Just wanted to post that I have just finished my first amp and it was a LM3875 derivative. I followed the RJM design and after I managed to find my wiring mistake
, it worked fine. My purpose for this amp is not full time listening, just for doing speaker measurements, but my initial impressions, jaded by the fact that its my first amp, are awesome. I have heard a few 'real' systems, and of course, they blow away this implementation, but now I at least feel like a DIYAmp newbie, instead of a wannabe!
After following the presented power star/signal star grounding scheme, the amp has no audible hum and it does attain way too much volume for its size. My heatsink, a modified aluminum TO220 design, is way too small if it were to be an amp for extended listening and I might even upsize it for measuring purposes. It gets hot to the touch when listening at around 92 dB from 3 feet on an 89 dB speaker. My supply is a +/- 12V EI from PE, a standard 35V bridge, Panasonic FC's and 1% metal film resistors and generic capacitors for the rest of the assembly. I used generic RS plastic boxes for the PS and the amp, as I plan on taking it around for testing speakers in different environments. My panel mount connectors were out of stock, so the umbilical does not exist right now and a temporary soldered connection is used to connect the two boxes. The DC offset is 86mV, making me think I have some better matching to do, but I just wanted to hear it to see if it worked and it did. Its not an optimized implementation like many of the posters here do, but it is interesting. I might heatsink it properly and try some listening tests, just to tell myself if I'm hearing through newbie rose colored headphones or whether it really does sound good.
This weekend, I hope to test for oscillation (using a 10ohm resistor and O-scope with a test signal???) and try to determine what needs to be changed to get the DC offset down. I'm not going to hook it up to one of my better loudspeakers until that's fixed.
Simply put, thanks for all the hype around the LM3875 and the wealth of resources provided by many members. I have all the parts for an Elliot P3A and many of the parts for Aleph 2's and I think I have the bug. Ironically, I just tested a million-dollar machine I designed at work and felt similar when this $50 chip amp worked. Engineers get weird highs!!!
Thanks. Hope I continue to learn and might contribute to the hobby like so many of you do!
Best Regards,
Sandy.

After following the presented power star/signal star grounding scheme, the amp has no audible hum and it does attain way too much volume for its size. My heatsink, a modified aluminum TO220 design, is way too small if it were to be an amp for extended listening and I might even upsize it for measuring purposes. It gets hot to the touch when listening at around 92 dB from 3 feet on an 89 dB speaker. My supply is a +/- 12V EI from PE, a standard 35V bridge, Panasonic FC's and 1% metal film resistors and generic capacitors for the rest of the assembly. I used generic RS plastic boxes for the PS and the amp, as I plan on taking it around for testing speakers in different environments. My panel mount connectors were out of stock, so the umbilical does not exist right now and a temporary soldered connection is used to connect the two boxes. The DC offset is 86mV, making me think I have some better matching to do, but I just wanted to hear it to see if it worked and it did. Its not an optimized implementation like many of the posters here do, but it is interesting. I might heatsink it properly and try some listening tests, just to tell myself if I'm hearing through newbie rose colored headphones or whether it really does sound good.
This weekend, I hope to test for oscillation (using a 10ohm resistor and O-scope with a test signal???) and try to determine what needs to be changed to get the DC offset down. I'm not going to hook it up to one of my better loudspeakers until that's fixed.
Simply put, thanks for all the hype around the LM3875 and the wealth of resources provided by many members. I have all the parts for an Elliot P3A and many of the parts for Aleph 2's and I think I have the bug. Ironically, I just tested a million-dollar machine I designed at work and felt similar when this $50 chip amp worked. Engineers get weird highs!!!
Thanks. Hope I continue to learn and might contribute to the hobby like so many of you do!
Best Regards,
Sandy.
any pictures?
also, my gainclone runs with +/-15v rails, and it never even gets warm, even at full volume. i use the OPA541, so i suppose its different to the LM3875.
also, my gainclone runs with +/-15v rails, and it never even gets warm, even at full volume. i use the OPA541, so i suppose its different to the LM3875.
Welcome to the club Sandy but be warned, this hobby is highly addictive! 
85mV does seem a bit high so you are wise not to use one of your 'better' speakers.
Did you build the inverting or non-inverting design?

85mV does seem a bit high so you are wise not to use one of your 'better' speakers.
Did you build the inverting or non-inverting design?
A little more
Matttcattt: I posted pictures on my website if anyone is interested. www.tshouston.net ,audio button, mic GC button.
