I know that the subject has already been discussed here but as I prefer to have the "good" answers and that I can not ask it to all the people here that I'm aware as being really competent (it would be a lot of pm), I decided to open a new topic.
I come to you for information when the "true" or "good" way to measure the reflective impedance of an opt. me, I have several ways to do it and recently, I entered into discussion with a member of here and we do not find the same results on the same opt. So, I try with two other opt which I know the impedance "displayed" and I fall just every time. either I do it with simplicity with a variac, a multimeter Vac and two or three resistances (4,8,16ohm) either I do with a generator BF in the range of 400 to 800Hz, always in AC, a multimeter Vac and the same resistances (4,8,16ohm) either I do with a generator bf with additionally a potentionmetre of 10 or 20k (according to the impedance estimate of the opt) according to the method of V1 / v1 / 2 and by raising the resistance of the potentiometer. and as I'm new in the tubes, I double my test first test = 100Vac in primary and secondary reading second test = 1Vac on each secondary level (one after another) and reading on the primary. then I meet both to see if it's okay. therefore, the formula or goal is to find the ratio of turns, then multiply it by its square, and then multiply the product by the impedance desired or estimated. please tell me if I missed something or if you did not understand my text.
I come to you for information when the "true" or "good" way to measure the reflective impedance of an opt. me, I have several ways to do it and recently, I entered into discussion with a member of here and we do not find the same results on the same opt. So, I try with two other opt which I know the impedance "displayed" and I fall just every time. either I do it with simplicity with a variac, a multimeter Vac and two or three resistances (4,8,16ohm) either I do with a generator BF in the range of 400 to 800Hz, always in AC, a multimeter Vac and the same resistances (4,8,16ohm) either I do with a generator bf with additionally a potentionmetre of 10 or 20k (according to the impedance estimate of the opt) according to the method of V1 / v1 / 2 and by raising the resistance of the potentiometer. and as I'm new in the tubes, I double my test first test = 100Vac in primary and secondary reading second test = 1Vac on each secondary level (one after another) and reading on the primary. then I meet both to see if it's okay. therefore, the formula or goal is to find the ratio of turns, then multiply it by its square, and then multiply the product by the impedance desired or estimated. please tell me if I missed something or if you did not understand my text.
it's you valvewizard !!
I am very happy to read you "in real life" and thank you very much for your site, I learned a lot with this one. can you explain why not use load resistance? I use it only to check if my calculations are good, but not to test the number of turns of the opt, I have a little trouble to express myself clearly in English.
I am very happy to read you "in real life" and thank you very much for your site, I learned a lot with this one. can you explain why not use load resistance? I use it only to check if my calculations are good, but not to test the number of turns of the opt, I have a little trouble to express myself clearly in English.
Oh, I thought you wanted to find the turns ratio! "therefore, the formula or goal is to find the ratio of turns, "
Exactly how are you trying to measure impedance? Use a variac, and measure the primary voltage and the primary current (e.g. measure the voltage across a 1-ohm resistor to calculate the current). Z = v/i
Last edited:
No load is best way to measure the turn ratio.....
The reason is that any current flow will introduce some error form copper resistance drop... The other error that can occur even when measuring with no load, is from the magnetizing current.... Also try to keep the voltages on the primary side high enough to keep the core out of the mud.... To low a frequency will have more magnetizing current based on primary inductance... Too high a frequency with be savaged by winding capacitance and leakages.. My personal preference is applying 400Hz .... Using mains frequency and variac should bring you fairly close.....If you measure the magnetizing current and have a fairly descent DC resistance measurement of the primary..you can roughly subtract the error to get you a little closer..keep in mind this is really a vector of imaginary and real components..
The reason is that any current flow will introduce some error form copper resistance drop... The other error that can occur even when measuring with no load, is from the magnetizing current.... Also try to keep the voltages on the primary side high enough to keep the core out of the mud.... To low a frequency will have more magnetizing current based on primary inductance... Too high a frequency with be savaged by winding capacitance and leakages.. My personal preference is applying 400Hz .... Using mains frequency and variac should bring you fairly close.....If you measure the magnetizing current and have a fairly descent DC resistance measurement of the primary..you can roughly subtract the error to get you a little closer..keep in mind this is really a vector of imaginary and real components..
Last edited:
maybe my method is 100% wrong...
but i do this way.
with a variac on secondary side of the OTP, i will set 1vac. then i measure the voltage at the primary.
(Vp/Vs)^2 = (Np/Ns)^2 = Zp/Zs
but i do this way.
with a variac on secondary side of the OTP, i will set 1vac. then i measure the voltage at the primary.
(Vp/Vs)^2 = (Np/Ns)^2 = Zp/Zs
Sometimes if you need a quick and dirty way to get a plate load measurement... If your OT has a common ground it works easier... Hook up a jumper wire from one of your heater lugs on power tube socket and connect the other end of jumper to the Secondary HOT terminal such as 8 Ohm terminal...Make sure no load or short is present at the output jack... Just pull the rectifier tube if applicable or turn on just the heaters ONLY.... Now you are applying half the heater voltage to the secondary ... no check AC voltage ratios and square times the ohms tap you are at..
