BigE's F5T build (Split from QA400) - diyAudio
Go Back   Home > Forums > Amplifiers > Pass Labs

Pass Labs This forum is dedicated to Pass Labs discussion.

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 25th November 2014, 11:41 PM   #1
BigE is offline BigE  Canada
diyAudio Member
 
Join Date: Aug 2008
Location: Toronto
Default BigE's F5T build (Split from QA400)

OK, so I'm going to select a few quotes from the QA400 thread that specifically deal with my F5T build. Those that are relevant to the measurement or use of the QA400 will be left in place.
---------------

OK,so I got this MARVELOUS shot by holding an 8" metal file wrapped in stranded 18 awg hookup wire connected between two alligator clips while in loopback mode.

Same at the back where the speaker posts are... I assume that at the rear, it is radiating from the unshielded power cord. I will borrow a shielded one and see if that helps.

This seriously looks like diode switching and harmonics from MUR3020 diodes.

Would snubbing work? Should I just follow Quasimido's rules? (Mark Johnson's bell ringing snubber).
Attached Images
File Type: png rectifier noise.png (437.0 KB, 476 views)

Last edited by BigE; 26th November 2014 at 02:03 AM.
  Reply With Quote
Old 25th November 2014, 11:44 PM   #2
BigE is offline BigE  Canada
diyAudio Member
 
Join Date: Aug 2008
Location: Toronto
Quote:
Originally Posted by 1audio
Simpler-
A coil of wire (with or without the file) from BNC hot to BNC shield. Connect to the QA400 input. Turn off the source side of the QA400 and start a continuous measurement. You should see 60 Hz and it will go up when you move the coul near an active power transformer for example. if the transformer has a rectified load you will see a lot of 120 Hz and harmonics. If loaded with an incandescent light bulb or a resistive heater it will be mostly 60 Hz.

These are ideal for this: Item # EP-101A, AC Probe-Axial-20 mV On Magnetic Shield Corporation and do pop up on ebay for lots less than $70.
__________________
  Reply With Quote
Old 25th November 2014, 11:52 PM   #3
BigE is offline BigE  Canada
diyAudio Member
 
Join Date: Aug 2008
Location: Toronto
OK four shots with that method, around an 8" file.

First is at the diode bridge at the front of the amp
Second is at the power cord at the back of the amp
Third is across the room at my workbench.
Fourth is in an enamelled steel can

There is no difference when the QA400 is placed in this can.
Attached Images
File Type: png rectifier noise2.png (430.5 KB, 453 views)
File Type: png power plug side.png (423.5 KB, 420 views)
File Type: png workbench.png (417.6 KB, 398 views)
File Type: png sensor in enamelled steel can.png (414.6 KB, 386 views)
  Reply With Quote
Old 25th November 2014, 11:53 PM   #4
BigE is offline BigE  Canada
diyAudio Member
 
Join Date: Aug 2008
Location: Toronto
Quote:
Originally Posted by 1audio
Time to look at the wiring around the transformer, bridge and caps. They should all be tightly twisted pairs minimizing the loop area so the radiation is minimized. There is a lot of harmonic energy around the rectifier. The LF falls off but the harmonics radiate further which may make sense.

Follow each current as a loop. Go from where it originates, e.g. power cord line to switch to fuse to transformer to power cord neutral. The area inside that loop need to be as close to zero as possible even if you need more wire to make it work. Transformer secondaries to rectifier to caps. Keep the centertap close to the other two wires or you have a 1/2 wave loop on each cycle.
__________________
Demian Martin
Product Design Services
  Reply With Quote
Old 25th November 2014, 11:53 PM   #5
BigE is offline BigE  Canada
diyAudio Member
 
Join Date: Aug 2008
Location: Toronto
It's all tightly twisted except that the center tap is independent of the other two voltages.

I am going to try to reduce the loop areas, but I am using a bunch of relays for the slow-charge circuit. So the wiring may not be as tight as would otherwise be expected.

I hope the issue is just at the diode bridge, though I may have to rewire that a bit. The center tap wiring heads straight for the cap bank, while the diode bridge is off to the side of the amp.

All of these circuits were mounted on the faceplate of the amp. Maybe they ought to be reorganized more sensibly? I will check through it with the coil of wire tomorrow, with the top off.

THANK YOU FOR YOUR HELP!

The replacement scope is on the way, so I will be able to snubberize the transformer.
  Reply With Quote
Old 25th November 2014, 11:54 PM   #6
BigE is offline BigE  Canada
diyAudio Member
 
Join Date: Aug 2008
Location: Toronto
Quote:
Originally Posted by 1audio
Wind the centertap with the leads from the transformer and on to the caps. Every 1/2 cycle the current goes through 1 diode, lead from the transformer and one lead to the caps, next 1/2 cycle through the alternate lead. Always returning through the centertap.

