Discrete 35W kit vs LM1875

This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.

i want to put together my first diy stereo amp and have these two options:

1. LM1875: QKits Electronic Kits: QK50, 25W Mono Amplifer Kit
2. a discrete 35W amp: QKits Electronic Kits: FK659, 35 Watt Stereo Power Amplifier Kit R1%

the power supply will be simple fully rectified dual unregulated with between 10K uF to 20K uF per channel.
Speaker for now is a single 8 ohm 8 inch crossover-free driver per channel. im happy with 25W per channel.

if this project goes smoothly, it will probably get upgraded to a bi-amped system with 4 identical amp blocks and this crossover: Linkwitz-Riley Electronic Crossover (the speaker will be polk audio rtia1 with crossovers removed)

Can someone please let me know the answers to these questions:

1. since i happen to have one, can i use a 35V center tapped transformer for kit 2? what kind of wattage should i expect per channel if this transformer can be used?
2. should i use an opamp buffer for the input stage? source is going to be a line level signal.
3. most importantly: given the above, which of these amps will sound better? im hoping for a low noise floor, clean mids and highs, and generally low strain at moderate volumes during prolonged use.
4. are these speaker protection units good enough: QKits Electronic Kits: FK650, Speaker Protector Kit (Stereo) (low pass filter threshold for the high-frequency channels will be changed if i upgrade to a bi-amp)
5. how will a stereo of bi-amp version sound compared to say a rotel, NAD 3125, etc?
5. finally, where can i read more about how to lay out an amp chassis to minimize noise?

Thanks a tonne!
1) +/-25V DC rails?
Same as the LM1875 kit , around 25W per channel or whatever it puts out.
The "tank" which drives your speaker is the Power Supply, and the "faucet" which controls the flow is the amplifier, so it won't feed what was not available there in the first place.

2) no.
A line source stage is both high level and low impedance, so it drives most "anything".

3) the 1875 is *excellent*
My crystal ball is a little cloudy today so I can't "see" the second one without schematic ;) , but it must be good too.
Power amp internal noise isn't usually a factor.

4) mmmmmhhhh, crystal ball still cloudy :(
Anyway, it should work as advertised, no Nuclear Science involved here :)

5) probably the same.

6) google is your friend.
Probably you'll find something here at DIY, Rodd Elliott has always excellent advice on his site, get one of the "good" amp design books (good investment in any case), etc.

Good luck with your project ;)
the discrete amp in link2 has a problem.
The output protection can blow the VAS transistor.
The VAS transistor needs a current limiter when ever the output has protection fitted.

As an aside the output protection has the same DC current limits as it has for transient current limits.
This may make the output protection audible, even when the amplifier is well within it's safe operating area.

Once you have learned how, you could modify the circuit to include these improvements.
The 35W amp in the second link looks like a disaster waiting to happen.

2 main problems (aside from what Andrew mentioned):

  • The driver transistors are TO92 devices with no heatsinking. They're going to burn as soon as you turn up the volume. When I saw that I thought the output devices might be Darlingtons, but no - they're TIP41 & TIP42.
  • The Vbe multiplier and driver transistors are nowhere near the main heatsink or each other so there's no temperature compensation. On the upside, the 1/2 Ohm emitter resisters at the output should prevent outright thermal runaway, but I'd expect crossover distortion and/or excesive heating of the output devices (if the drivers don't die first).

edit: ...and there's no supply decoupling capacitors.
Last edited:
Joined 2010
Paid Member
Don't forget folks, most Ebay schematics posted on the net are deliberately sabotaged so that the copyists among the kit sellers are confused for some time at least and need to do a bit of work to get it like the original. That's true to some extent of the designs of even ESP and probably many other kit sellers now.

So don't get flustered until you see what the client receives by email or posted upon purchase.
whoa! thanks for the heads up! the ups guy just dropped the package off. im going to try the #2 with an old car speaker and look for magic smoke. with any luck, its just false advertising! worst case, i solved my find-lots-of-identical-heat-sinks problem for a lm1875 stereo biamp project :)

i dont know nearly enough electronics yet to be able to fix the problems with #2 (or even fully comprehend them), but my goal is to be able to eventually design my own amp, so fixing this will be a notable milestone at some point
Joined 2010
Paid Member
Yes, it is a case of false advertising if we expect 35W RMS 8R from TIP41/42 reliably but it is at least not rated to do more at 4R. How one ensures that in operation is anybody's guess.
I did not suggest that the kit was better or worse than represented, simply that on recent history, we can't assume the schematic is correct or defines what you may receive. Making bald statements about performance based on what appears on the published schematic there is a mistake in my view, until it is confirmed that it is correct. It appears to be of US origin and probably subject to closer scrutiny than some other sources but the site kit prices generally seem to suggest otherwise.

We don't really know the accuracy of the information based on the schematic, is my point. Perhaps others can see more in the pics than I and perhaps not really.
By the way, +/-35V rails is *way* too much for 35W/8r .
45/50W would be much closer to truth, so there we find the first *mistake".

As Ian Finch suggests, probably the kit builders receive an "oops!" note with the kit, and/or are sold the proper transformer.

And for real 35W and TO220 outputs, TO92 drivers are fine.

In fact I sold tons of "small" 30/35W guitar combos driving TIP41/42 with BC639/640 .
0.8W dissipation .... at 90ºC case temperature :eek:

And if you use larger pads in the PCB, up to 1 W dissipation.
Also 80V/1A rating .
Rugged little beasts.
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.