I recently bought two UCD400 modules. Have the parts for a dual mono setup, but I hooked the two modules up to one supply for testing yesterday.
The modules are the "standard" version, NO MODS, so the inputs are AC coupled. (single ended)
I ran the modules WITHOUT heatsinks for about 10min, at relatively low listening levels. After this time, (I was in the kitchen) I heard a very loud **POP** - I ran to the living room, heard the right speaker still playing and saw the left diaphagm looking to pop out of the chassis.
I ran for the power cord, but didnt make it in time, the speaker fried first
I disconnected the blown speaker and repowered the amps. The amp LEDs still work, and the "right" amp/speaker combo sounds fine.
The right amp's heatsink gets warm, the left heatsink stays cold. However, I know that the modules get warm without Load, so I was a bit suspicious at this point.
Later I hooked up the working speaker (have no instrumentation at home) to the left amp.
The speaker outputs were apparently at PSU voltage, because the resulting pop and spark was quite energetic.
The drive transistor(s) are in fact shorted out. (Am at work now, have access to all necessary equipment) The amp must have gotten very hot, because the ESD pad i left the amp on had a melted imprint of the 5mOhm shunt resistor... this could also be from current AFTER the FET shorted though.
I suspect the amps got too hot, after running them without cooling surface.
Anyone with tips or experience? Is it reasonable to just replace the transitors or should I expect other problems? The MOS-drive transistor measures fine...
Thanks,
Mike
PSU:
Trafo: 40Vsec, 500VA
Caps: 2x Epcos Sikorel 10mF, bypassed w 100nF film
Dual rectifiers
Unloaded Voltage = ±55.4V
Load:
8ohm CSS FR125 10cm Fullrange, 7L sealed.
The modules are the "standard" version, NO MODS, so the inputs are AC coupled. (single ended)
I ran the modules WITHOUT heatsinks for about 10min, at relatively low listening levels. After this time, (I was in the kitchen) I heard a very loud **POP** - I ran to the living room, heard the right speaker still playing and saw the left diaphagm looking to pop out of the chassis.
I ran for the power cord, but didnt make it in time, the speaker fried first
I disconnected the blown speaker and repowered the amps. The amp LEDs still work, and the "right" amp/speaker combo sounds fine.
The right amp's heatsink gets warm, the left heatsink stays cold. However, I know that the modules get warm without Load, so I was a bit suspicious at this point.
Later I hooked up the working speaker (have no instrumentation at home) to the left amp.
The speaker outputs were apparently at PSU voltage, because the resulting pop and spark was quite energetic.
The drive transistor(s) are in fact shorted out. (Am at work now, have access to all necessary equipment) The amp must have gotten very hot, because the ESD pad i left the amp on had a melted imprint of the 5mOhm shunt resistor... this could also be from current AFTER the FET shorted though.
I suspect the amps got too hot, after running them without cooling surface.
Anyone with tips or experience? Is it reasonable to just replace the transitors or should I expect other problems? The MOS-drive transistor measures fine...
Thanks,
Mike
PSU:
Trafo: 40Vsec, 500VA
Caps: 2x Epcos Sikorel 10mF, bypassed w 100nF film
Dual rectifiers
Unloaded Voltage = ±55.4V
Load:
8ohm CSS FR125 10cm Fullrange, 7L sealed.
madstamm said:also, what are opinions on replacing the FDP42AN15A0 with an different part, possibly with lower RDSon?
How tolerant is the UCD drive circuit on drive transitor parameters such as gate charge?
Would the developers care to drop a line?
thanks,
mike
You could contact Hypex directly, in my experience they are very responsive. It is indeed a no-go to use the modules without any heatsinking. They do not need much but they do need some.
Best regards
Gertjan
I realize of course that I was not "supposed" to run the modules w/o heatsinking -
it is not a question if I have done the right thing or not, I certainly havent
I am curious if running w/o a heatsink will *typically* cause the output transistors to "melt" or not, and if I should be concerned that other parts have been destroyed as well.
again, i would think that hypex has some experience in thermally destoying their modules, (i did write them an eMail a few min ago), but has anyone else cratered their amps?
it is not a question if I have done the right thing or not, I certainly havent
I am curious if running w/o a heatsink will *typically* cause the output transistors to "melt" or not, and if I should be concerned that other parts have been destroyed as well.
again, i would think that hypex has some experience in thermally destoying their modules, (i did write them an eMail a few min ago), but has anyone else cratered their amps?
They shouldn't blow up if you're running them at lower power levels without a heat sink.
Something else probably happened, especially since your other amp still works fine.
