i am in the last year of electrical technology schol in italy, we use to calculate and measure transformers impedance every day for power grid use ( in the 100Kw to 3MW range) and we use a method called "industrial method" (the translation might be wrong...). we use a scalar voltmeter and a scalar ampmeter to ger the apparent power withour load in conjunction with a wattmeter to calculate the equivalent series impedance of the secondary winding that includes the primary winding reported to the secondary, we use this method to properly design the power distribution plants and the short circuit protection, so not the same use as this but it is very relayable and not vary complex to do. in this way you can also calculate the internal power losses due to eddy currents and hysteresis losses to make even accurate calcuations. if this can be helpful i can send to you the pages of my book with all the procedures ( in italian of course but since math in international......) hope to be helpful, 73 from IU3IRR 17years old operator :)
Hi Manuel, Your measurement process is very interesting and I would really like to have your procedure if it is not too much trouble to copy and send. My direct RUclips email is: theaudioshop@sbcglobal.net I have wondered how electric utility companies and large transformer manufacturers made these types of measurements.. Your comment is much appreciated. 73 WA4QGA
ElPaso TubeAmps Hi, it would be a pleasure to help you. I am going to do all the footage tomorrow, do you want also a video were I explain the process to make the Italian pages more clear?
Hi David, I really enjoy watching all the video's you post, even if I don't fully understand them. I am curious for instance when you measured the James transformer and it gave a reading of 20 odd Henry's compared to the older transformer which was around 42 plate to plate. What actual difference in construction (if any at all) would be necessary in building an amplifier if you were going to use one or the other in your opinion? Thanks, Tom
Hi Tom, I am sure no transformer expert so I can't tell you anything of value about how they are actually wound. I will say that, as I understand it, the higher values of inductance in the primary windings of these vacuum tube output transformers will provide better low frequency response. It sure seems like I am destined to build an amplifier (Williamson style) and use this Chicago Standard transformer. I recently bought it thinking it was an Acrosound but I didn't look really close until I got it home. I did measure it and it is a 10K primary with screen taps at 50% and not the usual 40% so... it looks like this is going to be my next project. Also, my grandson wants a vacuum tube headphone amplifier so that is also on the agenda. I plan on using the 1626 triode for output. There are a number of schematics of small 3 watt amplifiers using these triodes. The one I remember the most is called the, "Darling 1626 amplifier" and easy to find by an Internet search. It actually got its name because it was so cute and thus a "darling"... we shall see how it performs soon. Thanks for your comments.
Hi David, Thanks for replying so quickly. I am glad to hear that your grandson is taking an interest in tube gear... I think in a prior video you were concerned that you would have to sell it all, as no one (in the family) was interested in it. I had to look up the 1626 tube in my manual as I wasn't familiar with it. It said, it's typically run in class C for RF transmitting, an amplification factor of 5, with an output of 4watts. It would be great if you post a video of it when you build it, as it sounds interesting. Definitely build the Williamson style amplifier with the Chicago Standard output transformer, they are great amps, and you never fail to impress me with the quality of your builds. You are one of only a few, who actually measure and post the performance of your builds, trying to get the most out of these timeless designs, explaining why you change certain things and always backing it up with empirical data. I thank you for the time you take to make your videos and look forward to the next, Tom
As long as the impedance ratios are right, you cant do much about inductance, design wise - simply, MORE inductance gives you better bass extension at high power Even quite undersized OPT's can be ok if you dont turn your amp up full much (For guitar (BASS even???) The sound of an overdriven (undersized) transformer going into saturation can be .. interesting - not exactly Eddie Van Halen, a bit dirty, but fun to experiment with)
Very interesting method. Would it not be better to use the mean voltage from the Tek voltmeter, rather than the max? Doesn't that meter have a filter function to essentially give you a rolling average over X power line cycles or similar? Most bench meters have something like that.
