|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
Hi,I am playing with a UTC A-11 input transformer configured as 500R:PP grids. On the secondary side I have 47k resistors from ends to CT and they are feeding the grids of 6SN7. The picture is a 1 kHz square wave from the calibrator of my scope. My question is: How can I change the loading to dampen the overshoot?
Follow Ups:
Far be it from me to cast aspersions on the value of theory but some practical results might help..My preamp has a balanced output transformer feeding to the grids of the balanced inputs on my power amp and the terminating resistance is switchable from 600ohms (design spec of transformer) through to 500k (grid resistances of power amp). This is fairly closely analogous to the situation you're describing.
Do they sound different? - yes they do. Is there a measurable difference in square wave response? - ditto.
Is one clearly superior? NOPE. Dull recordings can sound "livelier" with higher terminating impedance (and therefore more ringing) but some material (esp CDs) become intolerable on these settings so I switch in more damping.
Remember the engineering rule - there is always one obvious solution to any problem and it's invariably wrong.
Mark,Just read your post to my question farther up the page. I spent a good part of the day today listening to music and measuring square waves with a variety of loads on the input transformer. I agree with your assessment 100%. Totally unloaded gave a nice sound and sense of air, but it was bright. Too low of a loading was dull. The switchable load sounds like a good idea to match source material and mood. What it came down to was a tie, soundwise, between the 47k secondary load I had started with and a 500 ohm resistor in series with the primary as suggested by VoltSecond in the post below. They both gave the same square wave response and about the same insertion loss. When I put a 1 megaohm rheostat across the secondaries the square wave response at attenuated levels was far superiour with the 500 ohm series resistance.
On the scope, try some series resistance on the primary until you get just the first overshoot with not much of an undershoot following it.The plate resistance of the tube driving the transformer may be enough series resistance to kill the overshoot completely.
Play safe and play longer! Don't be an "OUCH!" casualty.
Unplug it, discharge it and measure it (twice) before you touch it.
I tried a 500 ohm resistor in series with the primary and it gave really close to the same picture as the 47k pair to CT on the secondary. Much experimentation (well for one day at least) later I think I like those loads better than more complete damping. Things may change with familiarity, but works for me now.
Have you tried it unloaded? Try a staight in from the transformer secondary to grid with no shunt resistor. Then look at the waveform. If there is still ringing then you can try a capacitor and a resistor in series from grid to ground but this is going to have to be determined by experiment with a substitution cap and resistor boxes. James
> Have you tried it unloaded?How can a transformer operate correctly without a defined load? Allowing the operating impedance of the secondary to rise to the near-infinite impedance of the grid greatly reduces the effectiveness of the secondary's inductance. Are you seeing good results in terms of frequency response with unloaded transformers?
i'm not sure what you are talking about.how does the secondary inductance work into this all?
the unloaded secondary allows the load seen by the source to be the inductance of the primary. adding additional loading (in the form of a predetermined load resistor) requires the source to work harder, so if i can i, keep the secondary load as high as possible. the majority of the time my ears agrre with this approach.
dave
> the unloaded secondary allows the load seen by the source to be the inductance of the primary.That's not the proper load for your source. The source should see a resistive load reflected back from the secondary, not the inductance of the primary.
Allo TK,
Why should the source see a resistive load reflected back?Consider a choke loaded driver stage, B+ at 0v, dc coupled to the following stage... is this really that different to an non-resistively loaded 1:1 interstage?
I don't argue for one second that some resistive loading may be useful to damp the upper resonance, but I don't follow why the source must see a resistive load.
I see it as a trade off a (possibly) undamped HF resonance for better driver linearity from the higher, mostly inductive (+ driven stage miller cap) load.
What do you reckon?
