HI All,

I built a small single-channel amplifier using the LM675. I built it exactly as the Figure 2 non-inverting schematic in the data sheet. It seems to work fine with no oscillation but appears to clip the waveform at too low a voltage. I'm using ±9v split rails (tested with batteries and with power supply) but the output clips at less than 2vp-p. Gain is 20 just like the figure 2 example.

If I increase the supply voltage, the clip level also increases but wondering why its low. Figure 11 indicates that I should get >5v.

I'm using 8Ω high power resistor for the test load. I checked my signal source and its clean.

Here's the data sheet: https://www.ti.com/lit/ds/symlink/lm675.pdf

Thanks for any help!

 

I'm using ±9v split rails

What do you mean "split"?
Do you have one battery split to give ±4.5V or two batteries to give ±9V.
The output voltage you are seeing would suggest just one.

 

Ah, sorry for the confusion. Two batteries so the LM675 sees ±9v. I just double checked the voltages at pins 3 & 5 and they are correct. Bummer, that would have been an easy one to fix!

Thanks

Rob

 

The Output-Section of the LM675 is built just like any
other old-school BJT-Audio-Amplifier made with discreet Components.

It is not possible for the Output to get within ~2 to ~2.5-Volts of the Supply-Rails,
Plus or Minus, and even under no-Load conditions.

If your design can tolerate the amount of extra wasted Power created by this situation,
this family of Amplifiers can withstand a lot of serious abuse while providing substantial amounts of Current.
They need BIG Heat-Sinks.

If your project is expected to be Battery-Powered,
it would be a really good idea to learn about how to build a Digital-Amplifier,
they are more complex, in that You have to understand High-Frequency-Output-Filtering,
but they are incredibly efficient, and produce very little waste-Heat.
.
.
.

 

PP3 9V batteries are not meant to supply more than 50mA.
Measure the battery voltage while the circuit is actively driving the load.

 

Yes the heat sink does get hot. I went larger than would have been needed for my use (not music). This is for occasional communication while in a race car.

Here's the long story:

My ear plugs with built in speakers (actually are hearing aid transducers) can easily be heard but don't provide enough attenuation for the cars spl. I'm concerned it will affect my hearing.

I tried using my custom molded earplugs in my ears and the hearing aid transducers tucked up tight behind my ears. Put my helmet on, drove the transducers with my iphone and could hear it over the noise (and the total pressure inside my ears was less). Problem is the communications in the car distorts when trying to play at the same level the iPhone was able to achieve. Figured I'd build the amplifier as a temporary solution (the car will soon be quieter anyway).

When I bench test by driving with iPhone everything seems good other than the lower clipping voltage. I left the circuit DC coupled and didn't have a problem. I was attempting to keep this as compact as possible and it let me save the space that would have been consumed by two components. But, I think this caused a problem with the cars radio because the output from the amp was lower in volume (than direct from the radio, so gain less than 1) but through the amp it still could be driven louder without distortion than without the amp. I'm going to try AC coupling the input to the amp, maybe the radio has an offset.

Back to the first problem, with ±9v, I still think I should be able to get more than ±2v swing (actually I'm getting a little less than that). I could get power from the car but would then need to build dc to dc converter to generate the higher (and negative) voltages.

Don't know anything about digital amplifiers. Know of any beginner tutorials?

MrChips, will check that tomorrow but it also has the problem when using a 10A bench supply.

edit: thought I'd clarify that all my bench testing is with the iPhone, not the communication radio.

Thanks

Rob

 

I built a small single-channel amplifier using the LM675.

Where was this LM675 chip purchased?

 

think I got it from Digikey.

 

Then it should be good.

 

If we had at least a Block-Diagram drawing, ( Schematics of course would be the best ),
then we could more easily figure-out exactly what all the various sub-systems that
You are working with, are,, or may be,, doing,, or not doing.

As an off-hand guess ............
I would start with a Dash-Mounted "Audio-Mixer", with an Output-Channel for each person in the Car.
This would allow providing an appropriate "Pre-Amplifier" section which makes compatible
each different type of Input, and provide a level-adjustment for each.
( Maybe there's only the Radio and 2-Mics ??? )

You didn't mention what type of Microphone is being used for Voice, this is critical,
as is the Pre-Amp-Section for each Microphone,
each of which needs to contain ...........
High, and Low-Frequency-Filtering throughout the Circuit,
Super-Fast-Dynamic-Range-Compression,
and Moderate-Peak-Clipping with an additional high-Frequency-Cut.

There should be zero issues with running all this on a nominal~12-Volts.
No need for special "Split-Supplies", and no need for any high-power-Amps like the LM675.
.
.
.

 

Yes the heat sink does get hot.

Is this without any input signal?
Have you taken any voltage readings?
If the circuit is configured as in figure 2 of the spec sheet then pins 1,2 and 4 should read zero DC volts at idle.

 

In the car is one driver (no passengers), one radio, one microphone, and telemetry back to crew in the pits. No issues at all with the mic and transmit.

Went with the 675 for overkill. One driver has speakers inside their helmet rather than the earplugs with hearing aid transducers. They say the speakers are useless as they're always distorted. Figured the overkill design could be used to drive their speakers.

I connected the amp to several small speakers and it seems like it plays plenty loud enough (again when driven by my iPhone). I bet the apparent low gain has something to do with the interface to the radio. I need to read up again on calculating the input impedance of a non-inverting opamp. Maybe its too low and is loading down the output of the radio (hard to believe). I also need to look at the output of the radio to see if there's an offset. I'm not in the same location as the car / radio so will have someone else do it.

Idle current is low. I don't remember how low it was but I was happy as it wouldn't quickly kill the batteries. I went big with the heat sink as its hot inside the car, maybe 120F.

Will double check pins 1,2,4 for zero volts today.

Thanks!

 

The input impedance of the amp circuit is 22K, the value of the resistor from pin1 to ground.
Are you saying the amp functions OK with the iPhone but not the radio?
How is the radio connected to the LM675 circuit?

 

"Speakers" make terrible "Headphones", there's no way around this inside a Helmet.

Driving the usual Low-Impedance-Speaker is totally different than driving
the typical High-Impedance-Headphone.

How is the Output of the Radio marked, specifically ?

Is the Output of the Radio controlled by a Volume-Control-Knob ?

Has the Wiring of the Radio been "hacked-into" to obtain an Audio-Output-connection ?
.
.
.

 

The radio has jacks on the side for mic and headphones, yes volume control. The earplugs are around 30Ω. The amp driven by the iphone plays very loud through 4Ω speakers (and more than loud enough through the earplugs).

My buddy finally looked at the output from the radio on the scope. Bingo, it has a 2v positive offset. I'll put a cap on the input.

Thanks

Rob

 

Finally had a chance to play with this again. Added a 0.1mfd cap to the input before the 22kΩ resistor and it works great.