John Curl's Blowtorch preamplifier part II

Status
Not open for further replies.
Just the Corcom PEM. $5, and it has fuseholder, IEC, and switch built in. I don't measure power supplies (nor do I listen to them unless troubleshooting), I measure preamps and amps.

Never heard of it, probably a USA only company. And good power line filters do not necessarily have to be bulky and expensive.

When measuring anything audio, like it or not, you are measuring the PSU as well. Of what use is a great front end if fed by junk power?
 
This isn't gas, that's not a particularly good analogy. I use a simple Corcom line filter and my circuits are dead nuts quiet, enough so that visiting audiophiles often remark on that. The secrets are good power supply design, circuits with good PSR, and proper layout and grounding. No need for crutches.

Apparently it works well. Either I'm a genius designer or the concern is overblown.

Ladies and gentlemen , the Eagle has landed ........... :film:
 
Most scope users don't have that. If one blithely follows the original suggestion, the danger is extreme.

It's pretty easy to build your own low frequency differential probe at a low cost. It's just a divide by 100 (or 500) op amp with good common mode rejection, plastic case, battery operation, captive 10kV probe leads, captive BNC out.
 
Last edited:
diyAudio Senior Member
Joined 2002
Hi,

Regarding DNM slit foil caps:

I doubt any of you had the chance to do an A-B comparison.
Although it was about 20 years ago or thereabouts I once had the opportunity to participate in an A-B test comprising of a Naim amp with external PSU. One standard the other modded with DNM caps.
This was at one of the now defunct Hifi-Answers mag's shows BTW.

All I can say (and take it FWIW) is I was Wowed. The DNM capped supply was lightyears ahead of the standard one. So much cleaner and lifelike sounding.

Quite frankly if these caps would have been available at higher voltages, say 300 to 450VDC then I would not have hesitated one minute to use them.
This was way pre Black Gate or whatever tech post 199x something but I haven't heard such a marked difference from any power cap since, not ever.

There were literally dozen of other treasure tips from that same magazine. Some of which being widely accepted now, others still being discussed.

Examples being coupling cap tech, resistors and their sonic signature, idem ditto for metal conductors , solid core wire versus multistrand, copper versus silver or gold wires and to what degree of purity, pcb versus P2P ad nauseam.

While I'll leave the above open for discussion (although I have no doubt of the impact whatsoever), it still amazes me that people accept a tube/valve of the same type for what it is without prior measurement

I have measured thousands of valves and can assure you that most audiophiles are neither deaf nor stupid. They're just listeners without measurement gear. And yet they're more spot on than most gear will ever be.

There's no instrument more clever than its user. Think about it.

Best,

Frank
 
Last edited:
Member
Joined 2002
Paid Member
Just the Corcom PEM. $5, and it has fuseholder, IEC, and switch built in.

That’s an excellent form of line filter. And the insertion loss is good. I haven't seen it here.
5EHM1S pdf, 5EHM1S description, 5EHM1S datasheets, 5EHM1S view ::: ALLDATASHEET :::

I don't measure power supplies

Why not? Isn’t the source of DC power important for the pre or power amp?


Ray, a wideband active differential probe it’s not that easy to build as you say

George
 
Ray, a wideband active differential probe it’s not that easy to build as you say
George

We've done it successfully. It depends on how wide a bandwidth you want, but the results can be pretty good if you use the right resistors
and op amp, and make a pcb. You do have to tweak capacitors to match the two legs of the attenuator, if you need a wide bandwidth.
 
Why not? Isn’t the source of DC power important for the pre or power amp?

To an extent, yes, but I don't do low PSR designs. It is sufficient to calculate the basics for the raw supply, stick on a decent but not exotic active regulator, then sit back and enjoy. For folk who do designs with low or non-existent PSR, life is tougher, but I can't see why anyone would use those circuits except as a fashion statement or for very specialized use.

The power supply measurements I've done in the past mainly dealt with stability. There's lots of good data out there already to cook-book a supply with sufficiently low noise- Dietz's app note on the 317, for example. If I have a need for an ultra-low noise supply, I'm more inclined to ask myself why I have that need and what are my alternatives.
 
For folk who do designs with low or non-existent PSR, life is tougher, but I can't see why anyone would use those circuits except as a fashion statement or for very specialized use.

You are forgetting "fun". :)

I built Salas' "simplistic" phono preamp, which I assume has very low PSR and is dependent on the incorporated shunt regulator for its performance. It was a fun project to build, and sounds very good indeed. If the regulator has good line rejection, very low noise, and a very low output impedance throughout the audio band, then a low-PSR circuit can perform very well. Isn't that just six-of-one, half-a-dozen of the other? Make the circuit with high PSRR and don't worry about the supply, or build a circuit with low PSRR and build the rejection into the supply?
 
