Tripath EB 104A needs replacment power MOSFETS
Well ,I replaced the transformers on a amp made from a Tripath TA 104A evaluation board and used for subwoofer duty,thus raising the rail voltages from +/-70 V to about +/- 85V,(100 V caps) and at the beginning all vent well , no smoke , no excessive DC offset. However some additional work was made last night,a soft start was installed, and some better wiring in this "birds nest" of an amp.Didnīt like the AC filter quite close to the input DC blocking caps(Mundorf Silver/Gold , too large) . So I glued it with epoxy,on the edge,and may have changed the ground "pattern" a bit,as it seems to have some part in this.
I bought the amp ,who had it built for him by a friend,and then he modded it with audiophile stuff, like Black Gates. Even C-LC filtration!In no why a clean
creation, I have to doubt itīs reliability,but it made good service driving my Volt RV3143 woofers.
Last night, ready stedy, but upon measuring the offset, the one channel showing 2.6 V in mute condition OK according to the manual, the other almost -85 V DC !
And when unmuting,the negative rail dropped considerably,and I was glad I hadnīt conneced the amp to the quite expensive woofers. In the muted state,connecting a power resistor of 100 ohms to the failing otput,warmed it up considerably, and a speaker would have been burnt...
Obviously , at least one power Mosfet is out of duty,so replacing all of them with something more reliable seems to make sense!
Originally they are STW 38 NB20 , 200V VDs 38A devices, but they may be hard to obtain? So how about STW 50 NB20 , the same VDs rating,slightly higher Ciss (3400 pF compared to 2800 pF).
ID 50A instead of 38A.
Anything I can do to prevent a new failure? To get the best bass, I hesitate to use DC protection that includes relays?
Thanks in advance!
Strange,as the "old" transformers were custom made for a company, there were no information about anything on them.But the weigh of them is about 25 lbs/each, indicating maybe 800 VA types?But there was no softstart!
The new transformers are 500 VA. The rail voltage was raised from +- 70 V to +-85 volt.
The entire amp is somewhat non optimal built, and for the sanity of any perfectionist, I donīt publish a photo;)
What about Caddock thin film resistors for the input! And I am guilty of paralleling the "original" Mundorf Mcap Supreme Silver/Gold , with orindinary polyprops,which makes for a total of 4.2 mF ,but the gain of the amp has alsos been changed. In addition, I have thrown in some very big electrolytic caps for the rails. All this was made long before the failure.The value of the over current protection,has been change too,but there should have been a log, about all the changes made to this amp?
BTW, BBM, or deadtime , was changed(by the former owner), from the default 65 nS to "the slightly better sounding" 25 nS, the lowest alternative on the evaluation board.
Now I have measured some strange voltages for one of the power MOSFets, and I will order the STW50NB20īs . I will check a few things after desoldering both transistors on this channel.
As it is, itīs impossible to get the amp to unmute, and I hope that it is only some protection circuit, that has detected the high DC level on the output?
Iīm thinking of replacing this amplifier, even if I bring it back to operational status.
Lesson learnt: donīt mess with something thatīs working well.
My priority is a reliable amp, that also sounds good.
Yesterday I was driving one sub, with a spare channel on my Dynamic Precision A1 (about $11.000 ) Itīs a DC coupled amp.Listening to some Disco music, I was convinced that the Tripath amp may have a certain mightyness in the bottom octave, but canīt compete in terms of speed,precision and "organic"or integrated sound!:cool:
Class D amplifiers are complex electronics where things far beyond the understanding of any "audiophile tweaker" happen. Exotic modding may end up with smoke quite easily.
Don't you know how to test a MOSFET with a multimeter? You don't even need to unsolder it to ckeck whether it's shorted or not (unlikely according to the description that you gave).
|All times are GMT. The time now is 11:27 PM.|
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2016 DragonByte Technologies Ltd.
Copyright Đ1999-2016 diyAudio