Not exactly same MeanWell SKU but tomchr recommend use these for his new high wattage power amps and it takes some discussions over there between linear or SMPS, link: http://www.diyaudio.com/forums/vend...wer-amp-extremely-low-thd-20.html#post5355207
Thanks BRYTT, I'll go and have a look and see 'what's what' - I reckon Tom would have a pretty good grasp on the ins/out of the switcher supplies - interesting times, eh!
Agree and sometimes old stuff seems be best or at least more close to what we like by heart, but some of the new stuff also seems find its good use cases and think Tom is good on the ins/out from a scientific point of view, then we can take it or leave it if that old heart won't change opinion 😛
I have a couple of the 400mA 24v MeanWell PCB mounted AC/DC switchers in hand for my next desktop headphone amp. Have not tried yet but they look nicely made. Very compact.
Had a look on his website and the thread - seems well worth a try - i did notice he adds a smd choke and an electro cap on the outputs of his smd power supply - there must be a reason for it
I had a look on the Mean Well site for those RPS-200-24 units (200va, 24V, 8A) and it seems they're perfectly happy connecting 2 of them up to give the +/- 24 volt to use for the First Watt F6 & F7 amps - they're about U$60 ea from Mouser - not sure if adding the CMx would benefit anything but interesting to find out ...
I had a look on the Mean Well site for those RPS-200-24 units (200va, 24V, 8A) and it seems they're perfectly happy connecting 2 of them up to give the +/- 24 volt to use for the First Watt F6 & F7 amps - they're about U$60 ea from Mouser - not sure if adding the CMx would benefit anything but interesting to find out ...
If you want to use a rather large CRC or CLC filter after the SMPS then sometimes they auto shutdown due to current inrush. Perhaps a CL60 in series can solve that. I use amp Mx to reduce ripple, drop voltage, and provide soft startup ramp and CRC or CLC provides graceful shutdown. The cap Mx makes turn on pop a thing of the past. No relay with delay needed. If you don’t need voltage drop - maybe use without cap Mx. But low cost SMPS connected in series for dual rail Class A supplies can work very well.
Just be careful, as the usual way to attach a pair of power supplies has the ground/common of one being attached to the + of the other. That's fine if they are isolated, but a problem if the ground is a hard-ground (ie not floating).
Please ignore the . . . . in the crude ascii art below (necessary to make things line up)
supply 1
| |
| +--- positive ----------- positive rail
+----- ground ----+
. . . . . . . . . . . . . +------ common
+----- positive ---+
| +--- ground ------------ negative rail
| |
supply 2
If the ground of supply 1 is floating, you can attach it to the positive of supply 2. However if the ground of supply 1 is a hard ground (referenced to something like the case or the ground of the power in), attaching the positive from supply 2 will result in it shorting out. Did that make sense?
It’s a snubber so any value from 1ohm to 10ohms is fine (ideally needs to be viewed on O-scope and adjust value for best snubbing efficiency) - double your 4.7 for 2.35ohms is good too.
I've been thinking about using this cap multiplier along with a rail splitter to get a bipolar supply from a single SMPS and boost converter.
There are a couple of variations of the splitter here: Virtual Ground (A Regulated Rail Splitter)
My question is this: would it be better to split first and then have both +ve and -ve cap multipliers (plus CRC), or cap multiplier / CRC first followed by the splitter?
There are a couple of variations of the splitter here: Virtual Ground (A Regulated Rail Splitter)
My question is this: would it be better to split first and then have both +ve and -ve cap multipliers (plus CRC), or cap multiplier / CRC first followed by the splitter?
I had to use two SMPS - I don’t think you can do it with a single SMPS. Not sure how a rail splitter works. For a Class A amp the dissipation at 4 or 5 amps is kind of high - 20w. The high current draw on the DC-DC step up also seems to make it noisier. You can check out my attempt here:
Aksa Lender P-mos Hybrid Aleph (ALPHA) Amplifier
http://www.diyaudio.com/forums/soli...rid-aleph-alpha-amplifier-35.html#post5341014
http://www.diyaudio.com/forums/soli...rid-aleph-alpha-amplifier-39.html#post5342919
Aksa Lender P-mos Hybrid Aleph (ALPHA) Amplifier
http://www.diyaudio.com/forums/soli...rid-aleph-alpha-amplifier-35.html#post5341014
http://www.diyaudio.com/forums/soli...rid-aleph-alpha-amplifier-39.html#post5342919
In the application of high current use, Cap Mx maybe need a further discussion:
Too little capacitance – PALADIN
As you can see , by using Cap Mx, voltage ripple gone down, but the THD gone up.
