AM Modulator

The two mosfets at the far right are supplying a shootthrough current of over 700 thousand amps if configured like they are here on a board.... Something will blow instantly.... see the little dots passing straight through, regardless of any push pull.

@2ctiby At which node are you getting the shoot through current?

Shootthrough is when the current passes directly through both mosfets at the same time, rail to rail, ie when both gates allow ds current conducting in an uncontrolled manner instead of one being fully closed whilst the other is open.

They are low power and thus high resistance MOSFETs, not power MOSFETs.

It’s basically like a class-A MOSFET amplifier configured deep into the class-A region, but because the MOSFETs are low power, they don’t deliver much current even when shorted.

@jason9 ... Have you got a datasheet ref. for those, or a RDS(on) value in mind? ... I like the idea and your imaginitive setups, but it would be nice to clarify various points along the way.

You certainly are right about “imaginative setups”, because that’s all my circuits have been and probably all they ever will be. Not really a whole lot more.

Assuming RDS(on) is resistance when on, then with a gate voltage of 1V higher than the activation voltage the resistance measures to be at 160ohms when measured with a 1V ohmmeter (I just used the one provided by EC because I’m lazy).

@2cent, the two most right mosfets are a push pull stage which are in linear mode. There is never a SHOOT in this circuit. ITS NOT DIGITAL, ITS ANALOG!

Analog circuits can still have quite high power, and thus the ability to “shoot”, but this one is very low power.

@jason9 ... can we please avoid any diversion here and be given a datasheet ref for these mosfets?

@jason, Mr. @2cent has tested your circuit in a digital mode. I guess he even messed up the push pull stage to a complementary inverter stage to generate shoot-through which are totaly nonsense for a push pull stage in linear analog mode! Guess he is a little irritated mosfets can be used in analog circuits as well, even its not common.

@jason9 ... can we please avoid any diversion here and be given a datasheet ref for these mosfets?

I don’t have a data sheet. The best I can give you is: activation voltage of 430mV, a resistance of 160ohms at a gate voltage 1V above the activation voltage when measured with a drain-source voltage of 1V. Good enough? The rest you should be able to figure out, like that the gate capacitance is sufficiently low so as to not interfere with the operation of this circuit.

500pF gate-source capacitance gives -3dB attenuation. That’s probably good enough.

@2cent ... can we please avoid messup circuits

Thanks jason, but that info. is not what is required here.... Your mosfet settings do not comply to any realistic bought mosfet. I'm not here to critisise you or slam you down when you have not thought through a particular aspect. Hopefully, you will discover mosfet settings as you progress (which you have done already more than me in many ways)....good luck.

Mr @2cent latest circuit published includes a mosfet with Rds_on = 120 nanoOhm, very realistic, please let us know the specifucation and part no. Good luck.

@hurz, when discussing real components vs imaginary components, you also have to take into account when something is made some way for ease of showing some concept and the component doesn’t actual have any use in the main function of the circuit vs the component is a central part of a circuit that is meant to be real. In this case, all components are part of a circuit that is meant to be implementable so real vs imaginary makes a difference, but in @2ctiby’s circuit, the imaginary MOSFETs are meant to be convenient switching devices that could be replaced with many things, and he just used the imaginary components to aid in clearly showing a concept while the components themselves don’t have a part in the main function of the circuit.

Right, and you have used a convenient PushPull setup with 19.2mA quitesense current. No need to construct a shoot-through scenario which is by far not what happens in this circuit, cuz its not a complementare invertings stage!

@jason9 ... I have every confidence in you being able to make fair, sensible and academic assessments as you progress. Please continue to avoid the bias stupidities of someone who is keen to get you on his side regardless of any sensible discussion. (Your approx. 19.8mA as per your adjustments would not be reflected in shop mosfets placed in that position as shown, and your VTO does not replicate any standard shop mosfet).

Nice work... also it's large carrier, a little detail which some people doesn't seem to appreciate, might take a little more power at tx but makes rx way easier... nice job.

@2ctiby, if there are no available MOSFET’s with the proper specs, then the circuit can be adjusted as needed to account for that.

@jason9 ... exactly. Find the Rds fixed value and the Vth of a mosfet that you intend to try, then make appropriate circuit adjustments from there... we can't expect an unknown shop mosfet to conform to our pre-fixed circuit values. Then the necessary adjustments will be suitable.... but it demands a level of mosfet understanding to know what adjustments to make... no easy way round learning about ohmic/saturation and k value... that's why I wrote such a long drawn out many-page article for those who wish to work at that level....We can't just place a default EC mosfet for this level of work, nor can we adjust it willy nilly till it seems to fit our pre-adjustments.

To clarify your “700 thousand amps“ nonsense. Jason has setup the two mosfets as class AB analog amplifier. Not as complementary inverter! There is NO shootthrough problem. Anything you want to say about your very first comment?

Read my last comment again hurz ... we can't just alter EC mosfet settings to do as we wish... They and the circuit adjustments must conform to a real mosfet's parameters. It's no good getting our EC to look good with a 0.4v VTO if the shop mosfet placed there on the board is then say 1v VTO etc.... Try adjusting those two mosfets as though they were from a shop with say 1v and -1v VTO and full steam everything else (ignore approx 0.05 or whatever Rds for now), then measure the current passing due to shootthrough happening.... over 700kA ... there is nothing limiting the current (unlike the restriction implanted as seen here originally by purposefully adjusted internal mosfet settings).... that sort of possibility needs to be accounted for before simply placing unknown shop mosfets there.

http://everycircuit.com/circuit/6355824883793920

Why not just buy a low power MOSFET, adjust the component values in the circuit, and use that? And where’s your article on MOSFET Rds, VTO, etc.? I know VTO is the activation voltage, and that Rds is the resistance between the drain and source, but I’m sure there’s still a lot to learn, and I don’t know under what conditions Rds is measured (Vgs-VTO, and Vds (with the capital V for voltage and the two letters after the capital V being the first letter of two of the three MOSFET terminals)).

