Power regeneration questions

[analog_sa]

Today i built a very simple regenerator producing 2x12v rms in order to power a SS phono pre. I used 2xLM1875 and a simple low distortion 50Hz oscillator.

It didn't take me long to build but unfortunately it doesn't sound great. Pleasant and transparent mids and highs with less electronic hash but very poor and ill-defined bass and very bad subjective dynamics.

So, the 1875 has the job to generate a sine into a pair of 6800uF caps. Does it enjoy it? I still haven't measured the distortion but on the scope the sine looks good, whether the 1875 output is unloaded or loaded by the bridge and smoothing caps. The phono draws around 5mA per rail so the 3A max current of the 1875 seems sufficient.

My first impulse is to replace the 1875 with a more current capable chip, like an OPA549 but the sound is so lame that i doubt this will help enough.

The PS powering the LM is capable enough: 160VA with 2x10,000uF

Any ideas and suggestions?

[Peter Daniel]

Originally posted by analog_sa
The PS powering the LM is capable enough: 160VA with 2x10,000uF

Any ideas and suggestions?

How about reducing it to 2x1,000uF?

[analog_sa]

How about reducing it to 2x1,000uF

That's where it started from. As this is just a 50Hz amp i figured an increase in capacitance should provide more current reserve.

Of course you may have a point as what really matters is the output impedance not just at 50Hz but potentially at higher frequencies (load is a class AB amp, not to mention diode switching noise)

There was precious little difference (if any) between 2000 and 10,000uF anyway.

I still haven't tried it but hope that there will be no difference between mains cables at least

I intended trying various frequencies and multiwaves but this first attempt was really discouraging

[GregGC]

Originally posted by analog_sa
Today i built a very simple regenerator producing 2x12v rms in order to power a SS phono pre. I used 2xLM1875 and a simple low distortion 50Hz oscillator.

It didn't take me long to build but unfortunately it doesn't sound great. Pleasant and transparent mids and highs with less electronic hash but very poor and ill-defined bass and very bad subjective dynamics.

So, the 1875 has the job to generate a sine into a pair of 6800uF caps. Does it enjoy it? I still haven't measured the distortion but on the scope the sine looks good, whether the 1875 output is unloaded or loaded by the bridge and smoothing caps. The phono draws around 5mA per rail so the 3A max current of the 1875 seems sufficient.

My first impulse is to replace the 1875 with a more current capable chip, like an OPA549 but the sound is so lame that i doubt this will help enough.

The PS powering the LM is capable enough: 160VA with 2x10,000uF

Any ideas and suggestions?

Maybe the problem is in the other part of the circuit and not the PS you build. How does it sound with regular +/-12 v PS?

[analog_sa]

Well, i wouldn't offer this topic if i had any doubts about the rest of the circuit. It has been playing nicely (for basically an opamp circuit) for months; i know its sound very well and as it takes a low supply voltage and current it seemed a good candidate for a regeneration test bed.

[analog_sa]

Replaced the 1875 with OPA549 and problem solved. If anything the bass is now deeper and more solid than feeding the pre directly from a transformer. The sound is very similar to using batteries - a remarkable lack of electronic hash and grunge, very transparent and natural.

Whether this solution is cost/effort effective compared to building a better quality regulator i honestly don't know. It certainly seems to add more bulk though

[janneman]

Well, an obvious try would be to use a series resistor on the LM output to limit those nasty charge spikes when the bridge diodes open up to the (dynamically) short presented by the reservoir cap. That may limit the higher harmonics in the charging current. Like 10 Ohms, makes the LM think it powers a speaker, hah!

Jan Didden

[analog_sa]

Thanks for the suggestion Jan. Will definitely try it out tomorrow. Of course it will negate one of the advantages of regeneration - much lower driving impedance than a mains transformer. I'm also not so convinced the caps get significantly discharged by the few mA load. Anyway all seems to work fine with the OPA549.

On a different note i tried feeding the amp with various frequencies and the results are not what i expected (as usually )

There is not that much difference between 50 and 400Hz but strangely 50Hz is overall preferable. I went up to 15kHz which sounds by far inferior. Does there appear to be a logical explanation? There seems to be some distortion around the peaks at 10kHz on the scope.

[moamps]

Originally posted by analog_sa
I went up to 15kHz which sounds by far inferior. Does there appear to be a logical explanation?

Look at OPA's characteristic at higher frequencies. DF, PSRR.... goes down .....THD goes up...
How looks your preamp's PSRR?
EMI from generator 12V/15kHz to preamp (grounding issues, PFB)?

[analog_sa]

Look at OPA's characteristic at higher frequencies. DF, PSRR.... goes down .....THD goes up...

You're absolutely right. I was kind of hoping the good effects (fast recharging of the caps) will outweigh the bad. No free lunch this time.