SS Gyrator (Choke) Question about (References to Merlin's circuit but not only)

[kagliostro]

As far as I can understand, choke value in Henry is better if it is the higher possible

What impede the use of large Chokes (in Henry) is the weight, the resistence and the cost of this devices

because high Henry values call for large resistance value or high weight & price

I would like to know if a high value SS Gyrator (that has for sure a low resistance value if compared with real chokes)

can be considered a good thing also in project (like some guitar amps) that uses 10H - 15H or lower value chokes

On Merlin's PS book there are explanations about a SS Gyrator version of the wrapped chokes

this circuit can be build on the cheap and in a very small space, friends that have tried it are satisfied

also someone has simulated the circuit and say that also if it isn't a real SS Gyrator/Choke emulator, the circuit do his job

On Merlin's book there is a formula to calculate the values in Henry and in Resistance of the circuit

but the formula seems to apply correctly only to the H compute

The circuit to which I refer is this



So I would like to know your opinion about the fact that in practice, the value that interest is the Henry value (the higher the better)

being the resistance value a secondary aspect if it is low enough to don't affect much the B+ value

I hope I was able to explain adequately what I mean

Thasnks

Frasnco

[PRR]

> the formula seems to apply correctly only

Show your math.

Formula for R is simplified. It should include Rgs of the MOSFET with R3. To compute Voltage Drop it should include VgsOn of the MOSFET. But for any practical case, these errors are quite small. Use the formula and assume the real voltage drop may be up to double. The values shown give R of 20 Ohms. At 250mA this is 5V drop. In real life I might expect 10V drop. This difference is usually insignificant.

The major "flaw" of this "choke" is that it can not kick-up higher than the input voltage. If the right side goes to a large capacitor, as for B+ filtering, we do not expect (and do not want) much kick-up. It does not work (as shown) as an audio plate-load like a real choke could.

[kagliostro]

Hi PRR

Show your math

......

The values shown give R of 20 Ohms

My math says 10R

10R (R3) : [1 + (47000 (R1) : 47000 (R2))]

47000 : 47000 = 1

1 + 1 = 2

10 x 2 = 20

so 20R and not 60R as shown

instead

100uF x 47000R x 10R

100 x 47000 = 4,700,000

4,700,000 x 10 = 47,000,000 That I can assume as 47H as shown

So Henry OK but Resistance 

However, as you say, there is an insignificant drop in voltage and this, I think, is positive

--

The major "flaw" of this "choke" is that it can not kick-up higher than the input voltage

OK

If the right side goes to a large capacitor, as for B+ filtering, we do not expect (and do not want) much kick-up.

OK

this two things I think I've understand

here I didn't understand exactly what you are trying to explain me 

It does not work (as shown) as an audio plate-load like a real choke could.

I've a graph (if it can be of some use)



Thaks PRR

Franco

[PRR]

> not 60R as shown

Ah, I missed that.

[PRR]

Had a long struggle with the sim. Actually a long struggle with an idiot brain which forgot to put in the 10 Ohm resistor. Works a lot better when all the parts are there.

I essentially confirm Merlin's graph. His formula for "L" is in a correct form but the result is a bit better than calculated (calc=47H, acts-like=55H). His formula for "R" is obviously simplified to neglect MOSFET Gm and Vto. These omissions sorta complement each other if the drop on the "10 Ohm" resistor is a few Volts, which would be a good operating point.

We both show a resonance near 3Hz, which with 47uFd says 60H (a wee bit higher than the formula or the sim says). We both show -20dB near 10hz. My sim levels-out at -40dB while his levels-out below -65dB.

This makes sense. I modeled an IRF150, a very old HIGH-current (38A!) device with a Gate as wide as a drive-way and correspondingly huge Gate capacitance. With Miller there may be most of 1uFd stray C, which with a 100uFd Gate clamp cap is just -40dB. His MOSFET is surely more right-size, with less stray C, and deeper level-off.

Here are the effective DC resistances I simmed:
30r @ 697mA
45r @ 240mA
82r @ 97mA
261r @ 24.5mA

Note that the "worst" of these is only 6.4V loss. At very high current I sim 21V loss, 11V at nominal 240mA.

I also plotted a perfect 55H + 47uFd filter, which lays almost exactly on the higher parts of the Gyrator filter (excluding the odd peak).

[PRR]

> the odd peak

The "odd peak" is just loading. If the load is 1.2K (300V and 250mA) the peak is 1dB, just like the real choke. After my sim-struggle I grabbed a plot of the gyrator loaded with 12K, which does reasonably show a peak. At design 250mA tube-amp load the 3hz bump will be minor.

[kagliostro]

Many tahns for the simulation PRR

I really appreciate your effort 

Only one question, if possible

Few time ago you give me a council about a voltage dropper for G2 supply using an STP6NK90Z



So now I've in stock some STP6NK90Z, are this usable for this gyrator or is better I get some IRF820 ?

Thanks

Franco

[PRR]

> STP6NK90Z

Fine.

[kagliostro]

Many Thanks




Franco