UK Vintage Radio Repair and Restoration Powered By Google Custom Search Vintage Radio and TV Service Data

Go Back   UK Vintage Radio Repair and Restoration Discussion Forum > General Vintage Technology > Components and Circuits

Notices

Components and Circuits For discussions about component types, alternatives and availability, circuit configurations and modifications etc. Discussions here should be of a general nature and not about specific sets.

Closed Thread
 
Thread Tools
Old 19th Jul 2017, 1:22 am   #1
joebog1
Banned
 
Join Date: Jun 2015
Location: Mareeba, North Queensland, Australia
Posts: 2,704
Default Back EMF suppression.

A question for the boffins.

I am building another amp, and have chosen to use a choke input for the power supply. BUT, I also have chosen to use fast sand diodes rather than a valve rectifier ( possible peak currents of about 300 mA, 4 X 5B/254M aka 807 in loctal form). In the past I have used a RC shunt across the choke to limit the switching spikes that the diodes create, but have still had insulation breakdown of the transformer secondary windings. I have never been able to measure the exact value of the spikes, but it seems like they occur at switchoff, and consequent collapse of the field within the choke ( 5H 300 mA, 30 ohms DC resistance). Apart from using an GEMOV ( varistor) across the choke, does anybody have better/more elegant solutions?

With thanks in advance, Joe.
joebog1 is offline  
Old 19th Jul 2017, 8:56 am   #2
Herald1360
Dekatron
 
Herald1360's Avatar
 
Join Date: Feb 2007
Location: Leominster, Herefordshire, UK.
Posts: 16,528
Default Re: Back EMF suppression.

Back EMF happens only when something happens to stop the current in the choke from continuing to flow in the same direction. In valve equipment I would expect any stored energy in the choke to be dumped into the load before the heaters cooled.

If the problem is fast spikes you can't see, then a MOV is probably the best way to deal with them. Add a suitable RC as well.

It can be most instructive to play with choke smoothing circuits in a simulator- one thing in particular to be wary of is the fundamental resonance of the L and the smoothing C(s). This can cause some quite spectacular voltages if it gets excited- switch on/off and any step changes in load or varying loads near the resonance can do this. It is damped by the choke resistance but if it's a "good" choke you may need to add more damping (ac coupled to avoid large power losses at dc).

What value of smoothing C are you using?
__________________
....__________
....|____||__|__\_____
.=.| _---\__|__|_---_|.
.........O..Chris....O
Herald1360 is offline  
Old 19th Jul 2017, 9:00 am   #3
Craig Sawyers
Dekatron
 
Craig Sawyers's Avatar
 
Join Date: Nov 2010
Location: Oxford, UK.
Posts: 4,942
Default Re: Back EMF suppression.

A choke input filter needs a minimum current draw or strange oscillations occur. That is why a resistor is usually put across the capacitor - to ensure that the minimum load current is drawn even if the amp is not connected.
Craig Sawyers is online now  
Old 19th Jul 2017, 11:00 am   #4
G8HQP Dave
Rest in Peace
 
Join Date: Sep 2008
Location: Solihull, West Midlands, UK.
Posts: 4,872
Default Re: Back EMF suppression.

If at least the minimum DC current is being drawn then there should be no choke spikes from a choke input PSU as the rectifier diodes should cleanly swap duty from one to the other near the AC zero crossing. If less current is drawn then spikes at the choke are possible. Two solutions: add a resistor to draw minimum current, add a small capacitor to ground at the rectifier output.

However, you say that it is the secondary where you had problems. This can happen when the supply is giving more than the minimum current so the switchover between diodes happens with significant current flowing. This means either a sudden reversal of secondary current (bridge with one secondary) or sudden cessation of current in one half (full-wave with CT secondary). In either case spikes at the transformer are possible. I would guess that a CR snubber across the secondary or small caps across each diode might fix it.

Your first job is to determine whether you have choke spikes or secondary spikes.
G8HQP Dave is offline  
Old 19th Jul 2017, 11:44 pm   #5
joebog1
Banned
 
Join Date: Jun 2015
Location: Mareeba, North Queensland, Australia
Posts: 2,704
Default Re: Back EMF suppression.