Unfotunately it's level dependent what you measure. Try to simulate normal working conditions. No load will put the core in saturation. Better to measure plate-to-plate AC voltage @1kHz @half power and measure secondary voltage on a dummy load.
thank you for your answers, it comforts me in what I thought I knew. As I said, it all started with a member from here who kindly answered me when I asked him a question in pm. the question was about a pair (well two pair in this case) opt from a jukebox amp, and these opt have 6 or 7 secondary winding and when I wanted to determine exactly what were the winding that I could use to have an impedance thought of 6k to 7k, I did not find anything that corresponded to this range of impedance. So I thought that my (my) method was bad and so I tried on two other opt pair whose impedance I know and there I found just every time. I think that in this case, the answer is more complex because these opt are part of a circuit with several speakers and I never have the reflected impedance that I wish unless I redraw the whole circuit for 6v6 / el84 (10k) or kt88 (5k). what i do not know is if it's opt are worth it or if they are just good at making sound but not music.
No-Load will not cause saturation....
Saturation is a function of Flux-Density..
Flux density is directly based on the applied AC voltage and or DC current which is not the case here..
For example: I have a 1.75K Plate Load 100W OT.... I should be apply to apply up to 418 VRMS at 60 Hz .....if it is a descent OT... As long as you monitor the primary magnetizing current and see the inductance doesn't take a major dip...then your approaching saturation..
So does that mean that manufacturers overwind the secondaries so that when loads create resistance the tranny will meet its design specification? Every unloaded tranny I check has higher voltage. Does that give the true internal T/R? Then we must interpolate for design usable primary impedance? The design specs are all quoted as loaded operation.
That is a very good question....With power transformers ....they are speced and rated at full power....This accounts for core and copper losses..
With output transformers, usually not....
With output transformers, usually not....
Power transformers are specified in voltages on secondary with rated current load, that is usually 10% less than at idle. Output transformers are specified in reflected AC resistances, not voltages, so losses here do not apply. Of course, they would happen losing some useful power, but still reflected resistances depend only on ratios between wingdings, so you should measure no load voltages to find them.
I see that here too there is debate and I can tell you that you are not the only ones, I read some others on several forums in the world. I think it would be nice to refocus the question on:
reflected impedance (real)
reflected impedance (roughly)
reflected impedance (real)
reflected impedance (roughly)
yes, page 199 to 252.
the more I read, the more I realize that the problem is much more complex than it seems. the ratio of the number of turns is only the highest part of the tip of the iceberg and two different brand opt with a "sold" impedance of the same value, will probably not have the same behavior or the same resistance DC etc etc., etc
more iron is not necessarily better more turns are not necessarily better so either we do with what we have and we get an empirical result, either take all into account and adapt the diagram to the opt. I think two schools are opposed, the alchemists and the engineers. in the current state of things, I think I've learned enough to do no more nonsense and I'm like a lot here, able to get out of any amp with tubes, but the more I dig the subject and the more I realize that it has no substance and that the experience does for many in the tubes. I will continue digging, digging and digging again.
the more I read, the more I realize that the problem is much more complex than it seems. the ratio of the number of turns is only the highest part of the tip of the iceberg and two different brand opt with a "sold" impedance of the same value, will probably not have the same behavior or the same resistance DC etc etc., etc
more iron is not necessarily better more turns are not necessarily better so either we do with what we have and we get an empirical result, either take all into account and adapt the diagram to the opt. I think two schools are opposed, the alchemists and the engineers. in the current state of things, I think I've learned enough to do no more nonsense and I'm like a lot here, able to get out of any amp with tubes, but the more I dig the subject and the more I realize that it has no substance and that the experience does for many in the tubes. I will continue digging, digging and digging again.
I do not see any debate here... Wavebourn is one target..I am in agreement with him..We are focusing on Output Transformers..
Transformers will have copper and core loss....for various reasons....eddy currents, proximity effect, skin effect....ect..
The original poster is trying to measure his OT to figure the Plate load...
Plate Load is just a "feel good" number ...it is only valid at one discrete frequency when driving speakers..otherwise it is all over the place for various frequencies... especially for amps with no feedback.. Simple linear equations in the RDH book are not reality....
ok,good .
i resume
with variac 50 to 60Hz it's ok
with BF generator 1Vac 400 to 800Hz it's ok to determine the primary impedance, that's enough, the rest is another story 🙂
i resume
with variac 50 to 60Hz it's ok
with BF generator 1Vac 400 to 800Hz it's ok to determine the primary impedance, that's enough, the rest is another story 🙂
1Vac is too small...they BH loop is super non-linear down there....almost analogous to the sub-threshold region in FETS...
Get a variac and turn up to like 50V on the primary ...not knowing the power level of this transformer 50V on primary of un-loaded transformer should be fine.... Use two different AC Volt meters, One on Pri the other on Secondary.. Verify that the meters "track" beforehand...account for any error if any...
Last edited:
If to go deeper, here is how David Wafler wound his transformers, he had very few sections, but winding to different directions got very good HF frequency response. His patent is already expired.
- Status
- Not open for further replies.