I read you’re tutorials fun learning MOSFET’s 1-7, and they were quite helpful. From now on, when adjusting MOSFET parameters to get the desired Kn, I’ll just set W/L to equal 2 as you advised to keep things simple, and it doesn’t really matter if W and L are unrealistic because they just contribute to the Kn, the only important aspect of MOSFET’s when in saturation region (of course, you can’t forget about Vth and lambda, but they’re less important and harder to control in real life).

The output stage MOSFETs have a Kn of about 5mA/V^2. Is a 5mA/V^2 MOSFET with a capacitance of 500pF unreasonable?

Actually, a capacitance between gate and source of 200pF and an equal capacitance between gate and drain would be required for an attenuation of almost exactly 3dB. Is a MOSFET with a Kn of about 5mA/V^2 with a capacitance between gate and source of about 200pF or lower a reasonable MOSFET?

@1cent, you still dont get, that a class AB amplifier does not has a shootthrough issue! Its called quitesense current and is for jasons setup finetuned a little less than 20mA. Forget your “shootthrough“ bla bla bla ...

@jason9 ... nice to see that you are realising the importance of learning about mosfet paramaters. There is no simple one-off answer which solves all eventualities. I would suggest that you keep chewing over that 7page article until things fall in to place. Especially, I would advise that you shelve all things to do with capacitance whilst getting to grips with k. The Rds aspect is perhaps the most useful thing to learn and that can in fact be used instead of k in practice. I updated page 3 recently with a link regarding Rds ... Let me know if you find difficulty with any points mentioned in my articles.

@hurz ... You still don't get that we need to make external adjustments to any mosfet circuit in order to conform to the real mosfet's unalterable known parameters. That applies to any mosfet usage, whether it be for ohmic full switching or for various amplifier class type setups. jason9 originally created his setup by internally adjusting the EC mosfets without having considered that any bought mosfet would not necessarily conform to his adjustments.

For new viewers wondering what my 7 page mosfet article is about, it is here: http://everycircuit.com/circuit/5338883035496448

Not worrying about capacitance, would a MOSFET with a Kn of 5mA/V^2 be that hard to find? Worrying about capacitance, what is a normal capacitance value for such low power MOSFETs?

The datasheet Rds and Vth of a real mosfet are basically fixed values. In the EC mosfet they are not fixed, so I suggest that you setup the relevant required attached circuit parts, then alter the EC mosfet settings until the Vth and Rds read appropriately as per that datasheet. ie we have then fixed our KP, W, L (the combination and individual values do not matter once they are fixed to provide the correct Rds... since mathematically the k value is determined by those), but start first by setting the VTO. If after all that the circuit does not then work correctly, we should re-consider our circuit efforts rather than altering those mosfet settings to give a false Vth and Rds. By doing things as suggested, it means that we can then use a real mosfet in that EC schematic position knowing that the real Vth and Rds were simulated properly. Further adjustment of the external Vgs etc can then be undertaken as reqd. bearing in mind that the ohmic/saturation state may change, as per overdrive value. I can't help you with your circuit capacitance at this point since that is more related to your amplifier knowledge requirements... but I would advise keeping away from capacitance entirely whilst learning about mosfets. The internal capacitance certainly does not concern us since it is fixed unalterable.

Ok. What are the equations for Rds? You never mentioned Rds in your article, and because of the way the MOSFETs are set up, they’re locked in saturation mode, so is Rds even relevant in saturation mode since in saturation the MOSFETs are like a variable constant current source?

How do you measure Rds?

Fine, Mr 2cent agreed shootthrough are NOT an issue! Why does it take so long to explain how to measure Rdson ? Google out of order?

@jason, knows that these settings he made do not fit to mosfets he can buy in a shop. Anyway, you can use the default Vto as they do fit to mosfets for integrated circuits. So called low VT with threshold voltages far below 1V!

@jason9 ... You need to carefully learn about Rds and related aspects. I am not here to supply 'give me my answers now' replies to all of your queries. Rds is mentioned on my 7 page article page 3 ... look again and then also see here: http://everycircuit.com/circuit/6063244598050816

@jason9, here another alternative methode to setup Rdson and Vth, check it out http://everycircuit.com/circuit/5834820675698688

BTW, dont use the wrong Vgs to setup Rdson. @2cent does use 5V which is very strange and very uncommon, looks more an invention of Mr @2cent. Take 10V Vgs as most specification does use.

You can measure Rds by applying a small current to the drain and measuring the voltage and using ohm’s law to figure out the resistance, but the voltage has to be a few orders of magnitude below the overdrive voltage (Vgs - Vth) for best results, or else the resistance measured may be too high. Right? I also found out that Rgs varies linearly with the overdrive such that an overdrive of 7V will produce one 7th the Rgs as an overdrive of 1V, even though the saturation current various exponentially with the overdrive such that an overdrive of 7V will produce 49 times as much current in saturation mode as an overdrive of 1V.

The equation for Rgs seems to be (Vgs - Vth)/(2*Kn) where Kn is in A/V^2 and Rgs is in ohms.

Actually, the equation is 1/(2*(Vgs - Vth)*Kn)

Thats why i used 1mV and not the default 1V

@jason9 ... further info here: http://everycircuit.com/circuit/5922590962221056