diyAudio

diyAudio (http://www.diyaudio.com/forums/)
-   Chip Amps (http://www.diyaudio.com/forums/chip-amps/)
-   -   Alittle disaster story (http://www.diyaudio.com/forums/chip-amps/228620-alittle-disaster-story.html)

Sjef 25th January 2013 04:30 PM

Alittle disaster story
 
Today I bought myself a present, I thought.

I bought a second hand Yaqin SD-CD3 tube buffer with Sylvania 6SN7 tubes in it. It was affordable and I always wanted to try a tube buffer in front of my Gainclone amp. I'm using the gainclone amp in an active setup only for the midrange and was hoping for a bit more robust midrange with the tube buffer in between.

Anyways, long story short. I tested the buffer, seemed to work fine. Inserted it in between my DEQX and the gainclone. Loud hum and some music coming out of it. What was the case. The interconcts I'm using in my main system has WBT plugs on it and these do not fit on the cheap chinese plug on the Yaqin (same for the power inlet, way too wide, cable falls out when you just look at it)

Anyways. I tried to fix the humming by making a connection between the mass of the WBT plug and the mass of the Yaqin. This helped but I could not get a solid connection.

Then all of a suddon, when I was fiddling around, BOEM. The left channel of my gainclone exploded. The chip blew apart. And there was that funny burned electronics smell.

After analyzing what the hell went wrong I figured it out.
It's the bloody Yaqin. Why ? Well simple. The Yaqin in and output are AC coupled with coupling caps. But the Dman thing does not have any bleeding resistor on the input nor the output.

This means that when you have a short disconnection due to a bad connection the input and/or output cap charges up to about 28V DC !!!!
When the connection is reastablished there is a sudden discharge of 28V to zero via guess what, the input resistor of the poweramp, in this case my gainclone. I can tell you that a lm3875 does not like a sudden high DC on it's input, it instantly explodes.

Conclusion: 1) Be always aware of all these chinese crap. They may look nice but in reality nothing fits like it should. RCA, mains connection are all crap, and I mean real crap. Oh, did I mention they do look like it has quality, it has not, period. Third time I have experienced this myself with a chines hifi product. I will never ever again buy some chines ****.

2) Always use bleeder resistors on your input and output when working with high voltage tubes. Something like 100kOhm to 500kOhm after the output cap and before the input cap will suffice. This way the cap never has to discharge via your amp or source.

Finally, I ordered some new lm3875 straught away and in the men time I thought I just put a Trends TA10.1 I had in the closet in my system. Wow, this sound really bad, I mean really really bad compared to the gainclone. Well, could it be the Trends is just another chinese crap product. (had a lot of problems with it as well, plugs that just won't fit, bad soldering, looks nice, is crap crap crap crap crap crap,.

In other words:

DO NOT BUY ANY CHINESE HIFI CRAP FROM EBAY> IT'S ALL CRAP

balerit 25th January 2013 04:47 PM

I have to agree with you. I have just purchased a usb WASP 2018
PICKit2/PIC emulator programmer
and there is no user manual and no way of finding out which way to insert the blank pic in the ZIFF socket and I only get errors when I plug it into my laptop.
So I hope your nerves have returned to normal.

Sjef 25th January 2013 05:05 PM

Just soldered in some bledding resistors of 681k, wich was what I had lying around. DC on the outpu in steady state (without any load) went down from around 28V to 1.5mV. Now there is no danger anymore from a cap discharging into the input of your amp.

This thing is a time bomb in it's original state. I can hartly reckon to everyone who owns one (or any other chines tube crap) to solder in these bleeders. Without it it's just a matter of waiting when the crappy rca connectors fail (they will in time) and you will have fireworks all over. Even if your amp is AC coupled it could be dangerous. This kind of rapid discharge is lethal to any electronic circuit.

I really wonder why they where not in there. Is it really the 5 cents cost saving ?

CharlieLaub 25th January 2013 05:08 PM

So I assume the LM1875 amp (that you built?) did not have an input capacitor to block low frequencies? That might have helped, depending on the charge that was delivered from the buffer...

Like C5 in the pic below:

http://audio.claub.net/other_stuff/L...%20labeled.jpg

Better would be to use a set of diodes from the input to each PS rail as input protection.

Like this:

http://www.edn.com/ContentEETimes/Im...0OPAs/Fig3.jpg

-Charlie

cyclecamper 25th January 2013 06:05 PM

Don't buy any used homemade Gainclones from Amsterdam. No input isolation or protection! ..bloody...Dman thing...crap...all crap...I mean real crap...****...crap product...crap crap crap crap crap crap...

LOL just kidding...

Sjef 25th January 2013 07:04 PM

Well I would not buy any home made gainlones from Amsterdam for a completely other reason.

I have build this thing a little bit too solid. It's going to be a pretty hard job cahnging the chip, I have to desolder about everything in there. I always build very solid, in fact the very first amp I have build is playing for almost 20 years / 24 hours a day and is still going strong.

