View Full Version : Hot Toroids...
Hi!
My toroids are hot 27/7 dunno why, anyone got an idea?
Could it be that i have not shielded the dac well enough? (it lies 5cm from the truepath) Could it be picking up HF noise?
The wierd thing is that they are just as hot at full power compared to idle (not muted).
Pictures coming soon :)
Hmmm, seems like they are saturating for some reason. Just exactly how hot? Like burning hot?
------X-max. can be several meters on any driver, too bad it can be done only once...------
thomaseliot
16-Jan-09, 00:52
If you are using the Litz wire in the kit, and not all the wires are well scraped from coat and soldered, maybe just a few thin wires hold the whole high switching frequency.
audiophool
16-Jan-09, 02:07
Ivar,
In my experience, and I have built dozens of Truepaths, the coils don't run cool. Members have expressed concern before, and this is exacerbated by some members reporting cool coils, and it's always escalated by replies that "it must be a problem"...
It also should be expected the coils don't run cool because this would have added additional size and expense to Truepath's evaluation board.
If you are confident that you prepared your coils carefully and the amp works then I wouldn't worry about it.
To any members who have an amp with standard coils that run cool, I would like to verify this. Send it to me and I'll refund the postage and eat my words on this forum.
Rgds.
I agree with that. I have two Truepaths, both use the standard cores and litz wire and I believe that they are well scraped and soldered. Both of them run hot. At +-60VDC they are what I would term "very hot", down at +-50VDC they are very warm to hot. This happens even at low volume.
I also tried winding an inductor using the same core but with a single large wire instead of litz, with the same results.
I think that the people claiming cold inductors are probably not using T106-2 cores and/or are running low voltages.
Perhaps we need to use largee cores than T106. Maybe T130-2 might be worth trying. I'm not so sure about using MPP cores.
I got some T130-2 from 41Hz's shop.
I will try them out when I started to build my Truepath.
At work we have one of those Fluke digital thermometer "guns", perhaps it's a good idea to do some measurements with it.
I got T130, T160 and some MPP cores, only thing I don't have is a +/-60Vdc supply. I don't really like running the amp on the edge that much, this will certainly shorten it's lifespan. My Truepath is running at +/-47Vdc with standard toroids and I've indeed not seen them even warm yet.
I also have one working board with MPP cores, but haven't built it in a box yet...
I can order a +/-60Vdc lab supply to create a testbench to do the measurements. I shall also test some T130 standard perm cores...
------X-max. can be several meters on any driver, too bad it can be done only once...------
These look very promising as a solution. They are gapped ferrite 3C20 cores.
http://www.ferroxcube.com/appl/info/class%20D%20audio%20amplifier.pdf
http://www.ferroxcube.com/appl/info/gaptoroids.pdf
In their example class-d application they replaced a T106-2 cored output filter inductor with a gapped ferrite and measured less distortion. They only needed 13 turns compared with 37 for the T106-2. They should be cheaper than MPP and perform better as they have better frequency stability.
The core that they used was FERROXCUBE TN26/11-3C20-A113
I've no idea where to buy them though :-(
There are some IRS2092 amplifiers on ebay that use these cores.....
http://cgi.ebay.com/Class-D-Power-Amplifier-250W-X-2-500W-X-1-NEW_W0QQitemZ260347755000QQihZ016QQcategoryZ39783Q QssPageNameZWDVWQQrdZ1QQcmdZViewItem
I sent a distributor a RFQ for 100pcs. I will let you know how much they want for them.
Hey great find!
------X-max. can be several meters on any driver, too bad it can be done only once...------
Somebody try this amp with that new chip? It would be interesting to make a test with the truepath!
It looks nice, but I'm afraid it has hard to service parts, like for instance what's tha big block on top? Is that a "chip" or a heatslug? There are no pics shown from that side...
------X-max. can be several meters on any driver, too bad it can be done only once...------
Looking in the official web site of the chip, it seems that is a heatslug....that chip looks as it is a copy of Tripath? or similar?
It could perhaps be IRS2092S? The design looks much like their reference design...
------X-max. can be several meters on any driver, too bad it can be done only once...------
Yes, it is the IRS2092S. I'm sure it's a very close copy of the IRAUDAMP5. It's actually a very nice self-oscillating sigma-delta amplifier design.
I own two of the older IRAUDAMP1 boards (IR2011S) that are pictured in the first FERROXCUBE PDF link col_s posted. They are sitting on a shelf collecting dust along with various Tripath and ICEPower boards.