Member
Joined 2004
Paid Member
That’s an excellent form of line filter. And the insertion loss is good. I haven't seen it here.
5EHM1S pdf, 5EHM1S description, 5EHM1S datasheets, 5EHM1S view ::: ALLDATASHEET :::



Why not? Isn’t the source of DC power important for the pre or power amp?


Ray, a wideband active differential probe it’s not that easy to build as you say

George

If you look at the datasheet for the Corcom filter you will see 4 different filter types. The selection depends on what you are trying to deal with. There is no single perfect filter solution. The best noise isolation, by far, is a specifically designed double or triple box shielded transformer. Its also the most expensive and has poor regulation. Otherwise its the filters. I would look at the medical grade filters because they have low leakage- low current to ground through the caps if they are present.

If you have significant DC on the power line you will have a host of issues including buzzy transformers. It takes very little DC current to push a power transformer close to saturation. Fix the source of the DC, distribution transformers don't pass DC.

The isolation between outlets has been one of Richard Marsh's mantras and he had a presentation showing that cheap CD players were among the noisiest devices on your power network.

At high frequencies (above 10 MHz) the line impedance becomes that of the transmission line. Its imperfect but never low. The losses are very large so not much goes far. The power line communications systems need to pump amps at 10 MHz to get to the other end of your house with sensitive receivers. Add a line filter or surge protector and they almost don't work at all.

Sy must use balanced connections on the critical stuff or otherwise is isolating his signal ground from power ground. They can be tied together but only if the leakage currents are managed so they do not mix with the audio grounds. I cannot see a way if two boxes have filters. The Corcom filters in the link were introduced by HP in the late 1960's or early 1970's and have been a model for many others. That style is not used on modern products any more. Its way too expensive and switching supplies obviate the need for voltage selection for the most part.

Measuring line noise is better done with a transformer based device like this: http://gryphon-inc.com/Spec Sheets/Power Monitoring/917010A - ONEView.pdf Its much safer and you can easily look at the different noise modes. If you are looking at transients or surges the recommended practice is two Tek P6015a (http://4hv.org/e107_files/public/1277895156_2261_FT91689_tektronix_p6015a_high-voltage_probe_.pdf $2,000 ea) into a wide band differential amp. I cheat and use one P6015 into one of these THS720P (http://www.trs-rentelco.com/Manual/TEK_THS730A_Manual.pdf) because it has isolated inputs and seems to work OK with 6KV surges. Don't touch when the surge hits however. If you see those big surges on your service very often you have real issues to deal with.

An inexpensive solution that works for 120V is this QA190 Differential Probe It doesn't have the bandwidth of the above solutions but it has enough and a current clamp as well, for $79. Sorry, not for 240V.

Finally its not enough to look at the DC of the supply or the analog output in isolation. The problems surface when they are part of a system and looking at a system without becoming part of it is difficult.
 
Last edited:
Member
Joined 2004
Paid Member
To be specific, the Cyril Bateman papers show high distortion in Tantalum.

Will a power filtration with Tantalum capacitors cause harm to the signal?

The Bateman articles look at voltage distortion (really harmonic generation) across caps. Possibly looking at current distortion would be more relevant in a power supply. In a supply you have a low Z DC voltage and you are superimposing a varying current. The question would be the magnitude of the harmonically related currents. When the cap impedance is in the milliOhms at the frequency of interest the voltage is guaranteed to be very small. Tantalums have very low leakage and low esr up to pretty high frequencies. If you were using them in a crossover it could be a real problem because you are operating where they have significant reactance. In a power supply the reactance is much lower. But you could have high q resonant tanks with the multiple caps and power distribution connections. No easy answers.
 
Apparently it works well. Either I'm a genius designer or the concern is overblown.

There is always the fact that no two loactions are the same in terms of electrical hash.

If you live near the local transformer, you will by default have less of a problem. The further away from it one goes, the problem gets worse.

And if it should happen to be a new one, recently installed, your problem decreases still more.

These are things which we cannot control, they are a given, but they are far from being without influence.

But if it's as quiet as you say, then you muct be doing something right.
 
It's pretty easy to build your own low frequency differential probe at a low cost. It's just a divide by 100 (or 500) op amp with good common mode rejection, plastic case, battery operation, captive 10kV probe leads, captive BNC out.

That's one way to do it. The other is to own, or have access to, a Russian military grade 'scope from 1990 or later, which has some very nice add-ons supplied as standard.

Its only problem is that you shouldn't try to lift it, or you might get lumbago. And don't kick it, you'll break your foot. :p
 
On tantalum.

I have yet to see a better cap for low current level filtering. Years ago, I opened up a Beckman blood analyzer device - it was chock full of tantalums on each leg od each IC inside, and there are a LOT of ICs inside. Just one tantalum, 10 uF/25V, no usual 100 nF bypasses, nothing. Ditto for an HP oscilloscope I also saw opened up at about the same time.

But I would never use then inside an audio circuit. The few tests I did on them years ago clearly showed they tend to darken the sound and kill off a lot of detail.
 
Status
Not open for further replies.