|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
14.0.164.69
I have measured the OPT of the line magnetic LM219 integrated amp. The OPT is only 38H. The amp use 845 tube. What is the lowest bass I can get. I always find the bass is sort of short, and not very powerful. Is it anything to do with only 38H inductance?
Follow Ups:
To answer half of this question, we need to know the primary impedance as well as the inductance, and to answer the other half we need to know the rated power handling of the transformer and the frequency at which that rating is made. The first half tells you the small-signal frequency response; the second half tells you the power bandwidth.
That said, typical SET outputs have about 8 henries per Kohm impedanc, exceptiional ones about twice that, and acceptable ones half as much.
Thanks
10K primary
The whole tranformer is 11 pounds
114 x 55 stack
So that's 3.8 henries per Kohm, in the "acceptable" category but less than typical.
This assumes that the 38 henry measurement is accurate in the bass. It's remarkably easy to measure an unrealistic value for output transformer inductance. A typical meter with an inductance measurement works at 1000Hz and a few volts; this does not make enough magnetic flux to get the inductance to its stable value, at least for grain-oriented silicon iron cores. You need a lower frequency (below 100Hz) and a higher voltage (above 10 volts) to get there, in my experience.
Hi,
For measurement, I used auto transformer to adjust to 50V rms at 50Hz to apply on the primay in series of 10 ohm resistor.
By 2PiFL, 38H is calculated, what is the lowest frequency and output power in this low frequency this OPT can deliver
Excellent! That would work very well. Your measurement is quite accurate.
Hi Paul,
What would be the bass performance difference for 38H and 76H in this case
It would be typical of good 300B amplifiers (and will weigh about twice as much as 38 henries,for the same DC current capacity).
Such transformers (30 watts at 10K ohms) are quite rare in the market because they are so large and heavy.
"Such transformers (30 watts at 10K ohms) are quite rare in the market because they are so large and heavy. "
Sounds like you're describing the FS-100 Magnequest pair that I bought about 15 years ago. Don't recall the exact weight, but 20 or 22 lbs each sounds about right.
--------------------------
Buy Chinese. Bury freedom.
Probably more like 16 pounds for the FS-100. I don't have the specs, but a quick estimate is in the range of 40-60 henries at 100mA. (I have some proprietary information to base this on.) Since mass is roughly proportional to inductance for the same current, that would be 32 pounds at 80-120 henries to make an "above average" transformer.
Yes, it seems you're correct about the weight. I did a little digging and came up with a description from the original auction page, as well as a copy of the original instruction sheet.
I'm not sure the 40W rating in the auction description is correct. The instruction sheet states 30W, and I have run them at that level in a 211 A2 amplifier idling at 75-80 mA. They're fabulous transformers, maybe not "above average" as the term is being applied in this thread, but very sweet and clear, very satisfying to the ear, and absolutely not lacking in bass. I've never attempted to measure their inductance.
Incidentally, I've always wondered who seller "paulrichard" is/was. These were brand new when I bought them in the early 2000s, and they were packaged as though they might have come directly from MQ.
--------------------------
Buy Chinese. Bury freedom.
Thanks for the spec sheet - I'm adding it to my files :^)
With respect to the power handling, my calculations show 30 watts at 40Hz, following Mike's design rules as I understand them.
With respect to "above average", I am misusing that term when applying it to 845 amps and the like - it is rare to see more than 4-5 henries per Kohm at 30 watts.
In the 2A3/300B world, it seems 6 to 10 henries per Kohm is fairly common.
like
The measurement is done without the bias current
So the real L at low frequency, in this case, will be greater than the real use.
The best way to measure the OT trafo is driving it on secondary with a good ss amplifier with a 8 ohm or 4 ohm in resistors 50 watt in series.
The per primary close to the real Rp of tube.
In this way you can get the real performances.
This is fine for p-p
There is the possibility to inject the dc current to see what happen with bias current.
I will ahve all details and I am preparing a test set for this
In photo the proto of test set
Walter
This is single ended Walter.A lot better way to measure is in a real circuit*. We have the correct bias current and the correct source impedance.
Another interesting way to measure is with a ccs (see Walt Young for a good ccs).
A measurement with the source at the secondairy winding is not the way to do it.*Just with the endtube and a good tonegenerator
Edits: 04/08/21
I am not agree.
