Updated Mirage PI (Ex Hammerhead)- Mono Coil, DD Coil or Super D Coil Options

[Elliot]

It's difficult to compare these. If the preamp needs to settle to within, say, 500mV of ground to start sampling, then the coil needs to settle to within 1mV of ground is the preamp gain is 500. It's difficult to see this on a scope. It's easy, however, to look at the raw coil decay and think it is nice & fast, until you realize just how far down it has to settle.

It looks my my overall rig (including coil, coax lead, IRF840, etc.) produces a fly-back decay time of about <5uS (using my eye, not measured to 1mV above ground) once critically damped, so the impact of the clamping diodes and pre-amp is my next area of focus. I have noticed that when I increase the pre-amp gain, that the "clamped time" does increase a few uS, so the pre-amp is probably the priority area to work on. Green's suggestion of a two stage pre-amp is a good one. Can anyone point me to other posts that resolve this?

[green]

Yep, that is true and I take the pre-amp gain into consideration.

Still not a bad idea to look at the coil only decay.

Can see if coil is critical damped. Scope probe connected to coil hot lead can effect damping with a fast coil, I usually connect probe to hot lead insulation. Is it better to adjust damping looking at coil decay or amplifier out? I usually look at both, usually looks best when critical damped.

Didn't see Carl's reply, agree.

[Carl-NC]

Still not a bad idea to look at the coil only decay.

Sure, I completely agree, you should always look at the coil-only decay. And then look at the diode clamp signal. And finally, the preamp output. There are things to learn at each point.

[waltr]

It's difficult to compare these. If the preamp needs to settle to within, say, 500mV of ground to start sampling, then the coil needs to settle to within 1mV of ground is the preamp gain is 500. It's difficult to see this on a scope. It's easy, however, to look at the raw coil decay and think it is nice & fast, until you realize just how far down it has to settle.

Yep, that is true and I take the pre-amp gain into consideration.

Still not a bad idea to look at the coil only decay.

[Carl-NC]

On my HH2 it is the Pre-amp recovery that takes more than 10us whereas the coil decay is <8us.

It's difficult to compare these. If the preamp needs to settle to within, say, 500mV of ground to start sampling, then the coil needs to settle to within 1mV of ground is the preamp gain is 500. It's difficult to see this on a scope. It's easy, however, to look at the raw coil decay and think it is nice & fast, until you realize just how far down it has to settle.

[green]

Now I feel depressed! Looking at this screen shot, you can see the yellow gate drive signal (RX start point = Carl's cheat spot?) of the IRF840 and the blue RX signal after the pre-amp. [ATTACH]47372[/ATTACH]. Now look at the actual result [ATTACH]47373[/ATTACH]. Then see where the blue knee needs to start in order to achieve 8uS [ATTACH]47374[/ATTACH]. I suspect that I need to increase the damping resistor, which will reduce the over-damping and move the blue knee left somewhat. Short of that, I need a quicker search coil and lower capacitance coil driver (ie. IRF840B).

Start with a two stage amplifier and a MUR460 diode between the coil and mosfet. Then you need a fast coil.

[Carl-NC]

Now I feel depressed!

Haha! Welcome to the club! It appears you are at about 14-15us right now. I consider 8us to be quite difficult. You will need to look at everything... coil windings, shield, cable, FET switch, damping R, diode clamp, and preamp.

[waltr]

I also use the TX Gate OFF edge as the start time to measure decay delay.
The difference between Gate off to Drain off is very, very small so Gate is good.

I also look at the Coil output, before the clamp diodes, to determine if the coil decay is fast enough. On my HH2 it is the Pre-amp recovery that takes more than 10us whereas the coil decay is <8us.

This is why many have posted improved Pre-amp designs and there has been much discussion on which op-amp is best.

[Elliot]

Measuring the RX Response Time Window to commence at &lt;=8uS

Ideally it should start from the turn-off edge of the TX current, though I usually cheat and just use the turn-off edge of the gate drive. The TX current turn-off isn't abrupt, it usually has a transition of a microsecond or so, so taking the turn-off edge of the gate drive is pretty much a worst-case scenario.