Several related issues are at work here. First, not all sources are the same. The rated load impedance will vary, as will the degree to which the source relies on that value being resistive in order to maintain minimal phase shift and specified frequency response. For example, a source with a 10 ohm output impedance that is rated by the manufacturer for a 100 ohm load won't be much affected by a reactive load of 5,000 ohms. However, a source that derives its output from the anode of a 12AU7 will likely suffer audible anomalies driving such a load. Source devices are specified by manufacturers for a resistive load; failure to provide it places the integrity of the signal at risk.Second, signals couple between the primary and secondary of a transformer by several means and with numerous modifiers. In addition to the desired magnetic coupling, transformers exhibit capacitive coupling between windings. The effect of this capacitance varies as a function of both the frequency and the load. As the load impedance is increased beyond the transformer manufacturer's design specification, the energy coupled by means of the capacitance becomes a larger percentage of the total energy at the output of the transformer. This can create very large peaks and nulls in amplitude across the spectrum, as well as phase shifts approaching 180 degrees. It also plays a role in the exaggerated ringing that occurs when unloaded transformers are tested with square waves.
You are correct regarding similarities to choke loading, but there are significant differences. The most obvious in my mind is the lack of primary-secondary coupling and capacitance. Then, there's the current flow. For a transformer to operate correctly, AC current must flow through the primary and magnetically induce a corresponding current in the secondary. Not only is that impossible when the secondary is unloaded, but a choke has no such requirement. In the case of the shunt choke, the goal is simply minimal AC current flow through the (single) winding. Distributed capacitance and other strays must be controlled, but the design process and in-circuit application are otherwise considerably simplified.
Hello TK,Regarding the first point, I thought we were talking only about the secondary load. But take your point about a 12au7 having a tougher time than a 10ohm source for a given tx and load.
Completely agree with your point about inter-winding capacitance, which in conjunction with the miller cap of the driven stage is, to my mind, the prime engineering argument for using damping resistors or zobel networks.
But I still don't follow that there must be appreciable current flow for the tx to operate "correctly"; you don't define what you mean by this - do you mean that, for eg a 1:1 tx, that Vout will equal Vin (neglecting capacitive effects)?
Alternatively, what happens to the transformer action if the secondary is truly unloaded? I've been scratching my head on this one - does Faraday say the voltages will be the same despite differing current flows (magnetisation current, and 0, for the pri and sec respectively)?
All that neglects capacitance of course!
(TK) Then, there's the current flow. For a transformer to operate correctly, AC current must flow through the primary and magnetically induce a corresponding current in the secondary. Not only is that impossible when the secondary is unloaded...sure its possible, it just follow ohms law like any other situation. the load is there from the grid to ground resistance, and since we are only coupling voltage, only a very small current is "allowed" to flow.
i agree with Vry, that an unloaded sec need not be a problem. you just have ot realize there are two things at work here. the response of the transformer wrt load, and the load presented to the sorce, and each must be looked at and judged individually.
> sure its possible, it just follow ohms law like any other situation.That's self-contradictory. Ohm's law says that because the grid of an A1 stage represents an infinite impedance (for practical purposes), current flow is not possible. I know it's tempting to look at a transformer as a voltage transformation device, but magnetic induction and proper transformer action require current flow. What you hear when you use a transformer without a resistive load is anyone's guess, but it's likely to include numerous anomalies that don't accurately represent the original signal.
TK wrote:::::What you hear when you use a transformer without a resistive load is anyone's guess, but it's likely to include numerous anomalies that don't accurately represent the original signal.::::
Hi TK:
That statement... had me scrachting my head... and turning to some tehnical data for an input trans out of the Peerless archives....
