account if you compare these results to anyone else's.
From the 2 types of pack I have, the packs marked C (not sure if they are genuine) seem to saturate around 2.7ms of dwell, anymore is just wasted as heat. The clone packs are just about 2ms. I suspect there is fewer winding's on the clone coils.
I am using a car battery to power the system, so take that into account if you compare these results to anyone else's.
-
- Former Member
- Posts: 38
- Joined: Tue Jan 23, 2018 12:04 pm
- RX-8: 231
- Colour: Velocity Red
- Has thanked: 0
- Been thanked: 7 times
Re: Coil pack testing
Times between ignition cycles that I calculated and use in my testing.
1000 rpm = 60ms
2000 rpm = 30ms
3000 rpm = 20ms
4000 rpm = 15ms
5000 rpm = 12ms
6000 rpm = 10ms
7000 rpm = 8.57ms
8000 rpm = 7.5ms
9000 rpm = 6.6666ms
10000 rpm = 6ms
At 1000rpm one face of the rotor will have passed over the spark plugs in 60 ms which means the output shaft will have completed one full rotation. Working backwards, 1000rpm is the equivalent to 16.66666 revs per second. 1/16.6666 = 60ms.
If you spot a problem, please let me know
1000 rpm = 60ms
2000 rpm = 30ms
3000 rpm = 20ms
4000 rpm = 15ms
5000 rpm = 12ms
6000 rpm = 10ms
7000 rpm = 8.57ms
8000 rpm = 7.5ms
9000 rpm = 6.6666ms
10000 rpm = 6ms
At 1000rpm one face of the rotor will have passed over the spark plugs in 60 ms which means the output shaft will have completed one full rotation. Working backwards, 1000rpm is the equivalent to 16.66666 revs per second. 1/16.6666 = 60ms.
If you spot a problem, please let me know
-
- Former Member
- Posts: 5849
- Joined: Mon Apr 11, 2016 1:26 pm
- RX-8: 231
- Colour: Custom
- Location: West Sussex
- Has thanked: 54 times
- Been thanked: 367 times
Re: Coil pack testing
I think the following are usually of interest when measuring coils:
1) Volts in (coil dwell time is always voltage compensated, so the input voltage is critical).
2) Dwell time to saturation.
3) Energy out (basically voltage out x current out x burn time). This should ideally be measured as the integral of the power over the duration of the burn time.
4) Burn time itself (this is added to the dwell time to establish the duty cycle)
5) Voltage out (This needs to be sufficient to establish a spark in the most hostile conditions).
The best coil has a short dwell time, a high energy out, a long burn time, and 'sufficient' voltage (which depends on your engine and spark plug e.g. high compression ratios, turbos, and large spark plug gaps all require higher voltage out).
It would be a useful addition if you could drive your test rig from a stabilised adjustable voltage supply, and test the coils from around 6v up to around 15v. You will then have everything you need to define a full dwell table.
I should also say, great effort. I'd love to have access to a test setup like this. And I'd love to see the results for all the usual coils that get fitted to our motors. The perfect list might be:
- Random fakes
- First Mazda spec (before spec A)
- Mazda spec A
- Mazda spec B
- Mazda spec C
- D585 (RR version & BHR version)
- D585R
- RRP Mark 11
- MSD (as sold by RRP)
- AEM smart coils
I know that this would be a herculean task, but it would be an awesome database for us.
1) Volts in (coil dwell time is always voltage compensated, so the input voltage is critical).
2) Dwell time to saturation.
3) Energy out (basically voltage out x current out x burn time). This should ideally be measured as the integral of the power over the duration of the burn time.
4) Burn time itself (this is added to the dwell time to establish the duty cycle)
5) Voltage out (This needs to be sufficient to establish a spark in the most hostile conditions).
The best coil has a short dwell time, a high energy out, a long burn time, and 'sufficient' voltage (which depends on your engine and spark plug e.g. high compression ratios, turbos, and large spark plug gaps all require higher voltage out).
It would be a useful addition if you could drive your test rig from a stabilised adjustable voltage supply, and test the coils from around 6v up to around 15v. You will then have everything you need to define a full dwell table.
I should also say, great effort. I'd love to have access to a test setup like this. And I'd love to see the results for all the usual coils that get fitted to our motors. The perfect list might be:
- Random fakes
- First Mazda spec (before spec A)
- Mazda spec A
- Mazda spec B
- Mazda spec C
- D585 (RR version & BHR version)
- D585R
- RRP Mark 11
- MSD (as sold by RRP)
- AEM smart coils
I know that this would be a herculean task, but it would be an awesome database for us.