I've seen people say Cap Mx is only suitable for for low current application, and i saw some preamp or VAS stage use the Cap Mx technology but never power amp.
The question is: how much current is called "high"? 100mA? 1A?
Too little capacitance – PALADIN
As you can see , by using Cap Mx, voltage ripple gone down, but the THD gone up.
I've seen people say Cap Mx is only suitable for for low current application, and i saw some preamp or VAS stage use the Cap Mx technology but never power amp.
The question is: how much current is called "high"? 100mA? 1A?
Could be due to the clipping observed in the example tests when used with Cap-MX.
In with and without Cap MX tests, he is keeping the output level same and with same supply voltage... not good by not taking into account loss of headroom.
a better test would have been , to increase the supply voltage (trafo voltage) to take into account the loss of headroom due to cap MX
In with and without Cap MX tests, he is keeping the output level same and with same supply voltage... not good by not taking into account loss of headroom.
a better test would have been , to increase the supply voltage (trafo voltage) to take into account the loss of headroom due to cap MX
Why didn't he do another round of tests with a capacitor on the output of the 'Cap-MX' and as Prasi mentioned, either increase the rails to avoid clipping or just slightly reduce the gain?
Is there no voltage drop across the Cap-MX transistor - maybe I'm missing something obvious?
As to the use of a Cap-MX on a class AB amplifier,
Is there no voltage drop across the Cap-MX transistor - maybe I'm missing something obvious?
As to the use of a Cap-MX on a class AB amplifier,
@prasi @jameshillj
For Class A or Class AB, does the dynamic current jumps a lot when play a symphony of great dynamics? I sim Cap Mx in LTSpice, for a 60r load (say 400mA) to 30r load (say 1A), the output voltage goes down 3V without zener.
For Class A or Class AB, does the dynamic current jumps a lot when play a symphony of great dynamics? I sim Cap Mx in LTSpice, for a 60r load (say 400mA) to 30r load (say 1A), the output voltage goes down 3V without zener.
The Cap multiplier circuit is a voltage follower filter that reduces ripple, output impedance, etc but goes up/down according to current demand of the application and the supply design.
By adding that Zenner to the base of the transistor (1N5862, 28volt, 5 watts?!) you've fixed the voltage of the output of the pass transistor (IRL510x) so it has been converted to a simple regulator and the voltage across the pass transistor and the incoming transformer/diode supply has to be increased to allow for dips in supply created by current variations.
Then you have added another C-R-C with 2 x 0.47R in series that defeats all the best properties of the follower/regulator but consumes lots of components - it may even sound okay but why do it this way in the first place
Yes donovas,
Adding a bypass will slightly improve C9 but is dependent on the particular properties of that C9 - usually, you'd increase the whole C-Mx by using better base voltage control, some feedback mechanism, etc, etc but it isn't then a simple Cap multiplier circuit
There are plenty of more highly developed Cap multiplier filters around to investigate both low and high current use - have a look at the K-multiplier site for some good explanations about filter operating techniques.
By adding that Zenner to the base of the transistor (1N5862, 28volt, 5 watts?!) you've fixed the voltage of the output of the pass transistor (IRL510x) so it has been converted to a simple regulator and the voltage across the pass transistor and the incoming transformer/diode supply has to be increased to allow for dips in supply created by current variations.
Then you have added another C-R-C with 2 x 0.47R in series that defeats all the best properties of the follower/regulator but consumes lots of components - it may even sound okay but why do it this way in the first place
Yes donovas,
Adding a bypass will slightly improve C9 but is dependent on the particular properties of that C9 - usually, you'd increase the whole C-Mx by using better base voltage control, some feedback mechanism, etc, etc but it isn't then a simple Cap multiplier circuit
There are plenty of more highly developed Cap multiplier filters around to investigate both low and high current use - have a look at the K-multiplier site for some good explanations about filter operating techniques.