Amp stage Input Loading and Transformer termination

[guavatone]

I am confused with this.  Would both the transformer and the amp see 90,909 Ohms(the parallel calculation)?

my DMM sees 90.9K Ohms...

 

[shabtek]

wish I knew, there is the node of the opamp, which when powered up must present a path parallel to the Rs. nice simple question, somebody in the knowed will chime in and enlighten us

 

[guavatone]

another API loaded by our friend Bo.  No extra termination though and it seems like the Input XFMR would see 1M load

http://web.telia.com/~u31617586/API%20312%20DI%20mod..jpg

 

[3nity]

Looks like the 1M should be on the other side of the switch or relay! is that right?

 

[Bo Hansen]

Charlie,

On my API-312 DI mod. a ground reference resistor must be added on the op-amp input, because in the DI mode, the input have a coupling capacitor in series, and the op-amp must have a ground referense from someware.
And also, the resistor must be on the op-amp side whole time, to prevent big dc-shift between switched mic/DI mode.
(I think, the DI circuit in your schematic are a of clone of mine)

In the orginal circuit, the op-amp have the reference through the input transformer secondary to ground.

A rule of thumb for a termination are approx. 10 times up the secondary impedance.
So if the termination resistor are 100 kohm, and if you shunt this with a other 1 mohm, this is so little more, and not to worry about.

(some brands/types not need any termination at all, or some have only the RC-net to get nice square wave response without ringing)

--Bo

 

[pstamler]

With nothing plugged into the input, yeah, the tranny and the opamp will both see 90.9k.

But with a 150 ohm mic plugged into the input, the tranny will still see 90.9k, but the opamp looking back will see about 9.09k. The mic impedance will be stepped up to 10k, shunted by the 100k resistor (and the 1M, but that won't make enough difference to talk about.)

By the way, if you want the transformer to perform as Jensen intended, terminate it with 110k instead of 100k. Then, with the 1M resistor, you'll get an effective termination of ~100k.

Peace,
Paul

 

[PRR]

> the opamp looking back will see about

Neglecting winding inductance and the RC CC network.

Of particular note: at DC, the opamp will see ~~1K resistance. If the opamp has high bias current, this enters into your DC figurings. This is quite different from the capacitor-coupled case often found elsewhere in the audio system.

Though in this case, unless there is a transformer or wound-pickup in "DI OUT", the opamp must be happy either with ~~1K DCR mike or 1Meg DCR DI.

I do not understand why the transformer core and shield ground is disconnected in "DI" mode.

 

[CJ]

"I do not understand why the transformer core and shield ground is disconnected ..."
Translation: Why the expensive switch?

You could rig a 1/4 jack to do the switching for you.
One with a million solder lugs on it.

 

[guavatone]

As Alway, Thanks for the replies.

I originally thought about this issue because I laid out an EF86 setup with R-2 instead off DI-L(DI load resistor).
What I heard/read is that most tubes tend to like to see 1Meg at the grid and as close as possible to the pin.
-This thing is that DI mode makes fine sense for both the DI source and the Tube's grid. 

My guess is that we are choosing our battles.  Right?
-I mean that it's probably more important for a guitar's pickups to see a high Z input than it is for the tube grid to be
loaded at 1Meg.  My other guess is that transformer ringing is worse than a low ground reference for a tube or Opamp
in these situations.

With the EF86 circuit and a 1:7 I/P transformer and R-T=100K termination the mic will see 2041 Ohms from the tranfsormer, right?
-I can't seem to figure Paul's calc all the way to the mic.
-I used Turns Ratio Squared * Z to get 2041

NY Dave does it with the MILA as well.

 

[pstamler]

With the EF86 circuit and a 1:7 I/P transformer and R-T=100K termination the mic will see 2041 Ohms from the tranfsormer, right?
-I can't seem to figure Paul's calc all the way to the mic.
-I used Turns Ratio Squared * Z to get 2041

First off, the JT-110k tranny isn't 1:7, but 1:8.165 and some change. 8.165 squared is 66.6666...., and 66.6666...  x 150 = 10k. So that's what appears on the secondary of the transformer.

The opamp, or tube, is looking back at that 10k. It's also looking at 100k termination; 10k in parallel with 100k gives you 9.09k.

1/Rtotal = 1/R1 + 1/R2

So terminate the tranny with a 110k resistor, and hang 1M on the grid of the tube. Put the instrument input jack in between the 110k and the 1M resistors; that way, when the jack breaks the connection, the instrument will see 1M. But when there's nothing in the jack, the transformer will see ~100k (110k in parallel with 1M).

Peace,
Paul

 

[Kamel]

Hi Guavatone
I've done a P2P ef86 ecc88 pramp based on the j.hinson schemo, sounds really good with 1:7 sowter input,i use the switching jack trick, and i mesure in some case the output level of the pickup is too high for the ef86 input, i think it is usefull to use a swichable pad for the Hi Z input, maybe ala bo hansen in his Di Api 312 HiZ addon.

 

[guavatone]

Thanks Paul, I thought it had a parallel calc.


Kemel, I found the DI pad to be ineffective but I can't remember why.  Instead I customized the Feedback switch to have less gain.