I don't know if I agree with you or not :=)... but I hope your statement above is at least partially not true....
and the qualifier I was thinking is... do we need to look at this across some limited bandwidth.... we could argue about what that bandwidth should be....
well... let me just cite some stats and figures of a LIMITED range of tests done on one particular input trans... where they tested the unit with the secondary loaded and then unloaded.... and at two operating levels...
allow me to quote some text;
"Frequency response curves for four operating conditions, divided into two parts. The first shows response at maximum rated power level with the transformer terminated resistively and unterminated. Teh same conditions of operation are shwn in the second part except for the -60dbm level which is representative of micrphone output. Extreme performance stability is illustrated by these frequency responses taken at wildly different power leels and under two extremes of operations; that is, with open circuited secondary and with fully loaded secondary."
when you read the FR graphs... what it whous is that at max power (+8dbm) the reponse is approx down 1db at 10 hertz and aprox 2 or 2.5 db down at 100khz with the secondary loaded. Unloaded at this same power level the FR is approx -3db at 10 hz and about 5.5 or 6 db down at 100khz.
at -60dbm power level... the same response curves hold for each condition (loaded vs unloaded sec) except for visually the FR at 100khz with an unloaded secondary does seem to be down about 1db further than at full power...
but, what has me curious, is that across the audio band (say 20 hz to 20 khz) the response is essentially flat... perhaps a few tenths of a db down at close to 20 khz...
there is also a distortion graph done at +8dbm... with the secondary loaded vs the secondary unloaded... here there is a difference.... the distortion is about five times greater at 20 hz with the secondary unloaded.
Now... none of the data supplied looks at or evaluates square wave response (though the bandwidth and it's flatness of response might suggest it would probably do reasonably well) .but... across most of the audio bandwidth (say arbitrarily 20hz to 20 khz) the unit looks reasonable from these two perspectives (FR and Distortion) whether the secondary is loaded or unloaded....
a quick note... the greater distortion of the unloaded secondary test condition at around seventy hertz subsides to approx the same level of distortion as the unit with the loaded secondary.
My hunch is that it is best to load the secondary... and I have done some listening tests (informal) btwn line outputs driven by a 6J5 wherein we loaded and then unloaded the secondary... my subjective impressions at that time was that I liked different aspects of the sound each way.... and did not find either to be conclusively "better" sounding as much as I did hear some differences in how they sounded...
but, I guess my question to you TK (in your opinion) is response characteristics and sound quality "unpredictable" if we limit the bandwidth and pick a "high quality" (however we might define this) transformer whenever the secondary is unloaded....
it's not a "set-up" question... I don't know the answer to this myself...
MSL
> is response characteristics and sound quality "unpredictable" if we limit the bandwidth and pick a "high quality" (however we might define this) transformer whenever the secondary is unloaded....Good question! Sadly, I think the answer is "yes," good unloaded results can't be achieved reliably as a result of those two factors. Generally speaking, transformers fare best into an open load when they are A) designed for a relatively low operational impedance and B) transforming a relatively low frequency signal. Think about the test you read being performed on an FS-100. Sweep it with a couple mV through the 8 ohm tap and measure the response at the 10K output. Now, try it with a 10K cardbon resistor on the output. I'll bet it's as different as Cadillacs and Corvettes. :)
Hi TK,If current flow is necessary, then how come you can have enormous voltages induced in an unloaded secondary? Guess an alternative way of looking at it is that via ohm's law, the V must increase to make up for the lack of current flow (V potentially going to infinity with things like insulation breakdown, core losses?? tempering it).
Not trying to be confrontational, honest! Just trying to get my head around this.
I seem to recall a discussion on the tweaks asylum regarding MC carts, and whether they were voltage or current generators. Same idea really, except we replace the primary+primary ac with a wiggling magnetic field. Vaguely recall one of the physicists jumping in and saying both views were right. Or wrong...
VRY, I don't take this as confrontational. I know the tone doesn't always mirror the intent in technical discussions like this. It's just the nature of posting these short messages.Anyway, here's what I said earlier:
"For a transformer to operate correctly, AC current must flow through the primary and magnetically induce a corresponding current in the secondary."
