¿ªÔÆÌåÓý

  1. LTspice
  2. Topics

LT1171 model: syncronization not working

 

I am trying to make a negative buck regulator with LT1171, based on the circuit given on page 13 of the LT1170/71/72 datasheet. I am trying to add clock syncronization as shown on page 11 of the same datasheet (the FET-controlled version, for now)

According to the datasheet:

>?Synchronizing occurs when the VC pin is pulled to ground
> with an external transistor. To avoid disturbing the DC
> characteristics of the internal error amplifier, the width of
> the synchronizing pulse should be under 0.3¦Ìs.?
> ...
>?The transistor?must be capable of pulling the VC pin to within 200mV of
> ground to ensure synchronizing.

When I model this circuit in LTSpice, I find that although I pull VC ?below GND (the same potential as the LT1171 GND pin, actually at -18 V in this circuit) with repetition frequency 138 kHz and 0.1 us pulse width, the switching current doesn't synchronize with this signal.

I've uploaded my ASC file to the Temp directory as Neg Buck 20160803.zip.
?
If anyone has experience with this part, can you tell me,

* Does the LT1171 spice model include the synchronization feature?
* Or am I just doing something wrong?

Side question:

* Can anyone recommend a surface mount equivalent to the VN2222 FET recommended in the datasheet circuit? And is an LTSpice model available? (it seems that capacitive shoot through from the control signal to the VC pin can be an issue when using a poorly chosen FET here)

Thanks!

(For the search engines: this question might also be interesting to people working with LT1070, LT1071, LT1072, LT1170, or LT1172)
 

¿ªÔÆÌåÓý

On many of the LT switch mode controllers, the sync function is not implemented in the spice model. Don¡¯t know about that specific one.

Jim
'
James Wagner
Oregon Research Electronics



On Aug 3, 2016, at 10:40 AM, matt_bruensteiner@... [LTspice] <LTspice@...> wrote:

I am trying to make a negative buck regulator with LT1171, based on the circuit given on page 13 of the LT1170/71/72 datasheet. I am trying to add clock syncronization as shown on page 11 of the same datasheet (the FET-controlled version, for now)


According to the datasheet:

>?Synchronizing occurs when the VC pin is pulled to ground
> with an external transistor. To avoid disturbing the DC
> characteristics of the internal error amplifier, the width of
> the synchronizing pulse should be under 0.3¦Ìs.?
> ...
>?The transistor?must be capable of pulling the VC pin to within 200mV of
> ground to ensure synchronizing.

When I model this circuit in LTSpice, I find that although I pull VC ?below GND (the same potential as the LT1171 GND pin, actually at -18 V in this circuit) with repetition frequency 138 kHz and 0.1 us pulse width, the switching current doesn't synchronize with this signal.

I've uploaded my ASC file to the Temp directory as Neg Buck 20160803.zip.
?
If anyone has experience with this part, can you tell me,

* Does the LT1171 spice model include the synchronization feature?
* Or am I just doing something wrong?

Side question:

* Can anyone recommend a surface mount equivalent to the VN2222 FET recommended in the datasheet circuit? And is an LTSpice model available? (it seems that capacitive shoot through from the control signal to the VC pin can be an issue when using a poorly chosen FET here)

Thanks!

(For the search engines: this question might also be interesting to people working with LT1070, LT1071, LT1072, LT1170, or LT1172)
 

Hello Matt,

The sync-feature is only implemented in a very few SPICE models as?far as I know,?
?
Best regards,
Helmut
 

Matt, I just looked at it, and it looks as if SOME of the LT1171's
switching pulses are synchronized with the Vc input pin. Those ones
have a narrower pulsewidth than the ones around them.

Maybe it doesn't sync to every input pulse?

I can't get the datasheet (LTC's website seems to be down, or not
accessible from this side of the continent right now).

The pull-up on the Vc pin is interesting; makes me think that it is at
least partly implemented in the SPICE model, but that's no guarantee
that it is functional in the model.

Andy
 

To answer the OT question.
The VN2222 was a fairly generic ~5 ohm,60V vfet by modern standards.
Similar devices are things like the generic? 2N7000 series available from many vendors.

So an sot23 substitute device would be an NSD7002A from fairchild or other generic vendors for example.

The more modern part will have a much more detailed spec sheet than the supertex part.
Taking a quick look at specs they look quite similar.
If you know what specs are important to your application, there are likely better choices than the 2n7000 series.

Tim

1 - 5 of 5