| Electronics > Projects, Designs, and Technical Stuff |
| SMPS replacement for audio power amp transformer |
| (1/2) > >> |
| BlackICE:
I have a old Hafler DH-220 I'm haven't used for decades and thinking of bringing back to life. I have read about the aging of electrolytic caps and it wouldn't be too hard or expensive to replace them all with the exception of the large filter caps on the power supply. They are 10K uf at 75V. That got me thinking about what if I replaced the large heavy CT transformer that is about +-44V RMS at 4 amps with two 48V 5+A SMPS. The box would be lighter, more energy and efficient and I could choose the SMPS current output and size the "filter" caps depending on the peak current I wanted. I would bet with a higher current SMPS the amp could output much more than it's spec. 110w per channel for a few seconds before any thermal overloaded would occur. My main concern would be the noise of the SMPS causing havoc and showing up at the speaker output stage degrading the sound quality. Any opinions are welcomed. An example of a cheap (bad/dangerous?) off the shelf SMPS. It probably cost less than a new filter cap. https://www.ebay.com/itm/48V-4A-5A-200W-AC-DC-Power-Supply-Converter-Adapter-SMPS-Board-Voltage-Transform/264468367469?hash=item3d938b086d:g:oP4AAOSwpLpdgHja |
| German_EE:
There are two problems with your plan. Firstly, as you pointed out, switched mode supplies are noisy. Even with decoupling capacitors spread around like candies you will still hear background interference and any hum loops will increase this problem. Secondly, your amplifier transient response will suffer. 10,000 uf @ 75V is a lot of capacitance to lose and this will show in the response of your output stages. Please do not do this, audio power amps need big transformers and big capacitors. |
| BlackICE:
I wouldn't necessarily omit the filter caps. I could even use larger ones. Then by using a higher current SMPS the amp would have the same or better transient response as well as better intermediate response. However as we both said the noise issue still persist. Maybe shielding the PS and locating it remotely would mitigate the noise issues but that would be a packaging issue and little real use gain in my case, but the specs may look better. |
| MagicSmoker:
Meh, the claim that SMPSes produce too much noise to be used in audio is overblown, IMO. Sure ripple and noise from a hard-switched SMPS are going to be worse than the typical linear power supply, but given that the switching frequency will invariably be well above 20kHz it won't be audible even if PSRR of the amplifier has dropped to 0 by then. The real problem with SMPSes in this application is the potential for interaction between two control loops - the voltage regulation loop in the switcher and the negative feedback loop in the amplifier. Pretty much every car audio amplifier sidesteps this problem by running the switcher open-loop, and that's not a bad solution, really; I would suggest always using pulse-by-pulse current limiting, however. |
| David Hess:
--- Quote from: BlackICE on December 28, 2019, 11:17:20 am ---I wouldn't necessarily omit the filter caps. I could even use larger ones. Then by using a higher current SMPS the amp would have the same or better transient response as well as better intermediate response. However as we both said the noise issue still persist. --- End quote --- The power supply is sized for the average output current so additional output capacitance is likely necessary to handle the surge current. --- Quote from: MagicSmoker on December 28, 2019, 12:47:41 pm ---The real problem with SMPSes in this application is the potential for interaction between two control loops - the voltage regulation loop in the switcher and the negative feedback loop in the amplifier. Pretty much every car audio amplifier sidesteps this problem by running the switcher open-loop, and that's not a bad solution, really; I would suggest always using pulse-by-pulse current limiting, however. --- End quote --- The bulk output filtering capacitor damps any interaction. The reason the switcher is run open-loop is that it is not a switching regulator. The power amplifier does not require a regulated supply so instead an inverter is used to multiply the automotive DC voltage to a higher level allowing more power into the load presented by the loudspeaker. (1) Since an inverter is used instead of a switching regulator, a large energy storage inductor is not required and with a transformer inverter, ripple current at the input and output are low relaxing the requirements for the input and output capacitors. (1) This is also why 4 and 2 ohm speakers are commonly used in automotive applications instead of 8 ohm speakers. |
| Navigation |
| Message Index |
| Next page |