Designing a LM1875-based amp

[frasco]

I built a one-channel test circuit with minimal components. No pot, 22k at input to ground, 22k feedback, 560R to ground. 1500 & 0.1uF decoupling. I get ~26.5Vdc at the rails, ~10mV offset at output with speaker connected but no sound whatsoever. No hum, no popping, nothing. Could my chip be faulty? All connections are okay as far as I can tell. Are there any other measurements I could do to shed some light on what might be wrong?

edit: The chip doesn't even get warm fwiw.

[danielwritesbac]

Perhaps some technicolor will help resolve the wiring mess?
Beginner's Gainclone, HiFi LM1875, The Amplifier Board
This is not my very best project because it contains the error of omitting the NFB cap and the error of slightly too high voltage. . . so you might not want to propagate those two aspects. However, the phenolic board layout is quick and good.

[frasco]

That's a nice little walkthrough but I fail to see how it would help my case? Fwiw I have a bunch of different sized universal boards and I'll design the layout once I can decide on the schematic (and enclosure) to use.

Let me restate my question: If the chip is fried, wouldn't it be more probable to measure either 0V or 26.5V at the output instead of a steady 10.3mV? And if so, it's possible that the chip is ok, I just need to figure out where the problem lies.

Or are you trying to say that it has to do with the wiring?

[AndrewT]

Originally Posted by frasco

I built a one-channel test circuit with minimal components. No pot, 22k at input to ground, 22k feedback, 560R to ground. 1500 & 0.1uF decoupling.

can you show these components on the schematic?

[frasco]

Originally Posted by AndrewT

can you show these components on the schematic?

Resistors connect to power ground via signal star, speaker terminal connects directly to power ground.

[AndrewT]

You have a DC coupled amplifier.

I take no further part in your experiments.

[frasco]

The Behringer UCA202 I drive it with is AC coupled so I thought I'd fire it up and see how it behaves. Thanks anyways.

[pacificblue]

You are right that a fried amp will show rail voltage at the output in most cases, and the 10 mV you get, appear to be the DC offset of a healthy IC.
Your first step should be to measure continuity from the LM's pins to all other points tosheck for bad solder joints.
The next step should be to make sure there are no shorts in the circuit. Put some isolating material between the LM's pins and cover the unisolated wires, too.

Since your amp is DC-coupled, why don't you use DC as tracing signal? Disconnect the speaker, connect a small DC voltage to the input and see whether you get a high DC voltage at the output. Reverse the DC at the input and check if the output swings in the opposite direction. If you don't get the output, you can trace the DC voltage from the input until the LM's input pin and see, if and where there is a bad solder joint.

As an adjustable DC source you can use a trimmer or potentiometer connected to the rails or a battery. Make sure the trimmer has high impedance so that it doesn't get destroyed by heat dissipation.

[frasco]

Well now I'm embarrassed.

I hadn't connected the grounds at the rectifiers. I just measured 26V between V+ and V- and assumed I was good. After connecting the grounds I get 26V between V+/V- and GND. I was looking in the wrong places the whole time because of my lack of knowledge and experience. At least both of them went up now thanks to this. I feel so dumb though as I would probably have figured it out earlier if I hadn't just assumed the PSU was fine. Check, check and doublecheck everything the next time.

Thank you @pacificblue for your reply even if it didn't solve the problem directly. It will however be useful advice in the future!

@danielwritesbac, even though I don't agree with my beautiful project being a "wiring mess" I'm considering using your guide as a reference to building my finished amp on universal board.

And @AndrewT, you said you wouldn't participate in my undertakings anymore, however I feel like you don't have the option as your great advice is imprinted all over the forums.

Originally Posted by AndrewT

Do not use a chassis at this first build stage.
Be very aware that without the chassis, there is nowhere to connect the Safety Earth.
No pets, no children, no nagging wife in the same room.
Build up one channel complete from input RCA to Output speaker terminals.
Find out if your single channel is hum free and buzz free. Thoroughly check it.
When you are happy that it is perfect, build up your second channel and place it alongside.
It too should be perfect.
Now couple the Main Audio Grounds on each of the two parallel assemblies.
Check for hum and buzz. Attach first one channel of source to either input. Then to the other input. Then two source channels to both inputs. Is it still completely hum free and buzz free?

When you successfully get to this two channel hum free state, you are ready to start visualising how to package the perfect amplifiers into a single chassis.
Electrically everything must remain identical.
Physically (magnetic radiation) the "in chassis assembly" will be very different. But, because the electrical side is identical, you will know that any introduced hum or buzz can only be due to magnetic radiation from and to "loops" in the circuits.

The project continues, thanks again for all the replies! Here's an illustration of the aforementioned problem:

[frasco]

Made a preliminary layout with optional DC-blocking caps.

Signal star ground bottom left. Output with room for optional zobel up left. Power star ground between supply caps. Feedback resistor under the board for shortest possible loop.

Might be in trouble if I find that I need a lot more additional components but I'll cross that bridge when I get there.