Kerry Posted March 21, 2011 Report Posted March 21, 2011 Likewise, sorry to hear that. Let us know when you track down the issue what cause it. I should have the boards populated next week and I will be putting in the transformer order tomorrow, so I should be able to start testing the supply pretty soon (baring any forgotten components).
spritzer Posted March 21, 2011 Author Report Posted March 21, 2011 (edited) Thanks guys. Kevin took some time to think about it and he thinks it may just be a fubar cap or that it is incorrectly labeled. Makes sense since the PSU works except for the noises and excessive heat. It's a good thing I bought a box full of those caps so plenty of spares. Edited March 21, 2011 by spritzer
Craig Sawyers Posted March 21, 2011 Report Posted March 21, 2011 I've been turning on the T2 PSU in steps these last few days and it almost blew up in my face. The -260V supply went into meltdown mode and the 680uf/450V cap was burning hot after running for about a minute. Now comes the utter joy of figuring what went wrong... Jeez - sorry to hear that. I guess I got lucky with the power supply - it just fired straight up and hasn't missed a beat since. The amp - as most of you know - was a different kettle of fish entirely. It's got to be simple though - if the cap is roasting, and the rectifier bridge is OK - then it has to be the cap. Mislabelled and electrically in backwards. If that is the case, make sure the rectifiers haven't got cooked into the bargain. Try pulling the IXYS current limiters temporarily, and isolate the transformer, bridge and capacitor.
kevin gilmore Posted March 21, 2011 Report Posted March 21, 2011 has to be the cap or diode bridge. because it still makes the right voltage. my bet is the cap has the + and - mislabled. Caps get very hot this way.
spritzer Posted March 22, 2011 Author Report Posted March 22, 2011 If I have a spare bridge then I'll replace it as well but first I'll just replace the cap and see what happens. The rest of the PSU was just fine with no excessive heat generated. Now I really need some sleep before tackling the high voltage...
Craig Sawyers Posted March 23, 2011 Report Posted March 23, 2011 Now I really need some sleep before tackling the high voltage... You've got that right Spritzer - tiredness (and booze) are never good partners to HV work
luvdunhill Posted March 23, 2011 Report Posted March 23, 2011 Hm. If I had to choose between a mislabeled cap, or a board manufacturing error, I'd probably assume the latter.
n_maher Posted March 23, 2011 Report Posted March 23, 2011 Hm. If I had to choose between a mislabeled cap, or a board manufacturing error, I'd probably assume the latter. Yeah, like an internal short in the board...
spritzer Posted March 23, 2011 Author Report Posted March 23, 2011 Hm. If I had to choose between a mislabeled cap, or a board manufacturing error, I'd probably assume the latter. Turns out the cap wasn't mislabeled (unless the connectors were also mislabeled i.e. not likely) and after installing a new bridge as well the problem is still there. KG thinks the transformer is fubar and thus the AC is too high but I did test it before connecting anything and it was on spec. I'll test it again just to be sure but an internal short in the board or an incorrectly labeled resistor are all possibilities at this point. I'll probably fire up the +260V supply to see how it behaves since it uses most of the same parts. I can also swap the windings and see if it makes any difference.
Craig Sawyers Posted March 23, 2011 Report Posted March 23, 2011 Turns out the cap wasn't mislabeled (unless the connectors were also mislabeled i.e. not likely) and after installing a new bridge as well the problem is still there. KG thinks the transformer is fubar and thus the AC is too high but I did test it before connecting anything and it was on spec. I'll test it again just to be sure but an internal short in the board or an incorrectly labeled resistor are all possibilities at this point. I'll probably fire up the +260V supply to see how it behaves since it uses most of the same parts. I can also swap the windings and see if it makes any difference. That is bloody wierd. I'd really go back to basics. Pigtail the transformer leads into a bridge rectifier taped to the bench, and measure the rectified (ac) voltage - which will tell you if there is a problem with that particular winding. Connect up a 680uF cap and again see what it measures as smoothed ac. If all is OK, then there could be an interwinding short or something else peculiar in the way the -260V is generated from 300V using a dc offset from the -500V and -60V supplies. But again I'd go back to basics - even take the Tx out, and run each winding into a bridge screwed to a piece of wood (say), then add caps and see what the actual raw dc is from each of the three windings. Then check the voltage between each of the windings - they should be isolated. Birgir - do you have a megohm meter? The windings should be electrically well isolated to a rather high voltage. If there is an interwinding short, or low resistance, it should show up using a megohm meter with the voltage set to 500V. Do your transformers have interwinding insulation barriers?
kevin gilmore Posted March 23, 2011 Report Posted March 23, 2011 I did pay extra to have the boards electrically tested. But who knows what really happened. I just don't see how the electrolytic can get hot unless its backwards, or over voltage. No matter what kind of load it sees. As far as kerry's amp... Turn up the bias current to 2 amps per leg. Replace the output fets with the SiC fets that nelson is using. BIGGER HEATSINKS. And my test bench thing works absolutely great with +/-24 volt power supplys, and does about 30 pure class A watts into 8 ohms. Currently using switchers for the power, and each amplifier puts out about 200 watts of heat. Sounds really nice into vintage VOTT's. REALLY NICE. OK, so VOTT's are really more like 12 to 16 ohms...