I fried one of my 180 modules due to what I believe was a short from output to ground right at the fast on terminals. I checked some parts and fried more than just the output fets.
Something else probably happened, especially since your other amp still works fine.
I fried one of my 180 modules due to what I believe was a short from output to ground right at the fast on terminals. I checked some parts and fried more than just the output fets.
Hi madstamm,
I would suspect one of two things: 1) a wiring error (I know you'll swear it isn't so, but that's what usually happens to me), or 2) it's a grounding problem. It's easy to get ground loops with two channels being run off one supply.
I've seen both of these cause huge problems when first turning on new amps of all types. I have to admit though, I haven't heard many stories about new amps frying speakers.
Good luck with this.
Graeme
I would suspect one of two things: 1) a wiring error (I know you'll swear it isn't so, but that's what usually happens to me), or 2) it's a grounding problem. It's easy to get ground loops with two channels being run off one supply.
I've seen both of these cause huge problems when first turning on new amps of all types. I have to admit though, I haven't heard many stories about new amps frying speakers.
Good luck with this.
Graeme
thanks for the info regarding ESD modules. I will try to get info from hypex about this issue. Still seems strange to me that it could lead to problems... maybe the ESD pads start to get really conductive at 400kHz??
This is the response I got from Hypex regarding overheating, by the way:
"Overheating could be the case. That is the reason why the manual says you should provide for additional cooling."
...which is pretty much useless... "could" means next to nothing if you ask me...?
This is the response I got from Hypex regarding overheating, by the way:
"Overheating could be the case. That is the reason why the manual says you should provide for additional cooling."
...which is pretty much useless... "could" means next to nothing if you ask me...?
thanks for the info regarding ESD modules. I will try to get info from hypex about this issue. Still seems strange to me that it could lead to problems... maybe the ESD pads start to get really conductive at 400kHz??
This is the response I got from Hypex regarding overheating, by the way:
"Overheating could be the case. That is the reason why the manual says you should provide for additional cooling."
...which is pretty much useless... "could" means next to nothing if you ask me...?
Hi Graeme,
I connected my amps like this:
I did not earth any part of the test-setup, meaning the secondary side was left floating. Looking back this might not have been a good idea:
could not earthing the secondary "0V" net result in a large potential difference between signal and power ground in the amplifier? Or are signal and power grounds in the hypex modules DC-connected at some point?
thanks again for info so far.
mike
This is the response I got from Hypex regarding overheating, by the way:
"Overheating could be the case. That is the reason why the manual says you should provide for additional cooling."
...which is pretty much useless... "could" means next to nothing if you ask me...?
gl said:
Hi Graeme,
I connected my amps like this:
An externally hosted image should be here but it was not working when we last tested it.
I did not earth any part of the test-setup, meaning the secondary side was left floating. Looking back this might not have been a good idea:
could not earthing the secondary "0V" net result in a large potential difference between signal and power ground in the amplifier? Or are signal and power grounds in the hypex modules DC-connected at some point?
thanks again for info so far.
mike
madstamm said:thanks for the info regarding ESD modules. I will try to get info from hypex about this issue. Still seems strange to me that it could lead to problems... maybe the ESD pads start to get really conductive at 400kHz??
This is the response I got from Hypex regarding overheating, by the way:
"Overheating could be the case. That is the reason why the manual says you should provide for additional cooling."
...which is pretty much useless... "could" means next to nothing if you ask me...?
Hi Madstamm,
You could have saved yourself and other people here a lot of trouble just by reading the manual for instance. Just a few lines from the manual:
"...Altough Hypex UcD modules are based on very efficient Class-D technology, additional cooling is a MUST." The label (the big yellow one with the exclamation mark and starting off with 'Caution!') that holds the antistatic bag together says: "...Protection foam is for transportation purposes only. Remove foam before use."
Which part of these phrases are so hard to understand? The manual wasn't written in french now was it?
Maybe you should read the manual before firing things up and have a bit more patience than a kid unwrapping his gifts on christmas eve.
Cheers,
Jorrit
Hypex Electronics
Mr Mozes,
Im sorry to have riled you up - I said at the beginnning of the posts that I realize that I made a mistake. I assumed and the asumption was wrong. I now also realize I didnt read the big yellow label, Im tendentially interested in the present more than the package... and YES, unpacking a neat amp IS like unpacking xmas gifts, would not be good if it was otherwise
Normally the yellow label warns of ESD susceptibility only, so this is what i assumed it to say.
