I have built an LM1875 based amp more-or-less using the TI data sheet circuit for dual rail which does not have any capacitor in the output (between pin 4 and the speaker). Before connecting to speakers I measure 5-7VDC between pin 4 and ground and if I measure the current it shows about 1A. I am using 8ohm resistor as bench load whilst testing and notice (naturally with 1A) significant heat.
Is it required to put a 2200uF (seems most common value) or not in the output, if not, why might I see this voltage, and the fact that I measure 1A current its not a phantom or floating voltage.
Also I see some circuits with a 1 Ohm and series 22nF cap to ground, seems this improves stability? Correct? Required?
The DC split rail is 24-0-24V DC.
Any thoughts or guidance. I have powered for short periods so far.
Is it required to put a 2200uF (seems most common value) or not in the output, if not, why might I see this voltage, and the fact that I measure 1A current its not a phantom or floating voltage.
Also I see some circuits with a 1 Ohm and series 22nF cap to ground, seems this improves stability? Correct? Required?
The DC split rail is 24-0-24V DC.
Any thoughts or guidance. I have powered for short periods so far.
Last edited:
For dual rail, instead of an output DC blocking capacitor, you should use a capacitor
to return the NFB loop to ground. This sets the DC gain to unity (0dB).
An output Zobel to ground, as recommended by the mfr, is also required.
See the diagram here, follow it closely in all respects:
to return the NFB loop to ground. This sets the DC gain to unity (0dB).
An output Zobel to ground, as recommended by the mfr, is also required.
See the diagram here, follow it closely in all respects:
This schematic is exactly what i have, no variation even in the values. Still don't understand why i have DC voltage present on the output.
That TI application circuit has unity DC gain, so the output offset DC voltage should be no more than 15mV DC.
Either the LM1875 is defective, or there is a wiring error.
The 1A idle current perhaps indicates HF oscillation, so check (with a load) on a scope.
If you are not using the Zobel, install it now, it is necessary for stability.
Is the circuit built on a pcb with short leads and the recommended supply bypassing?
Either the LM1875 is defective, or there is a wiring error.
The 1A idle current perhaps indicates HF oscillation, so check (with a load) on a scope.
If you are not using the Zobel, install it now, it is necessary for stability.
Is the circuit built on a pcb with short leads and the recommended supply bypassing?
yes on a PCB with plenty meat for ground. I will work with the sig gen and scope later and check, but it looks that for some reason the devices maybe faulty (it was my first assumption) as not much else makes sense..
If there is a Zobel in your build, odds are the chip is questionable, but post a photo of both sides of the pcb.
The decoupling and the Zobel should all be connected close to their respective pins, with short grounds also.
Did you buy the LM1875 from a major distributor?
The decoupling and the Zobel should all be connected close to their respective pins, with short grounds also.
Did you buy the LM1875 from a major distributor?
Last edited:
I will, have to disassemble from the chassis/heatsink. I got the devices from RS Components in South Africa, very reputable.
This is solved, replaced both devices though could find no reason for the failure.
Odd, if they are all from the same source, maybe there was static damage from handling.
Could the supply voltages have been accidentally reversed initially, causing damage?
Could the supply voltages have been accidentally reversed initially, causing damage?