I am glad you liked this video as I am still very pleased with how it performs. I just recently added an outboard pot to it to reliability and very accurately measure inductance values down to 100 mH. As I am sure you know, measuring low L value inductors (mH and uH) values, it is easy to do with several simpler methods but the output transformers are in a class all their own. General Radio manuals are a very good source to read on how to deal with iron core inductors. When trying to get a consistent and accurate reading on inductance values in, for example, the 40H range of a 20-50 watt vacuum tube transformer, it is amazing how sensitive the core is to being magnetized just by simply moving it around on the bench. Trying to make these measurements on a metal (steel) bench is hopeless. I usually don't take the time to go thru the trouble of finding the max voltage on the Tek voltmeter anymore as I showed in the video, as I have found with more experience, it is really not necessary. With care, I can consistently and with a very acceptable accuracy, measure 250H or more. 250H from CT to one side is what you will get for a 1000H coil end to end and I have a transformer that is 1:1 at 1000H each side. This method is actually easier to use than the GR instrument and I have the highest regard for GR equipment. The GR 1650A is a beautiful instrument but it is based on 1KHz and 1KHz is not a good frequency to try and evaluate very high inductance coils. For capacitors I have found that using an external source up to 20KHz works good with the GR 1650A but not so with low values of frequency like 40 Hz which I have found necessary with high value inductors. I want to make another video on this instrument but I don't think many people are interested. Personally, I find these types of manual/mathematical measurement methods the way to evaluate the inexpensive digital instruments of today and not the other way around.
Thanks for the reply. I certainly found it interesting. Did you post your spreadsheet someplace? That would be helpful. I know you show most of the formulae in the video, but we electronics guys are a lazy bunch ;-) Anyway, I appreciate the effort you put into these videos. I have done some myself. I really appreciate your skill in constructing all the amps you make as well. I lack a lot of construction skill (I blame my lack of tools, lol, but seriously I need at least a drill press). Anyway, keep up the good work, and I really do hope I run into you on the air. You said 20m; so I will keep my ears out. If you do exclusively 20m does that mean you have a Yagi or something special for 20? I do mainly 75m phone, but some 40m and 20m if I am on during the day. Thanks for sharing your knowledge.
Yes, it sure will. Shorting the secondary and measuring the inductance of the primary is how we measure leakage inductance which is down in the mH or possibly even uH range. It comes from the fact that the primary impedance is a reflection of the load impedance on the secondary. If the secondary is shorted, effectively, so is the primary and the only inductance left in the primary is referred to as "leakage" inductance. You do have to put the proper load on the secondary to measure the impedance of the primary.
Quite! 8/4 ohms on might not be very instructive, but a dead short gives a measure of leakage inductance - hence transformer quality *either shorting secondary to measure primary, or vice versa multiplying/dividing by tfr's impedance ratio Oh and what you got on the bench? Tango's? Tamura's? They look nice! (Guess I should play the whole video lol) Meows from London, Pete & Stereocat, Gas Electronics UK)
Hi Mike, The little inductance meters do a good job with small inductance values from a few mH down to uH but they don't work very well, if they work at all, with inductance levels above a few Henry's. Even the nice Tektronix and Fluke meters on the ohms scale when attached across the primaries of these high inductance transformers will not give even a resistance reading - they just blink at you. I wish I had shown that in the video. The General Radio 1650-A will give you the same reading that I get with this setup in the video but even then, great care and patients are necessary. Quite honestly, I trust this method every bit as much as what I get from the GR lab instruments. I suppose ultimately this is even more complicated than using the GR instrument but I became driven to find a reliable method other than a lab grade instrument that would accurately and consistently measure these values of inductance. You could see just how sensitive these big inductors are to external energies when they are being measured at such low levels and a light bulb is close to them. I tried "biasing" the core with small amounts of current flowing in the secondary windings and things just got really weird. The first method that I posted that I refer to as the "square root of three method" is too unreliable and derived values can vary by 100%.
Not much at all anymore. Doug is a very talented fellow and has a lot of hobbies and I think he spends a lot of his time on vehicles and exploring the desert nowadays.
well we havent spoken in a while, hope you're doing well. Last year I did the heathkit IB-2A restoration with .1% components and compared it against that against my GENRAD1650 in one of my projects, this coming year will be a restored IM12 up against an HP 331A. Should be interesting. Happy new year
It turns out that you can write those little radio associated applications as a plugin for Excel as a mix of C++ and assembly language so the applications reads the data in from the excel cells does the calculations and spits out the answer back into one of the excel cells.