I didn't mean to imply that it's impossible to induce a difference of potential without current flow in the secondary. In fact, that's the first thing that happens when AC is applied across the primary. The lines of force cut across the secondary windings, inducing a voltage. However, the mere fact that the voltage is induced doesn't mean it will correspond to the turns ratio, or that the same voltage will be induced across the entire spectrum. Numerous other factors will influence those results, and all of them effect the output in a way that is directly related to the load on the secondary.
The capacitance between primary and secondary is only one factor. Capacitance distributed along the length of each individual winding is another. Then there's the resistance of the windings, the inductance, leakage inductance, core losses, etc. etc. The bottom line is that the transformer manufacturer must design the tranformer for a specific load, and the builder must apply that same load, or the transformer won't operate as intended.
My advice to anyone interested in this is to follow the manufacturer's recommendation for operational impedance (source and load) and forget about squarewave response. It's highly unlikely you'll be able to improve on the manufacturer's recommendations or the resultant performance in any meaningful way.
resistively, let's say for sake of discussion it is an 80k ct-ed winding, in goes a pair of as perfectly matched 40k resistors, right?Then we hook it to a pair of grids. Triode grids, that with Miller, offer something on order of 150 pF leakage to ground. Increase frequency up into the last octave we can hear and what has become of the load? gone to hell in handbasket I suspect( something like half or so ).
Considering what folks say about unloaded secondary, the capacitive grid load might just be enough for me. A large resistor might be a good place to start listening.
regards,
Douglas
Hi Vry: you wrote::::Not trying to be confrontational, honest! Just trying to get my head around this.::::
hey... this is some tougher stuff to understand and try to comprehend...but as I understand it on one level... AC current flow is necessary to achieve transformer action.... you must have ac current flow to "get the iron working"....
for years I too have been "trying to get my head around" many of these concepts in magnetics.... always a student I suppose....
I'm appreciative of many of the good posts on this board recently re: magnetics and their "workings".... they are thought provoking...
though I am also struck that we are far beyond "practical" diy understanding of iron and magnetics and how folks might best employ their ironware....
TK wrote:::::As the load impedance is increased beyond the transformer manufacturer's design specification, the energy coupled by means of the capacitance becomes a larger percentage of the total energy at the output of the transformer. This can create very large peaks and nulls in amplitude across the spectrum, as well as phase shifts::::
I often thought of what your saying above in the context of "why not to ratio" an output transformer upwards (i.e., say doubling the nominal primary impedance)... I just didn't know how to express my concern...in practice a 10K ohm output trans will use a different insulation packet than a 5K ohm OT. The 5K OT will not generally use as much insulation as the 10K unit (look at the voltage gradients as well as capacitive gradients of each trans)... ratio'ing up surely upsets these design 'balances' as you suggest...
which is another good reason to use a transformer at it's intended design center... and your rationale is an additional consideration on top of the fact that the ratio of R to L becomes much less favorable thereby creating large phase angles at 'lower' frequencies of interest.
.
Most of all the 1950s vintage mic preamps made by RCA and Western Electric used unloaded grid leads. The transformer was placed very close to the tube grid, as close as possible. Damping was not used unitl Neve got into the picture about 1968. I'n not saying that it should or shouldn't be. James
I didn't try that. I will do it with no resistive load and take a shot of that. Right now I am occupied with what the heck I did that brought my distortion way up. I decided to try a 5687, worse distortion somewhat to my surprise, and when I went back to the 6SN7 with the same operating points as before my 3rd order had about tripled and I am stumped.
New Page 1
Hi,
Here are the pics. The first is of the secondary unloaded, ie just
hooked to the grids of the 6SN7. By the way, all pictures were taken with
the amp actually running and using a 1x probe on one end of the secondary of the
input transformer only, that is I did not look at both ends. Any comments
on problems with the measuring setup are welcomed. Also, if you want any
other conditions I will give them a try if I have the stuff to do it. Nice
symmetry in the ringing is about all I can say about the unloaded condition.
Below is 94k from end to end, ie not connected to the CT.
