Forums

DIY hi-fi

Started by Tom1957 3 years ago18 replieslatest reply 3 years ago138 views

Hello,

Recently I built a tube pre-amplifier according to the Elector magazine from the late 1980's. It took me a while to collect all the necessary parts according to the specs and finally the unit is finished. After several initial difficulties, I must admit I am really glad I decided to do this as the unit sounds really above all my expectations. However, I still have several loose ends and questions. First off, the power transformer. I live in Europe, so it means I am strictly speaking about 230V operating voltage. The specs for the transformer secondaries are as follows: 2 x 360V / 50mA, 6,3 V / 1A, 13V /1.5A, 14V / 1A. I ordered a toroid transformer and what I received actually exceeded the expected values by 10% when ran in idle mode in the two 360V AC secondaries. The pre-amp power supply unit should deliver three different DC voltages: 12V DC, 350V DC and 310V DC. The company that made the transformer argued that it was normal and that everything would work well once everything is hooked up. So this is my primary concern. In my opinion, whoever designed the unit must have thought about it in the first place. Of all three DC voltages measured while the unit is fully operational only the 12V DC is OK, but that is probably because I can adjust it with a trimmer pot. Where I should get 350V DC I am getting 344V DC, and instead of 310V DC I am getting 288V DC. There is a loud buzz coming from the transformer and it is overheating considerably. Any help would be greatly appreciated.

[ - ]
Reply by JackCrensMarch 22, 2019

I'm a little surprised that the design requires so many different voltages. I've never heard of a design needing two different HV supplies. I'm also confused that there's a trimpot in the 12V supply. Sure that's not a "hum balance" pot?  The 12V is to supply heater voltage, so there should be no resistance at all between the transformer and the heaters.

About the buzz and the overheating, it would be easy to assume that the transformer itself is not up to the job. The loud buzz certainly sounds like the trannie is working too hard.

But I'm suspecting there's something wrong with the circuit. A preamp should draw miniscule current.  The fact that the nominally 310V supply is producing less than 290 makes me suspect that the circuit is drawing too much current. That's most likely a tube bias problem.

Does the schematic show what the voltages should be at the tube pins (or anywhere else)?

If so, you should be able to find out which tube is drawing too much current. If all else fails, just pull the tube and see if the trannie stops buzzing.

I'm guessing that you don't know much about tube electronics? Or troubleshooting?  If that's the case, welcome to the world of testing analog electronics. If you just built the preamp from a schematic, and plugged it in without testing the subsystems, that was a mistake.

If you were writing software, please tell me you wouldn't write the whole thing before testing the various functions.  That's called the Big Bang theory of testing, and it almost never works.

Ditto with electronics.  The right way to build a system is to build the first stage, test it, then build the second, etc.  

Fortunately, there are many old-timers here like me who've dabbled in tube electronics. We could probably help.


Do you have a multimeter?  A signal source (audio oscillator?).  An oscilloscope?  If not, things get harder.

[ - ]
Reply by Tom1957March 22, 2019
OK, I will try to answer the best way I can. The transformer was buzzing also in the idle mode when I first got it and hooked it up to 230V. The reason for so many different voltages is as follows: 14V is for a separate PCB that controls input micro-relays and a 12V relay on the power supply PCB which connects two high voltage lanes for the power supply tube (EZ80 or EZ81) after a 36 second delay. 13V is the tube heater voltage and the trimmer pot is there because according to the specs the output DC voltage should not exceed 12.6 V DC (I set it up to 12.0V). 6.3V is for the power supply tube. The reason for the two high voltages is because the pre-amplifier PCB has two separate stages. One is for the phono pre-amp and the other one is for the line pre-amp.

Unfortunately, the schematic does not show voltages at the tube pins. If you wish, I can upload all the schematics for the unit and post a link.

You are correct, I am no electronics whiz, far from it. I was given the magazine with the schematics, the parts lists and the instructions on how to build the PCB's. I've done things like this before and had no trouble, everything always worked. That's why I finally decided to do this.


I have the multimeter (three different ones actually), but I don't have an audio oscillator and I don't have an oscilloscope.


What really confuses me is that the unit worked for two weeks without any problems apart from the buzz from the transformer. There was no overheating at all. On the contrary. My power amp was way warmer (as usual) after hours of playing music. It all happened yesterday when I fired up the pre-amp. Immediately I noticed the heating and the smell from the transformer. I took the transformer out today and it immediately got hot even in idle mode. I measured voltages, and determined that both high voltage outputs were dead.

