deezums

iTrader: (3)

Beefy fets have lots of gate capacitance typically, right? Are you using a gate driver, or direct off a micro pin?

Leafy

iTrader: (1)

For 15x11s I think custom is your only way. I had Keizers, I know someone got Bogarts done. I forget what everyone else has. I really doubt a 275 street tire is as wide as a 275 hoosier so you're probably fine with 15x10s. To fit them you need to do the following, cut fenders and roll/hammer rear fender inner lining out/up, hammer front wheel well pinch weld flat, trim NB style shock mount flush where the tire hits or use flat billet style mounts, cut front bumper to fit. Hammer in wheel well where required. Adjust bump stop length/ I suggest you remove the spring from the shock but leave teh bump stop in while you figure this out so you figure out how much longer you need to make your bump stops once you get tired of hammering and trimming.

patsmx5

iTrader: (16)

Using a Beefy gate driver, or, what I thought was beefy. But something is off, cause the switching speed is measuring about 250nS to drive the beefy FET's gate to 12V. Going to redo my math, look at the datasheets again.

For now, I got it doing sweeps from 0-100-0 with a spal fan, and two 12V ebay fans. Ordered a current meter, will be here tomorrow.

My Kill-A-What meter shows I'm pulling 530W from the 120V wire that feeds my SMPS. If it's 90% efficient, that's 477W to the fans. I'm measuring 14.59V at full load with my Fluke, so that works out to 32.69 Amps into the fans.

I've been looking for a supply that will do 15V and all the amps, so far not finding much that isn't a thousand dollars. I did find a 12V one that will do lots of amps, but it isn't adjustable. Would really like to do all my testing at 15V.

patsmx5

iTrader: (16)

Found this guy:
Says it's adjustable to 15V and rated to 50A. So technically will work, though I would like to have some headroom as I want my fan controller to be rated to 50A, but tested higher. The gauge on the front goes to 16V and 64A so it may have some headroom.

patsmx5

iTrader: (16)

Bought that 15V 50A supply. It has pretty good reviews, and it will be good to have for testing this now, but also for future projects. I've had a few other things where I needed to run high amp 12V things, so I expect I'll use this from time to time in the future.

Been running this ~32A load doing sweeps for a total of 5 hours so far without issue.

Regarding the V/ns issue with the gate driver, I think my return path to ground was too skimpy. Next board, I'm going to beef that up, see if it helps. Also moved the bypass cap a bit closer to the driver, and ran larger traces on the power wires to get low voltage drop in/out of the gate driver.

patsmx5

iTrader: (16)

Ordered a set of 245/40/15 RS4's. Looking into wheels/lugs/rings RN, thinking the 15x10 deckagram wheels. Looks like that's about it for 15x10 wheel options?

patsmx5

iTrader: (16)

For the drag racing, I'm considering buying these.

https://www.jegs.com/i/Hoosier/522/18810/10002/-1

Those are 26.2" tall, 10.5" tread width, 11.7 section width. Drag Radial.

One reason to run these over a slick is so I can skip tubes. Also not sure if I can run a tube tire without a bead lock.

Not to mention, that's a really beefy tire! I think it would look awesome.

patsmx5

iTrader: (16)

Bought the drag radials above, and 6 of the 15x10 deckagram wheels, and lugs and valve stems. Wallet hurts! And not sure how much work it's going to take to get these tires to fit, but if I can get them to fit, it will look awesome. They are almost a foot wide!

Newaza

Get that bad boy leaving at 25+ psi off the tbrake with the tires above and that thing should fly.. I'm jealous!

patsmx5

iTrader: (16)

That's the plan! Wanna see what I can do, I always thought the car would go 6.7 in the 1/8th or faster at 30 PSI with a better launch.

I went out and looked at the rear wheel well areas just now. Looks like making these fit is gonna be some work.

patsmx5

iTrader: (16)

Update time! Videos of fan testing at bottom!

Got the 245 tires and the 15x10 wheels in. Dang, these wheels are huge. No idea how these are supposed to fit on a miata. I must say, I like the look of these in person more so than in the pictures. Looks like cleaning them is going to suck though, lots of detail in the design.

The 245 tires look good too. Being a 200 tread wear, they aren't as soft as the radials I'm used to. But they are wide, and have tread, and should make a great street tire. Hopefully something I can drive in the rain safely.

Also the ARP studs came in and valve stems. Just waiting on the Hoosiers now, should be here Wednesday.



Fan controller update too!

The 50A 8-15V power supply came in.

I did some testing running the PCB at 30A, PWM controlling 3 fans. Had it running at 4kHz, doing 0-100-0 PWM sweeps over a 30 second cycle. Did that for 6 hours without issue. However, as expected, running it at high duty cycles, the mosfet gets warmer and the flyback diode too.

So I got a new SPAL fan in, and wired it up with the new 50A power supply. Now I can apply a 50-51A load at 15V. This works out to 750 Watts of power! The power supply has a bit of headroom, it appears to drop out a few amps after 50A, but 50, 51, 52, it can maintain 15.00 V.


