Hi Z input buffer = Too much gain

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

elskardio

Well-known member
Joined
Sep 29, 2010
Messages
564
Location
Montreal - Canada
Hi Guys,

I'm using the following circuit for my Hi Z input on the preamp I'm working on (twin servo).

Its working great but with the preamp step up transformer I have way to much gain when I plug an instrument to the Hi Z input...

Could I use a simple voltage divider in front of the input cap? Any downside?

Thanks

Hi-Zbuffer.png
 
How much reduction do you need? Too much signal, possibly coming from an active instrument ? If indeed so,  you don't need Hi-Z.

Some of the various possibilities:

inject the signal after the TX

Skip the inverting stage, but keep C21 & R21 (connect left side to ground). This reduces level by 6dB, which might help but possibly not be enough.

The resistive divider indeed



Which intended IC-type?
 
Why convert to balanced? Why run it through the transformer at all? Just use one op amp and tap into the second stage. Skip the first stage. You don't need that gain. You just need the second stage to drive the OT.
 
squarewave said:
Why convert to balanced? Why run it through the transformer at all? Just use one op amp and tap into the second stage. Skip the first stage. You don't need that gain. You just need the second stage to drive the OT.

I have a switchable PAD in front of the input transformer... My first thought was: "Hey It might be handy to have the pad available to the HiZ input..."

Second thought: What you said  ;D
 
> voltage divider in front of the input cap? Any downside?

High hiss.

Guitar is more voltage than lo-Z mike. Your plan doubles that voltage. Pad in front of the first amplifier directly degrades signal/hiss ratio.

Assuming >9V supply and <$1 chips, go ahead and take the slight gain, and then waste-off a lot of signal on the way to the mike transformer. Penciling 2uV hiss at the hi-Z interface and 0.2uV hiss at the low-Z interface, these ratios roughly preserve your noise level.

 

Attachments

  • Hi-Zbuffer-2.gif
    Hi-Zbuffer-2.gif
    20.5 KB · Views: 56
The cheap obvious thing many companies do ,  and works totally fine in most cases, is break the  line after the xfmr secondary
insert D.I. in there , with or without  buffer
 
I understand the rational behind going into the transformer, that's sometimes part of the overall sound of the unit, and one reason why you bought it in the first place.
But, as you found out, you will need to drop the level.

Back when Toshiba 2SK170 J–Fets were about $0.20 each and easy to come by, I used both the circuits attached in the pdf below in several commercial designs.  I didn't always have any convenient voltage other than the 48V phantom rail so these are run from that.  But they are easily adapted to +/- 16V etc.

The 2nd circuit can be nulled for THD if that's desirable but, the first circuit works for Rock 'n' Roll and is what I'd probably use unless you wanted to go to the trouble of balancing and nulling the thing.
 

Attachments

  • DI Input Circuits.pdf
    144.4 KB · Views: 50
Elskardio --- note a detail in common on the two plans presented here. After buffering the instrument, both plans use a few-K to few-hundred Ohm pad to reduce level about 10:1.
 

Attachments

  • elskardio-2.gif
    elskardio-2.gif
    12.7 KB · Views: 48
You can eliminate +6dB of voltage gain by making IC1b a unity gain follower just like IC1a, but fed from input 0V. 

The passive pad that PRR suggested is typical for feeding mic input stages.

JR
 
PRR said:
Elskardio --- note a detail in common on the two plans presented here. After buffering the instrument, both plans use a few-K to few-hundred Ohm pad to reduce level about 10:1.
Yes exactly!  With the pads shown in both our schemes, mic amp gain for a passive guitar/bass will be in the 35 - 40dB vicinity for normal Pro line level out. 
 
Hi Guys,

sorry for the late reply... I was away on vacation  8)

PRR said:
Elskardio --- note a detail in common on the two plans presented here. After buffering the instrument, both plans use a few-K to few-hundred Ohm pad to reduce level about 10:1.

Thanks PRR for pointing that out. Looking at various schematics, I see indeed either a pad at the output or a step-down transformer (like 7:1 ratio).

I found a simple schematic using a single 2SK170 (seems like a popular choice for a Fet DI). I have a few in stock from a previous project... I'll give it a shot.
 

Latest posts

Back
Top