John Curl's Blowtorch preamplifier

Status
Not open for further replies.
I think you hit th e nail on the head. We should not worry about leaving a few pieces of audiophile gear on while outside there are thousands of folk running around on 5 litre Hummers, Porches, Pick-up's and other such gas guzzling monsters.

For hi fi th e rule is simple: Leave it on!

(BTW, when I was in produc t design full time, we used to leave the DVM's and precision signal sources on permanently as well . . . . better stability)
 
Only relevant if you drive a biggie yourself, wouldn't it be better to save that corpse for a presentation at an Ecovillage to an Incarcerate-HumWillies action group.

Plenty of regular people that have a bathroom fitted with an electrical heated floor, even a small unit accounts for several times the annual cost of an online BT equivalent.
For others, continuous 50 watts of preamp dissipation may add up to 1/5th of their annual electricity bill.
Keep it technical, the individual can choose between the pros and cons very well himself.
As do the ones shelling out +10K for a preamp without a power-switch, the BT wouldn't live up to it's name if it had one (imo).
 
Since the BT is a balanced circuit. How about forgoing the servos althogheter and use the technique utilised in the pass X amps?

I.e tying the opposing outputs to a middlepoint between the source resistors by 20-200K or so, this seem to give pretty good thermal dc stability.
 
Tazzz said:
...tying the opposing outputs to a middlepoint between the source resistors by 20-200K or so, this seem to give pretty good thermal dc stability.


Pretty good is in this case somewhere below 100mV DC at the output of a power amp. This offset at the output of a small signal stage would cause severe problems in the power amps and heavy switching noise in a pre´s circuit. I assume that a BT owner has an adequate speaker system in his place. The risk of damage in a zillion dollar speaker is pretty high:smash: :bawling:

Uli

:nod: :nod: :nod:
 
My servos drop the TOTAL offset, both common mode and differental to about 1mV or less. That works, and is good enough. The limitation is due to the internal offset of the IC selected to be the servo. If I had used a chopper stablilized IC, I could and have done even better.
 
Okay, call it 20W. I've got three young 'uns in the house and they prefer night lights while they sleep to keep the evil demons away. One or two more to illuminate the way to the bathroom and...(gimme a second, I left my slide rule at home so I've got to count on my fingers)...I'm counting at least four, maybe five night lights. At 4W each, that's 16-20W. Hah! So there's a preamp's worth of power, right there.
And that's not counting the porch light my wife leaves on so I can see when I get home at night.
Hell, all you've got to do is put a couple of bright LEDs on the front of a preamp and call it that room's night light.
In the overall scheme of things every watt counts if you're concerned about the environment or your power bill, but in real world terms 15-20W just isn't that much. At prevailing electricity prices around here, it would cost about a dollar to a dollar-fifty per month. If that bothers people, they can always put an inline SPST in an extension cord between the preamp and the wall and switch the preamp off when they're done listening. Just remember to turn it on ahead of any listening session.
Don't forget that in the winter that heat goes to counteract some of your heating requirement from your main heating system. It's only during the warmer months that it could be construed as any sort of problem.
As I said earlier, I can see fretting over leaving an amp on all the time, particularly if it's class A. But a preamp? That's a stretch from my point of view.
I confess that I still haven't managed to wrap my head around an earlier poster's question as to whether it was a marketing decision. That one completely eludes me. Cosmetics? Sure. Parts choice? Obviously. Weight, either in terms of shipping cost or consumer perception? Okay. I can see how all those things and more might factor into marketing related decisions. But the question of whether to turn it off and on? That's a little too deep for me. But I'm still working on it. Maybe there's some particularly subtle logic at work here that I just haven't picked up on.

