I'm having a hard time figuring out how T2 gives a 16 ohm output but the amp is lm1875 which is 4 or 8 ohms out according to datasheet.
I'm picturing T2 as an autotransformer but considering the amp is well designed for a 8 ohm speaker and there's a tap for 8 but but coming off the main output is 16 ohm?
If I only use 8ohm speaker, I can't just remove T2 completely right?
More than anything I guess I'm tring to figure the role of T2.
I'm picturing T2 as an autotransformer but considering the amp is well designed for a 8 ohm speaker and there's a tap for 8 but but coming off the main output is 16 ohm?
If I only use 8ohm speaker, I can't just remove T2 completely right?
More than anything I guess I'm tring to figure the role of T2.
Would this be an instrument amp of some sort?
The gain is abnormally high - 200k to 1.5k is a gain of about 42dB. The 16 ohm output is actually OK, you can connect a higher impedance load with a voltage swing penalty.
The role of T2 is to match the impedance (and the power output spec) into an 8 ohm load. C10 prevents DC from saturating the transformer. You can remove C10 and T2, and bring the 200k resistor down to about 20k for 23dB of total gain. This will also reduce the offset to very manageable levels: R8 and R10 should be as close in value as possible. These changes make it closer to a textbook implementation of the 1875, which this may not have been.
The gain is abnormally high - 200k to 1.5k is a gain of about 42dB. The 16 ohm output is actually OK, you can connect a higher impedance load with a voltage swing penalty.
The role of T2 is to match the impedance (and the power output spec) into an 8 ohm load. C10 prevents DC from saturating the transformer. You can remove C10 and T2, and bring the 200k resistor down to about 20k for 23dB of total gain. This will also reduce the offset to very manageable levels: R8 and R10 should be as close in value as possible. These changes make it closer to a textbook implementation of the 1875, which this may not have been.
Checks out, similar category. Yes, you could reduce the gain and omit T2+C10 if you want. C10 is not needed in the absence of T2, but its inclusion prevents DC from ever reaching the output. If you want that sort of assurance, you can upsize it a bit for 8 ohm speakers. I have done this in the past when I don't want to add a separate protection circuit.
The low voltage rating on C10 is because with a bipolar supply, there should be ~no DC on C10. It's there to keep T2 from creating a DC short on the output. Operating an electrolytic cap with more AC than DC will kill the cap before long. Normally a higher voltage cap would be used and a single ~48VDC supply.
Using 330uF for C10 means no bass below 30Hz for 16 Ohms, nothing below 60Hz for 8 Ohms, nothing below 120Hz for 4 Ohms, so I would use a larger cap, if any.
Yes, the transformer T2 is not a great idea for several reasons. But an 8 Ohm load will run the chip hotter so be sure it's well heat sunk. The LM3886 is essentially the same chip in a bigger package and better suited for 8 or 4 Ohms.
The strange input section and running the LM1875 at such high gain is a mystery, probably related to some custom application? Running the LM1875 at this high gain is a sound quality compromise so my guess is that this is not meant for music.
Very expensive transformers provide a fair sound quality, but not great. Cheap transformers are only acceptable for "communication" quality sound. Perhaps this is an "intercom" amplifier, but for music, transformers are to be avoided, and usually unnecessary.
Using 330uF for C10 means no bass below 30Hz for 16 Ohms, nothing below 60Hz for 8 Ohms, nothing below 120Hz for 4 Ohms, so I would use a larger cap, if any.
Yes, the transformer T2 is not a great idea for several reasons. But an 8 Ohm load will run the chip hotter so be sure it's well heat sunk. The LM3886 is essentially the same chip in a bigger package and better suited for 8 or 4 Ohms.
The strange input section and running the LM1875 at such high gain is a mystery, probably related to some custom application? Running the LM1875 at this high gain is a sound quality compromise so my guess is that this is not meant for music.
Very expensive transformers provide a fair sound quality, but not great. Cheap transformers are only acceptable for "communication" quality sound. Perhaps this is an "intercom" amplifier, but for music, transformers are to be avoided, and usually unnecessary.
I think its for an external control, obliviously someone drew it in and I've never seen it implemented
You're absolutely right, its a intercom for a loud and noisy steel mill with horns almost exclusively used.
What kind of compromise would running the 1875 at a high gain? What was the give away that it was used as an intercom?
I should also add C10 is a bipolar (electrolytic?) capacitor. I didn't know such things existed.
Gain is very high for Microphone level.
Transformer is for line matching to paging speakers.
Level or gain is very high for 16 ohm paging speakers.
If a 8 or 4 ohm paging speaker is used the levels would be too high.
So different taps.
More related to the original function of the paging system.
Likely someone did mods to the volume control, because the paging speakers
where to loud. Might have been connecting 8 ohm horns to 16 ohms.
Didn't understand the tap functions, who knows.
If your converting for music playback at home.
Then transformer and capacitor not needed.
Capacitor is as mentioned DC block for transformer, also small value to high pass the paging speakers.
No bass, voice only.
If converting for music, you can drive speakers directly, no cap or transformer needed.
The input transformer and transistor mute circuit also not needed.
That is all circuits for the paging function.
For extremely long cable runs in a paging system.
Also as noted the gain on the chip is set very very high for microphone levels.
Not line level stereo products.
Paging mics are muted, then you just hold a button to speak.
So all of that can be removed.
Just use the values in the data sheet for gain and the input circuit used for usual line level stereo applications.
If you hooked up a line level tape player or CD player to the input, it would be blasting loud and distorted.
Transformer is for line matching to paging speakers.
Level or gain is very high for 16 ohm paging speakers.
If a 8 or 4 ohm paging speaker is used the levels would be too high.
So different taps.
More related to the original function of the paging system.
Likely someone did mods to the volume control, because the paging speakers
where to loud. Might have been connecting 8 ohm horns to 16 ohms.
Didn't understand the tap functions, who knows.
If your converting for music playback at home.
Then transformer and capacitor not needed.
Capacitor is as mentioned DC block for transformer, also small value to high pass the paging speakers.
No bass, voice only.
If converting for music, you can drive speakers directly, no cap or transformer needed.
The input transformer and transistor mute circuit also not needed.
That is all circuits for the paging function.
For extremely long cable runs in a paging system.
Also as noted the gain on the chip is set very very high for microphone levels.
Not line level stereo products.
Paging mics are muted, then you just hold a button to speak.
So all of that can be removed.
Just use the values in the data sheet for gain and the input circuit used for usual line level stereo applications.
If you hooked up a line level tape player or CD player to the input, it would be blasting loud and distorted.
Last edited:
Intercoms use loudspeakers as microphones, and they boom like crazy at the cone resonance unless you either:
1: load them to the speaker impedance. I'm not sure 33 Ohms is enough, but I expect the maker tested and chose a compromise. Intercom speakers are often 45 Ohms.
or
2. use a 6dB (single) high-pass filter at about 1KHz to equalize the speaker response.
High gain at the LM1875 means very little feedback and therefore more distortion.
1: load them to the speaker impedance. I'm not sure 33 Ohms is enough, but I expect the maker tested and chose a compromise. Intercom speakers are often 45 Ohms.
or
2. use a 6dB (single) high-pass filter at about 1KHz to equalize the speaker response.
High gain at the LM1875 means very little feedback and therefore more distortion.