The amp will be very close to class A so current will be fairly constant after warmup.
I always fit bleeders across big caps ( 100 uF F&T ), but I have never calculated minimum current and usually fit 100K 2 watters. ( I got sick of the whack from discharging the caps into my fingers/hand )
I will have a look and see what I can come up with as far as minimum current goes.
Thanks to those that answered

Joe
joebog1 is offline  
Old 20th Jul 2017, 6:06 am   #6
Diabolical Artificer
Dekatron
 
Diabolical Artificer's Avatar
 
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 7,639
Default Re: Back EMF suppression.

I always put a 0.1u cap across each diode of a bridge, this suppresses the diode turn on spike which though only a few mV or so can be audible. Some diodes are worse than others as are some bridge rectifier's.

I know that has nothing to do with chokes Joe, was responding to Dave's comment above.

Andy.
__________________
Curiosity hasn't killed this cat...so far.
Diabolical Artificer is offline  
Old 20th Jul 2017, 8:26 am   #7
Craig Sawyers
Dekatron
 
Craig Sawyers's Avatar
 
Join Date: Nov 2010
Location: Oxford, UK.
Posts: 4,942
Default Re: Back EMF suppression.

The mechanism for spike generation (actually ringing) with a conventional capacitor filter is due to the reverse recovery of the diodes triggering a resonance of the mains transformer - the interwinding capacitance and leakage inductance. EI transformers are more prone to this (because of high leakage inductance and high Q resonance) than toroids (which although they have high interwinding capacitance have sub-uH leakage inductance and are inherently well damped)

The trick is not to put a capacitor across the diodes (although this does no harm) but to put a snubber across the transformer secondary.

There is an excellent test jig to measure this, and arrive at the correct snubber values here http://www.diyaudio.com/forums/power...-test-jig.html

There is also an excellent, partly threoretical, article on this by Jim Hagermann http://www.hagtech.com/pdf/snubber.pdf , but that assumes you have test gear to measure the transformer.

Again not directly relevant to an LC filter PSU, but fitting a transformer snubber even to this this may be no bad thing.

Craig

Last edited by Craig Sawyers; 20th Jul 2017 at 8:34 am.
Craig Sawyers is online now  
Old 20th Jul 2017, 9:39 am   #8
PJL
Dekatron
 
Join Date: Sep 2005
Location: Seaford, East Sussex, UK.
Posts: 5,997
Default Re: Back EMF suppression.

Switching off instantly removes the input current from the mains transformer primary resulting in back emf and possibly arcing on the power switch which will make matters worse. I would focus on suppressing the primary.
PJL is offline  
Old 20th Jul 2017, 10:17 am   #9
Craig Sawyers
Dekatron
 
Craig Sawyers's Avatar
 
Join Date: Nov 2010
Location: Oxford, UK.
Posts: 4,942
Default Re: Back EMF suppression.

That is why you arc suppress the switch with a capacitor or snubber such as Farnell 1187659 which is specifically designed for spark suppression.

Yes - you have to do the whole job. Use fast/soft recovery rectifier diodes, snub the transformer secondary to suppress ringing at several hundred kHz, spark suppress the switch, and if the transformer has >100VA rating use a soft start circuit to prevent massive inrush current, particularly with a toroid.

And while we are discussing this (and provided the valves are not directly heated) provide ten or twenty seconds delay between applying heater voltage and apply HT.

Craig
Craig Sawyers is online now  
Old 20th Jul 2017, 10:30 am   #10
G8HQP Dave
Rest in Peace
 
Join Date: Sep 2008
Location: Solihull, West Midlands, UK.
Posts: 4,872
Default Re: Back EMF suppression.

No need for power sequencing at the domestic power level.

Switch arcing occurs when the switch cuts the current, not when a diode cuts the current. However, a small X rated cap across the primary will deal with both causes of primary ringing.
G8HQP Dave is offline  
Old 21st Jul 2017, 3:18 am   #11
joebog1
Banned
 
Join Date: Jun 2015
Location: Mareeba, North Queensland, Australia
Posts: 2,704
Default Re: Back EMF suppression.

Excellent articles Craig!!

I can measure my leakage inductance for my transformers (I design and make them after all ) I always fit "spark suppressors across ( mains) switch contacts, but until I get to about 400 watts I don't bother with inrush limiting ( even with toroids) and I have found sequencing the power supplies to be more bother than worth, and have given this practice up. (big transmitters of course its essential).

The reason for my question was because I have a few toroids that have shorted turns or a shorted turn, with obvious flashover damage when I unwind them to find out why, it's immediately obvious that high voltage is the culprit, and the only reason I can figure is that the amp was switched off at zero crossover when magnetisation is at maximum, the collapsing field goes back into the transformer. With a conventional setup of capacitor input as opposed to choke input, the failures are more like reasonable failures- maybe one transformer fails in one hundred, but with large choke input the failure rate is more like it's going happen, rather than it may happen. I also think that half of the problem in solving this quandry, is that it's nigh on impossible to switch off the power at any point in the waveform cycle, and get meaningful results.

I do use varistors on every (high voltage) winding now in an effort to beat this bug. I also have a "bad habit" of making very quiet transformers by over winding them, or running them at very low flux densities, which results in massive inductances with so many turns on a toroid which is almost the ideal for efficiency to start with.

Thanks to all those that answered

Joe
joebog1 is offline  
Old 21st Jul 2017, 10:20 pm   #12
kalee20
Dekatron
 
Join Date: Feb 2007
Location: Lynton, N. Devon, UK.
Posts: 7,061
Default Re: Back EMF suppression.

Quote:
Originally Posted by joebog1 View Post
I have never been able to measure the exact value of the spikes, but it seems like they occur at switchoff, and consequent collapse of the field within the choke ( 5H 300 mA, 30 ohms DC resistance).
That shouldn't happen! On switch-off, the choke will try to keep its current flowing, so if fed from a bridge rectifier, it will drive the bridge into full conduction, and the diodes will clamp the voltage to two diode drops. If fed from a centre-tapped secondary, again, the diodes will clamp the voltage and (unless there is a lot of leakage inductance between the two halves if the secondary), the voltage will again be clamped.

Far more likely, is energy in the transformer's primary inductance. With a capacitor-input filter, any spikes caused by switching off at the point of maximum magnetising current (the zero-crossing voltage point), will be clamped by the secondary diodes into the low-impedance capacitor. But with a choke-input filter, the fast spikes have nothing to clamp them, because the choke after the diodes is high-impedance to spikes. So, the back-EMF of the transformer's collapsing field could easily zap the weakest item. If the transformer has an additional load, such as an LT winding feeding valve heaters, these will limit the back-EMF to a defined level, but I get the idea that this is not the case here. Why not just add a metal-oxide varistor across the primary, rated for 275V AC, as per your own suggestion?

Quote:
Originally Posted by joebog1 View Post
I also have a "bad habit" of making very quiet transformers by over winding them, or running them at very low flux densities, which results in massive inductances with so many turns on a toroid which is almost the ideal for efficiency to start with.
That bit doesn't make sense to me... Working at low flux densities, inductance is high, stored energy = 1/2 x L x I^2. So high inductance means low Imag, when you square the Imag its lowness 'wins' over the highness of the inductance, so stored energy drops with increasing inductance. Additionally, high inductance means lots of turns, meaning more winding self-capacitance, better able to absorb any spikes. (Though lots of turns does mean smaller wire diameter, with less thickness of the all-important enamel!)

Incidentally, I like designing transformers conservatively too, space permitting. Quietness and reduced switch-on surge is rather nice!
kalee20 is offline  
Closed Thread

Thread Tools



All times are GMT +1. The time now is 2:13 pm.


All information and advice on this forum is subject to the WARNING AND DISCLAIMER located at https://www.vintage-radio.net/rules.html.
Failure to heed this warning may result in death or serious injury to yourself and/or others.


Powered by vBulletin®
Copyright ©2000 - 2024, vBulletin Solutions, Inc.
Copyright ©2002 - 2023, Paul Stenning.