also called a Zoebel network. High Z low freq, relatively low Z at high freq. useful for amps with marginal stability that can oscillate or produce supersonic nonsense that the tweeter isn't happy about.
Prevent No-Load Failure of OPT (Outputs transformers) without speakers hooked
- Thread starter mroboto
- Start date
also called a Zoebel network. High Z low freq, relatively low Z at high freq. useful for amps with marginal stability that can oscillate or produce supersonic nonsense that the tweeter isn't happy about.
AmadeusMozart
Well-Known Member
Perhaps attach some spade connectors or some thick 2mm connectors if the holes are too big?
AmadeusMozart
Well-Known Member
You are changing the subject just because I do not agree with you.
Once again: A HiFi amplifier is not the same as a guitar amplifier and it was about protecting a Sansui HiFi amplifier that the question was asked. Different requirements.
Last edited:
Retrovert
Addicted Member
The very subject of the thread is, "Prevent No-Load Failure of OPT (Outputs transformers) without speakers hooked"
Flyback diodes protect against inductive kickback from "no-load failure" referenced by the OP. Doesn't matter if its HiFi or guitar. Kickback is kickback.
AmadeusMozart
Well-Known Member
reading does not appear your strongest point, the first and follow up posts specifically mention a Sansui HiFi amplifier. You are changing the subject, MOV protection for HiFi is enough. You cannot protect someone from plugging in a 115V mains transformer into 240V. So why protect a HiFi amplifier for accidental disconnecting speakers? Just use proper connectors and for the one in a lifetime occasion that a HiFi speaker goes open then the MOV is enough protection.
You "solution" is not the best, or only, solution by the way, I would do it totally different but that's me. At least I have made it clear to others that yours in not the be all, end all even if you do not seem to understand that.
Have a nice day.
Last edited:
That's otherwise known as a Zobel network, and can be a very important part of engineering an amp for optimal stability.
However, randomly adding it to an amplifier, can do more harm than good. I am speaking from experience. I've had cases where a marginally stable amp, went into full blown oscillation when an incorrect Zobel network was added. It's all about the high-frequency resonances, and the phase margin involved. As always- proper testing is required.
Regards,
Gordon.
You can hear transformers when they are terminated into dummy loads instead of speakers. With speakers connected you just can't hear transformers making sounds but they still are, especially at higher volumes.
Retrovert
Addicted Member
The very subject of the thread is, "Prevent No-Load Failure of OPT (Outputs transformers) without speakers hooked".
Note the "without speakers hooked". What could be clearer?
The OP did not request your opinion as to the perils of selecting the wrong mains voltage, and did not request your opinion as to the "proper" connectors for his speakers.
The OP, instead, posted a question about protection, asked me for information, and I answered the question, explaining the issues involved in selecting a particular value for the bleeder resistor, and further clarifying the criticality of its placement in parallel with the driver instead of on an unused transformer tape. The fact that you do not like that legitimate question does not mean that the OP is not entitled to an answer. I have provided such an answer and, unlike you, am not threadcrapping or hijacking the thread.
You should not be attacking me for answering a legitimate question with legitimate solutions based in well-accepted and black-letter electrical engineering. HiFi and guitar amplifiers have identical behavior when a speaker is disconnected and the output transformer is unloaded.
The issue at hand is removing inductive kickback when the speakers are disconnected.
The MOV is a sacrificial device, and solid-state diodes are, again, the better solution. The fact that YOU do not understand the difference between a diode and an MOV, or that a guitar amplifier and a HiFi amplifier have similar issues, is YOUR issue, not mine.
Again, Faraday's Law of Induction and Lenz's Law fully explain the physics involved in inductive kickback and Wikipedia will explain it to you.
Edit: Fixed Typo and Grammar Issue
Last edited:
Retrovert
Addicted Member
Likely a Gas Discharge Tube (GDT), a (generally) glass- or ceramic-enclosed device. I mentioned those in No. 47. This is a different approach than the flyback (clamping) diode. A spark-plug operates on basically the same principle.
The advantage of the GDT over the MOV or TVS diode is is higher voltage rating and higher current.
The disadvantage is slow speed, permitting high voltage to flow before the GDT strikes, and degradation with repeated strikes. The intended use is with relatively rare events, such as lightning strikes, and not the continual ignition-coil style arcing from an unloaded output transformer. The lifespan depends upon current passed, but is typically about 300 to 500 depending upon current. (I just looked it up to be certain.) Which, at 2 strikes per second, works out to about 3 to 4 minutes, at one strike second works out to about 5 to 8 minutes. And then the output transformer is not protected, arcs over, and is destroyed.