Here are the results in percent of accuracy: 22/7 is, 1.000402499434770000000000000000 % 355/113 is, 1.000000084913680000000000000000 % The Excel function of, PI() is, 1.000000000000000000000000000000 % Derive PI by, 4*ATAN(1) is, 1.000000000000000000000000000000 % I got used to having to derive PI because way-back-when I had to derive its value in BASIC programming as there is no PI value in BASIC. Thanks gentlemen...
From the models I have used in LTSpice it appears that the output transformer causes odd order harmonic distortion but not even order. Is that what you have experienced?
That is a good subject to discuss. I have owned SS amplifiers without output/auto transformers and the distortion specs are very good but for reasons I cannot qualify, they seem to sound "weaker" and with somewhat less "dynamic presence" compared to the output/auto transformer designs. Seems they lack the low frequency presence and power of the transformer designs. Possibly too subjective for a legitimate discussion. I think there is much more to the sound for the ear than simple THD. It is no secret that vacuum tube amplifiers have more distortion than SS designs but many prefer the sound of the vacuum tube over the transistor. An eternal discussion that I have no hopes of ending tonight... :-)
ElPaso TubeAmps I found an Earth Producer 440 for $30.00, '76? SS. I'm down that rabbit hole and I'm learning alot. I'm on a mission. I study you, Uncle Doug, Brad & Paul just to name a few. This fall this Amp head will sound awesome. It doesn't have a choice. Thanks so much for doing that thing that you do so well!
i am in the last year of electrical technology schol in italy, we use to calculate and measure transformers impedance every day for power grid use ( in the 100Kw to 3MW range) and we use a method called "industrial method" (the translation might be wrong...). we use a scalar voltmeter and a scalar ampmeter to ger the apparent power withour load in conjunction with a wattmeter to calculate the equivalent series impedance of the secondary winding that includes the primary winding reported to the secondary, we use this method to properly design the power distribution plants and the short circuit protection, so not the same use as this but it is very relayable and not vary complex to do.
in this way you can also calculate the internal power losses due to eddy currents and hysteresis losses to make even accurate calcuations.
if this can be helpful i can send to you the pages of my book with all the procedures ( in italian of course but since math in international......)
hope to be helpful, 73 from IU3IRR 17years old operator :)
Hi Manuel,
Your measurement process is very interesting and I would really like to have your procedure if it is not too much trouble to copy and send. My direct RUclips email is: theaudioshop@sbcglobal.net
I have wondered how electric utility companies and large transformer manufacturers made these types of measurements..
Your comment is much appreciated. 73 WA4QGA
ElPaso TubeAmps Hi, it would be a pleasure to help you. I am going to do all the footage tomorrow, do you want also a video were I explain the process to make the Italian pages more clear?
manuel minutello Wow! This is way over my head, but very interesting. I will keep studying and catch up! ;)
good video...I'd Like to try your spreadsheet
Hi David, I really enjoy watching all the video's you post, even if I don't fully understand them. I am curious for instance when you measured the James transformer and it gave a reading of 20 odd Henry's compared to the older transformer which was around 42 plate to plate. What actual difference in construction (if any at all) would be necessary in building an amplifier if you were going to use one or the other in your opinion? Thanks, Tom
Hi Tom,
I am sure no transformer expert so I can't tell you anything of value about how they are actually wound. I will say that, as I understand it, the higher values of inductance in the primary windings of these vacuum tube output transformers will provide better low frequency response.
It sure seems like I am destined to build an amplifier (Williamson style) and use this Chicago Standard transformer. I recently bought it thinking it was an Acrosound but I didn't look really close until I got it home. I did measure it and it is a 10K primary with screen taps at 50% and not the usual 40% so... it looks like this is going to be my next project.
Also, my grandson wants a vacuum tube headphone amplifier so that is also on the agenda. I plan on using the 1626 triode for output. There are a number of schematics of small 3 watt amplifiers using these triodes. The one I remember the most is called the, "Darling 1626 amplifier" and easy to find by an Internet search. It actually got its name because it was so cute and thus a "darling"... we shall see how it performs soon. Thanks for your comments.
