Switched mode constant current supply, huge spikes

I'm designing a buck converter using the HV9918 IC to drive some high current LED's.

I'm seeing some huge spikes at high bandwidth in the current waveform. Does anyone know where they're coming from?

These waveforms are across a 1.7 ohm current sense resistor. So the current in A is approximately 60% of whatever amplitude is shown in the pictures. The average current is about 250mA. I'm driving a 6.8V zener diode.

In all these pictures the bottom flat trace shows the 0V level for the top trace.

https://i.imgur.com/DePWP.jpg
The switching frequency is about 1MHz. There are huge spikes that are 700mA peak-to-peak!

https://i.imgur.com/QUpgI.jpg
Zoomed in. Those sine waves are about 300MHz. My scope is a 2465B 400MHz and I'm using a 400MHz probe, with the ground sleeve (no lead!) attached directly to one end of the through-hole current sense resistor and the probe tip on the other end.

https://i.imgur.com/J3x0t.jpg
With bandwidth limit of 20MHz on the scope, the current ripple looks much more reasonable. Only 30mA ish.

Should I even worry about it? Will the LED's "see" this high frequency current spike if a 20MHz scope doesn't?

Is it because of the inductance I introduce by having the LED load connected with alligator clips? In my final application I need it to drive a load through a cable anyway.

I've verified this on a 600MHz, 5GS/s digital scope too. https://i.imgur.com/DePWP.jpg

Why bother making links to all your pictures and not your chematic?

Interwinding capacitance will always let the edge dV/dt
blow through the inductor. You need lower ESR/ESL filter
caps, perhaps a better inductor (windings further apart,
single layer winding, ??? - I'm not a magnetics guy but
I bet there are options). You might post-filter with a
ferrite bead or another, maybe larger than the primary,
series inductor. Of course filter caps won't cut the
through-current, but will soak up a lot of it without it
banging the load. You may need a lot of little ones,
rather than a few big ones. Turn the spike FWHM into
frequency and you need an individual-cap SRF higher
than that.

Since it's supposed to be constant-currrent I'd favor
the inductor. With caps before it, a "T" low pass in
effect with the primary inductor.

The duty factor is pretty low so you probably don't care
in a thermal or average current respect. But every device
has a pulse power / current maximum even if they don't
spec it.

I'm not sure anyone has experience with these chips. I wanted to demonstrate the waveforms that I'm measuring. Here's my schematic though. I'm just following their application circuit pretty much exactly:

http://i.imgur.com/RI2w5.png

R4 is a short instead of 1k. This is tested on a PCB with very short trace lengths.

The inductor I'm using saturates at 1A: Digi-Key - 587-2376-1-ND (Manufacturer - NRG4026T100M)

Is this waveform typical of constant current supplies? All of the LED driver chip datasheets with application circuits never have any sort of post filtering on the output other than a capacitor to decrease ripple.

For illumination LEDs the phosphor persistence is so much higher
than the chop rate that (as long as the spike produces photons
more than heat) it's all good.

But those "spikes" are really bursts of ringing and won't make
you any friends when it comes time for EMI compliance
testing. That kind of thing probably wants a snubber across
the switch.

I am having a little trouble visualizing with the upside down buck arrangement, returning coil current to the B+ line, but I believe the ringing is the period between the point where the IC's MOSFET releases its conduction and the Schottky diode picks up conduction.

There is capacitance on the output of the MOSFET along with any PCB strays that prevents immediate voltage rise to point of diode conduction. This capacitance along with coil may cause the ringing during the rise time duration between MOSFET release and diode conduction.

Check and see if the ringing coincides with the MOSFET release.

You are probably making ringing worse with the C3 1 uF filter cap providing a high frequency return. Make sure your coil value is large enough to get you into continuous conduction on the coil. See fig.2 on page 5 of spec sheet.