zener diodes: can zener voltage drift over time?

Assuming that it is correct to say that zener diodes used for voltage regulation are constantly reversed biased and therefore constantly conducting at least some current and generating some amount of heat. Always on. Hours add up when considering a 30-40 year old amp.

Is it possible for the zener voltage to drift? If so, how common is it to see this with zeners in vintage amps?


.
 
My exp. with zeners is they either fail or go noisy (they rarely fail though). They should be replaced in vintage gear IMHO.
 
Last edited:
I've seen hundreds of power supply regulator failures over the years. 99% of the time, plastic packaged power transistors are to blame, and that's normally because their cooling arrangements are deficient. Zeners don't seem to drift or fail much unless they're also operating near thermal limits, which is unusual except in circuits where no power transistor is used, i.e. the simplest kind of shunt voltage regulator.
 
FWIW those zeners are used as shunt regulators - which means they dissipate the most power when no load is present, and if there is too much load (like a bad chip) the resistors could limit the voltage to below the zener voltage.
 
Thank you sregor

Actually, by coincidence, I think I learned some of that myself last night. Have been reading in my electronics text book about simple zener regulator circuits because I was trying to calculate the power that would be dissipated by these zeners. I came up with no-load current=.02A and no-load power=.28w [assuming that the source (main rectifier) voltage =34v and zener voltage =14v]

There is an example in the book showing what you describe. That when there is a load, the current through the dropping resistor stays the same but current through the zener decreases by the amount of current through the load.

Since I usually do not use tone controls and never use the phono preamp, which are the only things fed by this regulator, the load on these zeners will most always be at max.


If my .28w figure is correct (is it?) then with no load the zeners will be running at 56% of their 500mW rating. This is interesting to a newb like me...
 
Last edited:
Even though you aren't using them, the phono preamps and tone controls still are connected and will still draw current. My comment was more directed toward if something was awry, it might draw enough current (.02A per your calculation) that the zeners would no longer regulate voltage as no current would flow through them, resulting in lower voltages that might look like their voltage is drifting.
 
If my .28w figure is correct (is it?) then with no load the zeners will be running at 56% of their 500mW rating.

Correct: 0.02A * 14V = 0.28W dissipated in the zener diode.

56% of the power limit is reasonable, especially considering that dissipation is reduced by load current. If load current exceeds about 10~15mA, then the zener could start coming out of regulation. Be aware that diode heat dissipation is almost entirely through the leads. Heat has to flow readily into attached hardware for the diode to actually meet its dissipation spec.
 
Still, after 30~40 years active service, I wouldn't hesitate to replace them, specially when they are cheap and when you have to order other parts anyway. A suggested modern replacement could be 1N5244B from Fairchild (datasheet button on top of page), MRSP $0.0194. The size (DO-35) is probably ok for you; keep them off the PCB (1/2'') to enable some natural air draft around them for heat dissipation.

Edit, some additional thoughts... If these zeners are regulating power to input stages, notably the phono stages, then another thing to think about might be to select 'low-noise' types if you intend to replace them. Zener diodes can be(come) noisy, and in your amp they are at the very beginning of the power rails, that's the last place to start having noisy components. Granted, often there is further power rail noise filtering somewhere down in the tone control or phone pre-amp stages, but still... A number of high-end power supply circuits show often a small ceramic bypass cap across the zener to filter (lower) the zener noise (you could this between the elevated legs, or on the solder side of the PCB, or select a low-noise zener in the first place)

A quick google provided these low-noise zener diode series of Microsemi. 1N4108 would be 14 volt zener. At least this manufacturer provides noise numbers, often a missing property with other manufacturers.
 
Last edited:
@ sregor thanks for the clarification, I understand now
@ BinaryMike thanks for quantifying

@ Oilmaster
1) my final order has already been placed (my "current" final order :D) and it includes the Fairchild part you linked
2) searched Mouser and Digikey for the Microsemi low noise zener and they are not there
3) will get back to you to discuss the topics of bypassing zeners and also whether or not there is additional noise filtering in the phono preamp and tone boards of my project amp, there are some caps in there whose function I do not understand.

But first...



What are some thoughts as to the best way to mount zeners, or rectifiers, or any diode, to maximize heat dissipation:


.... Heat has to flow readily into attached hardware for the diode to actually meet its dissipation spec.

.... keep them off the PCB (1/2'') to enable some natural air draft around them for heat dissipation.....


Guys, which is the way to maximize heat dissipation? Short leads, which implies mounting close to the board and less air space? Or mounting higher off the board to give air space but which would require longer leads?

BinaryMike, when you say "attached hardware" do you mean the PCB tracks? Other than a solid solder joint, is there a way to maximize the mechanical connection?

.
 
Last edited:
oof, that's a good point... heat can be dissipated in various ways, such as through the PCB cupper tracks and/or through natural air draught. IMO opinion, SMD components use often the PCB cupper tracks, while through-hole components use often air draught, though I wouldn't exclude PCB tracks for zener diodes. Bon, brief, the first page of the referenced Microsemi diodes states 3/8'' up in the air.... I'm open to more views/opinions about this; it's a good subject.
 
think_smiley_56.gif


well... just drill a 1/4'' hole in the PCB and then solder it flat on the PCB, like that you have both !

think_smiley_55.gif
 
Well, if it comes to component mounting principles/habits by equipment manufacturers, then there is much more to it. First of all how much heat in each case needs to be dissipated vs. the selected component model. Inadequate heat dissipation lowers it's lifespan, which is a function of the actual component temperature (which I assume the equipement manufacture has tested/verified during prototyping phase). Then, if the equipment manufacturer considers for example 2000 hrs lifespan (of the entire unit) acceptable, and the poor component installation will (just) reach 2000 hrs, then he will not opt for better installation practices for certain components, notably when improved component installation practices are more expensive. And I think the costs price and installation of long lead components is more expensive. In the end, what you see in any piece of equipment, is that it all comes down to economics and lifespan choices, hardly ever to best technical practices in mainstream hifi equipement. that's where we folks step in.
 
I believe most diode manufacturers want the leads coupled to thermally massive objects near the body, but they also get their knickers in a knot if soldering heat is applied too close to the body. It's usually best to simply mount small power diodes in contact with the PCB, just like an axial-lead resistor, with no extra lead length. If you're designing a PCB, don't make the mistake of locating the current-limiting resistor so close to the zener that resistor heat adds to zener heat. I've seen this error several times in production gear. :nono:
 
Back
Top Bottom