[SOLVED] 75V to 25V buck converter - immediate restart after o/p short.

[Orson Cart]

Hello there, is there an ASIC buck IC (or buck controller) that can operate to 75VDC - but most importantly if the load is shorted - the converter should resume normal operation immediately the short is removed (ideally a 600mA current limit +/-20% on the output).

There are some chips that seem to provide this functionality but on close reading of the data sheet (and in practice) they enter a slow restart mode or hiccup mode after a short has been applied and are too slow restoring themselves to normal output operation.

Any help with circuits or KNOWN about IC's that REALLY provide the above functionality would be greatly appreciated.

 

[mtwieg]

Usually the soft start time can be changed through an external component value, usually a capacitor. Connecting the soft start pin to a pull up voltage with a low resistance should effectively disable the soft start functionality. That way the soft start time will effectively zero (but will still start from some sort of reset condition). Also you have to be careful that the chip doesn't break when trying to pull that pin low.

Could you specify some controllers you've been looking at using, so that I can see if methods like that will work?

 

[SeanM]

Hello there, is there an ASIC buck IC (or buck controller) that can operate to 75VDC - but most importantly if the load is shorted - the converter should resume normal operation immediately the short is removed (ideally a 600mA current limit on the output).

There are some chips that seem to provide this functionality but on close reading of the data sheet (and in practice) they enter a slow restart mode after a short has been applied and are too slow restoring themselves to normal output operation.

Any help with circuits or KNOWN about IC's that REALLY provide the above functionality would be greatly appreciated.

Hi Orson, The MIC28500 is good up to 75v input. It is a 4A switcher, so it may be a little overkill for your needs.

Regards,

Sean

 

[Orson Cart]

Hi Orson, The MIC28500 is good up to 75v input. It is a 4A switcher, so it may be a little overkill for your needs.
Regards,Sean

Hello Sean, the 28500 seems quite a good chip & can be made to work below 30V in - not easy to see how a sharp 600mA current limit could be implemented, Regards, Orson Cart.

Mtweig: Have looked at LM5115, LM5085, LM5118, LM5117, NCP1034(100V) some have 256 counters that then reduce current limit via a hiccup mode, others have a slow restart, even if the SS cap is set to 0pF the restart is not immediate.

What is required is a predictable current limit - 600mA +/-20% into a short circuit or other load value.

THe MC3063, 3064 is an ideal fit except it only goes to 40V and has a max duty cycle on the switch of 5/6, it does recover immed from a short to normal o/p
volts.

Regards, Orson Cart.

 

[grizedale]

i thought some of the linear.com buck chips have a high side current sense and can limit to your 600mA in event of short...they also can manage the 75V.

 

[mtwieg]

Hello Sean, the 28500 seems quite a good chip & can be made to work below 30V in - not easy to see how a sharp 600mA current limit could be implemented, Regards, Orson Cart.

Mtweig: Have looked at LM5115

Okay, this says it has a foldback current limit, but it only reduces the current limit by 20%. That doesn't seem like a big deal. Unless there's something else in there that makes it seem like it would be slow, I would give this one a try.

LM5085, LM5118, LM5117, NCP1034(100V) some have 256 counters that then reduce current limit via a hiccup mode, others have a slow restart, even if the SS cap is set to 0pF the restart is not immediate.

Okay I see what you mean with these. These will definitely give a delayed recovery from short circuit.

So you want true high side inductor or input current sensing, drivers good for 75V, and no sophisticated fault protections. That's a tough spec to fill. You might be forced to add an external bootstrap driver to a lower voltage step down converter that meets the first and third conditions.

 

[Orson Cart]

Mtweig, the LM5115 does seem like a possibility - pity about the current limit drop which we would prefer not to have.

We may have to do a fully discrete design to get the sharp current limit & instant recovery. Orson Cart.

 

[mtwieg]

Mtweig, the LM5115 does seem like a possibility - pity about the current limit drop which we would prefer not to have.

I have a hunch that if you were to get a controller that does exactly what you want with no foldback, you'd see some nasty things. Mainly excessive stresses on your FETs, and lots of overshoot during recovery from a short (especially if after the short there is a light load). Unless you can tolerate the overshoot after recovery, you should have some kind of limiting system, either on the slew rate of the output voltage, or some kind of foldback current limit (seriously consider this. If you set it so that it only kicks in with a short circuit load, say 10% of nominal output voltage, then it will probably have a negligible impact on recovery time while saving a lot of stress on the FETs).

We may have to do a fully discrete design to get the sharp current limit & instant recovery. Orson Cart.

What kind of recovery time are you looking for? What size is your output filter capacitor?

 

[Orson Cart]

mtweig, the output C is 20uF, the MC3063, 3064 does all that is required, beautifully, but is only 36V input, the recovery time required is that time taken to recharge 20uF to 25 volt with the 600mA current limit (or whatever current limit is used). The 3063 gives no overshoot recovering into a light load.
note: a proper current limit does away with any danger of FET overstress.

 

[mtwieg]

The MC3063 (you mean MC33063, right) might avoid overshoot since it uses a comparator as an error amp, but that has other drawbacks... I assumed you wanted a current mode controller or at least something with a linear regulation scheme.

And I've seen FETs cook under supposed current limiting conditions. Mainly in synchronous rectifier converters where very low resistance FETs and inductors are used. In that case there's very little to prevent current from climbing in the inductor uncontrolled, even when the on time is as short as the controller can do (like 100nS). But with a more lossy design that might not be an issue.

 

[Orson Cart]

NCP3063 / 3064 (yes On semi do an equiv with the extra 3 in front) - the point is fets can only cook under current limit if the heatsinking is not up to it, if the peak and average currents are not being limited - that is a different story. Most buck converters can control the duty cycle to 0% if required. Still the point of this thread is for engineers to offer suggestions as to IC's they have worked with that fit the requirements in post #1.