You can measure the s.e. and p-p really
This the only way to know the real performanceso of the trafo
Inside the circuit you have the tube charateristic that modify the results.
The main proble testing the OT is the large value of signal with a low distortion to test it in the right way.
With a goo ss amp that can deliver 30 volt rms on 8 ohms you can get for a ss.e.
For example, an OT for 300B with ratio of 1:20. For 8 watt rms the voltage is 8 Vrms on load but on primary there are 160 volt rms.
In which way you can get 160 Vrms applied on primary with a low distrtion to test it?
In attach two test on s.e. trafo, double C good level
One is the freq. response and other i a THD vs frequency
These are results of the trafo itself, and show a good performances.
As you see on THD as usual on low frequency the THD tend due the limit of the magnetic circuit.
These are real test and not simulation, that for a OT are not reliable.
I have done dozen of these tests on different types.
With my test set
Walter,The output impedance does not always follow Ohms' law. There is also capacitance and inductance in the source. That's why i like measure with a tube inside the circuit. Your way of measurement has the same problem, the solidstate amplifier is also limited, even more limited because most SS amplifiers are not so good above 50kHz.
BTW the plot you show is a PP transformer. It is not a very good transformer because the resonance is huge. Normally for a good transformer this resonance will be much lower and above 200KHz or better even higher.
Edits: 04/08/21 04/08/21 04/08/21
The plot are about single ended
The main problem is to test the trafo itself
Without the tube
And this is the easiest and efficent wayAnd with a properly interface it is possible to use a good sound card and ARTA sw
I have two Audio Precision plus more test equipments
And a differential probe that help for a large swingRegarding se amp you must have a good one
And the results show are fine mainly for thd
Walter
Edits: 04/08/21
Walter,The test you made is SE? 0,5dB at 10Hz? Well it is posible with OR a huge core (about 7-10Kg transformer for a 300B) OR a lot of copper loss(0,5dB or higher). But the resonance faillure is not acceptable. Arta is not good for high frequency measurements and as i mentioned before not many ss amplifiers are good enough for > 50kHz.
Actually i don't know any commercial ss amplier that can do 500Khz flat to do a good measurement.
Walter, What means dBr A in your measurements? Something creepy feelings I get with this. I wonder if you did the correct measuments???
dBA is not a lineair measurement and what dBr A is??? Tell us what kind of measurement this is?
Edits: 04/08/21 04/08/21 04/08/21
The trafo on test is a big one. On photo is one the right side
It is double C core amourphos laminate 0,023 mm
The resonance is fine at that frequency
But we are out in this thread. In case I will open another thread to continue
With a proper 24 bit 192 khz audio card you can read a lot of things until 96 khz
It is not necessary to reach 500 khz
Walter
The trafo on test is a big one. On photo is one the right side
It is double C core amourphos laminate 0,023 mm
The resonance is fine at that frequency
But we are out in this thread. In case I will open another thread to continue
With a proper 24 bit 192 khz audio card you can read a lot of things until 96 khz
It is not necessary to reach 500 khz
Walter
WalterBecause you use an amourphous core your core saturates much more earlier then a SiFe core so this core is not huge but just medium size. No, the resonance is still not acceptable, i have seen a lot better then this and it should be better because the phase shift is in the audio range.
Yes, a measurment till 500kHz is needed to see what the real performance is. We not want hf resonancence in the audio range.And to feel more comfortable: Tango is the referance.
Edits: 04/08/21 04/08/21 04/08/21 04/08/21 04/08/21 04/09/21
Hi
the Tango is fine but I never seen a heavy test as you can get with the method I say.
There is no way to get a great swing with low distortion, this is the fact.
The results I show is fine and the resonance is quite normal in a s.e. with that type of signal.
Regarding the trafo where I done the test, I assure you that is a very good stuff, 3,5 kohm / 8 ohm, ratio 21. Rdc 67 ohm. weight 3,8 kg (just for info)
In attach another test done on three trafos with a sw developped by F, Montanucci in Audioreview magazine where I have a collaboration and writing some article. The sw is propietary .
There is the plot of charateristic of impedance and phase . As you see there is a resonance at 2-3 kHz for each one ( with some difference of course). Until the resonance the modulus follow the inductive part of trafo, after resonance it follow the capacitive part.
