The schematic shows a 1Ohm resistor in series with a 0.22uF cap on the output and I know from experience that this RC circuit is necessary to get the amp working properly, but I'm struggling to understand why 
Could someone help me understand?
http://www.electroschematics.com/wp-content/uploads/2009/02/lm1875-amplifier.jpg
Could someone help me understand?http://www.electroschematics.com/wp-content/uploads/2009/02/lm1875-amplifier.jpg
To provide a defined load for the amplifier at very high frequencies.
Some amplifiers change their gain at HF if the load becomes near infinite. That change in gain makes the phase and gain margins go haywire and can result in oscillation.
Even with the speaker connected, the amplifier may still see a very high impedance @ HF on it's output due to inductance in the cabling and inductance in the speaker and it's crossover (if fitted).
Now that I am 6 years further on in my understanding, but still a long way off understanding AC circuit theory, I disagree (in at least with the explanation) with much of what KSTR stated in post7.
Some amplifiers change their gain at HF if the load becomes near infinite. That change in gain makes the phase and gain margins go haywire and can result in oscillation.
Even with the speaker connected, the amplifier may still see a very high impedance @ HF on it's output due to inductance in the cabling and inductance in the speaker and it's crossover (if fitted).
Now that I am 6 years further on in my understanding, but still a long way off understanding AC circuit theory, I disagree (in at least with the explanation) with much of what KSTR stated in post7.
Last edited:
To provide a dissipative load impedance (an impedance whose real part is a small number of ohms), at high frequency. This aids amplifier stability. It also helps enormously if/when the speaker is disconnected or the speaker-fuse blows.
At high frequencies, speaker+cable impedances tend to be reactive (real part of impedance is very high) rather than dissipative (real part of impedance is very low). The RC network, connected in parallel with the speaker+cable, corrects for this.
At high frequencies, speaker+cable impedances tend to be reactive (real part of impedance is very high) rather than dissipative (real part of impedance is very low). The RC network, connected in parallel with the speaker+cable, corrects for this.
Dual power supply avoids the o/p electrolytic capacitor (spk are directly coupled) resulting in flat freq. response,low distortion, tight bass & better damping. The only disadvantage is damaging DC in case of rail or ckt failure if no DC protection ckt is employed.
Single power supply requires the o/p electrolytic capacitor to block DC but is acoustically inferior due to capacitor being in series with spk.
Single power supply requires the o/p electrolytic capacitor to block DC but is acoustically inferior due to capacitor being in series with spk.
Mostly. I've come across a few designs recently that are single supply ~BTL type.
What they do, as much as I have been able to figure from the equivalent circuit schematics in the data sheets, is mirror the outputs with one being inverting. At idle both are at 1/2 power supply so no current flows. For example with a 12V supply they will both output 6V. As the signal come in, one swings from 6V-12V while the other swings from 6V-0.
No output capacitor needed but you can't of course tie the speaker to ground.
correct
A resistive load to stabilize gain and phase margins regardless of connected / unconnected loads. The RC network needs to be located close to the output devices ( chip ) and its return connection to power gnd ( PS bypass caps junction ) IE small loop area. Another reason for good HF caps (0.1 uF) for PS decoupling or the phase and gain margin could suffer without
The series cap value is such that at low frequencies the resistive element isn't dissipating excess output power, yet is purely resistive at the frequency where the closed loop gain is 1. IE Zobel network
Last edited:
if you answered for +vcc and -vee supply then i didnt get your point,please explain me.
As a student of electronics i know that op amp and differential amp only required +vcc and -vee supply then pls tell me does this ic contains any opamp or any differential amplifier circuit inside it?
i am not getting dual power supply use for lm1875.
I have been trying to get this message across for a long time and still I see Members designing Layouts that have the Output Zobel many centimetres away from the output devices and the Power Ground located even further away from the Output devices.
Yup
also the HF PCB decoupling caps ( SMD X7R are the best ) should be the 1st priority or parts placed out. heck on some high current CPU packaging they're located inside.
For linear power ICs ( chip amps ) , CFB op amps, and fast video drivers , the problem of power supply decoupling is key to prevent oscillations due to wiring / PCB parasitics. IE The PSRR rejection ratio is very poor at HF for most op amps.
Last edited:
Linear power amplifiers are, in fact, Opamps, both discrete and integrated circuit designs alike. Some designs can also be powered with a single power supply by connecting the Vcc to + rail and 0V to -rail. However, the input and output nodes then have to be biased to a voltage of half Vcc. That means adding a large capacitor at the output of 2,200 - 10,000 uF, rated to vcc at least. The capacitance value chosen depends on the speaker impedance and audio bandwidth required. Without it, you would have DC of half Vcc applied to your speaker =
Audiophiles believe that capacitors are simply bad for audio so they don't like this method. If you aren't so fussy, using large audio capacitors properly can be very good, and they can eliminate the need for speaker protection relays.
The IC application notes detail what you need to know about applying the IC in all the simple ways. Read them and compare - a single supply version is shown on page 3. http://www.ti.com/lit/ds/symlink/lm1875.pdf
As I remember at turn on the cap needs to be charged up to Vcc/2, meaning at turn on the speaker also gets a quite huge bump without a replay.
So Andrew, have you finally changed your mind about recommending that the network be placed on the speaker output terminals?
"thumps"
More obvious, is that amplifiers with no output caps can produce even larger and sharper thumps. Conversely, I also have built and still own several amps of both types without relays, even with singleton input stages, that don't "thump" noticeably.
It is a matter of design and tweaking whether an amplifier will thump, something well suited to DIY experimenting.
Yes, I have read that assumption but it's just not the case. The thump is caused at the input stage or even in the preamp, as the DC offset there swings wildly as power is applied or removed and the rails rise or fall out of synch. to equilibrium. The output cap is charged relatively slowly to its half-rail DC potential and the need to fit relays was just as much a matter of manufacturing consistency, avoiding the need to correct products that randomly thumped worse than acceptable. This was the problem before before ~1975, when most numbers of amps were still capacitor output. Products would be OK new but change over their lifetime, due to capacitor deterioration. Replacing caps with low ESR types can also make matters worse.As I remember at turn on the cap needs to be charged up to Vcc/2, meaning at turn on the speaker also gets a quite huge bump without a replay.
More obvious, is that amplifiers with no output caps can produce even larger and sharper thumps. Conversely, I also have built and still own several amps of both types without relays, even with singleton input stages, that don't "thump" noticeably.
It is a matter of design and tweaking whether an amplifier will thump, something well suited to DIY experimenting.
No.
The output Zobel must be placed right next to the output device terminals.
The R//L making up the remainder of the Output Thielle Network goes in the cable connecting the PCB to the chassis speaker terminals.
The Pi network Thielle with an extra R+C goes across the speaker terminals.
My recommended locations for these three parts has not changed.
Each part has a physical location that is optimised for the duty that each part performs.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.