john_ellis said:
Hi John,
I've often thought about the issue of using inductor input filters to increase the dwell time of the rectifiers, although not just in the context of Class A amplifiers. When one looks at the huge peak currents that flow for such a short period in the rectifiers (and as a result at the input of the power transformer, especially if the power transformer has some bandwidth), one is certainly tempted to find ways of spreading out this impulse of energy a bit. It would seem that doing so would reduce EMI emissions and also reduce a bit the deleterious effect of resistance in the copper of the transformer and also in the a.c. mains.
I've tried it, but with disappointing results, but I did not pursue it very hard. My idea was to use small amounts of inductance ahead of the reservoir capacitors with the idea of increasing dwell time of the rectifiers. I don't remember, but it might have been something on the order of one mH. I expected that, even though this was not really putting the circuit into input inductor mode of rectification, it might result in less rail sag under high current loading due to the increased rectifier dwell time and less influence of circuit resistance. I saw almost no effect, so I quit the idea.
Going into full input inductor mode with a larger inductor is quite another matter, and might have some good benefits (just like it did back in the good old tube days). In principle, better regulation and higher current capability would result, along with reduced EMI. In a solid state amp, we would still use equally large reservoir capacitors subsequent to the inductor to maintain strong current reserves for heavy signal transients. Ripple and RFI from external sources on the rails would largely be gone. The cost of the inductor is, of course, an impediment.
BTW, back in the tube days they used what were called "swinging chokes" sometimes, so as to keep the rectifier circuit in inductor input mode even under reduced current conditions (as would be the case for a Class-AB amp at idle). The inductance of the swinging choke was a function of current. It was large at low current, as needed to maintain inductor input operation, but then decreased at high current where less inductance was needed (probably through controlled core saturation). Maybe something like this would work in a solid state amp.
Cheers,
Bob
Hi Bob
I too tried small inductors and concluded that these don't do much. I also looked at swinging chokes, and size for size, it seemed to me that the optimum air gap in a choke operating at its design current always gave the most inductance - swinging chokes once approaching saturation never seemed to be able to match the inductance, but they had the advantage of higher inductance at lower current.
Hence for Class A it seems that you could arrange for an optimum ordinary choke to work well.
Perhaps I tried the wrong grade of iron, as I've always used M6. Maybe a grade with softer saturation characteristics would work better as a swinger.
cheers
John
I too tried small inductors and concluded that these don't do much. I also looked at swinging chokes, and size for size, it seemed to me that the optimum air gap in a choke operating at its design current always gave the most inductance - swinging chokes once approaching saturation never seemed to be able to match the inductance, but they had the advantage of higher inductance at lower current.
Hence for Class A it seems that you could arrange for an optimum ordinary choke to work well.
Perhaps I tried the wrong grade of iron, as I've always used M6. Maybe a grade with softer saturation characteristics would work better as a swinger.
cheers
John
Hi all
I assume this is relevant to the subject in general
http://www.ultracad.com/articles/esrbcap.pdf
Regards
George
I assume this is relevant to the subject in general
http://www.ultracad.com/articles/esrbcap.pdf
Regards
George
Snubbers
Apologies if this was referenced before, but I find this article from Jim Hagerman quite relevant: http://www.hagtech.com/pdf/snubber.pdf
Apologies if this was referenced before, but I find this article from Jim Hagerman quite relevant: http://www.hagtech.com/pdf/snubber.pdf
The pleasure is mine Mr. Cordell.
No, the article does not mention snubbers at all. Though it suggest another approach to alleviate the problem of multiple capacitor resonance, this does not mean that snubbers have no place under the sun either. To me, the strong point of this article is the description of the problem and not it's cure.
Regards
George
PSU Snubbers
Hi all
Talking about applying snubbers for reducing the PSU capacitors resonance (plus the set-up parasitics), I am affraid that the only effective way to do it, is:
Connect the PSU to it's dedicated load, apply power to PSU and probe with an RF Spectrum Analyser the wire connecting the PSU to the load (probe placed closest to the load).
This way, the main resonance frequencies for the whole set-up will be identified.