Up until now I had three pieces of Chinese hifi equipment. A Trends TA10.1, a Muse NOS 4*TDA1541 DAC and now this Yaqin CD3 buffer.
It's really funny to see how they d it. Everyone who sees these pieces think they lokk pretty good and solidly build. Upon closer look however, nothing fits. Specialyy these RCA plugs are really the very worst that are ever produced. I ones had one that was not even made of metaql. It was plastic with a nanofilm of gold (or something that looks like gold) over it. After unplugging it twice ther was only plastic left and no connection.

Soldering is very sloppy. Both the Trends and the Muze broke down wthing a couple of weeks due to bad soldering. Again, lokks fine at first. Upon clos inspection, guess what, crap allover again.

I can really bang my head I fooled myself again by buying this Yaqin.

I'm going to give it a try anyway. Got some spare RCA's and power inlet left of at least fitting quality. Gonna check every solder point etc etc. and will see if it's going to sound at least a bit decent.

Anyways, input protection is not needed if the preamp/cd player/ tube buffer or whatever stage in front of it has a proper cap charging resistor in the right place. Every circuit designer knows that and everyone does it for this very reason. Yaqin decided not to. Utterly stupid and dangerous.

Once again. In any AC coupled circuit ALWAYS put a resistor from the output side of the cap to ground. Even if it's a 1Mohm resistor that's fine.

Sjef 25th January 2013 07:48 PM

Just soldered in some bledding resistors of 681k, wich was what I had lying around. DC on the outpu in steady state (without any load) went down from around 28V to 1.5mV. Now there is no danger anymore from a cap discharging into the input of your amp.

This thing is a time bomb in it's original state. I can hartly reckon to everyone who owns one (or any other chines tube crap) to solder in these bleeders. Without it it's just a matter of waiting when the crappy rca connectors fail (they will in time) and you will have fireworks all over. Even if your amp is AC coupled it could be dangerous. This kind of rapid discharge is lethal to any electronic circuit.

I really wonder why they where not in there. Is it really the 5 cents cost saving ?

CharlieLaub 25th January 2013 07:50 PM

Quote:

Originally Posted by Sjef (Post 3342068)
Anyways, input protection is not needed if the preamp/cd player/ tube buffer or whatever stage in front of it has a proper cap charging resistor in the right place. Every circuit designer knows that and everyone does it for this very reason. Yaqin decided not to. Utterly stupid and dangerous.

Once again. In any AC coupled circuit ALWAYS put a resistor from the output side of the cap to ground. Even if it's a 1Mohm resistor that's fine.

I would say that you have just discovered THE reason why you should use input overvoltage protection! It's like a seatbelt in a car - you almost never need it, but when you do it may save you from smashing your teeth out on the dashboard (or worse).

OK, sure, IF so-and-so manufacturer had been more careful, blah, blah. There is certainly some blame that can be assigned there. On the other hand, it wouldn't matter IF you had just added $0.50 worth of input protection diodes to your amp... so it seems like about a 50/50 split as to who did not include certain components in their equipment design that could have prevented the damage from happening.

As a side note, I have also only had bad luck with around 50% of Chinese electronic parts and equipment that I have purchased from Ebay (some assembled chip amps for instance) so now I am more careful and trust less the fancy shiny case!!! :)

Anyway, I hope that you did not also damage speakers... if any were connected to the amp when it exploded who knows what was sent down the speaker wire.

-Charlie

Sjef 25th January 2013 08:10 PM

I hace been building amp/ preamp/ dacs for 20 years now and have never used dc protection on the input of power amps. Every single one of them is still working today and have never had any problem. DC protection is not needed if the preamp has a coupling cap AND A BLOODY CHARGING/BLEEDING RESISTOR. That's the place where it belongs. Even with dc protection on the input of the poweramp this could lead to serious problems.

Soldering diodes at the input of an amp is adding a highly non lineair device in a sensitive place. not my cup of tea. Still they only protect at voltages where the output DC has gone high enough to destroy your loudspeakers. It will protect your amp though. With a resistor in the right place, at the ouput of the preamp these things just don't happen.

I drive my tweeters with a class A DC coupled poweramp without any protection caps before the tweeter and no dc protection at the input. This is working that way for the last 8 years or so, never any problem. I fully trust my own builds, much more so than any factory equipment.

Just tried the Yaqin on my laptop/ desktop system. At least it works. noise is pretyy high though. Seams to dissapear when I squeeze the interconect by hand on the RCA. Again a sign of very bad RCA connectors. Apart from that, thers is no difference in sound with the Yaqin in place or not. But this is not excaclty what I would call a highres system.

CharlieLaub 25th January 2013 08:49 PM

Your argument is like saying:
"I've been driving around for 20 years without my seatbelt on and NO PROBLEMS until this guy comes out of nowhere and hits me. Now it's his fault that I broke my face!"

Use your protection.

I'm not at all saying that bleeder resistors shouldn't be used, just that sooner or later you connect a faulty component to your system, or an internal component fails and you get the rail voltage sent down the line, and you get this problem. Sometimes the amp will last long enough to destroy your drivers, too, so count yourself as lucky!


All times are GMT. The time now is 11:19 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