A big thanks to all who have replied.

[ - ]
Reply by JackCrensMarch 22, 2019

Ok, this changes EVERYTHING. If I understand you correctly, the transformer was buzzing even with no load connected to it. Then the circuit worked for a couple of weeks, but now the trannie is dead.  Right?

Sounds like the trannie is indeed dead, and was sick on arrival.  I'm afraid maybe the decision to go for a toroidal transformer wasn't the best idea. Methinks you might have paid a lot of money for a less than premium part.  Sorry.

[ - ]
Reply by Tom1957March 26, 2019

I didn't pay much actually, although paying anything for something that doesn't work is always too much. Around $30.00 in local currency. I went for a toroidal transformer because of space mostly. I  realized that today most hi-end hi-fi equipment companies used them, so I thought it would be the best choice. I did ask for a quote for an EI version and not only would it cost a lot more (if it works, I wouldn't care though) it would not fit the enclosure. However, now I am also considering the idea of actually ordering an EI transformer and put it into a separate box providing the replacement toroidal transformer I already got fails again.

[ - ]
Reply by mr_banditMarch 22, 2019

Find a hacker/makerspace near you. The best list is hackerspaces.org. Many hackerspaces in the US host meetings of Old Fart EE's That Love Old Radios - I assume there are such groups in your neck of the woods - and otherwise have EE's as members. Plus they have oscopes and other test equipment.

(You should also think about joining one, given what you like to do for "fun". Kindred spirits and fellow travelers, as it were.)

[ - ]
Reply by Tim WescottMarch 22, 2019

If it was working OK at first, and then stopped, then something broke.  It sounds like one or more turns shorted in the transformer, which certainly should not have happened if you were treating it right.

If you can safely measure the input current at idle, do so -- I think you'll find that it's way too high.

A careful visual inspection may be a good idea -- look for wires that have their insulation rubbed off, or for secondaries (or secondary taps) that you're not using that have gotten shorted.

[ - ]
Reply by Tom1957March 26, 2019

The transformer came completely sealed with a non transparent tape, so it's impossible to see the wires below. However, I am pretty sure I can locate the problematic spot as there's a difference in color there and it does look like something burned there.

[ - ]
Reply by stephanebMarch 22, 2019

This thread was started by OP on Electronics-Related, as it should have.

Despite the topic not being 'embedded', I took the liberty to cross-post it to the EmbeddedRelated forum where the community is much bigger and full of very competent EEs who might be able to help.  

[ - ]
Reply by Tom1957March 22, 2019

Thank you kindly.

[ - ]
Reply by CustomSargeMarch 22, 2019

I'd start with an AC ammeter to verify I'm not exceeding ratings. I think you are and by a fair margin - ergo the heat. I suspect the loud buzzing is a lack of coil epoxy allowing coils to move at AC frequency. I wouldn't chase any other issues until power is both stable and well within max ratings. Good Hunting... <<<)))

[ - ]
Reply by Tim WescottMarch 22, 2019

If you end up replacing the transformer, note that you don't need to get one transformer that does it all -- you can get separate transformers for each output voltage, or get them in whatever combinations you need.

Is a copy of the article on line?  There may alternatives to their schematic that don't require so many oddball voltages.

[ - ]
Reply by Tom1957March 22, 2019

You can find everything here:

http://www.jogis-roehrenbude.de/Verstaerker/SRPP-V...

Unfortunately, it's in German. But you have the circuit diagrams and the layouts and also the parts lists. However, I do have a pdf copy of the English release (spread out across two issues) and can send it or post it if you wish.

