Technical Ignition Coil Secondary Spark Energy Comparison (N54/N55, B58 and PR)

[V8bait]

@V8bait Confirmations of my thoughts and I'm quite surprised we're not seeing more IGBT failures. The higher inductance of the PR coil is contributing to very high flyback voltages of over 430 V. This exceeds the IGBT's rated collector to emitter breakdown voltage of 390 V. In the stock N54 coil & B58 coil the IGBT is able to quickly clamp the flyback but with the PR coil that isn't the case. If you look very closely at the falling edge of the light blue (IGBT gate voltage) you can see a small step near the bottom of the edge on the PR coil, that appears to be leakage (which make sense since we're above the rated breakdown voltage).
View attachment 49012

Stock N54 Bosch flyback voltage: 344 V (within maximums)
View attachment 49014

Eldor B58 Flyback voltage: 336 V (within maximums)
View attachment 49013

I will be testing more coils tomorrow and hope to have those results added into this post early next week.

Hmm, I think I see the issue then. The factory system has a capacitor for a couple hundred volts in the primary system, it's on the valve cover, I forget the nF/uF rating but that isolates the igbt from those voltages and helps pump the primary/secondary up, not in current (because it's a cap) but in voltage, I think it's technically called the condenser and that's literally the point of it. It captures breakdown voltage to increase primary (and subsequently secondary) potential, IE use all those nice mJ to jump a hot gap . To get 30-50kV you need the condenser to raise primary voltage into the hundreds momentarily to raise secondary to 30-50kV, so those flybacks make sense and I'd expect PR to make higher peak voltage as such. It all kinda fits in my head now.

I don't think the condenser functions properly testing output through zenners since the voltage is low and load is increased, but should work fine in free air tests.

edit- on second thought that cap is just for driving primary voltage up and doesn't see the igbt.

 

[Blue Lightning]

Without a condenser you're using the transistor (IGBT) limit as a pseudo-capacitor to drive voltage up and if you exceed its rated breakdown you're gonna have a bad time.

The question here is, why is anyone removing the condenser then ??.

What you said makes perfect sense. I built a wind turbine and fitted a crowbar dump load system on a voltage limiter to prevent overspending. I got horrendous flyback voltages from the turbine coil windings and kept destroying the IGBT's in the controller. Solution was to fin a humongous 82,000uF 100v Cap....... Problem solved no IGBT failures after that !.

 

[bahn]

Hmm, I think I see the issue then. The factory system has a capacitor for 500 volts in the primary system, it's on the valve cover, I forget the nF/uF rating but that isolates the igbt from those voltages and helps pump the primary/secondary up, not in current (because it's a cap) but in voltage, I think it's technically called the condenser and that's literally the point of it. It captures breakdown voltage to increase primary (and subsequently secondary) potential, IE use all those nice mJ to jump a hot gap . To get 30-50kV you need the condenser to raise primary voltage into the hundreds momentarily to raise secondary to 30-50kV, so those flybacks make sense and I'd expect PR to make higher peak voltage as such. It all kinda fits in my head now.

I don't think the condenser functions properly testing output through zenners since the voltage is low and load is increased, but should work fine in free air tests.

These flyback tests are free air tests and are not utilizing the zener diode load. I have one of those capacitors in hand from a spare engine harness. That capacitor is 0.1uF and is rated for only 100V (PN: 7551338 V3). The capacitor is between BAT+ and GND (not DME switched ground) and this tells me that the capacitor exists to mitigate RF frequencies by shorting it to ground to prevent radio interference.


See here from BMW:

I went ahead and put that capacitor into the circuit and retested flyback voltage with the PR coil. As expected (since the cap is between BAT+ and GND) there was no change to flyback voltage seen at the IGBT over the switched ground. Peak 424 V with a 11.5 uS duration.

As always I appreciate the feedback, any other ideas? I think this is simply due to the higher inductance of the primary side of the coil as flyback voltage is directly tied to the inductance (V= L* dI/dt)

 

[Traf]

Have you got around to testing the R8 coils ?

 

[nyt]

@V8bait any comments on the potential flyback issue with the pr coils?

 

[bahn]

Have you got around to testing the R8 coils ?

Yes I've tested them but I was waiting on an Eldor S55 coil to come in for testing as well so I could post the results for 3 new coils. S54, Eldor S55 and R8. To tide you over while I finish testing the Eldor S55 the R8 coil at saturation dwell (2.5ms) @ 14V made 126 mJ of energy with a spark duration of 1.70 ms.

 

[Traf]

Yes I've tested them but I was waiting on an Eldor S55 coil to come in for testing as well so I could post the results for 3 new coils. S54, Eldor S55 and R8. To tide you over while I finish testing the Eldor S55 the R8 coil at saturation dwell (2.5ms) @ 14V made 126 mJ of energy with a spark duration of 1.70 ms.

Thanks for the info ! I don't regret selling my R8 kit then, i'll hop on to the B58 ones at some point.

 

[V8bait]

@V8bait any comments on the potential flyback issue with the pr coils?

I think it's a good consideration to have but I wouldn't worry much about it. He's right about the cap it's primarily for interference. I seem to recall some diodes on the DME that protect the IGBT, or else there would be more failures if it was a concern, but it's nice to match things up with specs. B58 checks some real nice boxes overall, I'm still just a little curious as to how they manage the energy they do with lower inductance and lower current, but overall it's a solid option. I think the B58 and PR are definitely the best options available now, I'd probably choose based on if you want to relocate them for ST or not.