Also the phase switch at that time.
On the little window on top you can see the value for some freq. and at 20kHz you read capacitance and not H.
This is a problem in the real world and there aren't measurement of this type available, just to understand what happen inside the trafo.
In addition another test with a s.e. trafo without nucleus ( the same FIAT on previous image), as you see when it reach 2 kHz ( the resonance frequency) it became linear; until 2 kHz the effect of magnetic circuit is important.
The third diagram is a Fiat (same of the test explained) trafo for 300B measured with my method, 1 watt, the primary closed on 750 ohm ( like 300B); double C, laminated 0,1 mm, this is not amorphous
Last is the THD vs. Frequency test; good results
In few word driving the OT by secondary is the easiest way to understand a lot of things specially at different level of signal, that is the main goal
Walter
There is only 1 reason for a high power test and that is for low frequency.
For that we already now that 3500 Ohm need at least 30H, if you have less then automatic the distortion of the transformer will be a lot higher. Unless you are not interested in good bass (or need it for a midrange speaker).
Almost all signal in the transformer is lower flux or very low flux. The more realistic measurements are at low output power. There is a good reason why they measured for decenia at 1W. I know some guys who even think that that is already way to high.
Anyway the distortion of the tube is for many frequencies domminant.
In my opinion is always better to know how good is the OT trafo where you are playing.
This means that a complete test set is preferable.
And this method is the simplest way to know the quality of OT.
To read the perfomances at 1 watt ( as I do normally) is a pieces of the job.
Of course the distortion of the tube can be higher but the final THD will increase if the specs of the OT are not fine and the results is the sum of both aspect.
Walter
Walter, your test is not a common standard, to do your test you need a very stable very good designed ss amplifier with a frequency response of at least 1Hz - 200kHz, preferable better (to see hf problems). Those amplifiers are very very rare*.
So no, your test is not easy, specially is you also want a SE test.The Tango test is better if a good ccs is used.
Actually i really don't know which amplifier can do this job. Dos the Audio Presicion has an inbuild amplifier capable to do... 50W / 1Hz-200kHz / with ultra low distortion, driving at least 0,5 Ohm to 32 Ohm?
Edits: 04/09/21
To find a good ss amp is not complicated.
In the photo you can see on proto of test se a board that is a fet amp with the BW of 200 kHz, measured with AP. 120 watt rms. Around 50 euros, it is a kit, easy to install.
I am preparing an extensive report on Audioreview magazine in Italy about this method.
Where there are a lot of real measurement also of the ss amp involved.
It is a reprise on an article already publishe ( the diagram with three trafo is on it) of Montanucci who developped the sw you can see in the previous post.
And the setup is not dofficult, of course a little bit of background on DIY is necessary but who want to play with this stuuf I suppose is able to do this.
Walter
and the distortion of your Fet amplifier? Any link available?
Just a sample of one of my own measuremnts at a transformer with ordenairy SiFe core 0,05mm.It was cheap core i have bought some time ago.
You can see it has about -0,9dB at 10 Hz resulting in about -3dB at 3Hz.
High frequency is reasonable at -3dB at 500kHz with a small resonance at 300kHz (0,15dB).
Edits: 04/09/21
The way how i measure make a huge differance. The cabel i use has 100pF and when i measure with a 1:10 probe with about 15pF the results are different. The resonance is a lot bigger 0,4dB and the frequency response is also bigger, more then 1MHz. Still measured with the same output sourse, 1400 Ohm.
This transformer will not be used in my amplifier, it's just a sample for testing.
Please decribe the test set up.
What I said before that the set up I described is similar to the real use of the OT trafo.
This make a great difference.
Also regarding the signal test.
When you say that the Zsource is 1400 ohm is a partial information. Also are missing the specs of the trafo you are testinig and the voltage of test.
And ther isn't way to get the BW you show with a real OT trafo s.e or p-p due the parasitic components. As shown on diagram of the three trafos
When I test the OT, the differential probe I use ( due to the high voltage I get) has a 200 kHz of BW and a low capacitance.
Walter
Walter,
I didn't say it is a "regular" output transformer. I measured it with a tube and it is an SE transformer for a line output. The voltage output was 2V but it dos't matter if it was 10x less or 10x more, the results would be about the same.