And it worked... Sort of. It did work, but things were getting hot. So I slowed down the PWM ramp to take 10 min to do a 0-100-0 sweep. Then measured the temp of the mosfet and diode. Well.... Worst case temp on the mosfet was 266°F! The diode got up to 147°F. Terrible. Granted this mosfet is just soldered to the PCB, no attempt at heat sinking to another plane or anything. And no external heat sink.

However, if you don't PWM it, and just turn it on or off, the mosfet thermal performance is fantastic, about 5-10°F above ambient.

I went back over my math on the gate driver and found my mistake. Turns out 10 is not the same as 1. I missed a decimal and it's switching 10x slower than I expected. Using the actual values from the data sheet, I would expect this to take 200ns to switch, and it measured 200-220ns.

With a higher 7.8kHz PWM frequency, this works out to about 99.7% of the time it's on or off, and 0.3% of the time it's switching.

Still looking at parts, but my next iteration will be a board with a few different designs, to try different things.

Probably include two mosfets, and two gate drivers, to share the load between two pairs. Probably same mosfet, but with an attempt at thermal heat sinking to a plane on backside of PCB, and a 10A gate driver instead of a 3A gate driver.

Also going to try a PCB that has multiple 10A gate drivers driving the mosfet to get the switching speed down really low. Probably try designs with 2, 3, and 4 gate drives, just to see how much I can do and where the knee of the gate-drive-amps curve is.

No idea if I'm going to make this work, but the goal is to get the efficiency so good, that it makes very little heat. I'm shooting to have be able to run in a 125°C environment in still air at rated power reliably. It looks like the mosfet meets my specs, but it's not happy with the current switching speeds, so going to see if I can get that fast enough to keep it happy. My day job is with downhole tools for the oilfield, so I'm familiar with high temp and high reliability design requirements.

Here's a couple videos from testing!

patsmx5

iTrader: (16)

Wheels and slicks!

MetalMuffins

iTrader: (1)

I feel some "light messaging" might be necessary to get them installed. Looking forward to results.

Lokiel

patsmx5

iTrader: (16)

I think you guys hit the nail on the head.... Gonna need the hammer!

So, update time! Got a fan update, and another possibility that is in the works.

Been working on the PWM Fan controller. Got some good progress on it. The next design will feature two gate drivers rated at 12A each, in parallel, feeding the same mosfet. I'm expecting this to get the switching times much much quicker than before, hopefully reducing the switching losses to less than 0.1%.

In addition, there will be two sets of these! So four drivers, and two mosfets working to share the load. Also the next board will have considerable effort at thermal management to help spread the heat that is created out where it can dissipate. Also made several smaller improvements here and there.

Got the PCB layout going well, it's routed, all nets are connected. Gonna refine a few things, then thermal stuff, and then order it. Hope to have it ordered this week.

The other thing that's in the works... I'm considering finally getting a shop. Been wanting one forever. I may be getting one soon if I can find one for a fair price. Sure would be nice vs my carport.

patsmx5

iTrader: (16)

Update time!

Fan controller board is designed and ordered! I was hoping to get it ordered this week and that happened.

This prototype I'm calling V1. It has two mosfets, two gate drivers per mosfet, and five freewheel diodes. Thermal via stitching between top and bottom to help transfer heat away from the high power components. Traces that carry high current are wider than before, now routed top and bottom to double up area. In addition, there will be wires soldered external of the board to lower the electrical resistance to further reduce heat generated from voltage drop on the traces. Also went with 2oz copper instead of 1oz to again reduce voltage drop and improve heat transfer.

So a bunch of improvements! Hope this one works. It has about a dozen improvements so if it works, I'm expecting much better performance than the first test board I tried.

In my last post I mentioned I was shop-shopping. Well, I think I found one! It's a 1600 SF space, has a small office and separate bathroom, rest is open warehouse space. I have a verbal agreement from all parties regarding the basics, but we haven't finalized and signed anything yet. But it looks like I will have a shop soon. Long overdue. If this goes though, I'll be smiling for quite a while! If this goes through, I'll be bringing the miata there to finish building it there!

deezums

iTrader: (3)

You should do what they do with SMPS's and keep the high current traces unstopped/no mask, so you can pile mounds of solder on. Cleaner than external wires.

Or add some solder wick to the top of the traces, or solid copper wire, maybe that's what you mean?

18psi

iTrader: (76)

Pat,
I'm watching this insanity. I approve

patsmx5

iTrader: (16)

That could be a good idea. I will probably test that on my next revision and see how it compares to wires. My thought was wires would look nicer and produce less heat since it's copper and would conduct really well. But wires are a pain to install, soldering over would be quick.

Glad to see you here! Was wondering when you would chime in. You posted a few years ago something like "I wanna see what this thing would do with just the EFR7670". Well a few years later, you get your wish!

patsmx5

iTrader: (16)

Youtube video going over the new V1 PCBs!