 

[nyt]

I think it's a good consideration to have but I wouldn't worry much about it. He's right about the cap it's primarily for interference. I seem to recall some diodes on the DME that protect the IGBT, or else there would be more failures if it was a concern, but it's nice to match things up with specs. B58 checks some real nice boxes overall, I'm still just a little curious as to how they manage the energy they do with lower inductance and lower current, but overall it's a solid option. I think the B58 and PR are definitely the best options available now, I'd probably choose based on if you want to relocate them for ST or not.

I've been using the PR coils and doubt I'll change them. The dwell time is high but it pays off with the higher initial voltage.

 

[bahn]

I think it's a good consideration to have but I wouldn't worry much about it. He's right about the cap it's primarily for interference. I seem to recall some diodes on the DME that protect the IGBT, or else there would be more failures if it was a concern, but it's nice to match things up with specs. B58 checks some real nice boxes overall, I'm still just a little curious as to how they manage the energy they do with lower inductance and lower current, but overall it's a solid option. I think the B58 and PR are definitely the best options available now, I'd probably choose based on if you want to relocate them for ST or not.

Hey Justin, I'm on my phone right now so forgive the short reply. The B58 coils need more current since they're lower inductance. At 3.2ms its about 14 amps. This current is within the IGBT's rated range (and so is the flyback voltage generated). Another difference between coils is the material of the coils core and the distance between the primary and secondary windings. The closer they are the better the magnetic coupling and higher the flux. Of course to test that we would need to cut all of the coils open.

 

[V8bait]

I've been using the PR coils and doubt I'll change them. The dwell time is high but it pays off with the higher initial voltage.

Indeed. The numbers here really make ther B58 coil look great, I can't deny that. I'm really surprised to be honest, before I ditched bmw coils I tested this style to see if it was worth looking into and honestly, I saw little difference. Sure, it pulled a little more current and saturation was a little higher than n54 coils, but there were no tables available and the peak voltage wasn't anything like R8 or the mercury marine designs. I wrote them off.

I really want to applaud this testing here, it's very thorough. As somebody intimately associated with the PR setup (even if I'm not getting money thank you very much) I think I'll render my opinion-

B58 coils look like a major upgrade over N54. This testing not only confirms what I saw in terms of B58 coil reliability vs N54 but shows these things put out real power. Compared to PR, man it's tight. The PR in my early tests outshine the B58 on peak voltage just like the R8 did. At high cylinder pressures and moderate plug gap that's important- energy is only useful if you can span a gap. That being said, the B58 barely tickle the DME for load. Is that important? I'd say meh after torturing the PR coils, but it's definitely a thing. Win B58 there.

Straight up I think the B58 design will also address an issue with PR in regards to fitment- i mean I've never had an issue installing PR, but plenty have had difficulty getting the wires to engage. These shouldn't be an issue. Honestly, I think the question comes down to application. PR are barely more powerful but achieve higher peak voltage, so basically max effort on stock hardware, and some people want that. B58 are a really solid upgrade outside of that and if relocation isn't needed. Excellent work sir!

 

[Scorpstan]

Yes I've tested them but I was waiting on an Eldor S55 coil to come in for testing as well so I could post the results for 3 new coils. S54, Eldor S55 and R8. To tide you over while I finish testing the Eldor S55 the R8 coil at saturation dwell (2.5ms) @ 14V made 126 mJ of energy with a spark duration of 1.70 ms.

do you tested the r8 coils like they are installed?.....with the extra power lines?

 

[Lightning]

Also still wondering about the details of the R8 test results. From the little piece of information that we got, they are promising and might not perform close to the B58's. But this is just speculation with the little info we got.

 

[Traf]

Which dwell is set by the MHD tick option ?

 

[bahn]

Which dwell is set by the MHD tick option ?

The dwell table is not static as higher energy is not needed at idle and will instead reduce spark plug life. To find a good balance we ramp up dwell time above 2500 RPM peaking at 3.2 ms.

 

[seb.apprenti]

do we have the result of the eldor s55 coil test?

 

[bahn]

do we have the result of the eldor s55 coil test?

Just uploaded the data for Bosch S54, Audi R8 and the Eldor S55 coils. The Eldor S55 coil tested pretty much identical to the Delphi S55 coils. The Audi R8 coil performed well but I was actually a little disappointed with "only" 126 mJ, the B58 and PR coils are definitely an upgrade over the R8 coil. Sorry for the long delay in getting this data on here. We've had an overwhelming response from the community and have been focusing on shipping B58 kits out. I do believe I need to find another way of sharing this data as the original post is starting to become a flood of images as the number of coils tested increases.

Thank you everyone for your support and for sending me more coils to test. If you have a coil you would like tested send me a PM!

 

[Traf]

If i'm not mistaken the R8 would be using the IGBTs just as a trigger so not much current passing through right ?
Compared to the others it would keep the IGBTs cooler ?

 

[bahn]

If i'm not mistaken the R8 would be using the IGBTs just as a trigger so not much current passing through right ?
Compared to the others it would keep the IGBTs cooler ?

Correct as the R8 coil has its own IGBT internal. That said all coils tested except the PR coils are within the design limitations of the IGBT's used in the MSD8X DME's.

 

[Melilla1]

Hello I am interested in the B58 Kit for N54. I cannot order for Spain