kevin gilmore Posted June 23, 2009 Report Posted June 23, 2009 here is something that shoud autostart the gas tubes http://gilmore.chem.northwestern.edu/sdsgaspsautostart.jpg
kevin gilmore Posted June 23, 2009 Report Posted June 23, 2009 Any chance that Al has a bad tube in the amp section that is causing super high start up currents and causing the power supply to act up? I know in the Menace if you hot start it it'll go arcy-sparky in the rectifiers (overdraw). Completely different kind of thing. If the amp was drawing too much current then the gas tubes would never have enough voltage to light up ever.
Voltron Posted June 23, 2009 Author Report Posted June 23, 2009 This is mikhail, and anything is possible. But note the big power diodes on the side board. Wonder what they are for since the diode bridge is already mounted to the ES1heater board. I still think that 40ma to run 2 x WCF is pretty minimal, and in any case with the 0d3's is already over spec. Also the fact that the one channel is consistently popping tubes as soon as you put them in, points to much more current than normal. What i was thinking of for the auto-ignite is my stacked high voltage multiplier used to make the bias voltage for electrostatics. The doubled voltage would end up about 800 volts, and is perfect for starting 3 tubes in series. You add a large value resistor in series with a .1uf cap and a diode and it will definitely popstart the gas tubes. I wanted to respond to the bolded text because only one regulator tube popped on Sunday and that was the first and last time. The amp turned on and the first two regulators ignited and then I switched to four. The glow was much brighter in the front two than when the front four ignited, and the filament or whatever that you can see on the top of the rectifiers was much less bright on the rectifier in the middle of the amp (don't know which channel that is). The sound was not good upon startup, which is why I let it sit and turned on the speakers. While sitting there, I heard the pop and then turned everything off. If I recall correctly, I turned it all off and unplugged the headphones. I think I tried to turn it back on with everything the same, and the rectifiers lit, the amp lit, and then the regulators wouldn't do anything. There was no music playing either. If memory serves, if the switch was on but the tubes did not ignite, you could still hear music but I am not certain of that. Then I replaced the rectifiers with 5R4GYBs and replaced the front 4 regulators because the back ones had never been on. I tried starting up and I had K701s plugged in (more expendable) and there was no ignition and no sound. It did not pop the regulator at all, but there was a spark/arc in the rectifier so I shut it down. None of those tubes were popper, but the LED did light so power should have gone to them. One question I have is what are the four capacitors strapped to the top plate? They are in the last pic above. OK, all of that said, Kevin, are you willing to work on this thing? Do you want to build a new one in that box for me? If not, do you think you could readily direct someone how to do it, including the ignition circuit you added in? I appreciate your thoughts on this already, but I am thinking that the best solution is to rebuild it but I still don't know how to find the problem. I would love some advice, as I said above, regarding opening this thing up better. The tube sockets are wired to the top, in addition to the bundle of capacitors so you can only pull that up so far. The bottom has all the trafos and stuff connected to it, so it is unclear to me how to separate that.
kevin gilmore Posted June 23, 2009 Report Posted June 23, 2009 No way i want to work on that thing. If i would work on it, i would own it for the rest of all time. The fact that you can't even open it up shows how fucked up the thing really is. Its a wonder how he even got the thing together, unless the top unsolders with only a couple of wires. Are you sure that only tube sockets and no components are on the top panel? Better pictures any way possible would help me come up with some way to proceed. How many neutrik connectors are there on the back panel? At least with the ES1 units, i know what to do and how to proceed. Got to think about this mess a while. No idea yet what the 4 large caps bolted to the top do...
Voltron Posted June 23, 2009 Author Report Posted June 23, 2009 No way i want to work on that thing. If i would work on it, i would own it for the rest of all time. The fact that you can't even open it up shows how fucked up the thing really is. Its a wonder how he even got the thing together, unless the top unsolders with only a couple of wires. Are you sure that only tube sockets and no components are on the top panel? Better pictures any way possible would help me come up with some way to proceed. How many neutrik connectors are there on the back panel? At least with the ES1 units, i know what to do and how to proceed. Got to think about this mess a while. The top would only unsolder with all 8 tube sockets and those four bundled caps unsoldered. The only components on the top are those four caps that I can recall/see so far. I will work on it tonight, but who knows how to get this thing apart without fucking it up more. The back has two Neutrik connectors for the umbilicals. It would seem that one of them has a single blue wire connected to it. I don't recall how many wires go to the other neutrik. There are at least three wires twisted together coming out of the power transformer that have some heatshrink on the end, and there is one other wire with some crimped heatshrink on the end. Who the hell knows what is going on in there....
crappyjones123 Posted June 23, 2009 Report Posted June 23, 2009 well this is grade a doo doo. sorry about the whole thing.
