 
|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
|
|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
107.219.46.222
Hi,
I plan on making some changes on my 2a3 se amp. While on the bench I'd like to make some measurements B+, fil voltages, etc. While flipped upside down on the bench should I load the outputs with some resistors or is it ok to just leave them open while monitoring the voltages?
Thanks,
David
You can make static voltage measurements without a load, just don't run a signal through it! Better yet, short the inputs first.
It seems that we have a little disagreement on best practices in response to what I thought was a simple novice question: To load or not load the OPT while making simple voltage measurements on a SE 2a3 non feedback amp. Can others with experience care to chime in?
Regards,
David
As I imply below, any well designed audio power amp intended for domestic music reproduction should tolerate an unloaded secondary under normal use conditions without damage.
FWIW, when I run basic tests on an amp, I first do it with a load and if it looks good I run some tests unloaded, typically freq resp and sq wave at ~ 1W equivalent.
You run response tests unloaded?Is this a tube amp using global feedback?
And what do you mean by one watt? One watt into your 1meg scope probe?
Edits: 09/18/17
Yes, response loaded only by scope and DVM ~ 900K ohm.
Yes, tube amp with global feedback.
1 watt equivalent or ~ 2.8 V into probe load @ 8 ohm tap etc.
Also check square wave performance under these conditions.
Also check performance with 0.1-1.0uF as only load.
If amp exhibits problems under normal loading conditions, I won't run tests described above.
But what is your goal of unloaded measurements? That's not the amps normal operating condition and I would think the measured performance unloaded is unrealistically very good.
...while true that no speaker should present infinite impedance, an amp that is stable no-load is almost always stable under real world conditions of high impedance such as LF speaker resonance.IME, open circuit response is usually worse than loaded, esp if low/no feedback. All at 1W equivalent power. YMMV
Edits: 09/19/17
Do not leave the secondary at infinite resistance, if possible. It's why some amp companies actually ground the outlet, when speakers are momentarily disconnected.
Edits: 09/19/17
Based on the published theory of an unloaded transformer secondary I would definitely put a load on it. As Tre said below, even 1K is enough to prevent a dangerous voltage rise.But the gentleman from Bottlehead says he has run many 2 stage SET amps unloaded and has not had an issue. And I respect the Bottlehead group when it comes to SET amps.
So tough decision. If you have a 10 to 1000ohm resistor laying around there seems no harm in using it. If you don't then you might want to just go without.
Just a side note, I was putting some small snap on current transformers around my 60amp HVAC wires in the breaker panel for a DIY home energy monitor project. The spec sheet boldly tells you to keep a 100ohm resistor on the current transformer secondary or risk a nasty shock. Same principle here.
Yes, more comments appreciated.
Edits: 09/18/17 09/18/17
Since you brought up Bottlehead, every Bottlehead amplifier that has EVER been constructed has had its DC voltages check without a load on the output transformers...
safety and reliability issues aside... a few thoughts
First... why would someone ever want to see the behavior of an amp in a manner in which it will not be operated? The best suggestion here is to load the secondary not for safety but because that is how the device is going to be used. Loaded and unloaded behaviors are often quite different and unless the goal of the measurement is open circuit stability, I see no point in making open circuit measurements of a device intended to deliver current.
Second... lets pretend this is a transformer coupled linestage in place of an amplifier and ask if it is OK to operate the transformer coupled output into an open circuit? (by open circuit I mean the grid of a tube or a scope probe)
dave
Assuming this is a zero feedback amp, that's totally reasonable.
I'd you're going to measure hum at the speaker jacks, you don't really get the actual value without a load.
Hi CBP,
Thanks for the reply but a little clarification please. Yes, this amp is a zero feedback design, so is it "totally reasonable" to have the opt unloaded or loaded?
Regards,
David
Unloaded is OK for just checking DC voltages.
"Unloaded is OK for just checking DC voltages."Provided you don't send a signal through it. I would short the input jack if testing unloaded.
But what's the harm in putting a 10ohm 1/2w resistor across the outputs?
Edits: 09/18/17 09/18/17
It's just personal preference. I've run plenty of single ended amps with signal and no load *gasp* and haven't roasted an OT.
As always there are exceptions where you can get away with a non-recommend practice.But you do see how an unloaded OTP becomes the functional equivalent of an old CRT TV flyback transformer. And without the benefit of a damper diode!
