I was bored this afternoon, so I decided to try using a different amp to power my Speakerlab K speakers. I have a pair Monarchy Audio SM70 amplifiers that I bought a long time ago (close to 20 years??), and they have been sitting on a shelf for years. They are supposed to be class A with "Zero Feedback" (those words are proudly inscribed on the face plate of one of the two amps). So, I pulled out one SM70 amp, opened its top panel, and started probing around with my multimeter to make sure that all the voltages in the amp appear proper.
Then I found a problem. The SM70 uses an OPA2604 as the voltage gain stage to drive the output transistors operating in the class A mode (perhaps up to 15W?). I was a bit baffled by the measured rail voltages for the 2604: ~ +7.6V and -8.5 V, as provided by a 317/337 regulator pair. I was alarmed by the two voltages not only being quite different in magnitude and but also being rather low. In contrast, the voltages that go to the output transistors are around +-26V. I then pulled out the other SM70 amp, and its opamp rail voltages look even worse at +8.5V and -12.5V, while the voltages for the output transistors are ~+-30V.
My understanding of the design of this amp is that the rail voltages of the 2604 actually limit the output of the amp, and the OPA2604 was chosen because of its high max supply voltages (+-24V). The SM70 in the stereo mode is supposedly able to put out 25W, with a gain of 34 and a stated input sensitivity of 0.7V. The measured rail voltages are just way too low for the stated amp output, and the first SM70 amp I tested will go into clipping in the voltage gain stage with only <0.2V of input! I was then surprised and dismayed to find out that the output of each of the 317/337 regulators is set by a tiny trim pot. Those trip pots are super-sensitive to adjustments, and just putting a small screw drive on them could change the voltages by 1 volt or so. It appears that the aging of the trim pots caused the rail voltages to drift way off the intended values. It is rather disappointing to see that the opamp rail voltages, which are supposed to be critical in this design, are done in such a less-than-robust way.
I want to replace those lousy trim pots with resistors of fixed values. The problem is that I don't know what the correct rail voltages for the OPA2604 should be. I searched for the schematic of the SM70, but could not find anything. If you know the right rail voltage values, please let me know.
Kurt
Then I found a problem. The SM70 uses an OPA2604 as the voltage gain stage to drive the output transistors operating in the class A mode (perhaps up to 15W?). I was a bit baffled by the measured rail voltages for the 2604: ~ +7.6V and -8.5 V, as provided by a 317/337 regulator pair. I was alarmed by the two voltages not only being quite different in magnitude and but also being rather low. In contrast, the voltages that go to the output transistors are around +-26V. I then pulled out the other SM70 amp, and its opamp rail voltages look even worse at +8.5V and -12.5V, while the voltages for the output transistors are ~+-30V.
My understanding of the design of this amp is that the rail voltages of the 2604 actually limit the output of the amp, and the OPA2604 was chosen because of its high max supply voltages (+-24V). The SM70 in the stereo mode is supposedly able to put out 25W, with a gain of 34 and a stated input sensitivity of 0.7V. The measured rail voltages are just way too low for the stated amp output, and the first SM70 amp I tested will go into clipping in the voltage gain stage with only <0.2V of input! I was then surprised and dismayed to find out that the output of each of the 317/337 regulators is set by a tiny trim pot. Those trip pots are super-sensitive to adjustments, and just putting a small screw drive on them could change the voltages by 1 volt or so. It appears that the aging of the trim pots caused the rail voltages to drift way off the intended values. It is rather disappointing to see that the opamp rail voltages, which are supposed to be critical in this design, are done in such a less-than-robust way.
I want to replace those lousy trim pots with resistors of fixed values. The problem is that I don't know what the correct rail voltages for the OPA2604 should be. I searched for the schematic of the SM70, but could not find anything. If you know the right rail voltage values, please let me know.
Kurt
Last edited:
The SM70 was mentioned here with a schematic Monarchy SM-70 6 months report ;-)
I suspect the voltage rails are unmatched and are deliberately low voltages, to match the
requirements of the circuit.
Implementing fixed resistances instead of trims in 317 and 337 refer to relative datasheets
http://www.ti.com/lit/ds/symlink/lm317.pdf
http://www.ti.com/lit/ds/symlink/lm137.pdf
I suspect the voltage rails are unmatched and are deliberately low voltages, to match the
requirements of the circuit.
Implementing fixed resistances instead of trims in 317 and 337 refer to relative datasheets
http://www.ti.com/lit/ds/symlink/lm317.pdf
http://www.ti.com/lit/ds/symlink/lm137.pdf
Problem fixed
Thanks Chris for providing the links.
I think I've fixed the problem. The fix is very easy -- replacing each tiny trim pot with a fixed resistor of 87ohm. The opamp supply rails now are +19.8V and -20.0V.
To refresh my memory of how the amp sounds, I put it in my main system to drive the Speakerlab K speakers. The amp worked and sounded generally fine. The sound was somewhat "mellow" or polite, but there seemed to be a lack of transparency, and I kept feeling that it's not my cup of tea. When I put my LME49810 amp (build from amp boards from China) back into the system, the transparency and layered soundstage returned, and I breathed a sigh of relief. I don't think I can live with the SM70, not only because of its sound being not to my liking, but also because of the copious amount of heat it generates. It is a pity, now that I've fixed the SM70, it will be put back on the shelf.
Thanks Chris for providing the links.
I think I've fixed the problem. The fix is very easy -- replacing each tiny trim pot with a fixed resistor of 87ohm. The opamp supply rails now are +19.8V and -20.0V.
To refresh my memory of how the amp sounds, I put it in my main system to drive the Speakerlab K speakers. The amp worked and sounded generally fine. The sound was somewhat "mellow" or polite, but there seemed to be a lack of transparency, and I kept feeling that it's not my cup of tea. When I put my LME49810 amp (build from amp boards from China) back into the system, the transparency and layered soundstage returned, and I breathed a sigh of relief. I don't think I can live with the SM70, not only because of its sound being not to my liking, but also because of the copious amount of heat it generates. It is a pity, now that I've fixed the SM70, it will be put back on the shelf.
Last edited:
- Status
- Not open for further replies.