Opamp bootstrapping

[Athelionas]

Hi,

I'm currently designing a similar circuit to the one in the picture and I was wondering whether it is possible to achieve the same with mosfets instead of bipolar transistor as I have a handful of them lying around. If it is possible I'd highly appreciate a simple schematic too.

Thanks in advance!

[jcx]

what is the application? - bridging can nearly double your Vout, actual gain stage on output, floating supply output like the QSC, ect PA amps...


MOSFETs require higher G-S V to turn on so won't pull the amp supply pins as close to the rails

bootstrapping can also have interesting loop stability consequences - often not modeled properly by manufacturer's Spice models

[Athelionas]

I've experimented a little since my last post and got very strange results.

I'm using an opa1611 and with a +/-70V supply it should be able to provide approximately 120Vpp which I could not achieve this far. My preliminary design works on +/-50V with little more than 80Vpp without stability issues. This seems to be the upper limit, however, according to literature passing this boundary shouldn't be a problem, but if I try to increase supply voltage than it starts clipping no matter what I do. Maybe it has something to do with the spice simulation (TINA is buggy as hell) or maybe I misunderstood something and it resulted in a flawed design. I attached a simple test case if someone is interested.

My goal with this design is to be able to amplify a 5.6V rms signal to approximately 40V rms. Supply voltage is rather flexible at this stage, but it should be around +/-50 to 70v.

[geraldfryjr]

I have been wanting to try this myself,I have read many different opinions about this type of configuration.
although I have seen it in some commercial circuits as well,I am very interested in your outcome and results.

I will give it a try in circuitmaker and see if it I can simulate it with that program.

I have gotten mixed results when I have tried it before so I never took the time to wire one up for fear of getting hit in the eyeball from a piece flying of off the top of the opamp chip. he,he

jer

[jcx]

do you have the AD article? http://www.edn.com/contents/images/45890.pdf

one consideration is keeping the input V to the op amp within its common mode range over the full Vswing - why the mixed positive+negative feedabck is used

[geraldfryjr]

Thanks for the link,jcx,I have that issue but I have know idea which pile of boxes it is in. he,he

jer

[Athelionas]

Yep, I have that article and it proved to be quite handy. As far as I can tell common mode voltage of the opamp is not exceeded so I guess I get strange measurements because of the Spice model I use. I read somewhere that usually the power supply pins are not modeled properly in Spice thus you cannot simply design a bootstrap circuit in a simulator. I guess I'll have to build it so I can figure out if it really works or not (I guess it should).

[AndyM]

Hello Athelionas,

I can only second what jcx points out. I want to add, that the capacitors may turn the transistors fully on at power up, and this would destroy the Opamp. The capacitors should be parallel to the Zeners.


Best regards
Andreas

[Athelionas]

Hmm, you are right about the capacitor. I don't remember why I put it there as my actual circuit does not contain it anymore. Maybe it's a leftover part from a previous simulation but it really should not be there. Thanks for pointing that out.

[jcx]

Boyle macromodel absolutely fails to sim floating/bootstrapped PS effects

Analog Devices and TI have app notes on good macromodeling - but don't actually make sure that all (or even any in AD's case) of their op amp models use them

I have seen some ps effects sim OK with the TI/BB opa227 macromodel

the simplest op amp model is a vccs with compensation C load - make g, C huge and the output can be used directly in the sim