Nuuk: I built the RJM non-inverting version, as I wanted it to be stable under difficult situations and the minimized versions seem prone to oscillation unless done perfectly. I used a +/-12 EI transformer so 4 ohm loads wouldn't be critical (I hope!!!). I actually intend not to run it over a nominal 1 watt, as it is supposed to be for dedicated measuring. I have spare 3875's, so I will end up trying the non-inverted design soon, but for listening, not measuring. I want to check out the theory and measure my actual resistors to see if the 86mV is as dictated due to components. Its always good to try to understand the theory, instead of just tinkering. Either way, I'll probably put a trim pot in to zero it out.
Jam: You know what's in the works and its a deuce! Honestly, I think that's going to be a fall or winter project, though. Isn't that what class A is all about!?!?!? BTW: you might be interested in checking out the site and clicking on the home shop button. I guess I have no excuses now. . .
Thanks for the help/interest!
Sandy.
Matttcattt: I posted pictures on my website if anyone is interested. www.tshouston.net ,audio button, mic GC button.
Nuuk: I built the RJM non-inverting version, as I wanted it to be stable under difficult situations and the minimized versions seem prone to oscillation unless done perfectly. I used a +/-12 EI transformer so 4 ohm loads wouldn't be critical (I hope!!!). I actually intend not to run it over a nominal 1 watt, as it is supposed to be for dedicated measuring. I have spare 3875's, so I will end up trying the non-inverted design soon, but for listening, not measuring. I want to check out the theory and measure my actual resistors to see if the 86mV is as dictated due to components. Its always good to try to understand the theory, instead of just tinkering. Either way, I'll probably put a trim pot in to zero it out.
Jam: You know what's in the works and its a deuce! Honestly, I think that's going to be a fall or winter project, though. Isn't that what class A is all about!?!?!? BTW: you might be interested in checking out the site and clicking on the home shop button. I guess I have no excuses now. . .
Thanks for the help/interest!
Sandy.
Re: A little more
nice! looks better than mine, but the same quality pictures... 😎
Sandy H. said:Matttcattt: I posted pictures on my website if anyone is interested. www.tshouston.net ,audio button, mic GC button.
Nuuk: I built the RJM non-inverting version, as I wanted it to be stable under difficult situations and the minimized versions seem prone to oscillation unless done perfectly. I used a +/-12 EI transformer so 4 ohm loads wouldn't be critical (I hope!!!). I actually intend not to run it over a nominal 1 watt, as it is supposed to be for dedicated measuring. I have spare 3875's, so I will end up trying the non-inverted design soon, but for listening, not measuring. I want to check out the theory and measure my actual resistors to see if the 86mV is as dictated due to components. Its always good to try to understand the theory, instead of just tinkering. Either way, I'll probably put a trim pot in to zero it out.
Jam: You know what's in the works and its a deuce! Honestly, I think that's going to be a fall or winter project, though. Isn't that what class A is all about!?!?!? BTW: you might be interested in checking out the site and clicking on the home shop button. I guess I have no excuses now. . .
Thanks for the help/interest!
Sandy.
nice! looks better than mine, but the same quality pictures... 😎
Cheers!!Sandy😎
Looked at your site.....I made my tripods stands with 1 1/2 schedule 40 pipe.and filled them with sand..hahaha
Looked at your site.....I made my tripods stands with 1 1/2 schedule 40 pipe.and filled them with sand..hahaha
Re: A little more
I love this kind of DIY...
/Hugo 🙂
The "Cake box" Lab Supply is the best!!!Sandy H. said:I posted pictures on my website if anyone is interested.
Sandy.
I love this kind of DIY...
/Hugo 🙂
Having your cake and power supply too!
JOE DIRT: Yeah, mine are a little too flimsy for the heavy MTM's, but the 2-ways I originally had on them were fine. I'm still playing with placement etc, so I figured why bother making anything different for now. I think the next ones will be from a single pipe, not 3, or I will attach a gusset in the middle.
Netlist: Sadly enough, its still in that cake box. I got done with the aluminum part of the real case and loaned out my router, so the MDF part isn't complete. In the end, its probably best, as I have learned some better wiring practices and will re-wire it when I install it in the real box.
Sandy.
JOE DIRT: Yeah, mine are a little too flimsy for the heavy MTM's, but the 2-ways I originally had on them were fine. I'm still playing with placement etc, so I figured why bother making anything different for now. I think the next ones will be from a single pipe, not 3, or I will attach a gusset in the middle.
Netlist: Sadly enough, its still in that cake box. I got done with the aluminum part of the real case and loaned out my router, so the MDF part isn't complete. In the end, its probably best, as I have learned some better wiring practices and will re-wire it when I install it in the real box.
Sandy.
- Status
- Not open for further replies.
- Home
- Amplifiers
- Chip Amps
- Just finished