EAB MV63 Inductor Value

[Murdock]

One last question.
I simulated the whole circuit and get a good flat response. Gain switch also seems to work.
But what should I specify as load? Lets say, I take an output transformer with 8:1 step down ratio. 600 Ohm load. Would the load on the primary side of the transformer then look like 38400 Ohm (8²x600)?

Could you please post the whole schematic of the preamp?

Sorry, but I'm not allowed to  :-\
I had to sign a "contract" not to distribute the schematic...
I can give you the contact of the guy I got it from.

 

[abbey road d enfer]

One last question.
I simulated the whole circuit and get a good flat response. Gain switch also seems to work.
But what should I specify as load? Lets say, I take an output transformer with 8:1 step down ratio. 600 Ohm load. Would the load on the primary side of the transformer then look like 38400 Ohm (8²x600)?

Correct, however:
1) In real life, it's more likely to see 1.5-2k than 600.
2) This is correct for mid frequencies; at LF, the primary inductance will tend to dominate, at HF parasitic capacitance will interfere.

 

[Murdock]

Ok, thanks! Will try out some different transformers.
Still curious about the inductor...

As inductance is frequency dependent, should the inductor have 150H at 120Hz? And I suppose it doesn't need a gap, right?
And why is the 47k and the 900p cap there? I get a much better frequency response without it in the simulation...

The V41, which this preamp is based on, doesn't have the resistor and cap... It just goes straight into the 50nF cap from the plate.

https://drive.google.com/open?id=1jB_xmwrVZhMN0Ku-YsyvMpsZ0QkqXv8H

 

[abbey road d enfer]

As inductance is frequency dependent, should the inductor have 150H at 120Hz?

That's correct. depending on core material, inductance at 1kHz may be several times smaller. To add complexity, inductance also varies with level... 

And I suppose it doesn't need a gap, right?

Correct, no gap needed. I guess if it was gapped it would be enormous.

And why is the 47k and the 900p cap there? I get a much better frequency response without it in the simulation...

I can only conjecture. It looks like the designer opted for more or less matching I/O impedances, not exactly, still a 1:5 ratio, which is closer to textbook than now-current zero-input/infinite output. Which tends to justify the 47k. Now the 900p, I can only guess it's there to compensate some HF response loss in the subsequent stage. It amounts to about 2dB @20kHz.

 

[PRR]

> Now the 900p

There's easily 100pFd "hidden" in the inductor. I did not simulate that and get a mild dip. Perhaps with the 900p it is flatter. This will vary with the inductor. Much of this must have been worked-out on a prototype.

 

[CJ]

150 H on a small core will mean lots of turns of small wire which means lots of DCR,  maybe 5 or 10 K,

this will not be a factor at hi freqs as the 2 pi f L is going to be very large compared to the DCR,
but it might  affect the shape of the curve at low frequencies.

total Z will be a rt triangle with DCR on vertical and XL on horizontal, 

there also might be some phase shift going on, set probes at in and out nodes and compare, maybe the sim will show this.

there is a good chance the original inductor was done on a ferrite core

when you grind up the chrome  and bind it with epoxy, you create a natural air gap in the core material. 

this means that your inductance will be more linear vs frequency than that of a regular core made up of steel lams.

 

[PRR]

> lots of DCR,  maybe 5 or 10 K, .... but it might  affect the shape of the curve at low frequencies.

As I said (but did not show) above, adding 10k DCR shifts -3dB from 118Hz to 123hz without any large change of shape.

So now our "simple choke" has hidden resistance and hidden capacitance. Not as simple as it looks once the prototype is tuned and the choke-winder gets some consistency.