congo5 Posted July 25, 2016 Report Share Posted July 25, 2016 (edited) just WAG but assuming no wrong parts/missing or backwards, is the stn9360 in place under the boards? (I've forgot it) what are the voltages on the Cree? middle pin should be full unreg and to its left is the reg voltage GDS sorry really can't see much but what little I can looks good may have to wait for those who know more to chime in Edited July 25, 2016 by congo5 Quote Link to comment Share on other sites More sharing options...
s_r Posted July 25, 2016 Report Share Posted July 25, 2016 (edited) It's definitely the 10V LT1021. I'll double check the voltages on the cree in a bit. Edited July 25, 2016 by s_r Quote Link to comment Share on other sites More sharing options...
johnwmclean Posted July 25, 2016 Report Share Posted July 25, 2016 Have you tightened the vacant terminal block screws where your measuring from? Quote Link to comment Share on other sites More sharing options...
s_r Posted July 25, 2016 Report Share Posted July 25, 2016 Nope, the same is true on the other PSU board which measures just fine. As for the cree, D reads +559V, while G reads +17V (referenced to ground). Quote Link to comment Share on other sites More sharing options...
congo5 Posted July 25, 2016 Report Share Posted July 25, 2016 (edited) you can compare with the other side as they are the same circuit. it may be the CREE like you suspected or something wrong with the CCS 10m90/DN2540 check voltage across R1 on schematic should be close to what the working one is.. its too late for me to hook one up and measure, work hours are approaching. You'll have better help tomorrow anyway good luck Edited July 25, 2016 by congo5 Quote Link to comment Share on other sites More sharing options...
spritzer Posted July 25, 2016 Report Share Posted July 25, 2016 The Cree is well insulated at the hole so there should be no problems with using a steel screw. I would flip it over and look for any bad solder joints. That amount of difference with nothing blowing up could point to a bad connection. Quote Link to comment Share on other sites More sharing options...
astrostar59 Posted July 25, 2016 Report Share Posted July 25, 2016 (edited) Thinking about the thermal requirements of the Carbon Fets, is there a tried and tested way you guys are fitting the fets to the heat sinks? If done incorrectly it could spell problems. I have been digging around for information but there seems to be a lot of contradictory information on the web. I found this guide that seems to make sense: http://application-notes.digchip.com/081/81-43573.pdf Any thoughts? Edited July 25, 2016 by astrostar59 Quote Link to comment Share on other sites More sharing options...
johnwmclean Posted July 25, 2016 Report Share Posted July 25, 2016 Teaching your grandmother to suck eggs? Of course this has been covered here a bazillion times. We also deal with high voltage, most papers out their don’t cover this aspect (including the one you’ve reference). 1 Quote Link to comment Share on other sites More sharing options...
sorenb Posted July 25, 2016 Report Share Posted July 25, 2016 3 hours ago, astrostar59 said: Thinking about the thermal requirements of the Carbon Fets, is there a tried and tested way you guys are fitting the fets to the heat sinks? If done incorrectly it could spell problems. I have been digging around for information but there seems to be a lot of contradictory information on the web. I found this guide that seems to make sense: http://application-notes.digchip.com/081/81-43573.pdf Any thoughts? typing a keyboard also might spell problems ... Quote Link to comment Share on other sites More sharing options...
astrostar59 Posted July 25, 2016 Report Share Posted July 25, 2016 (edited) 1 hour ago, johnwmclean said: Teaching your grandmother to suck eggs? Of course this has been covered here a bazillion times. We also deal with high voltage, most papers out their don’t cover this aspect (including the one you’ve reference). No just want to know the best method. For example is it best to mount the fets directly to the back of the heat sink? Many here mount on a bracket, I am assuming as the legs are not long enough. Edited July 25, 2016 by astrostar59 Quote Link to comment Share on other sites More sharing options...
gepardcv Posted July 25, 2016 Report Share Posted July 25, 2016 If the case and PCB layouts permit, mounting transistors directly to a heatsink would work well. Often, though, the boards do not fully fit on the side of a heatsink. In that case, nothing wrong with an aluminum bracket and a layer of thermal paste connected to a heatsink. Aluminum conducts heat very well, which is why it's used for heatsinks, and also why steel cooking pans and pots often have aluminum bottom plates. Quote Link to comment Share on other sites More sharing options...
johnwmclean Posted July 25, 2016 Report Share Posted July 25, 2016 Both methods work well. If you want to rational a bracket adds thermal resistance between the device and the heatsink, theoretically therefore not optimum. But excuted well, like most builders here do, there are no problems with this method. 1 Quote Link to comment Share on other sites More sharing options...