Then, man can plug these frequencies into the formulas to calculate the proper snubber values.
After installing these snubbers into the circuit (closest to the load) , probing again with the Spectrum Analyser will show if the pills did their job.
My wife does not buy me a spectrum analyser. So, I am doomed to remain an amateur.
Regards
George
Hi all
Talking about applying snubbers for reducing the PSU capacitors resonance (plus the set-up parasitics), I am affraid that the only effective way to do it, is:
Connect the PSU to it's dedicated load, apply power to PSU and probe with an RF Spectrum Analyser the wire connecting the PSU to the load (probe placed closest to the load).
This way, the main resonance frequencies for the whole set-up will be identified.
Then, man can plug these frequencies into the formulas to calculate the proper snubber values.
After installing these snubbers into the circuit (closest to the load) , probing again with the Spectrum Analyser will show if the pills did their job.
My wife does not buy me a spectrum analyser. So, I am doomed to remain an amateur.
Regards
George
Apologies
Hi all
I read again what I have written above (post #270)
This was meant to be autosarcastic.
But I feel that it may have placed other amateurs in a defence state. I did't mean to.
I apologise for this.
The point of post #270 was to express my concerns over the efficiency of "blind try and err" method utilised by amateurs like me , who are not equipped with a Spectrum Analyser or an Oscilloscope of high bandwidth and high input sensitivity ( oscillations under discussion are of high frequency and low amplitude).
Regards
George
Hi all
I read again what I have written above (post #270)
This was meant to be autosarcastic.
But I feel that it may have placed other amateurs in a defence state. I did't mean to.
I apologise for this.
The point of post #270 was to express my concerns over the efficiency of "blind try and err" method utilised by amateurs like me , who are not equipped with a Spectrum Analyser or an Oscilloscope of high bandwidth and high input sensitivity ( oscillations under discussion are of high frequency and low amplitude).
Regards
George
Re: Apologies
George,
While I can agree with you on the general unavailability of high performance test equipment affordable to an individual, I think the situation is far better now than just a couple of decades earlier. I should have killed then for the kind of tools I have routinely now.
Good analog osciloscopes with at least 20 MHz bandwidth are available new for a few hundred dollars, and in my experience are good even for designing and debugging switching power supplies. The kind of HF artifacts bothering audio circuits are not very well defined at the usual 5 mV/div max sensitivity, but are fairly obvious to a trained eye.
And of course there is the refurbished equipment market where you can pick like new gear for a fraction of the cost.
And there is the PC based instrumentation offering far more flexibility than stand alone spectrum analyzers for example, with the penalty of bandwidth restricted to the audio card capabilities.
Again, good used spectrum analyzers are also available from scores of suppliers.
My experience with at least one vendor from which I got a Motorola Service Monitor was excellent, but a quick search now hints he is no longer in bussiness.
Rodolfo
gpapag said:
George,
While I can agree with you on the general unavailability of high performance test equipment affordable to an individual, I think the situation is far better now than just a couple of decades earlier. I should have killed then for the kind of tools I have routinely now.
Good analog osciloscopes with at least 20 MHz bandwidth are available new for a few hundred dollars, and in my experience are good even for designing and debugging switching power supplies. The kind of HF artifacts bothering audio circuits are not very well defined at the usual 5 mV/div max sensitivity, but are fairly obvious to a trained eye.
And of course there is the refurbished equipment market where you can pick like new gear for a fraction of the cost.
And there is the PC based instrumentation offering far more flexibility than stand alone spectrum analyzers for example, with the penalty of bandwidth restricted to the audio card capabilities.
Again, good used spectrum analyzers are also available from scores of suppliers.
My experience with at least one vendor from which I got a Motorola Service Monitor was excellent, but a quick search now hints he is no longer in bussiness.
Rodolfo
Re: Re: Apologies
These are very good points. There is a huge amount of very good analog HP and Tektronix test equipment out there. I did very well getting (and in some cases repairing) HP and TEK TM500 equipment on Ebay. Service manuals are also available for most of this stuff on the net.