For SE you need a bias current and for frequency measuments the standard* measurement methodes i use are very accurate. There is no need for an alternative methode as you discribe specially if it is a SE transformer wich you can't measure correctly that way (didn't see a bias current)
* tube + transformer or ccs + Resistor +transformer.
I repeat again.
The each trafo is a bi-directionl device.
The only way to understand the real performance of itself are:
1- a generator that can give the nominal Vac ( for different Rsource) on primary to get on secondary (close with a nominal load) until the max power available with a THD almost equal to 0 %
2- drive it properly from secondary to have on primary (closed with a R equal to Rp of tubes) a signal that varies until the nominal power
There no alternative way.
There is not anything better to do measurments at or near maximum power, it's just an academic measurement wich gives no information about what is better.
A performance measurement at normal listening levels gives me more information. I am not interested in how much heat is generated at low or high frequencies at maximum power although this can be calculated if needed.
So how do you solve the bias current problem in your set-up to measure a SE transformer?
I must measure all parameters from minimum to maximum.
If I spend money or I ask to have a custom trafo I need to know all specs.
Where is the problem?
I show you the simplest way to do this.
If you don't consider a trafo as a bi-directional device we are loosing time.
Of course, a minumum of equipment you must have, but a good ss amp+scope+tester+audiocard+sw and a attenuator you can read anything
I have some measurement in AI and AP extension to convert, you will read some other things
I will prepare a specifc thread on this method.
Walter
In your opinion it's the easiest way. I simply not agree with that and i explained why.
This minimum equipement you describe are or to minimum or a lot more difficult then you say it is. Btw an Audio-Precision is way out of budget for most people €15k?
That's why you are the only person who dos this and everybody else use a different way.
But no hard feelings, if you like your set-up its fine for me. Just fix the bias problem and maybe somebody else will use it too for his 300b transformer test.
My point of view is based on electronic.
If you are agree that a trafo is a bi-directional device you can agree with my test set.
You didn't explained why this test set is not correct, you have shown a test with a signal trafo that is a differente beast mainly for the signal threated.
That's all.
To be precise I have two Audio Precision, Sys one and Sys Two
But I never spoken about AP to do the test, I spoke about a good sound card with proper setting.
Regarding bias current, be patient! :)
Walter
First of all, I never said that a transformer is not bidirectional. That is only in your mind.
So you can test both ways BUT your method is just an alternative way to test, not better, not easier. It also has shortcomings.
For SE you must have a bias current, if you make measurements at high, higher volumes you will even make the wrong measurements if you do not have the correct bias current.
With a sound card the options are quite limited. Decent frequency measurements are only possible on a modest scale. For someone like me who wants a really good transformer it is unsuitable.
For simple and good equipment i prefer a Analog discovery 2
https://store.digilentinc.com/analog-discovery-2-100msps-usb-oscilloscope-logic-analyzer-and-variable-power-supply/
To be precise
I wrote more than one time that the trafo are bi-directional device without any comment by you.
Then, I wrote that, in my opinion, the test set I propose is the best to check in deep the OT trafo.
If it is possible to have a proper generator with variable Zsource and a range from 10 volt to 200 vrms from 1 Hz to 200khz ( as example) the standard method is also fine.
But the costs are very high and I don't think there is something available
For bias, I say again, please wait other info. Something interesting. I hope.
I repeat again, a good sound card and ARTA ( for example) says a lot of things about OT.
And not related to freq. response but mainly on THD vs frequency. That is the most important parameter.
The link you sent show a reasonable level equipment with some limitation ( of course, due the price)
I was not interested in your "bidirectional" question. This is common knowledge and not worth paying so much attention to.If you want a thorough check of your OT, there are better ways to do it. A large signal is not very interesting information. The ear is not very sensitive to high sound levels, instead low level performance is much more interesting. We can also see the benefits of better core material when looking at lower levels. So no, I don't need a generator 10V - 200Vrms.
I am also curious if you could measure a transformer like this picture above with ARTA or the AP????
Edits: 04/10/21
I forgot.
If you aren't interested on bi-directional question is not a problem for me, of course.
The fact that you don't answer is clear for me.
Then, your opinion on large signal is opinable and not correct under the technical aspect; what do you want forma an OT trafo? to get 1 mV?
