Jfets Id again
I'm powering up my F5T v2 with cascode. Jfets were purchased as matched BL grade from diyaudiostore.
On one channel I had to increase R5/R6 to 2k2 to be able to bias properly.
The second channel required even higher resistance - 4k7, but finally worked.
What worries me is unequal measurement of Id on that board. I'm measuring 35mV on R3 and 40mV on R4. The P3 was precisely set to have <5-10 mOhms difference on R3/R4.
I was able to bias it to 0.9A per output pair with DC about 5mV.
Should I worry then about Id difference?
I'm powering up my F5T v2 with cascode. Jfets were purchased as matched BL grade from diyaudiostore.
On one channel I had to increase R5/R6 to 2k2 to be able to bias properly.
The second channel required even higher resistance - 4k7, but finally worked.
What worries me is unequal measurement of Id on that board. I'm measuring 35mV on R3 and 40mV on R4. The P3 was precisely set to have <5-10 mOhms difference on R3/R4.
I was able to bias it to 0.9A per output pair with DC about 5mV.
Should I worry then about Id difference?
Hi, I am building F5-turbo monoblocks. I would like to test them out one at a time using +25V and -25V rails from two meanwell 24V supplies GST120A24-P1M. I have the supplies from my ACA amps. Is the F5-turbo biased specifically for +-36v rails? Will running them at 25v lead to anything suboptimal except lower power? Do I need to change the circuit if I run it from +-50V rails someday?
No. You are the person that will bias the amp and can bias it anywhere you like.
No.
Yes. You will need to re-evaluate the cascode setup and add more output device pairs to handle the additional dissipation.
I normally use a very low current supply, with 1A fuses just in case. My first bias point will be 100mA, and I play for an hour or so with test speakers to ensure the amp is working.
The F5T is a powerful beast and mistakes can be expensive.
The reason I know this is because I made many mistakes myself while building, destroyed lots of expensive components so I now build with some safety margin.
Bias is really just based on the constant current your output devices see. It doesn’t care what the rail voltage is. What is your power supply rated for? I wouldn’t be surprised if your existing PS is not up for the task in general if it came from an ACA. I believe you can cascode your jfets even if you use a lower voltage PS. If this is the case, and you plan on going higher voltage later, you should cascode. 25V rails seems very low for F5TV3 monos. It is a very expensive build. My opinion, spend the extra money and put a beefier PS in.
Fried MOSFET and interior redesign...
May bit lengthy, but I want to share my experience and hopefully someone will learn from my mistakes...
I have finished chassis assembly of my F5T cascoded v2. Pictures will be provided later as I have to redesign interior... in particular move transformer to the front to fight against AC mains noise.
Configuration: stereo F5Tv2, DIYA PSU, Speaker protection and soft start. 800VA 2x24AC transformer. 2x4 33,000uF PSU caps. 5Ux400 modushop chassis.
Bias set to 0.75 per device (0V370 across source .5R), amp was running in stable manner for a couple of days. The right channel had audible noise hum, though. I had to be really close to loudspeaker to hear it, but the left one was deaf silent. Another two days of scratching my head brought solution - the "active" loudspeaker wire of "problematic" channel was too close to transformer (located near the back panel). Moreover the wire was let alone without its ground brother twisted, what would help cancelling current induction. It's my first build... I learnt a lot so far and gained the respect as well.
I did some measurements with sound card (pretty decent Focusrite Claret) + FFT SW (RightMarkAudio and REW). THD was quite ok ~0.007, but overall noise poor - 50Hz and its harmonics at -60dB. It's unacceptable and has to be fixed, but I need to fix the amp first.
Yes, it broke
It happened when I was replugging RCA inputs (amp on):
1. The speaker protection engaged and its led kept on blinking.
2. Waited couple of secs and turn amp off.
3. To check which channel is causing this I disconnected DC power rails from left channel and turned it on again. This time speaker protection went off after couple of seconds. OK - the right channel was healthy.
4. Turned the amp off and reconnected left channel DC and disconnected the right one.
5. After turning amp on I had sparks and smoke from IRFP240 Mosfet.
Now problem debug and damage evaluation.
- both IRFP9240 are gone - GDS perfectly shortened, visually not damaged
- fried one 240.
- JFETs and rest of the stuff OK
Here is what I think what happened:
1. Replug of RCA input damaged (GS shortened) both 9240. Why????
2. Speaker protection detected DC and disconnected load from ground and whole -V current started flowing thru 240, but they survived.
3. When I disconnected other channel the PSU doubled it's capability focusing on one channel with shortened 9240 and this must have been too much for one of 240 which ended it's life in a "sparcstacular" way.
4. AC fuse didn't blow - 4A (230VAC) is definitely too high. BTW: dual mono designs are better in this aspect as AC fuse value can be halved giving faster response for overloaded single channel.
DIYAudioStore is out of stock of matched MOSFETs so I did matching myself. I matched devices mounted on small heatsink @0.8A and 45oC. Found 2 P and N pairs with 1-3 mV VGS variation. I hope it's good enough - please comment.