However, reading all the posts and checking and double checking everything for who knows how many times I may have stumbled upon the possible culprit. Again, I am not an electronics whiz, but I hopefully do have some sense of logic left in me, and my eyes do still serve me quite well. I have checked all the parts again this afternoon, all the resistors, capacitors, and then I checked the tubes. And here is what I have found. Luckily for me this forum supports images so I guess it would be best if I just publish them (two) and see if this might be the problem. So the pre-amp phono stage uses two tubes per channel, ECC83 (12AX7) and ECC81 (12AT7). This is the stage that measures lower voltage that it should - 288 against 310. Upon inspecting the tubes that are supposed to be ECC81 (12AT7) this is what I have found. Nowhere on the tubes does it say they are actually ECC81. Matter of fact all it says is this -  SIEMENS K 1007 CV 491 887. When I googled this out all the results I got were that this is actually ECC82. However, the inscriptions on the boxes they came in say they are indeed ECC81. But how can I be definitely sure? Everything I read on the internet about possible substitution definitely points towards the bias problem. I ordered these tubes from Buerklin Germany years ago and never gave it a thought. But maybe, just maybe, someone there placed the wrong tubes in the wrong boxes. Please check the attached images and tell me how can I be sure?20190322_214414_83638.jpg20190322_214555_30719.jpg

[ - ]
Reply by Tim WescottMarch 22, 2019

You said the transformer was showing problems when "idle"?  I took that to mean out of circuit -- is that not the case?

That if it is the ECC82 the behavior shouldn't be too much different; you could check the voltages at the plate of the bottom triode in each pair; if it's riding more or less in the middle (160V, +/- 30V) then you're probably OK.

You could measure the plate current directly by measuring the voltage on the 1K-ohm bias resistors -- 1V = 1mA, and there you are.

[ - ]
Reply by Tom1957March 23, 2019

Yes, exactly, out of circuit. It was buzzing more than I expected from a custom made toroid transformer.

As for measuring, I'll have to wait till the new transformer arrives as the old one's both HV secondaries got fried.


Update: I got hold of a tube that is certainly a Siemens ECC81. Upon visual inspection it is a bit different that the ones I have. I also measured the voltages as I got hold of a new transformer (through bribe I am ashamed to admit). The voltages with the alleged ECC81's  are 148V (left channel) and 141V (right channel). The voltage with the borrowed ECC81 is 151V.

[ - ]
Reply by Tim WescottMarch 24, 2019

Visual differences between tubes of the same type, even the same type and brand, is to be expected.

Did you measure the voltage across the 1k-ohm cathode to plate resistors (and, hence, the plate voltage) in each side?

[ - ]
Reply by JackCrensMarch 26, 2019

Ok, Tom, the tube type is extremely important. See the tube info here:

http://www.r-type.org/addtext/add070.htm

The three tubes are all dual triodes, and have the same pinouts, but their electronic characteristics are vastly different.  In particular, the ECC82 is a low-gain tube, with a mu of 17.  It's equivalent to a 12AU7.  The ECC83 is a high-gain tube, with a mu of 100. 

If you put an ECC83 in a socket meant for an ECC82, it's not going to have a plate load resistor maybe 5 times too low. Depending on the bias, it's going to be drawing _WAY_ too much current. That would explain the lower-than-expected voltages.

For the record, the ECC81 is equivalent to a 12AT7, which you don't normally see in audio gear.

The fact that one of your tubes can't be determined to be any of the three just complicates things even further.

One last thought:  The schematic only strengthens my curiosity as to why the design uses two different high voltages. I'm sure that last stage would run perfectly fine on the 350V supply. 

My guess is that the designer worked very hard to optimize the total distortion, which is impressively low.  Perhaps that explains the dual voltage supply.

Or not ;-)


[ - ]
Reply by Tom1957March 26, 2019

Hello Jack,

Thank you for your reply. The ECC83's are in their appropriate sockets for both channels, I made sure to that. My problem is that I am not sure if the ones I got from the vendor that are supposed to be ECC81's are actually ECC82's. I have no way to determine that as the inscription on the tubes does not tell that. All I got is what you can see in the images I posted. In any case, I have already ordered a matched pair of ECC81's from another vendor just to avoid any possible problems.

[ - ]
Reply by Tom1957April 2, 2019

So... the epilogue. A matched pair of brand new ECC81's arrived. I took out what seemed to be a pair of ECC82's I was sold by mistake as ECC81's. The brand new transformer, which is actually double the size of the one that fried, is totally silent and doesn't heat up at all. It also came with an aluminum foil protection (a completely different vendor obviously takes his job seriously) and all the voltages measured all over the unit are right on the target. The low 50Hz hum heard regardless the gain control also disappeared, so I assume one of the tubes I took out was the culprit for that. Or maybe both. Nothing overheats even after hours of listening t the music. Actually the only places I sense some heat (warmth actually) is directly above the tubes. A big thank you to all who contributed and helped me out with this.