For any add-on boards such as VR conditioners, optos and OEM interface boards
By ironmanisanemic
#36890
I have designed an external IGBT ignitor board that has 4 channels. It uses the ON Semi ISL9V2040D3ST. This seems to be a very robust chip. During testing I ran this very hard and didnt have a single failure.

I ran this using a signal generator directly into the board, and also using the speeduino v0.4.3 board feeding it a crank signal from my signal generator, and had zero issues either way.

Using a thermocouple and a function generator, i saw stablized temps of 33.9*C at a simulated RPM of 6000 with a dwell of 3.5mS. The chips themselves have a operating range of up to 175*C, so it is still well within its operating range. It could use stick on heat sinks to be on the safe side.

The IGBTs themselves are also logic level, so they can be driven off of 5v signals just like a smart coil.

Here is my project page for this board. There is no BOM as it is only the IGBTs listed above, but i have also attached the gerber files if you wish to have the board produced.

https://easyeda.com/ironmanisanemic/4-channel-ignitor

Also, If you see anywhere there can be any improvements please let me know. I am still very new at this and may have missed something obvious.

*EDIT*
I have a bunch of bare boards i can send out if anyone would like to test one. I only purchased enough IGBTs for 2 full boards and i am keeping both for current projects. So if you dont mind purchasing the IGBTs and doing some basic SMD soldering i will send you a board for whatever shipping costs. Send me a PM if you are interested. All i ask in return is you let me know how it works out and if you have any issues with the product.

Thanks!
Attachments
front.jpg
front.jpg (215.27 KiB) Viewed 7772 times
back.jpg
back.jpg (220.91 KiB) Viewed 7772 times
(66.78 KiB) Downloaded 266 times
IMG_20190804_132439.jpg
IMG_20190804_132439.jpg (1.54 MiB) Viewed 7773 times
IMG_20190804_130647.jpg
IMG_20190804_130647.jpg (1.67 MiB) Viewed 7773 times
User avatar
By PSIG
#36899
Nice! 8-) A very simple design, and wire strain-reliefs to boot. A couple thoughts are that stuff like this is typically sized for an appropriate available case or enclosure. Do you have one that fits this?

Your design is a much simpler version of this quad ignition driver @cx500tc designed, that can optionally have input resistors, output LED indicators, IGf/IDM (ignition failure/misfire feedback signal), etc. It is sized to 50x50mm for aluminum enclosures under $3, so the project is easy to finish. Being simpler, yours could be a smaller size that is common and inexpensive, and preferably aluminum for radiant cooling.

Finally, the TO-252 (DPak) IGBTs are generally limited to about 10 amps. Many COPs and high-performance CNP coils peak well above this. You might consider if this is a "universal" device, to allow TO-262 (D2Pak) or TO-263 (I2Pak) IGBTs for those hungry COPs that are trying to make-up for their small size. :lol: Also, test with all channels running under load, as they tend to heat each other, and more when enclosed.

Just trying to think of stuff. Hope that helps.

David
By ironmanisanemic
#36901
PSIG wrote: Mon Aug 05, 2019 12:40 am Nice! 8-) A very simple design, and wire strain-reliefs to boot. A couple thoughts are that stuff like this is typically sized for an appropriate available case or enclosure. Do you have one that fits this?


No i dont have a case in mind. I just tried to make it as small and compact as possible


Your design is a much simpler version of this quad ignition driver @cx500tc designed, that can optionally have input resistors, output LED indicators, IGf/IDM (ignition failure/misfire feedback signal), etc. It is sized to 50x50mm for aluminum enclosures under $3, so the project is easy to finish. Being simpler, yours could be a smaller size that is common and inexpensive, and preferably aluminum for radiant cooling.

I actually used that as inspiration for mine. I just wanted something very simple that did its job very well. Did the other board have a case spec'd already? I could try to fit mine to that.

Finally, the TO-252 (DPak) IGBTs are generally limited to about 10 amps. Many COPs and high-performance CNP coils peak well above this. You might consider if this is a "universal" device, to allow TO-262 (D2Pak) or TO-263 (I2Pak) IGBTs for those hungry COPs that are trying to make-up for their small size. :lol: Also, test with all channels running under load, as they tend to heat each other, and more when enclosed.

I am designing another variation for the TO-263 variation of the same chip. I am also adding mounting holes to add a extruded heat sink over the tops of the IGBTs. I will test this with all 4 channels going at once when i get some more free time. I was testing this with Ford wasted spark coil packs as this was the beefiest coils i have on hand. The whole bottom layer is battery voltage, and the whole top layer is the ground, so i believe the board has enough thermal mass to deal with the current.

Just trying to think of stuff. Hope that helps.

Thanks for the feedback, I always appreciate criticism positive or negative, its the best way to learn in my opinion.

David
Attachments
top.jpg
top.jpg (162.65 KiB) Viewed 7732 times
bottom.jpg
bottom.jpg (161.32 KiB) Viewed 7732 times
By fram
#36903
Finally, the TO-252 (DPak) IGBTs are generally limited to about 10 amps. Many COPs and high-performance CNP coils peak well above this. You might consider if this is a "universal" device, to allow TO-262 (D2Pak) or TO-263 (I2Pak) IGBTs for those hungry COPs that are trying to make-up for their small size. :lol: Also, test with all channels running under load, as they tend to heat each other, and more when enclosed.