JoaMat Posted July 25, 2016 Report Share Posted July 25, 2016 Amazing how good he is at getting responses! Quote Link to comment Share on other sites More sharing options...
kevin gilmore Posted July 25, 2016 Report Share Posted July 25, 2016 (edited) birgir is about to release a new bi-directional power cord system for people like astrostar59 its absolutely unique, nothing else like it, hand made in Iceland. both 110v and 220v versions available, solves once and for all the directionality of power cord issues. here is a picture of the 110v version bi-directional iec female not shown Edited July 25, 2016 by kevin gilmore 1 Quote Link to comment Share on other sites More sharing options...
JoaMat Posted July 25, 2016 Report Share Posted July 25, 2016 (edited) Great, will it cure excessive terrorizing posting as well? Edited July 25, 2016 by JoaMat 1 Quote Link to comment Share on other sites More sharing options...
spritzer Posted July 26, 2016 Report Share Posted July 26, 2016 Once one end has been inserted into the nostrils as instructed then it is bound to help. It is vital to achieve ultimate performance that all circuit breakers and fuses are bypassed. As we all know from the cable master Ray Samuels, nothing can strangle the flow of the larger electric molecules. Quote Link to comment Share on other sites More sharing options...
Hippyskygod Posted July 26, 2016 Report Share Posted July 26, 2016 Hi, I just joined because I bought my first Stax gear last week, an old SRM1 mk2 and some new L700 and I'm utterly mesmerised :-) Naturally I want to explore further. I would like a balanced input as my Pono sounds way better with the balanced mode with my ER4's. Can anyone help me get started on building a KGSSHV Carbon? Really what I need to begin with are PCB's or Gerbers. I don't expect spoon feeding though. While I'm getting parts together I'll read everything I can! Cheers Quote Link to comment Share on other sites More sharing options...
gepardcv Posted July 26, 2016 Report Share Posted July 26, 2016 Read everything on the DIY forum here. This thread, the old Carbon group buy thread, the KGSSHV thread, probably the T2 thread, the generic electrostatic amp build thread. You are in luck, as a new group buy for Carbon PCBs just started, see http://www.head-case.org/forums/topic/12671-group-buy-carbon-v70/ If you have never built an amplifier before, the Dynalo is a safer and easier first project. Quote Link to comment Share on other sites More sharing options...
HemiSam Posted July 26, 2016 Report Share Posted July 26, 2016 Ya can't make some of this stuff up.... HS 1 Quote Link to comment Share on other sites More sharing options...
wink Posted July 26, 2016 Report Share Posted July 26, 2016 2 Quote Link to comment Share on other sites More sharing options...
s_r Posted July 26, 2016 Report Share Posted July 26, 2016 I tried reflowing a number of joints around the 10m90s & the resistors around them. No change in behavior though. Also did some more measuring, the voltage across the 4.7kOhm resistor on both boards measure the same (about 1.77V). Pin 1 on the 10M90S right above the DN2540 reads +17V while pin 2 reads +559V, same as the cree. Not sure if those measurements are of much use though. Quote Link to comment Share on other sites More sharing options...
astrostar59 Posted July 26, 2016 Report Share Posted July 26, 2016 11 hours ago, johnwmclean said: Both methods work well. If you want to rational a bracket adds thermal resistance between the device and the heatsink, theoretically therefore not optimum. But excuted well, like most builders here do, there are no problems with this method. Thanks that is great. Quote Link to comment Share on other sites More sharing options...
JoaMat Posted July 26, 2016 Report Share Posted July 26, 2016 1 hour ago, s_r said: I tried reflowing a number of joints around the 10m90s & the resistors around them. No change in behavior though. Also did some more measuring, the voltage across the 4.7kOhm resistor on both boards measure the same (about 1.77V). Pin 1 on the 10M90S right above the DN2540 reads +17V while pin 2 reads +559V, same as the cree. Not sure if those measurements are of much use though. Long shot, but a flipped 1N4007 cross STN9360 will probably give the same result as you have. Quote Link to comment Share on other sites More sharing options...
johnwmclean Posted July 26, 2016 Report Share Posted July 26, 2016 Or a faulty LT1021. Quote Link to comment Share on other sites More sharing options...
mwl168 Posted July 26, 2016 Author Report Share Posted July 26, 2016 Also, double check the current limit resistor is the correct value. Last time I experienced similar problem it was due to the current limiting resistor being 1000 times the correct value (5.1K vs. 5.1R). Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.