Bob
ingrast said:
These are very good points. There is a huge amount of very good analog HP and Tektronix test equipment out there. I did very well getting (and in some cases repairing) HP and TEK TM500 equipment on Ebay. Service manuals are also available for most of this stuff on the net.
Bob
Snubbing PS Caps
I just revisited this text on "Solid State Power Amplifier Supply" design by Dejan Veselinovic, http://www.tnt-audio.com/clinica/ssps1_e.html and he uses RC snubbers across the capacitor banks.
From Part III: "Finally, an RC network, 1 Ohm/17W in series with 680nF, gets rid of any residual capacitor inductance. As far as I know, this was first used in 1972 by Prof. Matti Otala, in his legendary IEEE text on Transient Intermodulation. From that time, I have tried it many times, with I must admit varying results - sometimes almost insignificant, sometimes audible.
If audible, it tends to produce somewhat clearer, more coherent high tones, with benefits for ambience as well. If inaudible, it will quite surely make any amplifier more stable - the poorer quality its capacitors, the more benefit you will receive. But even with very high quality capacitors, there's always some residual inductance left over, perhaps small, but it's there - and it shouldn't be."
I just revisited this text on "Solid State Power Amplifier Supply" design by Dejan Veselinovic, http://www.tnt-audio.com/clinica/ssps1_e.html and he uses RC snubbers across the capacitor banks.
From Part III: "Finally, an RC network, 1 Ohm/17W in series with 680nF, gets rid of any residual capacitor inductance. As far as I know, this was first used in 1972 by Prof. Matti Otala, in his legendary IEEE text on Transient Intermodulation. From that time, I have tried it many times, with I must admit varying results - sometimes almost insignificant, sometimes audible.
If audible, it tends to produce somewhat clearer, more coherent high tones, with benefits for ambience as well. If inaudible, it will quite surely make any amplifier more stable - the poorer quality its capacitors, the more benefit you will receive. But even with very high quality capacitors, there's always some residual inductance left over, perhaps small, but it's there - and it shouldn't be."
Yes, the snubbers is a good point, and does not have to be expensive. I often use 1 ohm carbon film 1 Watt and a 0.1 uF ceramic capacitor in series. There is actually no real ongoing power dissipation in the reistor, and there is an advantage to having the physical size of the total snubber small.
On a positive note, the ESR of a typical aluminum electrolytic also naturally tends to provide some snubbing. So, for example, a 100 uF aluminum electrolytic across a 20,000 uF reservoir capacitor can help (although I would still add a discrete small-value snubber as well).
Bob
On a positive note, the ESR of a typical aluminum electrolytic also naturally tends to provide some snubbing. So, for example, a 100 uF aluminum electrolytic across a 20,000 uF reservoir capacitor can help (although I would still add a discrete small-value snubber as well).
Bob
I was under the impression that the snubberised bridge would have resistors in series with the diodes, and the cap would be placed across this series-pair. (Too lazy to draw a diagram now, but if you want it, I'll do it.) Your descriptions seem to imply that there is an RC series pair, and this pair is connected in parallel to the diode.
My topology is what I've seen in Randy Slone's books. I have no knowledge about which is right and which is wrong. But one consequence of Randy Slone's topology was to make the use of monolithic bridge rectifiers impossible if I wanted to put snubbers... I had to use individual diodes to build the bridge. Otherwise, I'd never be able to connect a resistor in series with each diode.
Evidently there are more than one way to snubber a cat, because the PDF file posted a few posts above mine discusses snubbers for rectifiers and adds an overall C as well as an RC series-pair, both in parallel to the diode. A bit complicated; I'm still trying to understand it.
My topology is what I've seen in Randy Slone's books. I have no knowledge about which is right and which is wrong. But one consequence of Randy Slone's topology was to make the use of monolithic bridge rectifiers impossible if I wanted to put snubbers... I had to use individual diodes to build the bridge. Otherwise, I'd never be able to connect a resistor in series with each diode.
Evidently there are more than one way to snubber a cat, because the PDF file posted a few posts above mine discusses snubbers for rectifiers and adds an overall C as well as an RC series-pair, both in parallel to the diode. A bit complicated; I'm still trying to understand it.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.