Circuits living dangerously when powered off? - diyAudio
Go Back   Home > Forums > Amplifiers > Solid State

Solid State Talk all about solid state amplification.

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 28th November 2005, 03:08 AM   #1
diyAudio Member
 
Join Date: Sep 2002
Location: Sweden
Default Circuits living dangerously when powered off?

A couple of days ago I asked myself a question I don't think I have ever asked before, and that I don't think I have seen anybody else consider either. The circuit I had in mind was a JFET buffer, but I think the question is equally relevant for a lot of other circuits. The question was:

What happens if the circuit is powered off and receives an input signal?

That is a quite normal situation, that most equipment experiences once in a while, or even very often. Maybe we switch on the CDP before the amp, or the power amp is switched off while we are listening in headphones etc. etc.

To illustrate my point, consider the JFET buffer shown below. What caught my attention was that a JFET is actually a diode and it will conduct like a diode if the gate voltage is higher than the voltage on drain and source. Suppose the buffer is powered off, that is, the rails are "open", and that there is an input signal. During the negative half of the signal, there is no problem since the signal voltage should not be expected to exceed the max rating for the FET. But during the positive half of the signal, J1 will act as a diode. It will conuct a current from gate to source and the input impedance of the following equipment will determine the size of this current. Assuming this output impedance is a reasonable value, we are not likely to exceed the max gate current rating. Should the output by chance be short circuit, we will likely exceed the max current though, but a short circuit is not a normal situation, though.

However, even if the rails are open, there might be decoupling capacitors, as shown in the schematic, so a current will flow from gate to drain and to ground via C1 . This is almost an AC short circuit, so we are likely to exceed the max gate current rating of the FET. The cap will get charged, thus giving the drain a positive bias that makes the gate-drain diode conduct only at the very top of the positive half cycle, thus limiting the current. However, that effect takes place only after the cap has already been charged, and the max gate current will be exceeded while charging the capacitor. I don't have a memory scope, so I couldn't measure the initial charge currents, only the steady state currents, which are (presumably) much smaller, but I fail to see what could limit these initial charge currents, unless we take measures to do so.

What this suggests is that although the buffer is safe for normal input voltages while powered up, it may be stressed beyond component maximum ratings when powered off!

A simple solution in this case could be a series resistor at the input, that limits the gate current.

I haven't considered other types of circuits yet, but similar things are likely to happen also in other circuits. For instance, the inputs of an op amp are usually not allowed to exceed the rail voltages by more than one or a few Volts. What happens if the op amp gets an input signal exceeding that limit when powered off?

OK, time for comments. For instance, do any of you ever consider this when designing circuits? Or is it maybe me making some mistake and imagining a problem that doesn't exist?
Attached Images
File Type: png jfetbuffer.png (10.6 KB, 230 views)
  Reply With Quote
Old 28th November 2005, 06:44 AM   #2
Stocker is offline Stocker  United States
diyAudio Member
 
Join Date: Dec 2003
Location: Austin
One more reason to use BJTs?

Seriously, that sounds like a pretty good question, especially if you spent a lot of time matching input pair devices....
__________________
Jesus loves you.
  Reply With Quote
Old 28th November 2005, 12:03 PM   #3
MikeB is offline MikeB  Germany
diyAudio Member
 
MikeB's Avatar
 
Join Date: Apr 2004
Location: GŁtersloh
Typically jfets are open between source <-> drain when left alone. You can measure that with simple dmm. In this case the charge to the caps would be immediately shorted. Also, the inputvoltage might be ~1v, the biggest charge would be ~1v, quite harmless.

Mike
  Reply With Quote
Old 28th November 2005, 12:31 PM   #4
diyAudio Member
 
Ultima Thule's Avatar
 
Join Date: Jan 2004
Location: Koskenkorva Land
Hi Christer,

I didn't know that the Gate is not fully isolated but have an effective diode in the G D direction...

But how about using a diode in series with upper JFET and an additional small decoupling capacitor from D to ground?

Highest frequency (20 kHz?) / Xc of decoupling capacitor determines the highest current we want to allow...

Cheers Michael
__________________
"If transistors are blueberries and FETs are strawberries, then tubes must be.. pears"
Michael 29th January 2010
  Reply With Quote
Old 28th November 2005, 12:34 PM   #5
diyAudio Member
 
Join Date: Sep 2002
Location: Sweden
Michael (B)

The d-s resistance is not very interesting, since it seems not to matter for the gate-channel diode when conducting. JFETs work as diodes between gate and drain and between gate and source. The SJ170, which I measured has a slightly higher voltage drop than an ordinary diode but seems not to have a very high forward resistance.

As for signal levels, it depends a lot of how and where one uses the buffer (or whatever circuit we consider). Line levels often have peak values of at least 2 or 3 Volts. So even if the average level will be much lower, there might be occasional high levels. One may also ask what it means in practice to violate the gate current max rating occasionally? Is the max rating absolute also for occasional peaks? We don't know, since the data sheet doesn't tell.
  Reply With Quote
Old 28th November 2005, 12:40 PM   #6
diyAudio Member
 
Join Date: Sep 2002
Location: Sweden
Michael K,

Quote:
Originally posted by Ultima Thule

I didn't know that the Gate is not fully isolated but have an effective diode in the G D direction...
JFETs work like that, and it is obvious if you check how they are made up. MOSFETS are a different matter.

Quote:

But how about using a diode in series with upper JFET and an additional small decoupling capacitor from D to ground?
Yes, that might work, but I would prefer a series resistance at the input, which wouldn't do any harm unless it is a very-low-noise application. I only meant that this is a problem that might be worth paying attention to and solve, not that it is difficult to solve.
  Reply With Quote
Old 28th November 2005, 12:43 PM   #7
MikeB is offline MikeB  Germany
diyAudio Member
 
MikeB's Avatar
 
Join Date: Apr 2004
Location: GŁtersloh
The 2k170 is specified with max gate current of 10ma, quite much i think ? You might add a 2k in series at the input, if you have an inputfilter you need it anyway ?

Michael (Ultima), jfets are VERY different to mosfets... they are even very different to bjts, they are something special. Their negative VGS-threshold is exceptional useful sometimes !

Mike
  Reply With Quote
Old 28th November 2005, 12:49 PM   #8
diyAudio Member
 
Join Date: Sep 2002
Location: Sweden
Quote:
Originally posted by MikeB
The 2k170 is specified with max gate current of 10ma, quite much i think ? You might add a 2k in series at the input, if you have an inputfilter you need it anyway ?
Yes, it is 10 mA. I agree it is simple to solve the problem with a series resistance. My point was just that such a resistor might be necessary, but it is seldom or never seen in schematics. And I was also making the more general point that maybe one should always, for all circuits, consider what happens when they are powered off but have an input signal.
  Reply With Quote
Old 28th November 2005, 12:55 PM   #9
diyAudio Member
 
jackinnj's Avatar
 
Join Date: Apr 2002
Location: Llanddewi Brefi, NJ
the input to some high end measurement gear (like the Boonton 1120) uses a 50 ohm reed relay and clamp diodes. of course the usual ccourse of action is for the relay to interrupt access to the first active device in an "over-voltage" situation -- i would assume that the circuit is "open" when powered down.

the Boonton 1120 is also employs a "belt and suspenders" protection scheme -- in addition to the relay and clamp diodes there are 1/16th ampere fuses on the input AND ferrite beads.

speaking of clamp diodes -- caveat DIY'r -- those built-in clamp diodes on your IC only clamp when the power is on.

any resistor in series with the source of the 2sk170 is going to add noise to the system.
  Reply With Quote
Old 28th November 2005, 04:51 PM   #10
mzzj is offline mzzj  Finland
diyAudio Member
 
Join Date: Jul 2005
Location: 65N 25E
Default Re: Circuits living dangerously when powered off?

Quote:
Originally posted by Christer


I haven't considered other types of circuits yet, but similar things are likely to happen also in other circuits. For instance, the inputs of an op amp are usually not allowed to exceed the rail voltages by more than one or a few Volts. What happens if the op amp gets an input signal exceeding that limit when powered off?

I have burnt my share of opamps when powering equipment on in wrong order. 4-20mA industrial current transmitter burned OP277's with ease (and short-circuited opamp resulted large current trough input to+rail and its connected 50ohms precision current shunt burned away)
  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
Dangerously wrong diagram at apexjr? xiphmont Chip Amps 24 25th July 2008 07:29 AM
Connecting Virtual Ground Circuits with True Ground circuits?? Minion Solid State 1 12th February 2007 01:28 AM
Research on tubes and circuits simple circuits Cazcotty Tubes / Valves 6 16th June 2006 12:50 PM
85 powersupply circuits 58 charge circuits gev Power Supplies 0 31st July 2005 11:27 AM
Living with (the) PMt lazyfly Multi-Way 1 31st March 2005 03:47 PM


New To Site? Need Help?

All times are GMT. The time now is 09:52 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