Grey
 
Incidentally, I took a moment and looked back at the Aleph-X thread. It's pretty much as I remembered it. People questioned the circuit's DC stability practically from the start due to their simulations. My prototype was stable, but several people kept telling me it shouldn't/couldn't be. There were also concerns about phase margins which may or may not have been based on simulations; I didn't take the time to read every post in detail.
No matter how you slice it, I think most people would agree that DC stability is something to be concerned about. Why the fellow who posted that he couldn't find anything about simulations didn't see those posts, I don't know. I don't suppose it matters, either way.
In the here and now, the Blowtorch uses servos. I tried a servo (discrete, not chip) for the Aleph-X, but was able to keep the DC offset (two different kinds, absolute and relative, just like the Blowtorch) under control without it, so I abandoned that approach. One of the tricks that doesn't show up in simulation is that you can use thermal mass to stabilize some DC problems. In the case of the amp I'm working with now--the front end of which is a cousin of the Blowtorch--I'm able to keep the DC within bounds by using thermal mass and resistive loading at the second stage.
If you approach DC drift as a thermal problem, you can try several things:
--Run everything so cool that it doesn't matter if air at ambient temperature blows across the device; it won't change the temperature significantly.
--Run everything so hot that ambient air is pretty much ineffective at changing the temperature. Obviously, you're going to get into reliability problems if you push this too hard, but nonetheless, it works.
--Use so much thermal mass that it averages the temperature over time and transient things like mere air currents are of no more concern to the circuit than individual waves crashing against a seaside cliff. In John's case (ahem), the chassis for the Blowtorch would, by virtue of sheer mass, go a long ways towards stabilizing the DC of the circuit. Add the servo and I imagine the thing is rock solid.
Interesting to me is the carved case that Charles Hansen is using. I have no idea how he's managing DC, but there's clearly a lot of mass involved in the MX-R. However, the circuit runs AB (yes? I don't recall seeing that he's biased class A), so the amp will get hotter as it runs harder. Mass or no mass, that's going to change things. In my case, I'm running class A, so the amp runs hot (well, warm, anyway...I'm using big heatsinks) all the time and that helps me keep DC pretty tame. I don't know if Charles is in a position to tell us anything about his DC compensation, but if he could say a thing or two I'm sure at least some folks would be interested.

Grey
 
Nearly all of our designs are true DC amplifiers. We've never used servos. Some products use capacitor coupling, and one used DC feedback. Here are the details.

Our first product was the V-3 power amplifier. It was all MOSFET, but not quite zero feedback -- there was DC feedback from the output stage to the folded cascodes. With power amps you have a little bit more latitude because you only have to worry about differential offsets and not common-mode offsets. I can't remember for sure, but I think the common-mode offsets over temperature and AC line voltage were somewhere in the 50 to 100 mV range, while the differential offsets were somewhere in the 5 to 10 mV range. Certainly not a problem for speakers.

Our next generation of products were true DC amplifiers (ie, the gain at DC is the same as the gain at 1 kHz). It's not all that hard to get rid of drift due to thermal tracking as long as you are using a complementary circuit. What was harder was getting things to be stable with varying AC voltage. This was simply due to a series of design decisions we made, but again we got the DC drift (both common mode and differential) down to a reasonable range for both the preamp and the power amp. This time we didn't need to use DC feedback.

Our third generation products were also true DC amplifiers but due to improved circuits and paying even more attention to detail, the performance has improved. In our preamp we typically see about 1 mV of differential drift over time and temperature, and maybe 2 or 3 mV of common mode drift. Our power amps have more gain and see a wider temperature swing but the differential drift is only about 5 mV and the common-mode (which matters not at all for the power amp) is only about 20 mV.

So we've got things really well under control with no servos, no coupling caps, and no feedback.

The only exception is in our phono stages. We don't even try for a true DC amplifier here. In the first place it wouldn't necessarily be a good thing with regards to LP playback. In the second place I'm not sure I could achieve a reasonable amount of offset when the gain at mid-band is 70 dB and the gain at low frequencies is 90 dB! I'm sure it could be done, but it would require some really fancy footwork, and I'm not sure it would be worth the trouble and expense.

So we just capacitor-couple the phono stage. It works well for tubes, and it's simpler than a servo. And the way I see it, the cap in a servo is going to color the signal at least as much as a coupling cap, plus you now have an op-amp in your signal chain that you have to worry about.

The only op-amps we use in our analog products are as DC sensors in our power amps. These shut down the amp in case of a fault condition. We use FET input op-amps in an inverting configuration and there is a 5 Mohm resistor from the speaker output to the virtual ground. I think the chances of the op-amp "polluting" the signal path is pretty remote with this arrangement.

Most of the information needed to make a DC-coupled amp is outlined in an article Erno Borbely wrote for The Audio Amateur over 20 years ago. He had a two part series. One was called the "Servo 100" and one was called the "DC-100". He may have those article posted on his website, I'm not sure.
 
I'm not sure that those exact articles are available. If I recall correctly, the ones available at this time are for more recent models and are pretty much here-it-is rather than this-is-why-I-did-it. Maybe someone has the older articles and can share. (For all I know, I may have the older ones buried somewhere, but I don't think so.)
I have used coupling caps with solid state before with pretty much the same logic you use--there've been some damned fine sounding tube pieces over the years that managed to survive having multiple caps. Obviously, it would be preferable to have no cap, but when it comes down to it you're going to have to choose between a servo, feedback, or a coupling cap...assuming that you can't design the DC out. I've used all three at one time or another. It tickles me when solid state folks get their feathers all ruffled when you mention caps because they're so programmed that Caps Are For Tubes.
I'm on a short tether at the moment--got to tidy up before third shift comes in--I'll try to get back to this as this DC thing happens to be very much on my mind at the moment.