Its a common mistake since it seems that there would be no current through the centertap but there always is. You can take the current clamp from the QA190 and clamp that around the leads and see what is happening. Use the waveform view on the QA400 to see the waveform. It will not be pretty.
  Reply With Quote
Old 25th November 2014, 11:55 PM   #7
BigE is offline BigE  Canada
diyAudio Member
 
Join Date: Aug 2008
Location: Toronto
OK, so I need to make a braid.

I wired the secondaries to a barrier strip so that I could feed a single pair of wires to the rectifier, instead of two pairs... you see my transformer has 5 Center tapped secondaries. I connect the secondaries at the barrier strip prior to going to the rectifier, and leave one set of CTapped secondaries unused.

I was trying to use this boat anchor of a transformer for something.....Please don't tell me I will need to buy more iron!
  Reply With Quote
Old 25th November 2014, 11:56 PM   #8
BigE is offline BigE  Canada
diyAudio Member
 
Join Date: Aug 2008
Location: Toronto
re are a couple of "before" shots, looking for problem areas, with about 20 turns of wire wrapped around a metal file, then the file is taped from end to end with electrical tape, in case I drop it.

The first shot is with the sensor away from the DUT. Note how low the magnetic field tester is reading.... this is "ambient".

The second is under the amp, at the left channel barrier strip... I have included a barrier strip between the secondaries and the slow charge circuit ( which lives on the faceplate of the amp, along with the rectifiers and cap banks.). The intent is to be able to easily remove the toroid in the event that the amp needs repairs.

The third is under the barrier strip that connects right channel secondaries to the rectifier.

From here, the secondary wiring goes to two diode bridges, one per channel.

Visual inspection shows that the right channel twisting of the secondaries to the barrier strip is not as tight as the left.

Also, since the circuit path is Toroid -> barrier strip -> rectifier ->blocking relay -> thermistor || bypass relay -> cap bank, I think I can replace the barrier strip with the blocking relay, thus losing quit a bit of open loop area.

The circuit would then go:

Toroid -> blocking Relay -> rectifier -> thermistor || bypass relay -> cap bank.

This will take some work....

THANKS!

Perhaps it is best to simply move the toroid and main power supply away from the amp?
  Reply With Quote
Old 26th November 2014, 12:04 AM   #9
BigE is offline BigE  Canada
diyAudio Member
 
Join Date: Aug 2008
Location: Toronto
Quote:
Originally Posted by ["bear"
the slow start relay, if that is what it is, is best positioned in the primary, not the secondary.

I'm unclear on what a blocking relay is. I presume the bypass relay parallels the thermistor?

Again I prefer to put all the soft start stuff in the primary.

Perhaps you might want to do that.
Also run a test with the relays and thermistors out of the circuit, use a variac
to bring up the amp slowly. Alternately, a lightbulb (or resistor) in series with the primary, with a hard switch to bypass it (a relay is fine).

Are you sure you have all the secondaries in phase ?
How do you know?
How hot does the toroid run?

Are you below the VA rating of the toroid when it runs the two channels?
Does the amp exhibit the extra "junk" when running only ONE channel?
How is the chassis grounded? Did you try a hard ground, just to see if that has any effect? (I think ur using the bridge + cap method)

The question in my mind is if this picked up "junk" is being induced into the output or is actually on the rails. Have you looked at the rails with a scope??

If ur scope has an output for the vertical amp, perhaps you might run that into the QA while you probe the rails. You won't get the full S/N, but you don't need it. You'd be looking for the "junk" on the rails... then you can look for the junk on the + and the - rails and see if they are in phase or out of phase. The noise you are seeing is looking a lot like in phase noise, UNLESS the source of the noise is HUGE and this is badly cancelled residual.

IF the junk is NOT on the rails, then the junk may be coming into the input. In which case if you short the input on the board and maybe detach the input wires (where do they run?? what are they??) and it goes away, then ur problem is suddenly known.

_-_-
__________________
  Reply With Quote
Old 26th November 2014, 12:08 AM   #10
BigE is offline BigE  Canada
diyAudio Member
 
Join Date: Aug 2008
Location: Toronto
Quote:
Originally Posted by bear
the slow start relay, if that is what it is, is best positioned in the primary, not the secondary.
The soft start is positioned on the primary. It is the DIYaudio soft start.

Quote:
Originally Posted by bear
I'm unclear on what a blocking relay is. I presume the bypass relay parallels the thermistor?
This relay blocks the voltage from the transformer from reaching the thermistor. The point of it is to provide a delay that will enable the thermistor to cool, in the event that the user power cycles the amp say 45 seconds after power on. At that point, the thermistor is hot, and will not lower inrush current if power is cycled.

