I own an adjustable linear bench supply (0-30V, 3A), with +12 and +5 fixed outputs.
I want to buy another one, so it is easier to get +/- rails for solid state circuits, and to be able to use in series with mine for reaching at least 60V dc (for use in audio amplifiers, mainly).
I lean towards linear supplies, just because they are quiter, but honestly I'd rather have a switching supply, for its reduced weight and size.
Considering the same budget, do you think is it OK to get a switching supply for audio stuff?
Thank you!
I want to buy another one, so it is easier to get +/- rails for solid state circuits, and to be able to use in series with mine for reaching at least 60V dc (for use in audio amplifiers, mainly).
I lean towards linear supplies, just because they are quiter, but honestly I'd rather have a switching supply, for its reduced weight and size.
Considering the same budget, do you think is it OK to get a switching supply for audio stuff?
Thank you!
I use linear regulators (+/- 16V-42V, some 8Amp). Big and bulky - yes. If an SMPS will perform significantly less depends on the SMPS and probably also the amplifier you use with it.
A compromize: Buy an SMPS and if you find it has too much ripple/noise at the output, make a capacitance multiplier yourself to be used downstream of the SMPS.
A compromize: Buy an SMPS and if you find it has too much ripple/noise at the output, make a capacitance multiplier yourself to be used downstream of the SMPS.
I have experience prototyping both linear and switching mode signal and power circuits.
The importance of ripple is relative.
One does not usually want a bench supply with HF ripple (switching mode) when prototyping signal circuits. When measuring with oscilloscope signals in the 1mV range the ripple can be found almost everywhere, unless extra effort is put into filtering and grounding optimization, not quite possible with breadboards. However, one does not usually need more than +/- 30V 1A when doing such work, many test circuits just draw dozens of mA, a too high current limit in a breadboard can turn a minor mistake into a major one.
For prototypes exceeding 30V 1A experimentation methodology (and safety) often dictates the use of two independent power supplies, one for signal stages and another for power stages. In normal applications there is almost always ripple at some frequency in supplies of power stages, a SMPS in that application does not make a difference.
A SMPS can be tweaked with additional filtering for making it lab friendly. There are 2 forms of interference: common-mode and differential-mode. Common mode (ground HF current) can be reduced below noise floor with extra CM choke stages, whose design does not have to be very accurate to achieve results. Differential mode is only measurable at the output itself, it does not disturb measurements, it can be reduced with LC filters, but accurate design is required to avoid disturbing the SMPS itself.
The importance of ripple is relative.
One does not usually want a bench supply with HF ripple (switching mode) when prototyping signal circuits. When measuring with oscilloscope signals in the 1mV range the ripple can be found almost everywhere, unless extra effort is put into filtering and grounding optimization, not quite possible with breadboards. However, one does not usually need more than +/- 30V 1A when doing such work, many test circuits just draw dozens of mA, a too high current limit in a breadboard can turn a minor mistake into a major one.
For prototypes exceeding 30V 1A experimentation methodology (and safety) often dictates the use of two independent power supplies, one for signal stages and another for power stages. In normal applications there is almost always ripple at some frequency in supplies of power stages, a SMPS in that application does not make a difference.
A SMPS can be tweaked with additional filtering for making it lab friendly. There are 2 forms of interference: common-mode and differential-mode. Common mode (ground HF current) can be reduced below noise floor with extra CM choke stages, whose design does not have to be very accurate to achieve results. Differential mode is only measurable at the output itself, it does not disturb measurements, it can be reduced with LC filters, but accurate design is required to avoid disturbing the SMPS itself.
I bought this power supply recently for signal circuit research.
Rigol DP832 Programmable Power Supply !!New Offer!!
Rigol DP832 Programmable Power Supply !!New Offer!!
Hi,
considerably cheaper than the Rigol, but an excellent performer is the Siglent SPD3303C. If You don't need programming functionality and TFT displays in a supply, then it is a great alternative. Its even basically programable from the front panel and via USB and dedicated software. Handling is simple and intuitive, the settings (in 10mV/10mA steps) are very precise and the voltages very stable, very low ripple and noise. Switching transients are quite fast and clean with no overshoot. Hardware-wise it features a more than sufficiently sized toroidal transformer. The fully internal heatsink is rather small and relies on forced air cooling by a temperature controlled fan. At idle and at lower power demands the fan remains almost undetectable. Build quality is of course to a price point, but the part and solder quality is nothing to really worry about. The display is a simple 7-segment arrangement -not as fancy as a trendy TFT- but easier to read, showing You just what You need to know. The buttons are arranged in logical groups, feeling a bit wobbly, but they act clean and the important ones also light up when activated (the Rigol looks weird to me .... a HelloKittyonSpeed design with an non-logical button arrangement, the costier 832A even worse than the 832). So far I'm very happy with my Siglent supply. It simply does an excellent job for it's price tag.
jauu
Calvin
considerably cheaper than the Rigol, but an excellent performer is the Siglent SPD3303C. If You don't need programming functionality and TFT displays in a supply, then it is a great alternative. Its even basically programable from the front panel and via USB and dedicated software. Handling is simple and intuitive, the settings (in 10mV/10mA steps) are very precise and the voltages very stable, very low ripple and noise. Switching transients are quite fast and clean with no overshoot. Hardware-wise it features a more than sufficiently sized toroidal transformer. The fully internal heatsink is rather small and relies on forced air cooling by a temperature controlled fan. At idle and at lower power demands the fan remains almost undetectable. Build quality is of course to a price point, but the part and solder quality is nothing to really worry about. The display is a simple 7-segment arrangement -not as fancy as a trendy TFT- but easier to read, showing You just what You need to know. The buttons are arranged in logical groups, feeling a bit wobbly, but they act clean and the important ones also light up when activated (the Rigol looks weird to me .... a HelloKittyonSpeed design with an non-logical button arrangement, the costier 832A even worse than the 832). So far I'm very happy with my Siglent supply. It simply does an excellent job for it's price tag.
jauu
Calvin
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.