Grey
 
Hi John,
Once a piece of gear settles in thermally, which takes an hour in the worst case for a calibrator that is infinitely more accurate and critical in DC drift than a piece of audio gear, there will not be a change in sound quality. There is something else at work there. understand that a piece of lab equipment is fan cooled and has lots of air flow, so it is harder to stabilize it compared to your sealed design. If leaving it on has that much of an effect on sound quality, you would have to shield it thermally from all other heat sources. Your amplifiers for example, that will generate extra heat when putting out power.

Grey,
My points against leaving equipment on was clearly made earlier and was purely technical in nature. I said absolutely zero about energy use. This was the first thing you brought up. I was more concerned with the dangers of excessive line noise and variations. I brought up amplifiers as a general point because I have seen speakers destroyed by amplifiers left running unattended. Some example did result in fire. These are facts and can happen.

My question stated:
Was there a specific reason the CTC Blowtorch needed to be left on, or was it more a marketing decision?
to which John replied that he felt the unit sounded better. This was a complete answer as far as I am concerned. You then referred to my earlier comment with:
I confess that I still haven't managed to wrap my head around an earlier poster's question as to whether it was a marketing decision. That one completely eludes me. Cosmetics? Sure. Parts choice? Obviously. Weight, either in terms of shipping cost or consumer perception? Okay. I can see how all those things and more might factor into marketing related decisions. But the question of whether to turn it off and on? That's a little too deep for me. But I'm still working on it. Maybe there's some particularly subtle logic at work here that I just haven't picked up on.
Was my question not valid? Often a manufacturer must use the part of the day (there is a cute French phrase for this) in order to sell his product. There is a never ending supply of arm chair critics that have suggestion on how to improve the sound of a product. The "leave it on 24/7" crowd would look upon John's decision to exclude a power switch as vindication for this generally stupid idea. Why are you inferring that there is more to my earlier question than there is? Note that I said "generally stupid idea". If there is a good technical reason to leave something on, it is not a stupid thing to do.

Case in point. In a Cal Lab, we leave the keep calibration equipment on unless the equipment is going out on an on-site calibration service call. This is to allow us to start working straight away instead of waiting the hour. When we do an on-site, the first thing that happens is the calibrators are set up and plugged in. We then set up everything else (computers and tools) so that minimal time is lost waiting. In the lab, we are ready to go straight away, so the equipment is left on, powered by a UPS (each piece of calibration equipment) that also provides AC line conditioning. All secondary equipment is turned off normally, including the computers (except the servers).

A long time ago in recording studios, MCI made a plasma meter bridge. They looked cool! Many were sold. Too bad the power supply had a design error. The result of this was that when you turned it on, there was about a 50 - 50 chance the meter bridge power supply would smoke. The fix? Leave it on. Mostly, recording studio stuff is left on because people can't be bothered to turn everything off and they don't want to wait for it to warm up. Remember that they are running a business and there is a TON of gear to turn on.

Early Bryston amplifiers sounded like **** when first turned on, also the power switches tended to burn out. These faults were all the result of poor design. The fix. Come on, guess Grey ....

Okay, I'll tell you. They instructed the users to leave the amps on. Problems from Bryston's perspective were solved. I've seen Bryston amps cause speaker fires. I think that put owners of this equipment at unreasonable risk.

Now, as I recall, you come from a telecommunications background. Those devices are meant to run "24/7" and have thermal sensors and other shut down devices installed. They are also not likely to have a DC fault that will cause damage to the connected equipment (telephones). More importantly, they were expected to work at any point in time, on demand. Can you imagine the telephone companies shutting down at night? Neither can I. That would be a huge safety issue, and is why telephone service is an essential service. An entertainment sound system isn't. You can't begin to compare them. That's the difference between "I want" and " I need".

In future, please consider comments made by other members before you take issue with them in this manner. John can take care of himself, and did. He answered my question.