kevin gilmore Posted June 23, 2009 Report Posted June 23, 2009 (edited) So a number of things are bothering me. Lets look at the main power transformers hammond 278x. Spec is 800 VCT @ 200 ma In a rectifier tube situation, that is 400*1.4 == 560 volts into the main capacitor less the drop of the 5u4, say 20 volts == 540 volts. also according to the 5u4 datasheet the maximum capacitance should be NO MORE than 40uf. Well i see a lot of caps 10 times that big. If you put 2 in series (where are the balancing resistors) that is still 5 times too big. Likely that is what the 4 caps bundled together on the top panel are. 540 volts for an amplifier that should never ever be run at more than 350 tops and a lot of extra power is going somewhere. Furthermore if the gas tubes really are the only regulators, then if you put more than .1uf on the gas tubes they turn into oscillators... Man this has fail all over it. How many people think i'm stupid enough to want to own the repairs to this thing for the rest of my life... Each 5u4 tube should have a minimum of 4 wires going to it, 8 wires right there between chassis and the top. The gas tubes are 4 wires, plus 6 wires to short out the tubes to the switch. Maybe if you remove the switch from the front panel first, there is enough length to pivot the top of the chassis out and over one of the sides. Otherwise how the hell did he put that piece of shit together. Spritzer has kane's es1 now, i'll let him show you the pictures he showed me... More fail... and more lies too, lots of expensive missing parts... What a train wreck. Trying to think of a way to get voltron out of this mess now... Hey voltron, what is your favorite B+ voltage?? 2 gas tubes == 150 volts?? This will influence what i do. On the other hand, i could just send him a pair of bench supplies, one for the high voltage, and another for the filaments... 0 to 300 at 200ma (one of those old heathkit supplies are perfect for this) and 6.3 VDC (one of those ultra cheap chinese things 0-20 @ 6amp) My guess for the umbilicals would be these wires arranged however GND, probably 2 wires, one per amp channel, and possibly one extra for filament ground B+ got to be 2 wires since there are seperated power supplies F+ the other filament wire. If its the ray samuels method, that would be 4 wires, but it sounds like you have 5... Edited June 23, 2009 by kevin gilmore
Voltron Posted June 23, 2009 Author Report Posted June 23, 2009 I most often ran 4 gas tubes, so 300 volts. Is there any way that only one wire should be attached to one of the umbilicals? I will check again, but I am pretty sure that there was only one wire on one side. I would be more than happy to start over with a new PS in that box if what is in there is really compromised. I also would never expect somebody who helped me out to own the repairs for life.
guzziguy Posted June 23, 2009 Report Posted June 23, 2009 [snip] Man this has fail all over it. How many people think i'm stupid intrigued enough to want to own the repairs to this thing for the rest of my life... [snip] I think that this is playing a large part in your feelings about this PS. As Al said, if you fixed it, you wouldn't own those repairs for the rest of your life. Besides, once you fix it, what can possibly go wrong?
Grahame Posted June 23, 2009 Report Posted June 23, 2009 Did the power supply come with the Pandora's box upgrade? At least there is hope (thanks to KG).
kevin gilmore Posted June 23, 2009 Report Posted June 23, 2009 I most often ran 4 gas tubes, so 300 volts. Lets try that again just to make sure. There are up to 3 tubes per channel. So if they are 0a3 they are 75 volts each. (red) If they are 0b3 they are 90 volts each. (orange) If they are 0c3 they are 105 volts each. (redish blue) if they are 0d3 they are 150 volts each. (purple) So 2 x 0a3 would be 150 volts... 2 x 0b3 would be 180 volts...
Voltron Posted June 23, 2009 Author Report Posted June 23, 2009 Lets try that again just to make sure. There are up to 3 tubes per channel. So if they are 0a3 they are 75 volts each. (red) If they are 0b3 they are 90 volts each. (orange) If they are 0c3 they are 105 volts each. (redish blue) if they are 0d3 they are 150 volts each. (purple) So 2 x 0a3 would be 150 volts... 2 x 0b3 would be 180 volts... You were speaking per side and I was adding them. I ran 2 OA3 per side (for a total of 4). Thought of that as 300 volts, but it is indeed 150 volts per side.
Voltron Posted June 24, 2009 Author Report Posted June 24, 2009 I finally said fuckit and unscrewed all the tube sockets, unglued the bundle of four caps and unscrewed a small binding post (or whatever it is called and popped the top off all the way. Here is the dog's breakfast in all its glory: I also created a flickr page with all the photos I took, most of which suck. Hopefully, this will give people who know something an idea of what is going on. I can confirm there is not anything I recognize as a FET or any other kind of semiconductor chip. SDS-XLR PSU Photostream Other than photography criticism, what do you think?
