Ian Finch,
I'm glad that you also hear the difference with regulated supplies, as most pundits say they should not make any difference whatsoever.
I do not notice a reduced HF response. HF sounds quieter, or not as gritty, which may make it feel the response is lower. I also don't see how a more stable power supply can reduce the HF response.
The possible downside is that a regulator introduces noises of its own, however, I think it is much lower than the power supply ripple anyways.
Hafler Transnova amps use regulated power supply for their input and driver stages.
my JLH class A amp with LT1083 regulated power supply
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
I can´t imagine how having cleaner steadier supply rails can affect "HF response".
Can you suggest what you think might cause such an effect?
If anything, a common "raw" supply can inject ripple and supply transients into the speaker out signal, when clipping, which happens on peaks all the time when system is used between 10 and 20dB below full power, simply because of regular music program dynamic range and normal Music transients .... personally I don´t see how could that "enhance" Music perception.
Same as above comment.
Put another way, my comment was that the effects described could equally be about a reduced HF response, so a little more qualification might have been helpful. That is a different matter to commenting on the technical benefits of regulation or any other means of noise reduction. This seems to have been unclear but I did not intend to imply causality there.
One point I was planning to make is that regulation itself is unnecessary for clean supplies and output - a capacitance multiplier circuit will clean up rails more effficiently if the amplifier's PSRR is inherently poor. Unless you use an undersized or EI power transformer, regulation should always be within the transformer's rated range of probably better than 8% with the common toroidal types used in medium power hifi applications. So I can't see (or hear) much benefit in tightly controlled voltage supplies at a greater cost when, as you indicate, clipping will still introduce noise, distortion and be a problem whatever supply rails are used.
Even so, according to Self, a power amplifier with a PSRR better than most affordable 3 terminal regs can manage, is not at all difficult to design and construct with standard components and well known techniques. I think this approach can and does also result in an amplifier that is as good as they get - just a little more efficiently.
I agree with you. It is largely a question of sonic priorities. More often than not mine align with the benefits regulation brings, yet i acknowledge there usually are trade-offs.
These trade-offs are not exclusive to power amps, nor to solid state. The power amp case is particularly demanding, as it is both hard and expensive to design a wide-band high power regulator.
Has anyone ever designed a high power, fast shunt-reg for a solid state amp? I believe there are commercial examples of such in tube amps.
When building point to point, minimizing the the number of wires saves a lot of time over designs with a lot of transistors that have good power supply rejection ration like the honey badger. Leak delta 70, armstrong 621, dynaco st120, had bad PSRR. Dynaco built a power supply regulator the PC14, that both limited voltage to 67 and limited current to 6.5 a, with two transistors. Unfortunately if a high gain modern transistor is substituted for a burnt up original, the current limit becomes 2 amps. So both djoffe and I have designed modern replacements for PC14 regulator, his in post 19 as the atikita 101. His uses 35 parts for LM3886 output board in the ST120 chassis, mine uses 25 for the original PC15 boards. My PC15 is modified with the djoffe idle bias control board, which improves the cold sound immensely.
In view of Andrew's statement post 3 that regulator outputs interact badly with amp board outputs, I think perhaps using slow slew rate darlingtons in my PC14 regulator is a positive advantage.
Using actual speakers the Peavey SP2-XT, this regulated power supply ST120 with improved bias control, sounds exactly the same at 1.5 Vpp as the Peavey CS800s amp which is unregulated and has no capacitors in the sound path. Hence my theory that the 3rd digit of HD accuracy only matters in sims, not in your living room. Test tracks included piano, bells & cymbals, bass drum hits. Peavey CS800s HD is .03% below the full power 1% HD rating.
In view of Andrew's statement post 3 that regulator outputs interact badly with amp board outputs, I think perhaps using slow slew rate darlingtons in my PC14 regulator is a positive advantage.
Using actual speakers the Peavey SP2-XT, this regulated power supply ST120 with improved bias control, sounds exactly the same at 1.5 Vpp as the Peavey CS800s amp which is unregulated and has no capacitors in the sound path. Hence my theory that the 3rd digit of HD accuracy only matters in sims, not in your living room. Test tracks included piano, bells & cymbals, bass drum hits. Peavey CS800s HD is .03% below the full power 1% HD rating.
Attachments
Hi, Why not use John Linsley Hood's design for the regulated power supply? Download the .pdf and look at the article on the 80W power amplifier.
https://www.americanradiohistory.com/Archive-Electronics-Today/80s/Electronics-Today-1989-05.pdf
Regards
Gary
https://www.americanradiohistory.com/Archive-Electronics-Today/80s/Electronics-Today-1989-05.pdf
Regards
Gary
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.