This is 47k from each end to the CT. It's a little hard to tell from
the picture, but the overshoot and the undershoot both decreased a little and
the strange jump up it takes right at the end of the wave is less compared to
the picture above.
This is 23.5k from each end to the CT. Rise time seems to have gone up
but overshoot is damped and undershoot is gone.
Taking a break from pictures and measuring and giving it a listen with the
23.5k setup. Can't say I hear much difference yet, just might be a little
duller sounding. Will have to pay more attention in a bit.
By the way, found the cause of the extra distortion. Had hooked the cap
from the CT of the OPT to the star ground instead of to the high side of the
cathode resistor on the 12B4s.
Michael.
How short is the cable between the transformer and the grid? It's best with fraction of inches. I'd start with a value about ten times the output impedance of the transformer and then come down with a substitution box. Try 25pf, then 50 pf in series and if that doesn't work go to 100pf and so on. Don't use any more capacitance than you have to. James
It's on a breadboard, so the leads are clip leads. The secondary is rated 50k end to end. The 23.5k loading looks pretty good to my eye, although a faster rise time might be nice. Will the caps decrease the risetime? Admittedly, this is only a 1kHz square wave, but my generator is in a snit at the moment, so this is the only clean square wave I have.
will also determine equivalent input noise. So first choose the damping resistor (just it in shunt) and try to measure noise with the primary loaded with a 500 ohm resistor. You might have to resort to headphones with the volume tuned up and then sub resistors if you have no sensitive dbm meter. Just pick the resistor that offers lowest noise and then start trying caps calues with that resistor in series with the cap. The cap just tames the ringing. James
Resistor shunt to ground, cap for ringing. Will need to get some low value caps and then will give it a try.
I will be quite curious to ear your listening impressions of the various connections.
I have seen enough of the rhetoric on the Asylum about how *only* a ringing secondary( or the connection is described in such a way as to be assured of ringing) has that proper sound. It looks from your measurements that a reasonable amount of damping is a good thing. I am as they say, all ears.
regards,
Douglas
My first listen with the 23.5k load (per side)left me with the impression that it might have been more dull than the 47k load was. I haven't listened to it unloaded yet, but I will give it a whirl. The length of time that it rang when unloaded worries me somewhat though. I think I understand the argument that a square wave is not a fair test of an inductor, but a transient, to my mind, is like a square wave with an infinitely short period. Won't it ring as well?I am still working on how a volume control would fit into this and one scheme that was suggested was a rheostat between the two grids after the input secondary. When I tried that I found that a 1 meg rheostat was better than a 250k for not having insertion loss. Maybe I will play with a few medium to high value pots in that position and see if two birds are killed with one stone.
I spent a lot of time trying to damp the square wave ringing in an ElectraPrint linestage OT. Using a resistor/cap snubber across the secondary, varying the values of both and watching the results on a 'scope.I built a fixture so that I could switch the stage inputs and outputs between the signal generator/'scope combo, and a CD player and the rest of my system while leaving the linestage powered up.
Unfortunately, what I found was as I suppressed the ringing (amplitude and/or duration), the sound became dull and lifeless. Transients and leading edges really suffered (note the rise times on the scope images in this thread). Damping apparently is a double-edged sword. I finally removed the snubber completely and have never gone back.
Later I had a long talk with Jack Elliano and it was his opinion that any audio transformer will ring when subjected to a square wave - the true test is music signals in the passband. His point is they are inductive and don't handle the sudden transition of square waveforms.
The only time I've loaded the secondary of a transformer since then is when I wanted a specific impedance reflected on the primary. Otherwise all my input, inter-stage and output trans have unloaded secondaries.
That and 65 cents will get you a small coffee...