Looser101 Posted June 24, 2009 Report Posted June 24, 2009 Can you look in behind this board, where its labeled G D S on the right side.
kevin gilmore Posted June 24, 2009 Report Posted June 24, 2009 wow, that is something alright... I do notice the 2 solid state diode bridges connected to the heater supply. 4 pins each, mounted to the chassis. This has to be solid state. The 2 x (2 caps in series with resistors) are obviously the main filter caps. Very hard to see, but i think i see at least one fet on the big board to the right next to the orange drop caps. the 3 pins, and it is probably soldered to the backside of the board. Actually labeled on the board as S D G (source drain gate) 2 diodes and 2 resistors in series hooked up to the thing... just stunning work... better pictures of that area of the board would help. edit: where did that second picture come from. Need better photos... DEFINITELY pass FETS... will update schematic in a bit...
Looser101 Posted June 24, 2009 Report Posted June 24, 2009 From here: SDS-XLR PSU - a set on Flickr ...but no shot of the rear of that board (which would be helpful).
kevin gilmore Posted June 24, 2009 Report Posted June 24, 2009 The hammond 266n12 is turning into a crispy critter too, just like the one in the extreme... And there is another filament supply on the main board, reason unknown.. Wonder where or if there is a second fet. sort of looks like the 2 tube rectifiers are wired together, and they shouldn't be.
n_maher Posted June 24, 2009 Report Posted June 24, 2009 The hammond 266n12 is turning into a crispy critter too, just like the one in the extreme... Not entirely surprising. Even if he's just using it for the 5U4's it's getting whacked for 6A and operating in a fully enclosed space. Now if he's running any of the other filaments off of it too...well, I think we know where this is going.
Voltron Posted June 24, 2009 Author Report Posted June 24, 2009 Yeah, I couldn't get a great shot of it, but that trafo is in meltdown. There are two windings (maybe?) and they are sliding out to the side in different amounts.
kevin gilmore Posted June 24, 2009 Report Posted June 24, 2009 (edited) Not entirely surprising. Even if he's just using it for the 5U4's it's getting whacked for 6A and operating in a fully enclosed space. Now if he's running any of the other filaments off of it too...well, I think we know where this is going. Shame on you nate... Go and look at the schematic i posted on this. The filament windings on the 5u4's are firstly 5 volts, and second, they are lifted up by about 500 volts. Those filament windings are part of the 278x transformers. One transformer per 5u4. OK, so voltron, you got the thing apart, lets go a little further, take the screws that hold the posts on, and tilt out the main circuit board, leaving it attached to the aluminum and take straight on pictures, and pictures under the board... What i'm now thinking is that mikhail lied about how the thing works... (no never would he do such a thing.) There may in fact be only one fet. So one of the gas tube strings is used as is without a buffer and just runs the 2 input tubes. and the other gas tube string is used with the fet buffer and runs the 4 x WCF tubes... So the 2 x neutrik power connectors are according to the pdf used to seperate things for maximum stereo imaging ?? Yeah that's it... Edited June 24, 2009 by kevin gilmore
n_maher Posted June 24, 2009 Report Posted June 24, 2009 (edited) Ugh, teach me to post while working on a deadline. Massive fail in that post there was. This begs the question, what's the amp tube compliment for this thing? I honestly don't know. Edited June 24, 2009 by n_maher
kevin gilmore Posted June 24, 2009 Report Posted June 24, 2009 (edited) Well the power supply is 2 x 5u4 and 6 x gas tubes plus 3 transformers and a pile of extra shit. The amp is balanced and is therefore 6 tubes, the usual mikhail plug in anything you want with adapters, but standard is 6 x 6sn7. 1 dual triode gain tube per channel, 2 dual triode output tubes per channel in WCF. So in picture dscn2645 is the semiconductor mounted to the chassis 4 wires or 3 wires... Sure looks like 4 wires, and therefore a diode bridge, tied to the second transformer, which makes it interesting because it sure looks like the 2 x 5u4's are actually in parallel which also makes absolutely no sense... Edited June 24, 2009 by kevin gilmore
kevin gilmore Posted June 24, 2009 Report Posted June 24, 2009 Yep the 2 x 5u4's are in parallel and the second transformer does nothing except make 1100 volts likely used as the gas tube starting voltage. and there really is only one fet... so the left channel, right channel seperated power supplies is pure gold plated texas cow dung.
Recommended Posts
Create an account or sign in to comment
You need to be a member in order to leave a comment
Create an account
Sign up for a new account in our community. It's easy!
Register a new accountSign in
Already have an account? Sign in here.
Sign In Now