Hi David, Thanks for replying so quickly. I am glad to hear that your grandson is taking an interest in tube gear... I think in a prior video you were concerned that you would have to sell it all, as no one (in the family) was interested in it. I had to look up the 1626 tube in my manual as I wasn't familiar with it. It said, it's typically run in class C for RF transmitting, an amplification factor of 5, with an output of 4watts. It would be great if you post a video of it when you build it, as it sounds interesting. Definitely build the Williamson style amplifier with the Chicago Standard output transformer, they are great amps, and you never fail to impress me with the quality of your builds. You are one of only a few, who actually measure and post the performance of your builds, trying to get the most out of these timeless designs, explaining why you change certain things and always backing it up with empirical data. I thank you for the time you take to make your videos and look forward to the next, Tom
As long as the impedance ratios are right, you cant do much about inductance, design wise - simply, MORE inductance gives you better bass extension at high power
Even quite undersized OPT's can be ok if you dont turn your amp up full much
(For guitar (BASS even???) The sound of an overdriven (undersized) transformer going into saturation can be .. interesting - not exactly Eddie Van Halen, a bit dirty, but fun to experiment with)
very interesting. I´m doing some simulations on ltspice with a modeled transformer but I found that the iron losses change the inductance value.
Very interesting method. Would it not be better to use the mean voltage from the Tek voltmeter, rather than the max? Doesn't that meter have a filter function to essentially give you a rolling average over X power line cycles or similar? Most bench meters have something like that.
I am glad you liked this video as I am still very pleased with how it performs. I just recently added an outboard pot to it to reliability and very accurately measure inductance values down to 100 mH.
As I am sure you know, measuring low L value inductors (mH and uH) values, it is easy to do with several simpler methods but the output transformers are in a class all their own. General Radio manuals are a very good source to read on how to deal with iron core inductors. When trying to get a consistent and accurate reading on inductance values in, for example, the 40H range of a 20-50 watt vacuum tube transformer, it is amazing how sensitive the core is to being magnetized just by simply moving it around on the bench. Trying to make these measurements on a metal (steel) bench is hopeless.
I usually don't take the time to go thru the trouble of finding the max voltage on the Tek voltmeter anymore as I showed in the video, as I have found with more experience, it is really not necessary.
With care, I can consistently and with a very acceptable accuracy, measure 250H or more. 250H from CT to one side is what you will get for a 1000H coil end to end and I have a transformer that is 1:1 at 1000H each side. This method is actually easier to use than the GR instrument and I have the highest regard for GR equipment. The GR 1650A is a beautiful instrument but it is based on 1KHz and 1KHz is not a good frequency to try and evaluate very high inductance coils. For capacitors I have found that using an external source up to 20KHz works good with the GR 1650A but not so with low values of frequency like 40 Hz which I have found necessary with high value inductors.
I want to make another video on this instrument but I don't think many people are interested. Personally, I find these types of manual/mathematical measurement methods the way to evaluate the inexpensive digital instruments of today and not the other way around.
Thanks for the reply. I certainly found it interesting. Did you post your spreadsheet someplace? That would be helpful. I know you show most of the formulae in the video, but we electronics guys are a lazy bunch ;-) Anyway, I appreciate the effort you put into these videos. I have done some myself. I really appreciate your skill in constructing all the amps you make as well. I lack a lot of construction skill (I blame my lack of tools, lol, but seriously I need at least a drill press).
Anyway, keep up the good work, and I really do hope I run into you on the air. You said 20m; so I will keep my ears out. If you do exclusively 20m does that mean you have a Yagi or something special for 20? I do mainly 75m phone, but some 40m and 20m if I am on during the day. Thanks for sharing your knowledge.
I"m curious; would placing an 8-ohm load on the secondary change the inductance seen at the primary?
Yes, it sure will. Shorting the secondary and measuring the inductance of the primary is how we measure leakage inductance which is down in the mH or possibly even uH range. It comes from the fact that the primary impedance is a reflection of the load impedance on the secondary. If the secondary is shorted, effectively, so is the primary and the only inductance left in the primary is referred to as "leakage" inductance. You do have to put the proper load on the secondary to measure the impedance of the primary.
Quite! 8/4 ohms on might not be very instructive, but a dead short gives a measure of leakage inductance - hence transformer quality
*either shorting secondary to measure primary, or vice versa multiplying/dividing by tfr's impedance ratio
Oh and what you got on the bench? Tango's? Tamura's? They look nice!