I received a reply from the distributor.......
quote:
Dear Sean;
Thank you for the request, unfortunately we do not have any of these in stock. Can you use anything else?
Regards,
Chris Spadafora, President
Eastern Components, Inc.
11 Union Hill Road
West Conshohocken, PA 19428
610-825-8610 ext. 104
610-825-6929 fax
www.eastern-components.com
cs1@eastern-components.com
I'll see if I can find another.
What kind of a reply is that!?
This can only be from a salesman...http://www.diyaudio.com/forums/images/smilies/xeye.gif
------X-max. can be several meters on any driver, too bad it can be done only once...------
here the link of the reference amp of IRF
http://www.irf.com/technical-info/refdesigns/iraudamp7s.pdf
That one in ebay it is based in the reference board, but different.
They look very promising! A ref. board can be had for a few hundred bucks, maybe I'll buy one soon...
------X-max. can be several meters on any driver, too bad it can be done only once...------
Yes, it looks like the Tripath technology, right?
Because I don't have time (And the knowledge may be also) to fix my truepath, I've ordered this IRF base amp ! It will arrive next week I think.
I have the amp10 to compare with it. Vbro, you offer kindly to help with my truepath, how we can do it? my personal email is abelma1@hotmail.com. Thx!
quote:Originally posted by abelma
Yes, it looks like the Tripath technology, right?
In short, no.
They are both Sigma-Delta designs with pre-filter feedback, but that is where their similarities end. The IR is self-oscillating where Tripath uses it's proprietary DPP "Digital Power Processing" to adjust the switching frequency.
Is that why dead time can't be lowered as much as with the IR amp? I can imagine the varying switching frequency has advantages but also down sides, like for the fets at 1,5Mhz switching it is harder to keep overshoots under control at very short dead time...?
------X-max. can be several meters on any driver, too bad it can be done only once...------
Damn, i almost forgot this tread. The rails are ca 56v, the toroids are hot, but touchabele, not much hotter at full power compared to low.
40 to 50 degrees C? i am pretty shure i have taken off all the enamel on each litz wire. But what i actually did was to solder the litzwires in the holes, then cut of the end of the litzwires ca half a mm over the board with a plier that cuts a flat cut, and then solder over this end, or top. So every wire is soldered "twice"
quote:Originally posted by 86Ivar
Damn, i almost forgot this tread. The rails are ca 56v, the toroids are hot, but touchabele, not much hotter at full power compared to low.
40 to 50 degrees C? i am pretty shure i have taken off all the enamel on each litz wire. But what i actually did was to solder the litzwires in the holes, then cut of the end of the litzwires ca half a mm over the board with a plier that cuts a flat cut, and then solder over this end, or top. So every wire is soldered "twice"
What's that method?
Do keep in mind that however solder is made of metal the conductivity is very poor, especially at places where lots of current has to pass through....
------X-max. can be several meters on any driver, too bad it can be done only once...------
audiophool
25-Jan-09, 02:45
quote:Originally posted by 86Ivar
...
The rails are ca 56v, the toroids are hot, but touchabele, not much hotter at full power compared to low.
40 to 50 degrees C?
...
Your inductors appear to be operating perfectly.
Rgds.
Hi guys,
I'm gonna try these soon:
http://www.snapdrive.net/files/361770/Truepath/Toroidal%20T130k.JPG
------X-max. can be several meters on any driver, too bad it can be done only once...------
Hi Vbro,
Are they T-130-2 kits ? I am currently building my first Truepath.
I have mounted everything except,
the toroids
TA-3020
the 4 mosfets and
the 2 input capacitors.
Just wonder how many turns I need to do on T-130-2....
Same number of turns as T-106-2 ??
I counted there are 15 strands of Litz wire, I remembered someone on this forum said 16-17 strands would be better suit for 130-2 cores.
Bigger cross-section and less turns ??
Well, I got them for AMP7, but requested T150-2 so I had them left over...
T130 need roughly 31 turns (says my meter...)
------X-max. can be several meters on any driver, too bad it can be done only once...------
Calculator here:
http://toroids.info/T130-2.php
it says 31.6 turns for 11uH, so 31 or 32 turns should be good
Powered by vBulletin® Version 4.2.0 Copyright © 2013 vBulletin Solutions, Inc. All rights reserved.