Edits: 09/18/17
...The TV flyback transformer is often cited as analogous to an audio OPT with an open secondary. The flyback transformer makes use of the extremely high dI/dT that occurs during retrace (voltage across an inductor = L*dI/dT). For an audio transformer passing a typical sine wave test signal at lower powers, I can't imagine the dI/dT factor would be so great as to stress insulation resistance of a stable and properly operating tube amp, secondary loaded or open circuit. I'm not sure even a square wave would be much problem at reasonable powers. Personal experience mostly supports this. The only situation where I've run into flyback like voltages on the pri of an audio OPT is when a tube arcs; a gross fault condition. Another potential damage situation might be where an amp is subjected to hard clipping and the load is suddenly removed as might occur with speaker burnout in a musical instrument amp but this is a special test condition.FWIW, many high quality amps from the golden age touted as a feature, unconditional stability and safe operation into any load including an open circuit. Marantz, Mac and the better Heaths come to mind.
Edits: 09/19/17
As an addition to the above... one needs to consider the types of transient sounds one can get through your speakers from plugging and unplugging interconnects when a system is on. How many tweeters have been sent to an early death by this practice? Now consider this situation on the bench without a speaker to tell you that things are getting scary and apply the same type of plugging and unplugging to an amp with an unloaded secondary and I see gussers concern.
I do it all the time and am willing to accept the results but I am typically not a belt and suspenders kinda guy. It is funny that I have never blown a transformer but I have taken out several PC scope / soundcard channels measuring my phono stage which is trasformer coupled. The quickest way to do this is to accidentially bump a clip lead connecting a tap on a LCR inductor and see how quickly a few hundred volts will show up at the secondary.
dave
...I assume gusser is referring to a steady state condition and not a transient/fault/error condition. I would expect sudden removal of a load under full power/clipping conditions to be problematic. Same for accidentally applying 10s of volts to a low level input as might occur during plugging/unplugging an interconnect into an input of a live amp.
For now my "theory" remains: it's primarily a dI/dt-dV/dt thing and much less so a loaded vs unloaded thing although there is interaction between them.
A continous signal with an open secondary is not an issue.
"A continous signal with an open secondary is not an issue."
But if a probe slips and there's a momentary spike, it's all over. Or if one happens to accidentally touch the input with the amp at full gain, and the output stage clips, look out! Placing a small resistor across the secondary is simply a good precaution. I wouldn't even think about risking my Magnequest outputs for the sake of a stupid 25¢ resistor!
--------------------------
Buy Chinese. Bury freedom.
With a 2 stage 2A3 amp, I'll assume the worst case scenario for this particular question and say that we have a CCS loaded 12AX7 feeding the 2A3. When applying signal to an unloaded amp, we are still limited by the bias voltage at the cathode of the 12AX7 and the mu of each amplifying device (as well as where grid current will come into play on the 2A3).
This also comes up when you want to measure output impedance, as you need to measure the voltage present when the OT secondary is unloaded. I have done this on lots of amps, and the increase in voltage on the secondary isn't particularly drastic.
https://www.google.com/search?q=unloaded+tube+amp&oq=unloaded+tube+amp&aqs=chrome..69i57.3791j0j8&sourceid=chrome&ie=UTF-8The overwhelming advice from many different sites is not to run a tube amp unloaded. And as I pointed out the plate voltage CAN become quite high either damaging the tube(s) or breaking down the insulation in the OPT.
OK so with some amps you can run unloaded. And some tubes and OPT's can handle kilovolt spikes for some time. And perhaps a flea powered 2 stage SET amp with modest B+ is immune for the most part.
But why risk it? Isn't a dirt cheap 10ohm 1 watt resistor worth the peace of mind especially for a novice? You have the experience to yank the cord when something goes haywire. A beginner may not and severely damage their amp.
Edits: 09/18/17 09/18/17
With a zero feedback SET amplifier, this just isn't an issue.
I still would slap a cheap resistor across the outputs first. That may be just me but it's what I would do.
Edits: 09/18/17
Many manufacturers of smallish tube amplifiers (like those used in tape decks and the like) used to include a resistor about 1K across the secondary. That was enough to dampen the Q and reduce the swing if the customer accidentally disconnected the speaker.
--------------------------
Buy Chinese. Bury freedom.
1) You must have a load on any output transformer coupled amplifier. Tube or solid state. An open secondary transformer has very high to virtually infinite primary impedance. That means you have a very large inductor on the OPT tube and the plate can rise to several KV with applied signal. Destruction of the OPT is a definite possibility.2) if you are to perform meaningful voltage tests, you want the amp to be in a typical operating condition. The very least is an ideally non-inductive resistor at the impedance of the OPT. More ideally there are some good speaker simulators you can build that get you closer to the non-ideal speaker load.
Edits: 09/18/17 09/18/17 09/18/17
Always load a tube amps outputs. In the short term you're probably okay, but better safe than sorry.
Dan Santoni
This post is made possible by the generous support of people like you and our sponsors:
Top of Page ]
[ Contact Us ]
[ Support/Wish List ]
[ Copyright © 2025, Audio Asylum, all rights reserved ]