I am designing another variation for the TO-263 variation of the same chip. I am also adding mounting holes to add a extruded heat sink over the tops of the IGBTs. I will test this with all 4 channels going at once when i get some more free time. I was testing this with Ford wasted spark coil packs as this was the beefiest coils i have on hand. The whole bottom layer is battery voltage, and the whole top layer is the ground, so i believe the board has enough thermal mass to deal with the current.
For IGBT design, you might probably considere a single side pcb with aluminium substrat. It's cheaper and could be a good choice because of the heatsink properties of this aluminium substrat.

Take care of the traces isolation, and try to have litz wire soldering pads as close a possible to the IGBT because of high current into the pcb. (Despite of the width of your traces, keep in mind pcb copper thickness is currently just 35µ, so the shortest trace is the less resistive!)

my 2 cents!
Franck ;)
Attachments
a.JPG
a.JPG (89.49 KiB) Viewed 7718 times
By JHolland
#36906
The 10A limitation is continuous current, the peak can be much higher, when calculating average power dissipation you need to remember to add the power dissipation during clamping. The datasheet doesn't state thermal resistances for each package but its always best to measure them on the actual PCB because they are never the same as the datasheet value anyway.
I would put provision for gate resistors to ensure that the IGBTs aren't ever directly connected to the processor output pin, SMD resistors don't take much space.
User avatar
By PSIG
#36913
I would usually differentiate between peak and continuous current; however this unit does not have pulsed current data specifically, and instead has inductive load limits to show why there is no pulsed or peak current limit. The way I am reading the datasheets :roll:, the graph below shows why it cannot use peak amps above continuous rating without failure based on level of inductance.

Just for reference, the common Ford ZZ coils are 4.7mH to show where the limits are in one real-world case that is not particularly powerful; if this unit is "universal". While most applications would allow higher peak current (e.g., I regularly drive little LEDs with no resistor and 12V PWM ;)), the inductance of ignition coils throws a different spin on the issue. Thoughts or corrections welcome.

David

ISL9V2040 dtasheet:
ISL9V2040_I-vs-L_limits.png
ISL9V2040_I-vs-L_limits.png (35.45 KiB) Viewed 7670 times
Last edited by PSIG on Tue Aug 06, 2019 2:24 pm, edited 2 times in total.
By JHolland
#36933
These are not high power devices, the mJ rating is a good indicator of that, I have used some 450mJ devices in a diesel application with some peculiar spark plugs and a high energy coil. I'll see if I still have the part number for those; I'm fairly sure that they weren't logic level but I don't see that as an issue, some proper buffering would be good, in order to protect the micro without slowing switching times.
By ironmanisanemic
#36949
Ok, so I have taken everyone's suggestions in mind, and have made some adjustments.

First, i went looking for a heat-sink, and i found one 60x60mm Heat-sink. With this heat-sink in mind, i sized the board to be 60mm by 60mm. This works out because of the larger package size of the TO-263AB. So now endurance and larger current spikes shouldn't be an issue.

I also eliminated the bottom power layer, and now coil power is routed separately. This lets both sides to provide current carrying capacity. Also this allowed me to space things better so there is more isolation between traces and ground plane.

I added gate resistors as well. As it was pointed out this is to protect the arduino from over-current. I sized it at 150ohms, this should allow about 30ma of max current draw in the event of a short.

It is still at the same project page linked to above, only now im up to version 0.3.

Let me know what you guys think.

On another note, what would be a good coil/set of coils to stress test this board with? I would like this to be as universal as possible, so i need to be able to do extensive testing.

Thanks!
Attachments
top.jpg
top.jpg (244.49 KiB) Viewed 7544 times
bottom.jpg
bottom.jpg (245.14 KiB) Viewed 7544 times
By JHolland
#36952
ironmanisanemic wrote: Wed Aug 07, 2019 2:30 am Just came across this IGBT, from what i can see it can handle much higher currents. According to the Datasheet it can take sustained 44 amps of collector current. Does it seem like it may be more suited for a universal application?

https://www.digikey.com/product-detail/ ... ND/4963437

You can put 41A through it at a case temperature of 25C but if you do the temperature will soon rise so the 25.6A value is more realistic. What you also need to watch out for is the SCIS value which is the energy the device can handle when clamping the flyback, in this case 300mJ ,;this energy is a function of inductance and current, the plot shown previously is just a different way of showing that. If its a fast high energy coil that relies on the dead time to limit the current then you have to allow for the possibility that the dwell time could be excessive if there is a fault or a programming error. In one application I did the coil was dwell limited to 15A with a diagnostic shut-off at 25A so the IGBT had to be rated to handle the higher value. Ultimately you probably won't be able to satisfy everyone so its not really worth putting too much effort into it.

The gate resistor is not so much to protect the Arduino from short circuit but from the gate voltage during flyback. If you study the datasheet you will see that the clamp consists of two zener diodes that turn the IGBT on when the flyback voltage exceeds a certain value, the Gate-Emitter breakdown voltage is rated as 14V typical, so your 150 Ohms resistor is only going to limit the current to around 94mA which is far too high, the max. breakdown voltage is not quoted so you will need to allow some margin for that.

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?