ST-120 LM334 burnout

Hi all,

I noticed yesterday that I've got low output on one channel of my ST-120.  This has happened twice before — not always in the same channel — and both times replacing the LM334 regulator did the trick.  I've ordered a new part, and I'm sure that's the problem this time.  It's an easy enough fix, so I don't mind the occasional replacement.  But I do wonder if I should be looking for the underlying cause.  The unit does got hot, and it may be that the problem is triggered by long sessions where the amp is on all day and gets hotter than usual.  But it sits on a top shelf in free air, so maybe that's not the issue.  Any ideas?  

— Gideon

Hello Gideon,

I relate to your problem, and admire your dedication to maintenance - well worth it! Here are some thoughts, from easy to hard.


1) Hot air rises, since it is less dense than cooler air. As proof, some people use ceiling fans to distribute hot air near the ceiling.

You can locate a small quiet fan near the amp to help reduce this heating effect, assuming you cannot move the amp.

2) Install the LM334 under the PCB to shield the device. Use taller feet to raise the chassis and thus subtlety improve air flow.

3) Install a small TO-92 heat-sink on the LM334 using thermal grease. (Both available at Mouser and/or Digikey). This will help the LM334 dissipate more heat.

HTH, David

Thanks, David. Excellent ideas. Will try them all.
Gideon

Gideon, you are welcome.

David

grosen wrote:Hi all,

I noticed yesterday that I've got low output on one channel of my ST-120.  This has happened twice before — not always in the same channel — and both times replacing the LM334 regulator did the trick.  I've ordered a new part, and I'm sure that's the problem this time.  It's an easy enough fix, so I don't mind the occasional replacement.  But I do wonder if I should be looking for the underlying cause.  The unit does got hot, and it may be that the problem is triggered by long sessions where the amp is on all day and gets hotter than usual.  But it sits on a top shelf in free air, so maybe that's not the issue.  Any ideas?  

— Gideon

My question is: Why use the LM334 and not develop a long tail phase inverter or similar with resistors and a capacitor? The price to pay when the LM334 fails or burns out can be high. LM334 is delicate and senslible to temperature. In an old school amp, it shoudn't be there! Wrong choice from developers!

DavidT1 wrote:Hello Gideon,

I relate to your problem, and admire your dedication to maintenance - well worth it! Here are some thoughts, from easy to hard.


1) Hot air rises, since it is less dense than cooler air. As proof, some people use ceiling fans to distribute hot air near the ceiling.

You can locate a small quiet fan near the amp to help reduce this heating effect, assuming you cannot move the amp.

2) Install the LM334 under the PCB to shield the device. Use taller feet to raise the chassis and thus subtlety improve air flow.

3) Install a small TO-92 heat-sink on the LM334 using thermal grease. (Both available at Mouser and/or Digikey). This will help the LM334 dissipate more heat.

HTH, David

Hi Guys. Sorry, but it seems a lot of paletive solutions for a bad phase inverter project for a old school amp.

If the problem were common we'd hear more about it here.  So it must be rare.  But for those who have it the easy permanent fix is Pavel's tiny board with a surface mount LM334 and the recommended protection circuit -- available at audioamp.eu.   Cheap and straightforward.  (In my case it also stabilized channel balance, which can wander a bit if your LM334s are getting too hot.)  -- Gideon

For my upcoming MK-3 Project I seem to always want to fix something that isn't necessarily broken. I decided to look into the LM334 issue a little closer. Using LTSpice I modeled both the Uncompensated and Compensated LM334 circuits
The graph om the left is the Temp vs Ma of the Uncompensated circuit and the graph on the right is the graph of the Temp vs Ma of the Compensated circuit.
The Uncompensated circuit goes from 7.1ma to 9.4ma for a delta change of 2.3ma.
The Compensated circuit goes from7.4ma to 7.6ma for a delta change of 0.26ma.
Per the T.I. data sheet the max operating current is 10ma. Now these numbers represent the extremes but I can see where with different tubes and tolerances the max current my get exceeded with the Uncompensated circuit. It seems to me that the Compensated circuit may be able to withstand a wider range of tolerance.
Just food for thought. Since the circuit from Audioamp was out of stock, I decided to layout a version using thru-hole components.