spritzer Posted March 23, 2011 Author Report Posted March 23, 2011 (edited) That sounds like a plan. My transformers are the same SumR spec as most of the other T2's built to KG's spec. No megaohm tester I'm afraid but I'll wire up a bridge and a cap tomorrow. Edit: I'm with KG though, it's only the cap and the parts around it (the copper conducting the heat?) that get hot so this is bloody strange. If this was some stupid mistake on my end I'll be so pissed... Edited March 23, 2011 by spritzer
Kerry Posted March 23, 2011 Report Posted March 23, 2011 (edited) ... As far as kerry's amp... Turn up the bias current to 2 amps per leg. Replace the output fets with the SiC fets that nelson is using. BIGGER HEATSINKS. And my test bench thing works absolutely great with +/-24 volt power supplys, and does about 30 pure class A watts into 8 ohms. Currently using switchers for the power, and each amplifier puts out about 200 watts of heat. Sounds really nice into vintage VOTT's. REALLY NICE. OK, so VOTT's are really more like 12 to 16 ohms... Very cool I was thinking of adding some local feedback on the output stage to keep the gain to about 10x (on the output stage). This will help keep the gain on the drains of the input stage closer to 1x amplification. Right now with all the negative feedback it is really low (hard to measure even on the scope). What do you think? Did I mention this is very cool! Edited March 23, 2011 by Kerry
kevin gilmore Posted March 23, 2011 Report Posted March 23, 2011 Well with a set of linear supplies, it tips the scales at 800 watts of heat. I would say its kinda hot
Craig Sawyers Posted March 23, 2011 Report Posted March 23, 2011 If this was some stupid mistake on my end I'll be so pissed... Well, my latest errors have been (a) a solder bridge that vapourised two tracks on a pro power amp I was fixing. Um, re-fixing as it turned out; and ( missing soldering a component at all in the BH. I was so paranoid about the T2 I did a 100% solder inspection under a magnifier - and I still missed an unsoldered joint (so 99.95% inspection...)
livewire Posted March 24, 2011 Report Posted March 24, 2011 Thanks for posting your DIY trials and tribulations guys. Now I dont feel so bad. Spritzer I hope you get it sorted soon.
chinsettawong Posted March 25, 2011 Report Posted March 25, 2011 Will there ever be another run of the T2 PCB? I'm so very interested in building one. Wachara C.
spritzer Posted March 25, 2011 Author Report Posted March 25, 2011 Well, my latest errors have been (a) a solder bridge that vapourised two tracks on a pro power amp I was fixing. Um, re-fixing as it turned out; and ( missing soldering a component at all in the BH. I was so paranoid about the T2 I did a 100% solder inspection under a magnifier - and I still missed an unsoldered joint (so 99.95% inspection...) I do hate it when that happens. I was having some random issues with the first Egmont clone I built and it was naturally due to unsoldered joints. I had looped and clamped the wires (all done P-P) but didn't realize the error until I had stripped the amp for a complete rebuild. Here I was thinking the cheap Pfhier pots were acting up...
spritzer Posted March 27, 2011 Author Report Posted March 27, 2011 (edited) It appears I have good news on the T2 front. Today is my first day off in a month and a half so I'm running at approximately 50% which is way more than is the norm. The first thing I did was to go over the PSU and everything was as it should be, all resistors matched the schematic and everything was properly soldered. So that wasn't the issue I checked the transformers and lo and behold, I didn't get any resistance readings on the 485 and 285 windings... Turns out SumR fucked up the data sheet so violet*/gray were listed for the 485VAC winding but it is supposed to be gray/blue. Ditto for the 285VAC windings, they listed blue*/green but it really is violet/green. Here are some pics of the other transformer under test but the resistance on the offending transformer is the same: Needless to say I'm not happy but I'm more angry at myself for trusting the datasheet... Now I just have to fix the leads and hook everything up for testing. Edited March 27, 2011 by spritzer 1
kevin gilmore Posted March 27, 2011 Report Posted March 27, 2011 (edited) So lets see, about 650 volts on a 450 volt cap. Yep that is probably going to cause it to get stinkin hot... And the regulator circuitry does not care, as it can handle 800 volts easy. Edited March 27, 2011 by kevin gilmore
spritzer Posted March 27, 2011 Author Report Posted March 27, 2011 Good to know. I'm rearranging the wires now so back with some news shortly
spritzer Posted March 27, 2011 Author Report Posted March 27, 2011 It´s alive!!!! I had a minor scare when I forgot to tighten one of the 485VAC wires (so the PSU made no sense at all) but it was quickly fixed. No noises and no smoke and the same goes for the + side as well. Now I just need to finalize the wiring and turn my attention to the amp chassis...
kevin gilmore Posted March 27, 2011 Report Posted March 27, 2011 Make sure you change out that electrolytic again, and throw the overvoltaged ones in the trash can. To everyone, power up the transformers on the bench to make sure you know how they are wired before you hook them to the boards.
n_maher Posted March 28, 2011 Report Posted March 28, 2011 Not the first time SumR has sent an incorrect data sheet. They've done it to me before too.
Voltron Posted March 28, 2011 Report Posted March 28, 2011 That's great Biggie! I hope you get another day off in the next 90 so you can finish it.
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