The cooling time for the thermistors to become sufficiently protective for the 88,000 uF per rail per channel is 1 minute. A 555 timer circuit closes the blocking relay after one minute so that the secondaries can provide power to the thermistors(1 per channel). The secondaries are wired to the NO contacts.

A second relay bypasses the thermistor after another minute. LTspice was quite clear that the second inrush caused by bypassing the thermistor would take another minute delay to get under 20 A. So, after a minute the thermistor is bypassed, the amp is ready to go.

Quote:
Originally Posted by bear
Also run a test with the relays and thermistors out of the circuit, use a variac
to bring up the amp slowly. Alternately, a lightbulb (or resistor) in series with the primary, with a hard switch to bypass it (a relay is fine).
Not to sound flippant, but what for? The size of the cap bank requires a slow charge circuit. It *must* work with the slow charge circuit installed.
EDIT: OK... it will clearly show that the problem is in the slow charge or not.

Quote:
Originally Posted by bear
Are you sure you have all the secondaries in phase ?
How do you know?
How hot does the toroid run?
Yes. I measured their relative phases. Not hot.

Quote:
Originally Posted by bear
Are you below he VA rating of the toroid when it runs the two channels?
Yes. The amp has been on now for a solid week.


Quote:
Originally Posted by bear
Does the amp exhibit the extra "junk" when running only ONE channel?
I don't understand what that test will show. The measurements above coincide with a visual inspection showing that the right channels twisting of the secondaries is looser than the left. Also, that the center tap is nowhere near the secondaries, as it connect to the PS ground directly.

I will likely be addressing this wiring problem tomorrow.

Quote:
Originally Posted by bear
How is the chassis grounded?
Third prong/safety earth.

Quote:
Originally Posted by bear
Did you try a hard ground, just to see if that has any effect? (I think ur using the bridge + cap method)
I am using a ground loop breaker from the zero volt line to the safety earth at a single point -- the safety earth. Hard ground is noisy. This was the first issue with the build that was solved.

Quote:
Originally Posted by bear
The question in my mind is if this picked up "junk" is being induced into the output or is actually on the rails. Have you looked at the rails with a scope??
The scope is in transit... it arrived DOA a few weeks ago, and a new one was sent out late last week.
The scope that is coming is a Rigol DS1102 I don't think is has a vertical output.

Given the charts from the DIY magnetic probe, it appears that the junk is induced on the rails. The junk looks the same at the barrier strip as it does the output. It is worse near the rectifiers. Once the scope arrives, I will add an R-C snubbing network.

I think there are two issues here. The ultra low frequency noise from the wiring, and the switc hing noise from the diodes.

Why do the caps not smooth the noise out?

Quote:
Originally Posted by bear

If ur scope has an output for the vertical amp, perhaps you might run that into the QA while you probe the rails. You won't get the full S/N, but you don't need it. You'd be looking for the "junk" on the rails... then you can look for the junk on the + and the - rails and see if they are in phase or out of phase. The noise you are seeing is looking a lot like in phase noise, UNLESS the source of the noise is HUGE and this is badly cancelled residual.

IF the junk is NOT on the rails, then the junk may be coming into the input. In which case if you short the input on the board and maybe detach the input wires (where do they run?? what are they??) and it goes away, then ur problem is suddenly known.

_-_-
Thanks! I will short the input and measure..... if I recall, that did nothing. Each input is a well twisted pair of 20 AWG solid core wire. Each are about 1.5" long.

I am going to try a couple of things... one is to move the relay that implements the 1 minute delay from power on to in front of the diodes, which then replaces the barrier strip, and ought to clean up one section of wiring.

When I do this, I will twist the wiring everywhere into much better shape -- It will probably require splicing to the secondaries.

The real issue may simply be that the toroid is too darn close to the slow charge relays, and that is where the noise enters. It may be smarter to make this a two box affair, and keep those relays far from the toroid. The toroid is a potted, but not electrostatically shielded Plitron.

Thanks bear, I hope I have explained everything, if not, please ask!

Answering your questions helped me see that there really are two issues here..... wiring and snubbing.
  Reply With Quote

Reply


Hide 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
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
QuantAsylum QA400 1audio Equipment & Tools 1561 Today 07:38 PM
FS QuantAsylum QA400 audio analyser jan.didden Swap Meet 3 29th September 2014 06:10 PM
What to do.. Build Aleph JX or F5T V3 monoblocks? sibbeli Pass Labs 22 2nd March 2014 03:00 PM
My F5T V3 build buzzforb Pass Labs 91 27th October 2012 10:01 AM


New To Site? Need Help?

All times are GMT. The time now is 07:47 PM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2