-
- Forum Moderator
- Posts: 9053
- Joined: Fri Aug 04, 2017 4:56 pm
- RX-8: Turbo
- Colour: Nordic Green
- Location: cornwall
- Has thanked: 97 times
- Been thanked: 539 times
-
- Spin Doctor
- Posts: 32537
- Joined: Sat Sep 06, 2008 10:56 pm
- RX-8: 231
- Colour: Lightning Yellow
- Location: Groomsport, Co Down, NI
- Has thanked: 444 times
- Been thanked: 2080 times
Re: Coil pack testing
Powering from a 12V car battery doesn't replicate any real-world scenario.
When starting the engine, only the battery is providing power but it is also running the starter, so the voltage could easily be down at the sub 8V mark, though the engine revs are also low @ 180-350 rpm.
Once the engine is running then the "12V" system is actually soon up at 14.5V, which is what the coils should see on their input most of the time.
-
- Former Member
- Posts: 38
- Joined: Tue Jan 23, 2018 12:04 pm
- RX-8: 231
- Colour: Velocity Red
- Has thanked: 0
- Been thanked: 7 times
Re: Coil pack testing
The battery was available and voltage provided for information only if anyone is comparing these results to others on the web/forum.
Fully understand the voltage changes. Having a fixed reference when comparing the coil packs is important. I do not currently posses a bench power supply that can handle up to 20 amps hence a battery.
Fully understand the voltage changes. Having a fixed reference when comparing the coil packs is important. I do not currently posses a bench power supply that can handle up to 20 amps hence a battery.
-
- Former Member
- Posts: 4726
- Joined: Sun Jan 10, 2016 10:14 pm
- RX-8: 231
- Colour: Velocity Red
- Location: Leicestershire
- Has thanked: 265 times
- Been thanked: 359 times
Re: Coil pack testing
Park your car in your kitchen and hook the coils up to the alternator
-= Rotormelon Money Pit Build of Tasteless Questionable Bodgery =-
Experimelon 183whp rebuild | Rotary Revs D585s | PZ Suspension | Hawk & Yellowstuff pads | Extra steering lock | Custom exhaust | Added rearward flames of glory | Unnecessarily repurposed ball gags-= Clicky click for build thread =-
-
- Former Member
- Posts: 38
- Joined: Tue Jan 23, 2018 12:04 pm
- RX-8: 231
- Colour: Velocity Red
- Has thanked: 0
- Been thanked: 7 times
-
- Committee Member
- Posts: 5275
- Joined: Sat Apr 13, 2019 6:40 pm
- RX-8: 231
- Colour: Custom
- Location: Cider Country
- Has thanked: 11 times
- Been thanked: 280 times
Re: Coil pack testing
Just remove a wall into the kitchen first tell her its for remodelling purposes Im sure she will be fine with it.
Club Treasurer
Kentaro The Red - Sadly fallen...
Star Platinum - Work in progress
For whats pending... See here: viewtopic.php?f=204&t=86071
Kentaro The Red - Sadly fallen...
Star Platinum - Work in progress
For whats pending... See here: viewtopic.php?f=204&t=86071
-
- Former Member
- Posts: 38
- Joined: Tue Jan 23, 2018 12:04 pm
- RX-8: 231
- Colour: Velocity Red
- Has thanked: 0
- Been thanked: 7 times
Re: Coil pack testing
Some results. I wound up the pulse rate to 10000rpm. The C packs get very warm where the rectangular block of plastic protrudes near the connector. The clones get a lot hotter, as in can't touch.
Suspect the clones have fewer turns and a higher impedance mosfet for coil drive.
Suspect the clones have fewer turns and a higher impedance mosfet for coil drive.
-
- Former Member
- Posts: 38
- Joined: Tue Jan 23, 2018 12:04 pm
- RX-8: 231
- Colour: Velocity Red
- Has thanked: 0
- Been thanked: 7 times
Re: Coil pack testing
Okay I am back and with some results to share after improvements to my setup so it is very consistent.
I have tested 5 coil packs of various types and documented the results anonymously for now. Let's have some fun with this and see if you can order the packs in terms of quality/performance.
I have tested 5 coil packs of various types and documented the results anonymously for now. Let's have some fun with this and see if you can order the packs in terms of quality/performance.
You do not have the required permissions to view the files attached to this post.
-
- Former Member
- Posts: 38
- Joined: Tue Jan 23, 2018 12:04 pm
- RX-8: 231
- Colour: Velocity Red
- Has thanked: 0
- Been thanked: 7 times
Re: Coil pack testing
No takers? Guess this is getting too obscure.
There is 1 failing coil pack in there which I was quite surprised with as the car did run with some very big backfire on gear change.
There is 1 failing coil pack in there which I was quite surprised with as the car did run with some very big backfire on gear change.
-
- Former Member
- Posts: 5849
- Joined: Mon Apr 11, 2016 1:26 pm
- RX-8: 231
- Colour: Custom
- Location: West Sussex
- Has thanked: 54 times
- Been thanked: 367 times
Re: Coil pack testing
The results are interesting, and thanks for doing the test and putting the results up. I'd just like to see which coil is which, to turn it into really useful info. Sorry to be a killjoy!