Pete
Hi Pete:you had initially written (in part):
:::...it was his opinion that *any* audio transformer will ring when subjected to a square wave - the true test is music signals in the passband. His point is they are inductive and don't handle the sudden transition of square waveforms.::::
and in a suceeding reply to Voltsec you wrote;::::Your point about not all transformers ringing is well taken.::::
I wanted to make a few comments on the text in the first passage... but these are not directed to you in particular... and I felt that another view could be considered as regards "square waves" and "ringing"....as voltsec also points out... it is not a *necessary* truth that *all* transformers will ring when passing square waves....
and I had one of those deep technical discussions via the phone with voltsec last evening re: square waves and ringing....
both intuitively and empirically I know that many well designed transformers can pass 10khz square waves very, very well. without uncontrolled ringing or excessive amplitude overshoot...
the passage I quoted above suggests that inductance limits achieving good square wave performance... but it's not OCL that limits square wave response (typically) but rather the amount of leakage inductance at some higher frequencies and the amount of capacitiance (stray, shunt and load capacitances) and their interaction which often creates the conditions necessary to have large peaks or spikes in the square wave response....
but... importantly... both the capacitances and the leakage inductance are factors that can be controlled by the deisgner of magnetics... and the interplay of the two do have some bearing on the sound quality as well as technical (electrical) performance of the unit.... of course, the Q of the transformer must be designed properly as well...
the point I am aiming at (with or without success :=)... is that if you control your leakage, your capacitances and the Q of the transformer you can very well have a unit that looks good on 10khz square waves....
it is not the case that the "inductance" of the transformer makes good square wave response unobtainable.
MSL
Hi Mike,Any misunderstandings or false assumptions from my post are entirely my own. Your explanation (while most excellent), went mostly over my head, as did a large part of my conversation with Jack Elliano.
You said "I wanted to make a few comments on the text in the first passage... but these are not directed to you in particular..."
I appreciate that, thanks. :-)
And to be fair, no implications should be directed at Jack either. Its best to assume I'm paraphrasing Jack poorly since I'm sure he's as good at controlling his stray capacitance and leakage inductance as you are. Someone reading my post and your response might get the wrong impression, and we wouldn't want that. You are both outstanding designers and builders of fine audio transformers.
My point was that square wave response is not the only measure of how well an audio transformer is behaving in a given circuit. I think you, I and Voltsecond all agree on that. :-)Best,
Pete
Sometimes I ask myself if I should address a particular topic or post... as I sense my response might be misunderstood and/or misapplied. I'm happy to see that you took it in the spirit it was intended. No aspersions were intended to any living or dead human...
just wanted to address the technical issues that had been raised.IIRC, I think Norman Crowhurst may have written an article on square wave testing of transformers.... if the subject interests you... look up some of Van der Veen's articles that he has written (I forget if they are on the amplimo site or the plitron site) where I am pretty sure that he addresses the issues of leakage, capacitance, and Q. His tranneys have very well controlled Q's by the way and very small leakage and etc....
and you wrote, further;
::::My point was that square wave response is not the only measure of how well an audio transformer is behaving in a given circuit. I think you, I and Voltsecond all agree on that. :-)::::
Absolutely. I don't consider any one factor of design to be paramount or the holy grail... and in my experience the designs which usually measure and sound the best are the ones that optimized for a wide range of performance and design considerations.
thanks again,
Thanks for posting.I don't quite agree with the statement:
> ". . .any audio transformer will ring when subjected to a square wave. . ."
Any transformer can be made to ring when driven with a squarewave, but a transformer does not have to ring when driven with a squarewave.
I do agree that:
1. Damping can be a double edged sword; however, double edge swords aren't necessarily a bad thing when you need a sword with two edges.2. Music is the most important test.
If it didn't sound good with the RC installed, you did the right thing in removing it.
Some transformers cannot be RC damped and work well in the circuit because they are too far off in their design, they are not suitable for the circuit they are used in or our expectations are too high. We also don't have to completely kill the ringing to do good. If you drop the overshoot from 60% to 15% we probably are at the "Quit and Eat Dinner" point. An "all real root" transient response, which has zero overshoot, is probably overkill in most, but not all, instances. Even a Bessel low pass filter has a bit of overshoot on a squarewave.
