[CaptBubba]

What I am planning to do kinda a stopgap amp. Baisically it will be either a multi-loop amp (META42 style) or a CHA 47 depending on which way a pair of DPDT swtiches are thrown. I am doing this because I have the parts for another CHA-47, but want to do a multiloop amp, and am too impatient to wait on some 2001s to come in.

Ok, I am going to make this amp run off mains power. The problem is that for some reason I just can't get regulators to work right, the either introduce a hiss or just sound wrong in every amp I have tried to do a regulated PS in. So I am just thinking about using a boat-load of filter capacitance along with some inductors to make a PS. The question is: how much filter capacitance do I need? I have a bunch of low inductance electrolytics: about 17,000uF worth, but because I want to make around a +/- 7.5V supply I will be forced to double up some of the 10V caps so they don't blow, so just say 10,000uF to be safe. I also have oodles of low-rating film caps (1-130nF) just begging to be used for decoupling.

Would this be sufficent for a CHA47/multiloop amp?

I will either be running it with a rad-shack ac-ac wall wart or a small 350mA tranny. Would a ac-dc wall wart be a better choice because it may have some filter capacitance already in it?

 

[zzz]

Ask yourself how much ripple (Vr) is acceptable at the maximum current your amp is going to consume (I). Then if my logic is not flawed, you have CVr = It, where t = 1/120sec if you're using bridge rectifier. So C = It / Vr (in F).

 

[puppyslugg]

Hi CaptBubba,

Are you using the lm317/37? 78/79XX?

 

[ppl]

Most Reg's do not like lare capacitors on there output my solution is place a small series resistor 1-2 ohms in series with the Reg.s and prior to the caps. also include rhe resistor within tthe Feedback loop of the reg so that resistor that goses from the adjt pin to the output should now go from the output of the series resistor rather than the reg,s output.

 

[puppyslugg]

Hi ppl,

Do you prefer resistors or zener to set the voltage of the regulator?

 

[ppl]

I like Zeners for low impedance and drive the Zener from a fet connected as a current source about 5 Ma. But some like resistors

 

[puppyslugg]

Quote:

....and drive the Zener from a fet connected as a current source about 5 Ma.

ppl,

I'm having trouble visualizing that. Do you have a schematic or something?

 

[CaptBubba]

Quote:

Originally posted by puppyslugg Hi CaptBubba, Are you using the lm317/37? 78/79XX?

I've tried both. I get better luck with the 78s, but they still don't sound right. The lm317 series has earned my hatred because it inevitably buzzes in some way.

ppl, I'll have to try that, I'll let you know how it turns out (I have a few lm317s laying about that I can try it with. The hard part will be finding a 1-2ohm resistor.

 

[ppl]

I have never liked those three term regs for sound quality thay all sound like crap However the LT-1085 & LT-1033 while spendy are the best of the Bunch IMHO a simple series pass Transistor with a Zener from the Base to gnd to set the voltage and driven from the Regs output with a curent source is Best You get true complimentry topology and none of the instability of the IC regs. Some of the New low dropout regs are nice but no Neg. part is available.

 

[CaptBubba]

Quote:

Originally posted by zzz Ask yourself how much ripple (Vr) is acceptable at the maximum current your amp is going to consume (I). Then if my logic is not flawed, you have CVr = It, where t = 1/120sec if you're using bridge rectifier. So C = It / Vr (in F).

What effect (if any) would adding a choke to the input have?

 

[ppl]

puppyslugg> I am sory for the delay in posting what you needed see attached

 

[puppyslugg]

ppl,

Thanks for the schematic.

Are there particular fet's I should be using?

In Q1, is the source and gate tied together and connected to the output of the lt1085? And the drain is tied to the zener?

In Q2, is that the source or drain is tied to output of the lt1033?

 

[puppyslugg]

Hi ppl,
Quote:

Most Reg's do not like lare capacitors on there output my solution is place a small series resistor 1-2 ohms in series with the Reg.s and prior to the caps.

Would that apply to the lt1085/33? Does using a resistor at the ouput, affect the output impedence of the regulator?
Quote:

a simple series pass Transistor with a Zener from the Base to gnd to set the voltage and driven from the Regs output with a curent source is Best You get true complimentry topology and none of the instability of the IC regs.

Do you have a complete schematic of the above?

Thanks....again!

 

[Tomo]

Hello,

Note in OUR case, current draw is comparably small compared to the PSU for speaker amps like PASSAMP. So you do not need to have "&*^&"-load of capacitance. Neither do you need super high regulation.

I would say 4700uF~10,000uF depending on your pocket. Difference is so subtle because too large a capacitance or too many capacitor simply introduces parasitic characteristics. So in the end, you "might" end up with same results.

Also, note that opamp or buffer based amps are very very very resistant to supply ripples. So you do not need super high ripple rejection either. (To all, you guys should differentiate PSU designs for high current amps and opamp based amps. Because their needs are fundamentally different. Thus, designs should be different as well. I mean, what good does KG PSU does to battery powered CMOY amp? If you can hear that difference, you are not a human. Think somewhat reasonably.)

In fact, my "reference" (as in one of the best amps I have ever attempted or I have ever tried as in commercial amps) opamp-based amp has surprisingly small pre-regulator capacitance of 3300uF and ~200 uF of post-regulaltor capacitance which include chip bypass capacitors. And regulators are LM317/337 pair. But then all of the caps are ELNA and BG

I guess quality over quantity for us headphone amp builders, indeed!

In terms of regulators, I would not recommend Linear Technology regulators. This is primarily because it is way too expensive. If you care that much, I would use a battery PSU. This is inherently better simply because you don't have to mess with AC at all.

HOWEVER, if you are talking about amps with low supply rejection or low supply independence and require high current, there is advantages in using LT spendy regulators. (But then some people even avoid regulators.)

Sometimes subjective,

Tomo

 

[ppl]

Tomo> I agree with you on most points. I also do not like regulators. Batteries are indeed the the most quiet. also true the LT regs offer only slight improvement over the LM-317 & LM-337. While Regulators do reduce Ripple to low levels thay introduce noise and this is harder to get rid of. By you using Pre and post regulation you can tone down the capacitence But i have heard improvement using 25,000uF on the output of an AC Headphone amp over the more modest 3,000uF and these are quality caps. Also this amp used pre and post regulation. powering EL-2009/AD843. I also hear improvements in lowering all forms of noise on the power supply it shows up as a blacker background and this is with Opamp/Buffer based designs this is why i use additional Power supply circuits on evean my Battery operated amp's. But since my Eye sight is bad maby my Hearing has improved to a point better than normal, and I think i am human, however maby could have come from another planet. Equipment that sounds good to most people sound awfull to me.

puppyslugg> Yes in The Neg reg. the gate /source connected to gether and drain is going to Zener. on the positive the Gate/Source gose to zener. in-order for the drain to be the output on Both Polarity's Complimentry P and N Ch devices will be needed.