|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
Hi Guys,
How do you measure the output impedance of SE amp ?Thanks
Follow Ups:
the old standby is to use a sig gen, a dmm, and a dummy load.first, hook up the sig gen, set it to zero signal, attach NO LOAD to the output terminals of the amplifier, but hook your dmm set to ACV across the output.
power up the amp.
slowly--CAREFULLY--increase sig gen output until you read something like 5vac across the output terminals. DO NOT READJUST YOUR SIG GEN OUTPUT FROM THIS POINT ONWARDS.
power down the amp.
now, strap your dummy load across the output. for illustration let's assume it's 8r.
power up the amp.
read the resulting Vac across the output terminals. let's assume it's 4.5vac.
power down the amp.
what have we learned?
with no load attached, you are reading the open circuit output voltage of the amp, with no current flowing. with the load attached, you experienced a drop in output voltage. the drop in output voltage is proportional to your internal amplifier's resistance... the higher the output Z, the higher the voltage drop.
with the load attached, what is the current flow in the output section of the amp? ohm's law v = ir, we get 4.5v = i * 8, or i = 0.563a.
now take the DIFFERENCE in the two measured voltages: 5.0 - 4.5 = 0.5v
ohm's law again to find the internal amplifier's resistance which solves for both a current flow of 0.563a and a voltage sag of 0.5v...
0.5 = 0.563 * r
r = 0.89 ohms
I believe you are measuring total losses not just output Z. The output transformer has insertion losses (copper and core) that maybe as much as Z out. Some folks say to use the same methode but at 10 ohm and 8 ohm loads this way some of the insertion losses are included. At each load your not going from no loss to full loss but some loss to some more loss. Not perfect but better.I measure transformer insertion loss buy putting a high voltage (30 or more rms)on the input of the transformer and measure the output, loaded and unloaded with the input adjusted to be the same voltage with or without load. If the Zout of the generator is low 75 ohms or less the losses should be transformer only losses not Zout losses.
Subtract insertion loss from your methode to get Zout.
What maybe the best methode but I never tried or heard of anyone trying is to measure a fullrange speakers Qes with a solid state amp then the tube amp in question and determining the affects of Zout on Qes to get Zout.
Z out on tube amps is very confusing with a lot of complexity of impedance and phase it would be a good subject for a book with lots of different methodes tested. Good luck.
scholl,i see what you're saying, but i don't necessarily agree that the transformer losses aren't part of the output Z to begin with, and shouldn't be considered.
when i look at the equivalent circuit of a transformer (sec 5.2.i, fig 5.9 in the RDH4) i see that the resistances of the primary and secondary windings are in series with the primary and secondary respectively. here they will indeed increase the output Z wrt the ideal case, and if they are measured as such then all the better, since it is a more true representation of the behavior of the amplifier.
the core loss is represented as a shunt across the ideal transformer, on the primary side in parallel with the primary inductance, which would serve to lower the secondary impedance. this effect may cancel a good portion of the increase in impedance offered by the copper losses, or it may not. (i suspect that in a well designed tranny the two would be very evenly matched, since they both represent a drop in overall efficiency and performance.)
It's very hard to say if insertion losses are to be considered in the Zout. I still think the effects of Zout on the speakers Qes maybe the best way to find out.
Ken,I'm a bit uneasy about having SE amp ON without load.
Would not it create some very high transient voltages or can even damage an amplifier ? All tube amp manufacturers warn again using their amps without any load.Thanks
hey sergey,you're right, of course--it IS a bit dangerous. that's why i emphasized that you should CAREFULY increase the sig gen voltage while keeping an eye on the vac of the secondary. :)
with a 5v output signal, your primary voltage swing will be well within limits.. think turns ratio--if it's a 5k to 8r winding, thats 25:1 turns ratio, so you're talking 25 * 5v or 125vac. this should be fine for any reliable, working condition transformer.
there are other methods of determining output z--including even using another power amp to actively drive the output of the amp under test. unfortunately this requires a separate amp. for this reason, the one i described is the one i use (and incidentally is not relegated to only output stages, but can be used on any stage in the amp).
Hi,If one feel uneasy about use a open curcuit as one reading (I can see reasons to be, though not over conserned) on can use - say- a 4ohm, a 8ohm and a 64ohm for readings.
It will be a bit more complicated to solve, but basically what we are interested in is the cange in voltage given a change in current.
I've been doing it that way for years with tube equipment and I've never had a problem - just keep the signal levels low. -jd
This post is made possible by the generous support of people like you and our sponsors: