Messing with Electronics
New Member
- Joined
- Feb 2, 2023
- Messages
- 4
- Likes
- 29
Hello everyone,
I am a vintage audio enthusiast who enjoys designing, modifying, and repairing tube amplifiers. I'm also an objectivist and a long-time reader of ASR, but I've always wished there was more tube-related objectivist content on this forum. As we all know, tube amplifiers are far from "ideal" amplifiers, with their high distortion, high noise levels, and low damping factor. But, in my opinion, this is what makes them enjoyable -- tube amplifiers are bad enough to actually sound different, whereas competently-designed solid state gear is all pretty much the same.
Anyways, I bought an E1DA Cosmos ADC and an SMSL M500 MK II so that I could do proper measurements (using REW) on all the gear I design and restore. Please let me know if anything in my graphs looks off; I am relatively new to measurements with REW.
Without further ado, the first amplifier to cross my bench since is a Heathkit A-8A tube integrated amplifier, which I restored and subsequently tested. I thought it would be fun for ASR users to see what kind of performance you could expect from a relatively cheap 1950s-era amplifier. Although, to be clear, there were much cheaper -- the A-8A uses nice components and is generally well-designed. It's also quite a rare amplifier, and I doubt I'll be able to find another for stereo use.
The above picture shows the amplifier, post-restoration, in the configuration I tested. For any tube fans out there, from left to right: NOS 6SJ7, NOS 6SJ7, JJ 6SN7, and two factory-matched Tung Sol reissue 5881s. The rectifier is a used Tung Sol 5U4GB, not that it makes a difference. I chose these tubes because they offered the best measured performance (although the difference was only a few db THD).
As a disclaimer, many factors can affect the performance of vintage electronics. Even with a restoration, it's hard to know whether the results you're seeing reflect the design's optimum performance. With this amplifier, I believe the distortion results I'm seeing are representative. However, I suspect that rebuilding the amplifier with a better grounding scheme, shorter wire runs for the input jacks, and new connectors would reduce noise considerably.
Here are the measurements.
Spectrum @ 5 watts (the output level Amir uses for power amp comparison):
As you can see, THD is -50.2 dB and THD+N is -49.6 dB.
At lower power levels, noise dominates. There is an audible hum from the speaker with no input signal, but it's not really noticeable during actual listening.
Here's the spectrum at 15 watts, which is close to the maximum you can get with reasonable distortion (hard clipping begins at 17-18 watts):
Below is the frequency response. I'm still new to REW, so the units are wrong, but I verified with my oscilloscope that output power during the sweep was 1 watt:
As you can see, there's some treble roll-off, but nothing too severe inside the 20-20k range.
For those interested, here's the square wave response at 10 khz with an 8 ohm resistive load and a .1 uf capacitor shunting the load. As the frequency response graph would suggest, it's not pretty, but the amplifier is quite stable -- adding the capacitor does basically nothing to the waveform.
Anyways, hopefully this interests some of you. There's a longer write-up with pictures and additional measurements on my website, but I don't want to make this post excruciatingly long.
Subjectively, I enjoyed this amplifier, although it's clearly not the pinnacle of hifi. It pumped out way more bass from my small bookshelves than my solid state receivers, probably thanks to the low damping factor and strong LF response.
I am a vintage audio enthusiast who enjoys designing, modifying, and repairing tube amplifiers. I'm also an objectivist and a long-time reader of ASR, but I've always wished there was more tube-related objectivist content on this forum. As we all know, tube amplifiers are far from "ideal" amplifiers, with their high distortion, high noise levels, and low damping factor. But, in my opinion, this is what makes them enjoyable -- tube amplifiers are bad enough to actually sound different, whereas competently-designed solid state gear is all pretty much the same.
Anyways, I bought an E1DA Cosmos ADC and an SMSL M500 MK II so that I could do proper measurements (using REW) on all the gear I design and restore. Please let me know if anything in my graphs looks off; I am relatively new to measurements with REW.
Without further ado, the first amplifier to cross my bench since is a Heathkit A-8A tube integrated amplifier, which I restored and subsequently tested. I thought it would be fun for ASR users to see what kind of performance you could expect from a relatively cheap 1950s-era amplifier. Although, to be clear, there were much cheaper -- the A-8A uses nice components and is generally well-designed. It's also quite a rare amplifier, and I doubt I'll be able to find another for stereo use.
The above picture shows the amplifier, post-restoration, in the configuration I tested. For any tube fans out there, from left to right: NOS 6SJ7, NOS 6SJ7, JJ 6SN7, and two factory-matched Tung Sol reissue 5881s. The rectifier is a used Tung Sol 5U4GB, not that it makes a difference. I chose these tubes because they offered the best measured performance (although the difference was only a few db THD).
As a disclaimer, many factors can affect the performance of vintage electronics. Even with a restoration, it's hard to know whether the results you're seeing reflect the design's optimum performance. With this amplifier, I believe the distortion results I'm seeing are representative. However, I suspect that rebuilding the amplifier with a better grounding scheme, shorter wire runs for the input jacks, and new connectors would reduce noise considerably.
Here are the measurements.
Spectrum @ 5 watts (the output level Amir uses for power amp comparison):
As you can see, THD is -50.2 dB and THD+N is -49.6 dB.
At lower power levels, noise dominates. There is an audible hum from the speaker with no input signal, but it's not really noticeable during actual listening.
Here's the spectrum at 15 watts, which is close to the maximum you can get with reasonable distortion (hard clipping begins at 17-18 watts):
Below is the frequency response. I'm still new to REW, so the units are wrong, but I verified with my oscilloscope that output power during the sweep was 1 watt:
As you can see, there's some treble roll-off, but nothing too severe inside the 20-20k range.
For those interested, here's the square wave response at 10 khz with an 8 ohm resistive load and a .1 uf capacitor shunting the load. As the frequency response graph would suggest, it's not pretty, but the amplifier is quite stable -- adding the capacitor does basically nothing to the waveform.
Anyways, hopefully this interests some of you. There's a longer write-up with pictures and additional measurements on my website, but I don't want to make this post excruciatingly long.
Subjectively, I enjoyed this amplifier, although it's clearly not the pinnacle of hifi. It pumped out way more bass from my small bookshelves than my solid state receivers, probably thanks to the low damping factor and strong LF response.
Last edited: