DIY Audio Projects Forum

Calling All Serious Tube Experts
Page 1 of 8

Author:  gofar99 [ 20 Mar 2017, 21:48 ]
Post subject:  Calling All Serious Tube Experts

Hi Folks, Every once in a while something comes up that defies correction. Even with 50+ years of tube experience I get stumped on occasion and I'm not too proud to ask for other opinions. Anyhow it involves red plating on power tubes. It seems that my UK connection is having trouble with it. He is a licensee of some of the Oddwatt designs including the KT120 power amps. He has four assembled and all behave the same. The power tubes red plate at anything over about 50% dissipation. After having trouble with KT120 tubes before it was suspected that the tubes were at fault as the design has been successfully built many times in the diy channels and there are a large number of commercial ones out there as well. None have reported this problem. He returned more tubes to the supplier than most folks will buy in a lifetime. Some are a little better than others (humm), but none function correctly. I have a difficult time thinking all are bad or seconds. BTW I run them at about 85% of dissipation in the posted designs and have a pair of daily use ones that I'm testing for longevity at about 5% higher dissipation. Well past 2000 hours and no red plating. They play and test perfectly.

We have done every imaginable diagnostic you can think of and all indicate proper values. So I'm looking for something either so subtle or so odd that you would not normally look for it. The amps are not oscillating, voltages are stable, resistance measurements are correct and wave forms are what is expected.

Any ideas? Serious folks can PM me for more information as this has a lengthy history. Thanks.

Good listening

Author:  Geek [ 21 Mar 2017, 00:33 ]
Post subject:  Re: Calling All Serious Tube Experts

My PM is off here, so I'll toss out bad sockets or cold solder joint... got me on one of my own builds, TWICE! >_<

Author:  ILoveHiFi [ 21 Mar 2017, 01:41 ]
Post subject:  Re: Calling All Serious Tube Experts

Not a tube pro my self, but heres some advice.

Try decreacing tube current, and increace the ammount of tubes you are using for ouput.
If its a push pull config measure and set both output tubes to have the same operating point (both tubes are driven equally hard)

If you built the same amp and ran it without any plorbems.
Then it must be some kind of circuit errror. I'd sugest giving patience checking the circuit .
Also measure the current input to all tubes and checking the operating point.

Author:  Suncalc [ 21 Mar 2017, 06:29 ]
Post subject:  Re: Calling All Serious Tube Experts


I know this may sound stupid, but have your guy disconnect the output transformers and make sure that the screen tap and B+ line are not reversed in the transformer. This can be easily missed when checking voltages with no signal. The reversal may even sound ok at low to moderate volumes. However, as the signal increases the screen gets run at higher AC voltage than the plate which can really drive up plate current when the design says it shouldn't.

I apologize if you already checked this, but I'm thinking about faults that might be missed even by an experienced builder.

Author:  gofar99 [ 21 Mar 2017, 14:45 ]
Post subject:  Re: Calling All Serious Tube Experts

Hi, Everyone, so far we have tried those things. I wish the amps were here so I could put them on my bench....unfortunately there is about 2000 miles of ground followed by a like amount of ocean between us.

Good listening

Author:  blackdog [ 21 Mar 2017, 18:33 ]
Post subject:  Re: Calling All Serious Tube Experts

Things that come to mind.
Bad coupling caps.
Fogot to install the grid leak resistors on the kt120 grids

Is he using the lm317 bias? Are the regs working properly.
You could easily eliminate the kt120's by temporarily install kt88's and see how far you can push them.

Author:  Gio [ 21 Mar 2017, 20:24 ]
Post subject:  Re: Calling All Serious Tube Experts

Mainly guesses, but maybe one will be lucky.

Given that there are 4 amps behaving the same, the problem may be outside of the amps. Examine the power source. Is this behavior only at the dealer shop? Try isolation transformer. Is the power transformer ok with 50Hz (If that is what they use overseas?). Is it a capacitive issue with the interconnect/speaker wires or a strange speaker load. Is it the same behavior with different speakers and wires. Also check the signal source and try a different one.

On the circuit, I would start by looking at the regulators also mentioned by blackdog above. Are they authentic HV? LM317 being passed off for LM317HV? A bad chip batch, could cause a problem common with the four amps.

Good luck.

Author:  Juancho [ 22 Mar 2017, 04:06 ]
Post subject:  Re: Calling All Serious Tube Experts

Thanks for the tips so far Guys. The regulators came from different suppliers in two different countries, and cathode v is very stable, so I think that rules them out. We've used different sources-vinyl, digital-and different speakers all with the same results. The power supply measures perfectly, and even with 50hz, these are among the quietest amps I know: they are virtually noiseless with my ear jammed up against my 95db sensitive 15ins bass units! 2 of the amps are In Belgium and two in the UK, so the mains supply is different.

There are two similarities between the pairs of monoblocks: we opted for a longer, thinner chassis compared to Bruce's more square one and the four transformer sets came from the same Edcor production batch.

Have also tried a known set of KT88s with the same result: they work fine in another amp dissipating 36 watts in Class A but redplate at the recommended bias for these amps.

Author:  Geek [ 22 Mar 2017, 04:58 ]
Post subject:  Re: Calling All Serious Tube Experts

Hrm, perhaps....

I had a customer who kept blowing tubes. The amps worked fine, other than at his house! His house had built-in audio distribution from manufacture. He replaced the speaker cable as a last ditch effort and voila!

Turned out there was a painting hung in the wrong spot - nail through the in-wall cable.

A second customer had a similar problem, but the cable showed OK. Turned out to be a HUGE roll of leftover jammed in the drywall creating enough inductance to put otherwise perfect amps into oscillation.

Just a thought.

Author:  Peter W. [ 22 Mar 2017, 06:56 ]
Post subject:  Re: Calling All Serious Tube Experts

Might as well jump in with both size 13s and no life jacket. In no particular order, and based on my direct experiences:

a) If all four amps were assembled in sequence, there is every possibility that the same 'mistake' was made on all four units. There is no shame in this as if the instructions are not absolutely clear, the first mistake is made. And the amp passes the most basic preliminary tests - NEXT! It may be 5,000 miles between *you* and him - but it is likely not even 200 km between him and another decent tube guy. What I am suggesting is that a 3rd party go through the amps and the schematics.One who has never seen them before. This person will have no preconceived notions and no expectations. Writing for myself, a neighbor had put together a tube guitar amp - which was behaving very badly, even though all voltages were present and correct, the bias set properly and the amp was dead-silent with no input. He fought with it for almost two weeks. Then he called me over.... I looked at it for less than 5 minutes and asked him about a specific pin-bridge that was not in the schematic - though others were. That was it. One clip with a pair of mini-cutters and he was in business. Like that.

b) Wire Dressing: Are the wires to be shielded properly done? *Twisted Pairs* are typically to be connected only at one end. Make sure that the are not connected at both ends. And are any shields missing? Are adjacencies correct? This may take a bit of experimentation - and in that case do only one (1) amp at a time until it is right (if this happens to work) before applying to any of the others.

c) Pinched wires - you would be shocked and awed at how often this happens. And if a chassis layout is particularly open to such a problem, no reason at all it could not happen four times. Really. The Dynaco ST120 had a real problem that way as one common example.

d) I am working on a Scott LK150 at this moment that pins the bias voltage meter on the left channel - even though the reading at the test point is correct, and the amp plays perfectly well. I am working backwards from the meter - but here is a situation where appearances are not what they suggest. I expect to find out a mis-connected wire, or a wire that looks good, but it is the insulation holding it to the pin - the broken conductor being hidden. Anything wired with solid wire will often have this issue. So check *EVERY* connection. Use a dental pick or similar to pull on the connection if deep in the chassis.

e) Look for common parts to _all_ the amps, and that are in the relevant circuits. Even if they came from separate suppliers, even in different countries, if they all came from an ultimately common source as already noted they could all be equally defective.

Repeat: In no particular order!

I have my "weirdness" stash. One is a GE 5AR4 that tests perfectly well on a very high-end tube tester (Hickok 539B), and will play perfectly well in an amp - for about 15 minutes. Whereupon it goes cold. No heater at all, cold. I have reheated the pins and all that. But, there it is. Give it about 2 hours to cool down - repeat. Does not matter in which-of-three amps.

I have a 20uF@500 V, 105 C (Panasonic) cap that operates at about 400 V in-circuit. It will run for hours in a vintage radio at low volume (6F6s in PP output). If I boost the output it goes open - and that can be ugly! Lower the output, it comes back. Don't ask me how long it took to figure that out.

So, don't be surprised if you find something strange.

Best of luck!

Page 1 of 8 All times are UTC - 6 hours [ DST ]
Powered by phpBB® Forum Software © phpBB Group