-Chris
 
GRollins said:
a marketing decision, cosmetics, parts choice, weight

The paradox :
the CTC preamp could have been called GreenTorch.
Just think ; it's built to last forever, has a less ageing cosmetic design, and is also close to fully recyclable.
Compared to several other pre models, the BT even has a low power consumption figure. (apart from the lot getting confirmation on the main bias levels in the CTC)
No volume attenuation or source selectors that require a couple of watts to run, no crystal displays, no R/C, even useless switches are left out.
The few digit competitors of the BT burn at least twice as many watts/hr.
Green always costs, the bio stuff we eat overhere costs a multitude of the Soylent Green cookies at the mall but it just tastes so much better.
Forgive me father, i leave my full metal jacket preamp on all the time and i live on gas-guzzler avenue, i left my hippie hair lying somewhere a long time ago.
 
anatech said:



My question stated:

to which John replied that he felt the unit sounded better. This was a complete answer as far as I am concerned. You then referred to my earlier comment with:

Was my question not valid? Often a manufacturer must use the part of the day (there is a cute French phrase for this) in order to sell his product. There is a never ending supply of arm chair critics that have suggestion on how to improve the sound of a product. The "leave it on 24/7" crowd would look upon John's decision to exclude a power switch as vindication for this generally stupid idea. Why are you inferring that there is more to my earlier question than there is? Note that I said "generally stupid idea". If there is a good technical reason to leave something on, it is not a stupid thing to do.

Case in point. In a Cal Lab, we leave the keep calibration equipment on unless the equipment is going out on an on-site calibration service call. This is to allow us to start working straight away instead of waiting the hour. When we do an on-site, the first thing that happens is the calibrators are set up and plugged in. We then set up everything else (computers and tools) so that minimal time is lost waiting. In the lab, we are ready to go straight away, so the equipment is left on, powered by a UPS (each piece of calibration equipment) that also provides AC line conditioning. All secondary equipment is turned off normally, including the computers (except the servers).

A long time ago in recording studios, MCI made a plasma meter bridge. They looked cool! Many were sold. Too bad the power supply had a design error. The result of this was that when you turned it on, there was about a 50 - 50 chance the meter bridge power supply would smoke. The fix? Leave it on. Mostly, recording studio stuff is left on because people can't be bothered to turn everything off and they don't want to wait for it to warm up. Remember that they are running a business and there is a TON of gear to turn on.

Early Bryston amplifiers sounded like **** when first turned on, also the power switches tended to burn out. These faults were all the result of poor design. The fix. Come on, guess Grey ....

Okay, I'll tell you. They instructed the users to leave the amps on. Problems from Bryston's perspective were solved. I've seen Bryston amps cause speaker fires. I think that put owners of this equipment at unreasonable risk.

Now, as I recall, you come from a telecommunications background. Those devices are meant to run "24/7" and have thermal sensors and other shut down devices installed. They are also not likely to have a DC fault that will cause damage to the connected equipment (telephones). More importantly, they were expected to work at any point in time, on demand. Can you imagine the telephone companies shutting down at night? Neither can I. That would be a huge safety issue, and is why telephone service is an essential service. An entertainment sound system isn't. You can't begin to compare them. That's the difference between "I want" and " I need".

In future, please consider comments made by other members before you take issue with them in this manner. John can take care of himself, and did. He answered my question.

-Chris


I find your position curious...and self-contradictory. On the one hand, you go to great lengths to tell everyone how experienced you are with electronics and that you are well aware of the benefits inherent in leaving equipment on all the time. On the other hand, you ask rather disingenuously if leaving the Blowtorch on all the time was a 'marketing decision,' then go on to say it's a 'stupid idea,' advanced by 'armchair critics.' Is there some reason that you believe that audio gear operates according to different principles than test equipment?
So which is it? You can't have it both ways. Either equipment benefits from being left on or it doesn't.
Or are you advancing that the idea that the Blowtorch is somehow different from other preamps (viz. your acceptance of John's saying that the Blowtorch benefits from being left on, but then clearly showing your opinion of people who leave things on) and that for some mysterious reason the Blowtorch likes being left on, but other preamps don't...
You then go on to speak of amplifiers that burned up speakers. I thought the Blowtorch was a preamp, but perhaps I was mistaken. Or was it, perhaps, the preamps associated with those amps that caused the burned speakers with the amplifiers as innocent bystanders? I'm also not clear as to how owner attendance would have prevented the problem, except perhaps if the speaker went on to catch the house on fire; maybe if the owner leapt for a fire extinguisher he could put out the speaker before the flames spread to the drapes and carpet and from there to the rest of the house.
Owner attendance certainly didn't help the fellow who had the Kinergetics amp on the floor next to my bench. He was there when it popped and sent 11Vdc and infinite current into his top o' the line B&W speakers, thoroughly toasting the woofer.
In short, yeah, I'm kinda having trouble understanding your position. None of the pieces of the puzzle fit.

Grey

P.S.: 'In future, please consider comments made by other members before you take issue with them in this manner.'
This from a guy who thinks I was in telecommunications? Try mainframe computers, as I clearly stated in the Hafler modification thread...and something you went to great lengths to abuse me over...
Sigh
 
Status
Not open for further replies.