So you heard the "pop" like because the GDT took just long enough to fire that some over-voltage hit the speaker, rapidly moving the cone with a high excursion.
Edit: Fixed Typo.
Last edited:
Phillips did this as i stated early in the thread.
@AmadeusMozart : What kind of amp is irrelevant , trying to inject crap into this thread ? Or just trying to make noise ?
Retrovert: You are correct. The transformer fails by way of an overvoltage situation across the primary when the output connections flap in the breeze. This can happen either if the circuit design is marginally stable and the amp takes off, (usually nearly rail-to-rail at high frequency-ugh!) and loses it with no load, or if the amp stays stable with no load (my home-made ones do-I tested for that) and musical peak voltages not loaded down cause the overvoltage situation to occur. In my particular amplifiers, the tweeter saver circuit is to prevent O.C. music peak damage. I have them installed for that purpose. Oscillation is not an issue in my case. With no tweeter saver and no load, my pair do not oscillate. I do use about 23dB of negative feedback in them. I would not drive them hard with music in that condition and expect them to escape that unscathed, however. It's best practice to have that RC in there regardless of amplifier gain & phase margin. I sport a pair of them in mine. It's possible that maybe some brands of tube could make the amps unstable, and I have not tested for that. Since I do not intend to produce & sell the amps, I won't do the necessary validation testing on that. Also, the amplifiers are circlotron type bridge amps with only 800 or so ohms primary impedance in my home made output transformers using quad build insulation on the windings, and that helps minimize an O.C. overvoltage problem. I certainly would take your advice on my other conventional push pull tube amps. 2 pair of KT77 tubes are used in each bridge type amp.
My Ampeg V4 (as noted by GordonW) has two strings of flyback diodes from each plates to ground. Originally, there were two diodes on each side but one side was replaced with three 1N4007’s so I guess they served their purpose...
My Traynors high wattage have thyrectors or TVS on schematics. They are gone on my YGL3 but I plan to add two strings of flyback diodes. I’ve also removed the internal drivers in this combo so as a protective measure, I added a wire to ground to the ext. 1/4 shorting jack. Also, two 220ohms@10 watts in series (440 ohms for a 4ohms tap) across the secondary.
Now, I’ve just bought a Traynor YBA-3 this week (around 120w power) and the thyrectors/TVS are still present.
The only bad thing I heard is that flyback diodes could short, rendering them useless.
My Traynors high wattage have thyrectors or TVS on schematics. They are gone on my YGL3 but I plan to add two strings of flyback diodes. I’ve also removed the internal drivers in this combo so as a protective measure, I added a wire to ground to the ext. 1/4 shorting jack. Also, two 220ohms@10 watts in series (440 ohms for a 4ohms tap) across the secondary.
Now, I’ve just bought a Traynor YBA-3 this week (around 120w power) and the thyrectors/TVS are still present.
The only bad thing I heard is that flyback diodes could short, rendering them useless.
Last edited:
Retrovert
Addicted Member
Yes! I totally forgot about that application. GDTs to protect the CRT from excessive voltage conducted by shorts from particles flaking off the cathodes. One for each of the three cathodes (at least in a color TV, one for monochrome) and the three plates. Really shoddily made as I recall. Maybe ceramic capacitors with a notch cut into the plates? Sound familiar? Not the variety which looked like a glass diode. I have no idea where my manuals for an late 1960s Heathkit TV are, or I'd take a look. Dude, we are sooooo old.
I do not believe this was a regular firing situation because the gap would be degraded, with both erosion of the electrodes and carbon trails (aka carbon tracking) leading to more arcs. Instead, this covered temporary shorts from particulates which would then burn out.
Ahhh, the kids will never experience the sphincter-tightening joy of using a high-voltage probe (ribbed for our non-electrocution pleasure) to verify the HV in a CRT TV whilst simultaneously trying to not electrocute themselves. That was one of the few tests which always gave me the willies. Mostly because all the old-timers told me stories about being knocked across the room by high-voltage discharges from contacting the CRT's anode after it had been discharged and should have been "safe". (Dielectric absorption aka soakage.)
Edit: added DA to soakage.
Last edited:
Retrovert
Addicted Member
Yes, shorting is a risk. So before you add a string of diodes, I suggest purchasing a single diode of sufficient voltage capability.
The reason is that in the event the diode is used the first diode shutting off will take the brunt of the voltage. This eventually burns out that diode. Either sharing resistors or a higher-voltage diode is therefore required to avoid failure.