Then , regarding your test, please let me know the brand of your trafo and I will buy it for test; I have a good number of trafo tested left in a corner. I have also company that make for me the stuff on request ( the ones in the photo)
Last, of course the AP will test at the best the stuff and it is reference test set; surely your test set is not at the same level.
With Arta and sound card, as wrote more tha one time, you can understand lot of things.
Walter
Actually, this rejected transformer causes problems in a tube circuit.
And the real answer is no, we couldn't find the problem if we used an AP or ARTA because of their bandwidth limitations. Fine equipment in itself but good at other things.
But for that problem we could have used a scope and some other skills. Just out of curiosity I made a frequency curve.
This is a SE transformer 3500 Ohm tested with a EL34 in triode mode. Kathode bias and the low frequencie is a bit limited by the electrolitic capacitor.Would be nice to have a 10dB NF and see what happens, 3Hz -300kHz must be possible for SE. With a 300B even slighty better.
Edits: 04/11/21 04/11/21
Please decribe the test set up.
will buy this trafo for my test, please send details.
PS= it is funny, you show a diagram of one trafo seems to be fine but, then after my request to buy it, you wrote that this one got problem in a circuit.
It is a joke?
Something strange.
I repeat I am ready to buy your OT, to verify with a certified test set.
Walter
Walter,That was just a example of a transformer ( pp 5k) and what could be wrong and what happens if the test equipment has limitations. That perticular transformer i don't have anymore, it went in the trash bin.
The latest picture i posted of the 3500 Ohm SE i can make but not for sale, as we all know, i don't sell here anything. So don't send pm.My test equipment is good enough, the "certified" (what ever that means) test set would not make any differance.
Edits: 04/11/21
Ok, I understand.
At the end of the story you can't give any addtional specs or photo on the stuff you mentioned.
It seems a virtual world.
If I have the OT like you description I will be in hurry to sell them!!!!!
In attach the photo of two OT for SE, on the left is the one custome made from F.I.A.T. Roma for me, it is double C, 0,01mm, with a classical bell, for 300B s.e. Involved in a circuit the performance are good and the sound very fine.
The other is the one I spoke before, comes from Onori always in Italy, amourphos, 0,023 mm.
As you see I show the stuff I am speaking, this make a great difference.
Last, the AUdioreview magazine is insithe EISA so the test set is compliant with worlwide dierection.
Bye
Walter
Ohh, you like pictures of my projects?
The differance between you and me is that I actually design transformers.
Edits: 04/11/21 04/11/21 04/11/21 04/11/21 04/11/21
Good!
It is possible to buy or not?
not allowed here Walter
Ok
If not here
Where?
Just friends.....
I will repeat again.
On OT trafo you don't reach that type of answer. I mean only trafo not involved on a circuit
And please explain which type of signal you are using and the circuit; for a 2,7 volt on 8 ohm, 1 watt, with a ratio of 27 (in your case) you need 67,5 volt rms on primary. This is an acceptable start point
In addition a practical example of performance.
In photo a pair of monos built by me years ago, double p-p pf 300B
Then the response with and w/o FB (10 dB)
As you see w/o FB the answer is fine and wide
Then a CCL test, where I sent by AP some signal to reach the max power changin the load.
Normally with a poor configuration the curve tends to be horizontal because the quality of OT and the proper configuration.
In this case the magnetic circuit is big and the configuraion of OT is little bit different, single secondary at set at 5 ohm. (this is another aspect of the test set I use)
Lowering the load the power tend to increase ( with a good ss normally the curve is more vertical) and this is fine.
Of course the signal are continuos and impulsive.
Due to the huge Power bank the lines are similar; normally with poor capacitance ( less Joule) the curve are dissimilar, the impulsive is better, of course.
Walter
Really well done! I love seeing the power tubes up front. I'm surprised you didn't need to shield the small-signal stages.
--------------------------
Buy Chinese. Bury freedom.
Thanks
I have done some stuff also with different tubes as KT120 and KT150, same aesthetic.
The driver stage ia ECC88 + 6H30; very strong and very linear.
The amps are quite also with 300B with the filaments in ac.
The OT are custma with songle secondary set at 5 ohm
Walter
In addition there is another consideration
Each trafo is a bi-directional device
Walter
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: