Maker Pro
Maker Pro

Need cheap buffer for audio device

jeff kloth

Dec 16, 2015
8
Joined
Dec 16, 2015
Messages
8
Greetings, This is my first post. I know a fair amount about electronics but am still clunky at times. My project needs help. The project is a musical organ device. The object is to replace a magnetic wheel with pickup per sound or voice with an electronic file player. The original organ has 91 wheel/pickups that play through switch contacts to allow the sound out (wheels are always spinning). The "handful" of tones that let the sound through as music is played arrive on a buss that simply feeds the primary of a matching transformer. The combined tones (or single toones) leave the transformer and go on to more amplification, etc.

In my design, I have made digital recordings of the wheel outputs and put them each on a micro sd card. THey are played by a stereo player circuit from china that is probably only a PIC of some kind streaming the wav or MP3 data converted to PWM. Each circuit is only about a buck and replaces two of the tone wheel outputs. The problem I have is that as the outputs of these travel through the key contacts to the transformer, the combined signals interact to the various circuit board player's outputs, loading them down and causing volume to decrease.

The original wheel pickups are low impedance and the input to the transformer primary is low impedance. The wheel pickups do not exhibit the loading problem. I assume the output of the circuit board player does not like the near short circuit (the tranny input is only about 20 ohms measured).

So I thought making some kind of cheap and easy buffer was in order. I tried an LM324 quad op amp because it seemed good, being able to be single-side powered from the same 5v rail as the circuits. I need the buffer to be cheep and easy because there are like 78 channels of them to build! The LM324 did not work. I tried a small mosfet but no configuration worked. The player board does have surface mount caps on the two outputs but they are not electrolytic, etc.

So the help I need is something that can buffer each output, hopefuully with 5v single-sided supply, and only provide unity gain (the outputs are already at good volumes) but not be unhappy when stright-wire bussed. In other words the outputs will not interact with each other level-wise. I hope this makes sense enough for some ideas from you great forum members!
Jeff
 

GPG

Sep 18, 2015
452
Joined
Sep 18, 2015
Messages
452
I tried an LM324 quad op amp because it seemed good, being able to be single-side powered from the same 5v rail as the circuits.
Schematic? What impedance are the circuits happy to drive ? (one output)
 
Last edited:

jeff kloth

Dec 16, 2015
8
Joined
Dec 16, 2015
Messages
8
The circuit of the "player" output is meant to drive earbud style headphones--so basically an iPod style connection consisting of right/left and common ground. The recording I am playing outputs at around 21mv which is similar to the original wheel pickup outputs. I have played it on one output with the original device on a key of the organ next to it and there is no difference in level or timbre of the tones. Again, my problem is the "loading" anomaly when the china player circuit outputs are allowed to be multed together, resulting in a level drop which is no good in this application.

The LM324 circuits I tried were inverting or non-inverting voltage follower as seen here: http://www.electronics-tutorials.ws/opamp/opamp_3.html

In my case the gain was not unity, but was higher and sometimes had distortions in it. If I try to put a voltage divider (volume control resistor) before the LM324, the china circuit freaks out and sometimes distorts or won't play until reset. I should mention that I have tried using a 10 ohm resistor between the output of the player and the LM324 stage to approximate a headphone load. If I don't, the signal to noise is generally too high and I would rather not lower the recorded signal for that reason.

I reasoned that like a passive mixer front-end, if I took a high resistance resistor from the outputs and multed that to the input of the op amp, it might work, but it did not like that either.

To reiterate the goal, I would like to have both outputs of each player buffered such that they do not interfere with each other's level and the resultant output of the buffer stage would be unaffected by multing into the low-impedance of the transformer's primary. Speaking to that, the original wheel pickups (which are similar to electric guitar coil wound magnet systems) probably can't sink much current either.
 

GPG

Sep 18, 2015
452
Joined
Sep 18, 2015
Messages
452
Add as many R1 as needed. Since the output is low impedance the value is not very critical 10K will do. Set the gain with R2 The other resistors are to set the operating point to 1/2 V for single supply, not needed with dual supply. Is this a Hammond Rotary?
 

Attachments

  • adder-bmp.jpg
    adder-bmp.jpg
    11.4 KB · Views: 70
Last edited:

jeff kloth

Dec 16, 2015
8
Joined
Dec 16, 2015
Messages
8
Indeed! The mighty Hammond organ like B3, etc. The goal: to remove the heavy tone generator assembly for a portable unit being built from an "organ donor"...

In your provided schematic, the 2 "R" divider at the positive input-- one end goes to ground and one to +5, yes?

Finally you said "add as many R1 as needed..." Do I assume you mean for me to employ just one buffer, in this case as a "mixer buss" rather than one per player output? That would be cheep and efficient if it is the case...
 
Last edited:

GPG

Sep 18, 2015
452
Joined
Sep 18, 2015
Messages
452
In your provided schematic, the 2 "R" divider at the positive input-- one end goes to ground and one to +5, yes?
Yes. But + to the ground if you have bipolar supplies.
Finally you said "add as many R1 as needed..." Do I assume you mean for me to employ just one buffer, in this case as a "mixer buss" rather than one per player output? That would be cheep and efficient if it is the case...
Yes. as long as the combined input of 10 keys ( I only have 5 on each hand) doesn't over voltage the amp, and the impedance the amp has to drive is so far unknown, otherwise only one amp
 
Last edited:
Top