kevin gilmore Posted June 29, 2014 Author Report Posted June 29, 2014 300v pulls too much current for the kind of thing used in a srd. Besides which very low noise is rather important.
eggil Posted June 29, 2014 Report Posted June 29, 2014 88 degrees Celsius on the heatsinks seems a bit excessive. Anybody else has measured the temperature of the heatsinks? I don't know how accurate these cheap Radio Shack IR thermometers are.
luvdunhill Posted June 29, 2014 Report Posted June 29, 2014 How long can you put your hand on the heat sink before you have to pull it away?
eggil Posted June 29, 2014 Report Posted June 29, 2014 (edited) May be a second. No more. Currently measuring 85 degrees in one of the heatsinks. Edited June 29, 2014 by eggil
nopants Posted June 29, 2014 Report Posted June 29, 2014 Mine are also running fairly hot, probably also around 80C last I checked. I haven't run into any issues running up to 12 hours at a time
RudeWolf Posted June 29, 2014 Report Posted June 29, 2014 60 - 70°C is what I consider to be the long term safe margin for non-tube electronics. Remember that lower temps yield a longer lifespan. Can't really remember, but the rule of thumb was like double the hours for every ten degrees or so.
spritzer Posted June 29, 2014 Report Posted June 29, 2014 I'm considering tricking mine out with another hv supply for the 300v line, anyone know of appropriate low power (ie reasonably sized) transformer? how does the srd generate bias in that small package? It doesn't make any sense to use a PSU that is not better than the KGSSHV unit as you'd just degrade the performance. Dropping the voltage with a resistor for the front end doesn't have that big of an impact though. For a transformer you might get away with an oversized 230VAC unit. The margins are very small and I'd use a 250V unit instead but it could work without dropping out of regulation. The SRD's use an input limited voltage multiplier. It can't take any load though which means they are all but impossible to measure.
luvdunhill Posted June 29, 2014 Report Posted June 29, 2014 (edited) May be a second. No more. Currently measuring 85 degrees in one of the heatsinks. It's too hot then, doubly so since it isn't cased yet. If you leave it as-is you need to consider how these devices fail and what overall effect that will have on the entire circuit when that happens... Edited June 29, 2014 by luvdunhill
Victor Chew Posted June 29, 2014 Report Posted June 29, 2014 Enrique, what your cathode resistor value?
eggil Posted June 29, 2014 Report Posted June 29, 2014 (edited) Victor, I used 1K resistors. I think I will modify the PSU per RiStaR's idea: 1- removing the heatsinks, 2- bending the transistors back, and 3-using an aluminum bracket, connect them to larger off-board heat sinks, 4-plus making sure there are large vents in the case. http://www.ebay.com/itm/171015163803?_trksid=p2059210.m2749.l2649&ssPageName=STRK%3AMEBIDX%3AIT Edited June 29, 2014 by eggil
Victor Chew Posted June 30, 2014 Report Posted June 30, 2014 Ik wlll get you more power. It is going to pull a lot of current and the psu is going to get quite hot. Offboard sounds like a good idea but I am not sure how much current those tracks are rated for, how much they can take and for how long they will last. For that reason, I have decided to stick to specs. and to use 1.5k.. But of course, with 1k you are going to get more power and arguably/perhaps "better sound". I have heard Kelvin's, using 1k, and I must say it sounds very good to me. No way to A/B compare until I am done with mine.
eggil Posted June 30, 2014 Report Posted June 30, 2014 (edited) I guess it may be easier to replace the 1.5K with 1K than offboarding the 8N80's. Edited June 30, 2014 by eggil
Victor Chew Posted June 30, 2014 Report Posted June 30, 2014 (edited) You mean replacing the 1k with 1.5k. Off boarding will certainly be much more work than resistor swap. Can you let us know the temperature if you decide to go the 1.5k route. Edited June 30, 2014 by Victor Chew
eggil Posted June 30, 2014 Report Posted June 30, 2014 (edited) Just changed the resistors from 1K to 1.5K. (Thank God for the Hakko 808). Everything works but won't be able to leave it on long enough to measure temperature tonight. CORRECTION: With the 1.5 K resistors: temperature in the mid 80's after 1 hour of listening. With the 1 K resistor: 91 degrees! It is still too high. Edited July 1, 2014 by eggil
kevin gilmore Posted June 30, 2014 Author Report Posted June 30, 2014 (edited) this should handle the power angle bracket to big heatsink and 3 x fet per side http://gilmore.chem.northwestern.edu/kgsshvpsminiexternal.jpg Edited June 30, 2014 by kevin gilmore 4
eggil Posted July 1, 2014 Report Posted July 1, 2014 (edited) OK then. Now to buy a chassis with a big external heatsink to offboard the transistors after removing the 4 small heat sinks, similar to the offboard KGSSHV built, using aluminum brackets. Edited July 1, 2014 by eggil
spritzer Posted July 1, 2014 Report Posted July 1, 2014 It's also possible to just use the chassis as a heat spreader to a sink close by. We aren't talking about too much heat here...
eggil Posted July 2, 2014 Report Posted July 2, 2014 (edited) Don't know yet. Spritzer, you mean just use offboard heatsinks inside the chassis? I guess a real big one would work :)I haven't found a chassis I like Edited July 2, 2014 by eggil
johnwmclean Posted July 2, 2014 Report Posted July 2, 2014 (edited) Angle the devices and sink to the chassis floor. The pcb already has holes just for this occasion. If you still worried about heat, up the thickness of the base panel, I'm thinking 6mm. Edit: Maybe your pcb doesn't have screw holes, but you can still do this with a bit of care. Edited July 2, 2014 by johnwmclean
luvdunhill Posted July 2, 2014 Report Posted July 2, 2014 My experience suggests that you have more heat than is appropriate to couple them to the case. I defer to the designers though...
spritzer Posted July 2, 2014 Report Posted July 2, 2014 Simply using the chassis won't be enough. Add some thick aluminum plates as heatspeaders and then some heatsinks will work. Something inline with this: You can use internal heatsinks if there is enough airflow for them to function properly. There are no chassis like this so you need to fabricate everything.
eggil Posted July 6, 2014 Report Posted July 6, 2014 (edited) EDIT: I should have listened to luvdunhill earlier: This is what he mentioned before about removing heatsinks: 1. remove the device 2. use a heat gun and heat up the heat sink. Use a heat shield to avoid the PCB. Get 'em hot 3. thread a piece of string through the hole so you have something to pull 4. Use Chip Quik 5. pull They pop right off I struggled removing the first 2 heatsinks, then I remembered I had some Chip Quik and the next 2 heat sinks popped without difficulty. I used a regular solder gun, but still worked great. That stuff is magic! My PSU is now ready for bigger heatsinks. And in case anybody is wondering how to connect an external pot, here are some pics. Edited July 6, 2014 by eggil
Peleus Posted July 7, 2014 Report Posted July 7, 2014 Dear all, I've got an idea to using T2 psu in the megatron plan. Is it an good idea, or not? It's nearly impossible to get 2SC3840 & 2SA1684 in local resource, and, fortunately, I got six 2SC3675！ So, I wonder to modify the T2 psu by adding more one fqp8n80c..... Is it work fine for megatron ?
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