Now I feel depressed! Looking at this screen shot, you can see the yellow gate drive signal (RX start point = Carl's cheat spot?) of the IRF840 and the blue RX signal after the pre-amp. . Now look at the actual result . Then see where the blue knee needs to start in order to achieve 8uS . I suspect that I need to increase the damping resistor, which will reduce the over-damping and move the blue knee left somewhat. Short of that, I need a quicker search coil and lower capacitance coil driver (ie. IRF840B).

[Carl-NC]

Thanks Carl. That IRF840B does indeed look very good for low output capacitance! I noticed an error in the Vishay datasheet though. They have the VDS listed as both 550V (product summary table) and 500V (Maximum rating table). Am pretty sure it's just 500V.

They're rated at 500V but the ones I've used avalanche at ~550V. Same with the IRF740, they tend to avalanche at ~450V, but rated at 400V.

On a separate note, when measuring the decay time to start sampling a pulse, do you start your measurement from the peak back EMF point (e.g. 450V peak), or from the start of the truncated RX waveform (if you get my drift)? Many state that a <=8uS is a good goal to try and achieve, but I haven't seen any uniform definition of when the window actually starts.

Ideally it should start from the turn-off edge of the TX current, though I usually cheat and just use the turn-off edge of the gate drive. The TX current turn-off isn't abrupt, it usually has a transition of a microsecond or so, so taking the turn-off edge of the gate drive is pretty much a worst-case scenario.

[Elliot]

There is an IRF840, IRF840A, and IRF840B. Coss gets progressively (and impressively) lower. The 840B is significantly better than the IRFBC40LC.

BTW, the same is true with the popular IRF740... there are A & B variants with lower Coss.

Thanks Carl. That IRF840B does indeed look very good for low output capacitance! I noticed an error in the Vishay datasheet though. They have the VDS listed as both 550V (product summary table) and 500V (Maximum rating table). Am pretty sure it's just 500V.

On a separate note, when measuring the decay time to start sampling a pulse, do you start your measurement from the peak back EMF point (e.g. 450V peak), or from the start of the truncated RX waveform (if you get my drift)? Many state that a <=8uS is a good goal to try and achieve, but I haven't seen any uniform definition of when the window actually starts.

[Elliot]

The mosfet you mention looks good Elliot. I am using the IRF 840LC VIS which is also a low C version in my MPP E at the moment and am happy with it.

Nice to see someone else wondering about dielectric strength of the insulation in the cable with those flyback pulses jumping about. Thought I'm the only nutter ...

I am still searching for a really good cable for PI. The ones I have right now are a bit on the skinny side for the core conductors, prone to breakage.
Somewhere I have a memo-writeup to myself with wants and no-gos for a great PI cable.

How is your experience mechanically with the Canare DA202 you mention?

Thanks Polymer. No nutters here, just thinkers and problem solvers! I haven't been using the Canare DA202 cable very long, so I can't give you a definitive answer on expected mechanical life, but I can tell you that I'm happy so far. The shield is a flexible mesh (not a solid foil), and when you fold the cable in on itself, it bends to a comfortable working radius of about 3.5cm. It also has two internal PE filter rods to maintain a constant 110 Ohm impedance, and they are as tough as boot nails to try and break! Just use one core wire and the shield. Don't under any circumstance parallel the two core wires together, otherwise you'll almost double its capacities to 27pF / foot! The only down side to the cable I can see is that it's light blue colour will get dirty through use - a first world problem for sure!

[Carl-NC]

There is an IRF840, IRF840A, and IRF840B. Coss gets progressively (and impressively) lower. The 840B is significantly better than the IRFBC40LC.

BTW, the same is true with the popular IRF740... there are A & B variants with lower Coss.

[Polymer]

The mosfet you mention looks good Elliot. I am using the IRF 840LC VIS which is also a low C version in my MPP E at the moment and am happy with it.

Nice to see someone else wondering about dielectric strength of the insulation in the cable with those flyback pulses jumping about. Thought I'm the only nutter ...

I am still searching for a really good cable for PI. The ones I have right now are a bit on the skinny side for the core conductors, prone to breakage.
Somewhere I have a memo-writeup to myself with wants and no-gos for a great PI cable.

How is your experience mechanically with the Canare DA202 you mention?

[waltr]

That IRFBC40LC looks good. Just added it to my 'parts ordering' list to try.

The lower C would be a benifit even if not pushing the fly-back Voltage.