Please do not confuse these posts as an accusation. Im trying to find out WHY the amp broke and WHAT parts might all be affected, not trying to claim a poor product. In contrary, I was very well pleased, for 10min at least
I still do not have an answer to my questions:
1. Are melted output transistors a typical thermal failure
2. What other parts might be in trouble
YOU might save yourself a lot of trouble by thoroughly reading posts (and eMails) as well.
cheers,
Mike
Im sorry to have riled you up - I said at the beginnning of the posts that I realize that I made a mistake. I assumed and the asumption was wrong. I now also realize I didnt read the big yellow label, Im tendentially interested in the present more than the package... and YES, unpacking a neat amp IS like unpacking xmas gifts, would not be good if it was otherwise
Normally the yellow label warns of ESD susceptibility only, so this is what i assumed it to say.
Please do not confuse these posts as an accusation. Im trying to find out WHY the amp broke and WHAT parts might all be affected, not trying to claim a poor product. In contrary, I was very well pleased, for 10min at least
I still do not have an answer to my questions:
1. Are melted output transistors a typical thermal failure
2. What other parts might be in trouble
YOU might save yourself a lot of trouble by thoroughly reading posts (and eMails) as well.
cheers,
Mike
BWRX!
This was a UCD400! It has a significant idle loss, ~8 W as I remember, and this power on such a small surface causes a very high temperature without forced cooling (could be >150 C). Failure was a quite normal consequence.
madstamm!
You can't expect more. They haven't seen your config, so possible alternative reasons of failure cannot be eliminated.
This was a UCD400! It has a significant idle loss, ~8 W as I remember, and this power on such a small surface causes a very high temperature without forced cooling (could be >150 C). Failure was a quite normal consequence.
madstamm!
You can't expect more. They haven't seen your config, so possible alternative reasons of failure cannot be eliminated.
Pafi said:
I reacted because I asked 2 questions and got an overly short response to only one of them. I suppose this is unprofessional conduct on my part, I apologize.
I had the modules running with ±24V in the lab the day before "the experiment", and they got only handwarm during several hours of ON-time. (by the way, this was with the esd pads ON) In the flight of anticipation I disregarded the consequence of doubling the voltage rails (quadrupling the power). The ESD pads were attached during the hours of ON-time in the lab by the way! NOt like Id recommend this, but I think that means in my case ESD pads were probably not the cause of failure.
Im not sure about the 8W melting the output FETs by the way. I'd guess 10º/W heatsink-ambient resistance of the UCD400 heatsink, given still air (very little convection cooling, mostly radiation). This gives (very roughly) 100º heatsink temp, and probably not more than 125º die temp. the datasheet specs operating temp up to 175º.
At die=125º:
max current = 20A
Ron = 1.8x normal value... hmmm this one might do it!
okok, the rising Ron probably killed the FETs. Power dissipation would increase ~proportionally with Ron, which in turn would increase temp again, and so on.
madstamm said:
this might help....uhhhmmm you say you put the modules under an ESD pad?? arent ESD pad/s conductive??? pls test it with an ohmeter.... if it played, and ONLY at low volume settings, I believe you have caused some -shorts- by way of THAT pad...
-OR-
have you been playing with the deadtime trimpot?
ESD pads are conductive, just like any almost other material... but their conduction/resistance is controlled. In post 8 i mentioned "measuring" (if one could call it that the pad. I used two pointy ohmmeter probes and got ~10MegOhm with the tips 5mm apart.
i also ran the modules on the pads for a few hours w/o anything happening. (at low supply voltages)
Im quite sure now that it was in fact the output transistors getting too hot - i did a quick calc and it looks like running the modules w/o an additional heatsink and without any forced cooling generates enough heat to increase the fet temperature, effecting some important parameters adversely. the degrading of the parameters in turn generates more heat, this degrades the fet more, etc etc... runaway thermal bomb, with the clock quite possibly being at 10min.
so idling the modules at recommended PSU voltages will likely kill them. not exactly sure, depends a little on just how good the heatsink radiates...
and no playing with any pots maybe later
the pad. I used two pointy ohmmeter probes and got ~10MegOhm with the tips 5mm apart.i also ran the modules on the pads for a few hours w/o anything happening. (at low supply voltages)
Im quite sure now that it was in fact the output transistors getting too hot - i did a quick calc and it looks like running the modules w/o an additional heatsink and without any forced cooling generates enough heat to increase the fet temperature, effecting some important parameters adversely. the degrading of the parameters in turn generates more heat, this degrades the fet more, etc etc... runaway thermal bomb, with the clock quite possibly being at 10min.
so idling the modules at recommended PSU voltages will likely kill them. not exactly sure, depends a little on just how good the heatsink radiates...
and no playing with any pots
maybe later - Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.