Announcement

Collapse
No announcement yet.

How to test output transformer performance for overperforming amp

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • How to test output transformer performance for overperforming amp

    Hmmm, not sure what information can be gleaned from this scenario but I built this amp and it uses the Hammond 278CX PT which gives me about 590V B+ with four power tubes. My OT is Hammond 1650T which is listed as 120W output tranformer.

    My question is, this amp with 590V at the plates does about 150W at clipping. So by the numbers it is exceeding what the OT is made for, however it doesn't get hot or anything. It seems to withstand it without a problem. I don't think it will fail, but my question is, is there a way to determine how good its performance is other than just swapping OTs to see if there's a significant sound difference.

    I ask because I am considering buying a Heyboer 150W OT with the same 1900 ohm primary to see if the amp will sound better, fuller, better bass response, etc, when it is played at high volume.

    The amp really will ONLY be played at high volume because that's what it's made for. The amp seems to get maxed out at about 6 on master volume (like most amps usually seem to do). it gets kind of grainy and trebley sounding and adjusting tone controls starts to have less effect. Maybe the amp is must maxed in general, or maybe the speakers are maxed and since they're putting out full volume they can't really control the sound as much anymore. Maybe we are reaching power amp distortion and negative feedback is feeding a distorted signal back to the PI, unlinearizing the amp in general. I dont' know. But I wonder about the OT. Any thoughts other than just swap the OT and see?

  • #2
    Originally posted by nsubulysses View Post
    ...My question is, this amp with 590V at the plates does about 150W at clipping. So by the numbers it is exceeding what the OT is made for, however it doesn't get hot or anything. It seems to withstand it without a problem. I don't think it will fail, but my question is, is there a way to determine how good its performance is other than just swapping OTs to see if there's a significant sound difference...
    One test that you can do is to measure the full power bandwidth of the amp. Or just measure the maximum clean output at various frequencies. Your Hammond 1650T OT is rated at 120W, 30Hz to 30kHz. The 120W is not an absolute limit at all frequencies. If it is able to put out "150 W at clipping" as you stated, that is not necessarily exceeding its rating. We also don't know how much clipping you are allowing and that makes a significant difference in your measured result. I assume that your test frequency is the typical 500 Hz to 1 kHz that many people use. If you lower the test frequency I'll bet that you will find that the max clean power drops to 120W or less as you approach 30 Hz. You would need to make sure that the dive signal remains the same because, if there is somewhere else in the signal chain that attenuates the amplitude, that will obscure your test of the OT. From your description it seems that the amp is working pretty well and I would not put the OT at the top of the list of culprits that could be causing the sound characteristic that you don't like.
    Last edited by Tom Phillips; 02-26-2017, 12:38 AM.

    Comment


    • #3
      What Tom said,
      The 1650T is regularly used by DIY guys for 150W Bass Amps and 200W Guitar Amps.
      When you are NOT trying to reproduce 30Hz then the power handling goes up. (The Power Handling is mostly in the Low Bass Frequencies)

      For guidance: On a HiFi Monoblock with 1650T and a quad of KT88 (B+ at 510V) I found that:
      at 18Hz core saturation commenced at 50 Watts
      at 30Hz it would core saturate at its rated 120 Watts
      To me that says that at 60Hz (and above) it could handle >200Watts with ease.

      Fundamental of low E on your guitar is 81Hz, on a 4 string bass then low E is 40.5Hz but Bass Amps often have significant cut at these low frequencies (-6dB at 40 Hz is very typical).

      Cheers,
      Ian
      Last edited by Gingertube; 03-07-2017, 01:25 AM.

      Comment


      • #4
        Sorry I fell off from this thread. I had to do a bunch of repairs and of course could not work on/test my own amp

        Anyway, I did this test incorrectly and have not had time to do it again yet. I could get basically full power at all frequencies but I have to boost the input signal a lot. I guess the true way to do this test is to test all frequencies with the same standardized amplitude input signal, maybe like 100mV or something. And that I should input this signal at the PI rather than the input of the amp to test the frequency response of the OT/output section rather than the preamp and the rest of the amp. Is this correct?

        I tune my guitar to A standard so almost an octave down

        Also from Gingertube saying -6dB at 40 Hz is typical, it seems like getting a 150W spec'd OT would not help my low frequency response at high volume?

        I will do the test with standardized amplitude input signal at different frequencies later today and post my results. thanks !

        Comment


        • #5
          If you can get full power at all frequencies, then you don't need to worry about the OT. You need to tailor the frequency response of the rest of the amp.
          Unless I am misunderstanding the issue?
          Originally posted by Enzo
          I have a sign in my shop that says, "Never think up reasons not to check something."


          Comment


          • #6
            I tested it two ways -- signal at input of preamp and signal at PI

            1V pk-pk signal at input with master volume on 10 and gain volume set at point of clipping (gain volume on about 2) with 1K sine wave as my bench point. Without adjusting any amp volume settings and making sure 1V pk-pk was accurate through the frequency settings I got

            4 ohm load

            1KHz -- 24.5V -- 150W
            900 -- 23V --- 132.5W
            800 -- 19.5V -- -95W
            700 -- 16.6V -- 68.9W
            600 -- 14.9V -- 55.5W
            500 -- 12.9V -- 41.6W
            400 -- 11.2V -- 31.4W
            300 -- 10.2V -- 26W
            200 -- 8V -- 16W
            100 -- 4.24V -- .25
            80 -- 3.18V
            60 -- 2.25V
            40 -- 1.26V
            20 -- .46V


            With signal input at PI I needed 8V pk-pk to bring the amp to clipping with 1K sine wave. 8V pk-pk pretty much did the same for all the other frequencies too

            it sums up as this

            1K - 200hz -- 24 - 24.5V -- 144 - 150W
            100hz -- 23.9V -- 142.8W
            80 --- 23.9V -- 142.8W
            60 --- 23.7V -- 140.4W
            40 ---- 22.5V -- 126W
            20 --- 19.3V -- 93W

            Hmmmmmmmmm

            If I add more bass in the preamp the sounds is not so in your face. Maybe it is what it is. and the speakers can't take that volume and still do the full frequency range
            Last edited by nsubulysses; 03-07-2017, 10:03 PM.

            Comment


            • #7
              Something strange happens though. When the signal is applied at the input jack the signal at the input (purple) will start to distort when the amp reaches clipping at output (yellow). All the way up in volume until the output section begins to clip, the signal at the input jack is a perfect sine wave

              Click image for larger version

Name:	scope of input clip.jpg
Views:	1
Size:	1,013.2 KB
ID:	844920
              Click image for larger version

Name:	scope input clip.jpg
Views:	1
Size:	1.02 MB
ID:	844921
              Last edited by nsubulysses; 03-08-2017, 03:23 AM.

              Comment


              • #8
                Originally posted by nsubulysses View Post
                ...I guess the true way to do this test is to test all frequencies with the same standardized amplitude input signal, maybe like 100mV or something. And that I should input this signal at the PI rather than the input of the amp to test the frequency response of the OT/output section rather than the preamp and the rest of the amp. Is this correct?...
                Yes. An also verify that your PI to PA coupling caps are large enough capacitance that they are not rolling off the frequency response. You could do that by monitoring the AC amplitude feeding the PA grids.

                Comment


                • #9
                  Originally posted by nsubulysses View Post
                  ......Maybe it is what it is. and the speakers can't take that volume and still do the full frequency range
                  Good point. What are you using for speakers and what's the intention here. For example and some extremes, an open back guitar cabinet isn't going to put out the low end that a ported sub cabinet is. Maybe (and speculating) you need to explore a speaker setup and/or cabinet that is more efficient in the low end.
                  "I took a photo of my ohm meter... It didn't help." Enzo 8/20/22

                  Comment


                  • #10
                    It is a sealed 4x12 with internal baffle that divides it into two 2x12s. I have 2 weber grey wolves and two Celestion G12K 85s. The Celestions are definitely the more trebley, middy ones and are probably also a little less loud than the Webers.

                    If I replace them with two Eminence EM12s (EVM-12L inspired I guess) I bet it would be much better. They are supposedly very clean and full and loud.

                    I upped some of the capacitors that parallel the grid resistors in the preamp in the mean time and it's a bit better. like from 470pf to 1000pf or more. At least I know the OT is not the choke point

                    Comment


                    • #11
                      If you could find some EV Force 12's, those would be a good choice, also. And, about 1/2 the weight. Well, half the weight of the EVM-12L anyway. I guess only slightly lighter than the Eminence.
                      "I took a photo of my ohm meter... It didn't help." Enzo 8/20/22

                      Comment


                      • #12
                        You're not going to find a speaker that will reproduce much below 50 Hz. Subwoofers tend to be high-passed around 40 Hz, even. An average ported bass cab is down 3dB by 50-60Hz. With a sealed bass cab that's closer to the 80-100Hz range.

                        The bigger issue here is that your preamp is cutting a ton of bass. Your test where you injected signal right at the PI shows that the PI, power tubes, and OT are doing a fine job for a guitar or even bass amp. Heck, even a Hi-Fi amp... so the loss of clean bass - or early clipping of the bass - is happening before the PI.

                        I looked in the other thread and I don't know which preamp schematic you went with, but you might want to either try increasing the bass early in the preamp or cutting more treble late in the preamp.

                        Comment


                        • #13
                          Originally posted by nsubulysses View Post
                          I tested it two ways -- signal at input of preamp and signal at PI

                          1V pk-pk signal at input with master volume on 10 and gain volume set at point of clipping (gain volume on about 2) with 1K sine wave as my bench point. Without adjusting any amp volume settings and making sure 1V pk-pk was accurate through the frequency settings I got

                          4 ohm load

                          1KHz -- 24.5V -- 150W
                          900 -- 23V --- 132.5W
                          800 -- 19.5V -- -95W
                          700 -- 16.6V -- 68.9W
                          600 -- 14.9V -- 55.5W
                          500 -- 12.9V -- 41.6W
                          400 -- 11.2V -- 31.4W
                          300 -- 10.2V -- 26W
                          200 -- 8V -- 16W
                          100 -- 4.24V -- .25
                          80 -- 3.18V
                          60 -- 2.25V
                          40 -- 1.26V
                          20 -- .46V
                          You are NOT testing the OT response but the full amp, which to boot being a guitar one has huge bass cut and top boost.

                          With signal input at PI I needed 8V pk-pk to bring the amp to clipping with 1K sine wave. 8V pk-pk pretty much did the same for all the other frequencies too

                          it sums up as this

                          1K - 200hz -- 24 - 24.5V -- 144 - 150W
                          100hz -- 23.9V -- 142.8W
                          80 --- 23.9V -- 142.8W
                          60 --- 23.7V -- 140.4W
                          40 ---- 22.5V -- 126W
                          20 --- 19.3V -- 93W

                          Hmmmmmmmmm
                          Now we are closer, but even here you have coupling cap (amp input and power tube grids) time constants, a fancy name for bass cut, that are in the path.

                          If you want to measure OT response in a Lab, youŽll probably drive it with an SS amp, simply because these can easily supply a flat frequency response, are constant voltage output, very low distortion, etc.

                          If you want a more realistic frequency response measurement, then drive it with tubes but use 10X the standard coupling cap values, simply to put them out of the picture; then you are certain that any variation comes from the transformer itself.
                          Tubes, of course, are flat from DC to several MHz or at least, worst case, tens of kHz..

                          If I add more bass in the preamp the sounds is not so in your face. Maybe it is what it is. and the speakers can't take that volume and still do the full frequency range
                          Juan Manuel Fahey

                          Comment


                          • #14
                            I'm still recuperating from some plumbing repairs so if I sound foggy, take it with a grain of salt and tell me I'm drooling.

                            The limitation on power transfer ability in a transformer is ultimately either thermal (something melts in the fervent heat) or impedance (it can't let enough power into the M-field gyrations to get it into the secondary.)

                            The melting we understand. The impedance issues are more subtle. At high frequencies, the amount of power the transformer will let in is limited by the series leakage inductances. At low frequencies. the primary inductance shunts incoming signal through thr primary without ever jiggling the M-field to get to the secondary. Between these two extremes, it's mostly thermal. So a transformer power rating must ***always*** include frequency limits to make any sense.

                            The power rating of the transformer INCREASES from the power max at its lowest frequency. This is because the impedance of the primary doubles with each octave higher signal. A transformer that can do 100W at 30Hz can usually do 200W at 60Hz and 400W at 120Hz. That's the reason for those teeny, tiny transformers in power supplies. You don't need a big transformer at 100KHz to 1MHz.

                            Because of this, I seen no conflict between what you saw and the transformer's specs.
                            Amazing!! Who would ever have guessed that someone who villified the evil rich people would begin happily accepting their millions in speaking fees!

                            Oh, wait! That sounds familiar, somehow.

                            Comment


                            • #15
                              Originally posted by nsubulysses View Post

                              If I add more bass in the preamp the sounds is not so in your face. Maybe it is what it is. and the speakers can't take that volume and still do the full frequency range
                              Don.t forget the feedback network
                              "If it measures good and sounds bad, it is bad. If it measures bad and sounds good, you are measuring the wrong things."

                              Comment

                              Working...
                              X