(Guess I should play the whole video lol)
Meows from London, Pete & Stereocat, Gas Electronics UK)
Would you be willing to share the Excel spreadsheet?? Not everyone is great at Excel programming.
Absolutely - I just don't know how.
@@ElPasoTubeAmps if you want to email it to me I could post it on a few of the amp blogs. kevinbourke81@gmail.com
You can use this spreadsheet to follow the calculations from the video:
www.delatsch.com/wp-content/uploads/2019/05/Inductance-Spreadsheet.zip
Cant we use an inductance meter or is that maybe not accurate?
Hi Mike,
The little inductance meters do a good job with small inductance values from a few mH down to uH but they don't work very well, if they work at all, with inductance levels above a few Henry's.
Even the nice Tektronix and Fluke meters on the ohms scale when attached across the primaries of these high inductance transformers will not give even a resistance reading - they just blink at you. I wish I had shown that in the video.
The General Radio 1650-A will give you the same reading that I get with this setup in the video but even then, great care and patients are necessary. Quite honestly, I trust this method every bit as much as what I get from the GR lab instruments. I suppose ultimately this is even more complicated than using the GR instrument but I became driven to find a reliable method other than a lab grade instrument that would accurately and consistently measure these values of inductance.
You could see just how sensitive these big inductors are to external energies when they are being measured at such low levels and a light bulb is close to them. I tried "biasing" the core with small amounts of current flowing in the secondary windings and things just got really weird.
The first method that I posted that I refer to as the "square root of three method" is too unreliable and derived values can vary by 100%.
Gotcha. I have one on my Sencore LC-102 capacitor/inductor analyzer but also you can change the frequency in which you test at on it.
Hey, you hear anything from uncle doug?
Not much at all anymore. Doug is a very talented fellow and has a lot of hobbies and I think he spends a lot of his time on vehicles and exploring the desert nowadays.
well we havent spoken in a while, hope you're doing well. Last year I did the heathkit IB-2A restoration with .1% components and compared it against that against my GENRAD1650 in one of my projects, this coming year will be a restored IM12 up against an HP 331A. Should be interesting. Happy new year
PI=(22/7) is very close to actual value.
355/113 is way closer - write 113355, split in half, put the larger number on top - brilliant approx to pi :), only about 0.1 ppm out.
It turns out that you can write those little radio associated applications as a plugin for Excel as a mix of C++ and assembly language so the applications reads the data in from the excel cells does the calculations and spits out the answer back into one of the excel cells.
Excel has a PI() function.You don't have to write your own.
Here are the results in percent of accuracy:
22/7 is, 1.000402499434770000000000000000 %
355/113 is, 1.000000084913680000000000000000 %
The Excel function of, PI() is, 1.000000000000000000000000000000 %
Derive PI by, 4*ATAN(1) is, 1.000000000000000000000000000000 %
I got used to having to derive PI because way-back-when I had to derive its value in BASIC programming as there is no PI value in BASIC.
Thanks gentlemen...
Even using a good quality core the output transformer causes a distortion and logical loss of frequency response.
From the models I have used in LTSpice it appears that the output transformer causes odd order harmonic distortion but not even order. Is that what you have experienced?
Perhaps the the odd order harmonic distortion is why some amplifier designs try to get rid of the output transformer all together (OTL designs).
That is a good subject to discuss. I have owned SS amplifiers without output/auto transformers and the distortion specs are very good but for reasons I cannot qualify, they seem to sound "weaker" and with somewhat less "dynamic presence" compared to the output/auto transformer designs. Seems they lack the low frequency presence and power of the transformer designs. Possibly too subjective for a legitimate discussion. I think there is much more to the sound for the ear than simple THD. It is no secret that vacuum tube amplifiers have more distortion than SS designs but many prefer the sound of the vacuum tube over the transistor. An eternal discussion that I have no hopes of ending tonight... :-)
ElPaso TubeAmps I found an Earth Producer 440 for $30.00, '76? SS. I'm down that rabbit hole and I'm learning alot. I'm on a mission. I study you, Uncle Doug, Brad & Paul just to name a few. This fall this Amp head will sound awesome. It doesn't have a choice. Thanks so much for doing that thing that you do so well!