Also when using RC damping on a transformer, we have to remember that the input impedance of the transformer will be changing as we add the RC damping. Too much damping and the load gets too hard to drive for some circuits. If we have to add more than 3 times the circuit's self capacitance to get the ringing under control, we need to re-think the circuit. When we add 3 times the circuit's self capacitance as an external capacitor for RC damping, the ringing frequency (upper -3 dB bandwidth) is halved when the series R is shorted out.
On low level transformers where load damping and power aren't a big concern, sometimes adding a series resistor in the primary is a better choice than an RC on the output.
If you used the CD player to make the squarewave, the CD player's output is ringing on its own because of the truncated sine series used for the sampling. By the time you roll that off, the bandwidth can be seriously restricted.
Did the frequency of the ringing change when you added the RC without the R (adding the cap alone)?
If it did not, the RC was in the wrong spot.I started some web pages on this topic a long time ago. Not too many people showed interest so I've never gone back and tried to improve them. If you have comments, email me through Asylum E-mail. (Note: I only take Asylum E-mail from registered users.)
Here are the two relevant web pages.
D amping ringing in audio and power transformers (148k, 09/22/02)
I'm glad to see that you are experimenting and listening. If you do any more experimenting. Please post again and/or drop me an E-mail.
That's all for now.
Play safe and play longer! Don't be an "OUCH!" casualty.
Unplug it, discharge it and measure it (twice) before you touch it.
Hey VS,I hope I didn't come across as being an expert or having the last word on the topic. My results were empirical and subject to 101 variables that might affect the outcome. Vote for me and I'll lower taxes. ;-)
Your point about not all transformers ringing is well taken.
To answer some of your questions...
1. I used a square wave generator, not a CDP. The generator had a 600ohm output impedance driving a 50Kohm attenuator in my linestage.
2. The EP OPT was purpose designed for the application, and I was using 6J5GT's to drive it. This tube was the design target for this xformer. The linestage was "in circuit" the whole time - meaning it was driving a power amp (with dummy speaker loads o' course).
3. Yes, the freq varied with only a cap across the secondary.
Saving the best for last, I started this whole experiment after reading a couple articles on the web by some guy who's a Dr. Pepper drinker...;-) The methodology should sound familiar to you, and I'm fairly confident that (for my situation) the results are valid.
Here's my two cents (bringing our total to 67 cents) - its a pretty safe bet that if you use a well-made audio transformer in a circuit/application that it was designed for, that transformer ringing just isn't worth worrying about. On the other hand, when forcing PP iron to act as a parafeed OPT, etc., then all bets are off.
Best,
Pete
Oops - forget the part about PP transformers as parafeeds. There are much better examples of xformer abuse that might cause ringing. Not enough coffee...
Damping ringing in audio and power transformers (148k, 09/22/02)LC tank Q ver 1.2
could you post the url's for these?
thanks,
On of the links.http://www.siteswithstyle.com/voltsecond/LC_TANK_Q/LC_TANK_Q.html
Play safe and play longer! Don't be an "OUCH!" casualty.
Unplug it, discharge it and measure it (twice) before you touch it.
well put pete...
The correct resistance value is the one specified by the manufacturer of the transformer. Using that value may not entirely remove the ringing, but that's OK. It will provide the specifications and performance (including frequency response) that the manufacturer intended for the transformer.
You might want to try it with a capacitive load like the tube will present. Don't forget Miller...the capacitive load of the tube might do the trick, you'll have to see. If it were warmer in the shed, I'd be trying some of this, but my fingers don't stay functional long out there....
regards,
Douglas
Hi Douglas,That shot was taken with the transformer hooked up to the operating amp.
This post is made possible by the generous support of people like you and our sponsors: