Go Back   Home > Forums > >
Home Forums Rules Articles diyAudio Store Blogs Gallery Wiki Register Donations FAQ Calendar Search Today's Posts Mark Forums Read

Tubes / Valves All about our sweet vacuum tubes :) Threads about Musical Instrument Amps of all kinds should be in the Instruments & Amps forum

DCPP - 'notch' in frequency response
DCPP - 'notch' in frequency response
Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 21st October 2019, 03:25 PM   #11
tristanc is offline tristanc  United Kingdom
diyAudio Member
 
Join Date: Dec 2015
Location: Oxford
Quote:
Originally Posted by jackinnj View Post
Oscillating at 128kHz?
Is that from estimating the period atop the square wave?

I'll stick a scope on the speaker terminals with no signal and see if there's anything going. Then put 1kHz and focus in on the shorter timescale. Perhaps for a FFT up to 200kHz and see if there's a spike anywhere.

Quote:
Originally Posted by DF96 View Post
I notice that the frequency response plots in post 1 show a slight rise in the left channel at the same frequency as the right channel has a dip. It may be the same effect, but of opposite polarity.
Good spot. So something oscillates at some frequency when either left or right channel is excited at ~40kHz.

Quote:
Originally Posted by Alllensoncanon View Post
Change C9 and C15 value might reduce the right channel ring a bit (ie: Change it to 1000pf. See if that reduce the 40KHz notch.
That should be possible - can clip in 100p then 220p without removing the board to test.

Thanks for all the input.
  Reply With Quote
Old 21st October 2019, 06:23 PM   #12
jackinnj is offline jackinnj  United States
diyAudio Member
 
jackinnj's Avatar
 
Join Date: Apr 2002
Location: Llanddewi Brefi, NJ
DCPP - 'notch' in frequency response
Quote:
Originally Posted by tristanc View Post
Is that from estimating the period atop the square wave?
Yes

I am working on one project at the moment, but will put the DCPP on the bench in a day or so.
  Reply With Quote
Old 21st October 2019, 06:32 PM   #13
baudouin0 is online now baudouin0
diyAudio Member
 
Join Date: Mar 2019
Must be a small phase difference between the two drive paths. Since you have a dual channel scope, disconnect the NF, apply say 20KHz to the input and measure the phase difference first on the plates of the phase splitter and then move to the output stage.

The circuit is slightly unusual in the R30/C7 R31/C10 seem to form some sort of bootstrap for the first stage. Certainly I would not be placing caps across the output transformer primary. Maybe repeat the test without these present.

The input needs to be driven from a low impedance to stop miller capacitance.

Last edited by baudouin0; 21st October 2019 at 06:40 PM.
  Reply With Quote
Old 21st October 2019, 09:13 PM   #14
tristanc is offline tristanc  United Kingdom
diyAudio Member
 
Join Date: Dec 2015
Location: Oxford
Quote:
Originally Posted by jackinnj View Post
Yes
I went looking with the FFT - plots attached. With the amp in situ (speakers connected, preamp connected and muted) I hooked a scope probe tip to the +ve and ground to -ve connections at the speaker. I.e., after a few meters of cable. What do you reckon of the spikes? I couldn't see anything obvious going up and town the timescales at max vertical sensitivity.
Quote:
Originally Posted by baudouin0 View Post
Must be a small phase difference between the two drive paths. Since you have a dual channel scope, disconnect the NF, apply say 20KHz to the input and measure the phase difference first on the plates of the phase splitter and then move to the output stage.

The circuit is slightly unusual in the R30/C7 R31/C10 seem to form some sort of bootstrap for the first stage. Certainly I would not be placing caps across the output transformer primary. Maybe repeat the test without these present.

The input needs to be driven from a low impedance to stop miller capacitance.
Thanks - this is really useful. The pico's AWG output impedance is 600ohm in to the amp's 100k. My preamp has a really high input impedance (>1meg) and ~200ohm output.

Here's what Pete wrote about those components:
Quote:
There is local feedback around the output tubes, in the form of plate-to-grid feedback (ala Schade) of about 10%. It's implemented as resistors from the plates of the output tubes to the plates of the driver tubes (the 220k resistors R29 and R47 above). This linearizes the output stage and provides a reasonably low output impedance of about 2.8 ohms (without loop feedback). You see small caps (820pF) from plate to GND to optimize square-wave response in the feedback path.
I can do some scoping without removing the board (which would be a pain but seems inevitable). In order, taking sweeps / view square wave after each step, I can:
  1. Clip in an extra 100p across each cap in turn, then both
  2. Disconnect NFB
  3. Scope driver plates and compare phase
  4. Scope outut grids and compare phase
  5. Reflow the joints around the components in question
  6. Remove/replace capacitors based on 1
Also attached is a gut shot from yesterday. Additions to Pete's layout / design:
  • Inputs / outputs / IEC moved to a rear panel
  • Toroidal PT (the Edcor buzzed too much and got way too hot for my liking)
  • Tag strip takes heater supply and rectifies for LED power indicator
  • Inputs moved next to each other, so one run is longer than the other
Attached Images
File Type: png FFT-nosignal.png (28.8 KB, 115 views)
File Type: png FFT-nosignal-200khz.png (30.3 KB, 111 views)
File Type: png FFT-nosignal-100khz.png (29.7 KB, 111 views)
File Type: jpg IMG_0405.jpg (449.5 KB, 105 views)

Last edited by tristanc; 21st October 2019 at 09:38 PM. Reason: Adding image
  Reply With Quote
Old 22nd October 2019, 10:11 AM   #15
baudouin0 is online now baudouin0
diyAudio Member
 
Join Date: Mar 2019
Sorry understud its local NF rather than bootstrap. I would have isolated the 820p from the transformer by splitting the 220k resistor. Anyway disconnect all feedback and look at the phasing of both drive signals. Just wondered if the 820p might resonate with the leakage inductance of the transformer. You may be better placing a very small cap across the 220k instead and have a zero - don't know.


Does the notch move up in freqeuency if you make the 820p say 470p.




Yes a high impedance into the first stage may affect the HF response - however its a pentode so the miller cap should not be an issue.

Last edited by baudouin0; 22nd October 2019 at 10:26 AM.
  Reply With Quote
Old 22nd October 2019, 06:18 PM   #16
tristanc is offline tristanc  United Kingdom
diyAudio Member
 
Join Date: Dec 2015
Location: Oxford
What a frustrating day.

Quote:
Originally Posted by jackinnj View Post
Yes
I measured the ringing - it came out around 41kHz

Quote:
Originally Posted by baudouin0 View Post
Anyway disconnect all feedback and look at the phasing of both drive signals. Just wondered if the 820p might resonate with the leakage inductance of the transformer. You may be better placing a very small cap across the 220k instead and have a zero - don't know.

Does the notch move up in freqeuency if you make the 820p say 470p.
Thanks - I did what I could, described below.
  1. Disconnect GNFB Ringing still here - attached
  2. Clip in an extra 100p across each cap in turn, then both Hardly any difference with 100p clipped in across C9 & C15, then C9 only, then C15 only. I then clipped in 470p across C9 & C15.
  3. Scope driver plates and compare phase Voltage too high for picoscope I think...
  4. Scope outut grids and compare phase Couldn't see any difference in phase looking at a 1khZ sine wave at the grids...

Getting annoyed I tried one last thing before firing up the soldering iron - a Zobel on the OT secondary. Used an RC decade box I twiddled the knobs to try and stop the ringing. Had a little success with 1k & 1u in series with no GNFB. Then connecting the GNFB made the ringing more obvious. Plots attached.

Struggling to interpret what's going on...
Attached Images
File Type: png NFB-off.png (52.1 KB, 88 views)
File Type: png 100p-added.png (56.1 KB, 15 views)
File Type: png 100p-added-C9only.png (54.4 KB, 15 views)
File Type: png 100p-added-C15only.png (54.5 KB, 15 views)
File Type: png 470p-added.png (56.5 KB, 12 views)
File Type: png Zobel-1k-1u-NFB.png (53.9 KB, 12 views)
File Type: png Zobel-1k-1u-noNFB.png (51.2 KB, 14 views)
  Reply With Quote
Old 22nd October 2019, 06:39 PM   #17
Mark Tillotson is offline Mark Tillotson
diyAudio Member
 
Join Date: May 2018
Location: Cambridge UK
The slew rates are very different for the two channels, something is very different between them, perhaps a wrong value resistor, or dry solder joint, very different bias points, bad tube.
  Reply With Quote
Old 22nd October 2019, 06:45 PM   #18
baudouin0 is online now baudouin0
diyAudio Member
 
Join Date: Mar 2019
Disconnect all NFB on both channels. Then work through from input until you see a difference - you can use a cap in series with your scope and 1M to ground if DC voltage too high or go on grid of final valve.

The only other thing maybe a faulty transformer having tried everything else could you swap them over?

Last edited by baudouin0; 22nd October 2019 at 06:50 PM.
  Reply With Quote
Old 22nd October 2019, 06:56 PM   #19
baudouin0 is online now baudouin0
diyAudio Member
 
Join Date: Mar 2019
Rather than worrying about the frequency response the THD on the right channel climbs as the signal level reduces badly. This is probably crossover distortion. If only one half of a output stage is properly working you may well see ringing. Check TP6 and TP3 but also check the resistors. They can go O/C if a valve went short. Check R9 and R10 too.

Last edited by baudouin0; 22nd October 2019 at 06:59 PM.
  Reply With Quote
Old 22nd October 2019, 08:33 PM   #20
tristanc is offline tristanc  United Kingdom
diyAudio Member
 
Join Date: Dec 2015
Location: Oxford
Quote:
Originally Posted by Mark Tillotson View Post
The slew rates are very different for the two channels, something is very different between them, perhaps a wrong value resistor, or dry solder joint, very different bias points, bad tube.
Are you using the pictures in post 7 to determine the slew rates? DCPP - 'notch' in frequency response

I think there's a period of desoldering / reflowing ahead. Will also try another set of driver valves.
Quote:
Originally Posted by baudouin0 View Post
Disconnect all NFB on both channels. Then work through from input until you see a difference - you can use a cap in series with your scope and 1M to ground if DC voltage too high or go on grid of final valve.
Yeah - desoldering is needed to completely remove the NFB so will likely do it in stages.
Quote:
The only other thing maybe a faulty transformer having tried everything else could you swap them over?
That might be an easier one to test first off...
Quote:
Originally Posted by baudouin0 View Post
Rather than worrying about the frequency response the THD on the right channel climbs as the signal level reduces badly. This is probably crossover distortion. If only one half of a output stage is properly working you may well see ringing. Check TP6 and TP3 but also check the resistors. They can go O/C if a valve went short. Check R9 and R10 too.
Thought it would be worth posting my voltages - pretty close between channels and mostly in line with Pete's apart from my driver plate voltages are higher.
Attached Images
File Type: jpg dcpp_sch_power.jpg (498.6 KB, 37 views)
File Type: jpg dcpp_sch_voltage-left.jpg (550.7 KB, 35 views)
File Type: jpg dcpp_sch_voltage-right.jpg (557.2 KB, 29 views)
  Reply With Quote

Reply


DCPP - 'notch' in frequency responseHide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump

Similar Threads
Thread Thread Starter Forum Replies Last Post
Frequency response Pygmy Multi-Way 20 4th April 2018 05:28 AM
in-room frequency response & transient response MCPete Room Acoustics & Mods 1 8th October 2011 07:31 AM
We know that Frequency Response isn't the end all be all... so what else is there? JRunior Multi-Way 219 5th February 2011 07:43 PM
DIY Frequency Response mic? Boscoe Equipment & Tools 7 15th October 2010 04:11 AM
Doubts on Phase Response and Frequency Response dumrum Multi-Way 11 5th April 2004 11:39 AM


New To Site? Need Help?

All times are GMT. The time now is 09:24 PM.


Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2019 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 14.29%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2019 DragonByte Technologies Ltd.
Copyright ©1999-2019 diyAudio
Wiki