I wrote all of that up here, including citations from learned tomes:
Lavane
Lunatic Member
That's the ones. In the 70's I worked for an electronics wholesaler. We supplied TV and radio repair shops for three counties. I used to have to deliver CRT's and pickup the dead ones to be rebuilt. We had pallets of old tubes to back to the factory. Those "rejuvenated" CRT's never seemed to last all that long. I remember delivering a few of the first microwave ovens back then also.
Retrovert
Addicted Member
Following up on my comment in No. 100 that a single higher-voltage diode should be used.
The Vishay GP02-40, as an example, is rated for 4,000 VDC @ 0.25 A continuous, 15 A peak for 8.2 mSec (120 Hz). The GI250-4 is similarly rated. Both are under fifty cents apiece from Mouser. (No connection to Mouser, merely selected it for the relatively convenient user interface.) I believe this device to be suitable, but urge readers to take their own counsel on selecting a device using some rational analysis and calculations.
Here is my reasoning.
The energy in the Primary is obtained from the power supply, so if the Primary has its flux converted to voltage (Faraday's Law and Lenz's Law) and then discharged through a diode each audio cycle, the maximum energy pumped into the system limits the voltage excursion to 2 × B+ @ the maximum B+ current per cycle. I note that the maximum current obtained could be higher if energy were repeatedly pumped into the system on multiple low-frequency cycles without a protective device, but the very design of the flyback circuit limits the energy input per cycle.
In the typical audio amplifier B+ is a maximum of about 250 mA, and that saturation level of current is rarely used, even with a high-transient peak. Plus, most of the energy in audio is in the lower frequencies, so the peak current cycle should be roughly where it is needed.
I therefore conclude, back of the envelope, the 250 mA of the diode should be more than adequate to discharge the energy in the Primary all day every day.
Selecting a higher-voltage device as the flyback diode avoids the issue of exceeding a 1,000 VDC peak for a conventional lower-voltage diode and the ensuing failure leading to a lack of protection or a short.
And, as always, I suggest fusing the B+ at the Secondary prior to the rectifier and again immediately following the rectifier, prior to the filter capacitors. (The filter capacitors will, of course, integrate and thereby remove the distortion component from fuse heating. See my quoting elsewhere on AK of Douglas Self's analysis of that issue. While I believe the effect is real and measurable using instruments, I remain agnostic as to whether or not human hearing can detect it.) This protects both the Secondary and the expensive tube rectifer which, in addition, should have diodes added to eliminate the PIV stress. While the intent is to protect against upstream failure in the event of an output tube short, the event of a flyback diode short — given the above, the probability should be reduced to the realm of the highly unlikely — would also be protected against.
YMMV.
The Vishay GP02-40, as an example, is rated for 4,000 VDC @ 0.25 A continuous, 15 A peak for 8.2 mSec (120 Hz). The GI250-4 is similarly rated. Both are under fifty cents apiece from Mouser. (No connection to Mouser, merely selected it for the relatively convenient user interface.) I believe this device to be suitable, but urge readers to take their own counsel on selecting a device using some rational analysis and calculations.
Here is my reasoning.
The energy in the Primary is obtained from the power supply, so if the Primary has its flux converted to voltage (Faraday's Law and Lenz's Law) and then discharged through a diode each audio cycle, the maximum energy pumped into the system limits the voltage excursion to 2 × B+ @ the maximum B+ current per cycle. I note that the maximum current obtained could be higher if energy were repeatedly pumped into the system on multiple low-frequency cycles without a protective device, but the very design of the flyback circuit limits the energy input per cycle.
In the typical audio amplifier B+ is a maximum of about 250 mA, and that saturation level of current is rarely used, even with a high-transient peak. Plus, most of the energy in audio is in the lower frequencies, so the peak current cycle should be roughly where it is needed.
I therefore conclude, back of the envelope, the 250 mA of the diode should be more than adequate to discharge the energy in the Primary all day every day.
Selecting a higher-voltage device as the flyback diode avoids the issue of exceeding a 1,000 VDC peak for a conventional lower-voltage diode and the ensuing failure leading to a lack of protection or a short.
And, as always, I suggest fusing the B+ at the Secondary prior to the rectifier and again immediately following the rectifier, prior to the filter capacitors. (The filter capacitors will, of course, integrate and thereby remove the distortion component from fuse heating. See my quoting elsewhere on AK of Douglas Self's analysis of that issue. While I believe the effect is real and measurable using instruments, I remain agnostic as to whether or not human hearing can detect it.) This protects both the Secondary and the expensive tube rectifer which, in addition, should have diodes added to eliminate the PIV stress. While the intent is to protect against upstream failure in the event of an output tube short, the event of a flyback diode short — given the above, the probability should be reduced to the realm of the highly unlikely — would also be protected against.
YMMV.