I'd like to understand why RCA replug (Focusrite disconnect and connection of my preamp - that os also DC style) could've caused immediate damage?
Any clues?
Also my respect to the 200W heating technology powered up by 800VA transformer has risen significantly
and I want to take all security measures to prevent from (or at least limit) unexpected disastrous effects. I studied this thread carefully and that's what I fished:
1. I'll increase 47 mosfet gate resistors to 200.
2. 3n9 polypropylene caps across R7/8 feedback resistors
3. 100 ohm Jfet drain resistors before cascode. That's not pretty as requires "aerial" mount.
4. Fuse DC rails conservatively. 4 individual fuses to be able to use the lowest possible fuse value. Any downsides of this solution - in particular when either + or - V is disconnected by blown fuse loosing the V symmetry?
To be clear: I don't think I had oscillation issue, just being a paranoid.
Did I miss something?
And again- what could be the reason of the damage, that definitely happed during RCA input connection?
To be continued...
May bit lengthy, but I want to share my experience and hopefully someone will learn from my mistakes...
I have finished chassis assembly of my F5T cascoded v2. Pictures will be provided later as I have to redesign interior... in particular move transformer to the front to fight against AC mains noise.
Configuration: stereo F5Tv2, DIYA PSU, Speaker protection and soft start. 800VA 2x24AC transformer. 2x4 33,000uF PSU caps. 5Ux400 modushop chassis.
Bias set to 0.75 per device (0V370 across source .5R), amp was running in stable manner for a couple of days. The right channel had audible noise hum, though. I had to be really close to loudspeaker to hear it, but the left one was deaf silent. Another two days of scratching my head brought solution - the "active" loudspeaker wire of "problematic" channel was too close to transformer (located near the back panel). Moreover the wire was let alone without its ground brother twisted, what would help cancelling current induction. It's my first build... I learnt a lot so far and gained the respect as well.
I did some measurements with sound card (pretty decent Focusrite Claret) + FFT SW (RightMarkAudio and REW). THD was quite ok ~0.007, but overall noise poor - 50Hz and its harmonics at -60dB. It's unacceptable and has to be fixed, but I need to fix the amp first.
Yes, it broke
It happened when I was replugging RCA inputs (amp on):
1. The speaker protection engaged and its led kept on blinking.
2. Waited couple of secs and turn amp off.
3. To check which channel is causing this I disconnected DC power rails from left channel and turned it on again. This time speaker protection went off after couple of seconds. OK - the right channel was healthy.
4. Turned the amp off and reconnected left channel DC and disconnected the right one.
5. After turning amp on I had sparks and smoke from IRFP240 Mosfet.
Now problem debug and damage evaluation.
- both IRFP9240 are gone - GDS perfectly shortened, visually not damaged
- fried one 240.
- JFETs and rest of the stuff OK
Here is what I think what happened:
1. Replug of RCA input damaged (GS shortened) both 9240. Why????
2. Speaker protection detected DC and disconnected load from ground and whole -V current started flowing thru 240, but they survived.
3. When I disconnected other channel the PSU doubled it's capability focusing on one channel with shortened 9240 and this must have been too much for one of 240 which ended it's life in a "sparcstacular" way.
4. AC fuse didn't blow - 4A (230VAC) is definitely too high. BTW: dual mono designs are better in this aspect as AC fuse value can be halved giving faster response for overloaded single channel.
DIYAudioStore is out of stock of matched MOSFETs so I did matching myself. I matched devices mounted on small heatsink @0.8A and 45oC. Found 2 P and N pairs with 1-3 mV VGS variation. I hope it's good enough - please comment.
I'd like to understand why RCA replug (Focusrite disconnect and connection of my preamp - that os also DC style) could've caused immediate damage?
Any clues?
Also my respect to the 200W heating technology powered up by 800VA transformer has risen significantly
and I want to take all security measures to prevent from (or at least limit) unexpected disastrous effects. I studied this thread carefully and that's what I fished:1. I'll increase 47 mosfet gate resistors to 200.
2. 3n9 polypropylene caps across R7/8 feedback resistors
3. 100 ohm Jfet drain resistors before cascode. That's not pretty as requires "aerial" mount.
4. Fuse DC rails conservatively. 4 individual fuses to be able to use the lowest possible fuse value. Any downsides of this solution - in particular when either + or - V is disconnected by blown fuse loosing the V symmetry?
To be clear: I don't think I had oscillation issue, just being a paranoid.
Did I miss something?
And again- what could be the reason of the damage, that definitely happed during RCA input connection?
To be continued...
Last edited:
that's simply no-no ....... plugging and unplugging with amp ON
even with much less sensitive (not fragile, sensitive in meaning of responsive) amps , you simply don''t do that
it's the same as you trying to seat in your car while your lovely wifie is driving it , at least 30mph
go figure
from same reasons , you have powering up sequence - source , preamp , amp ......... and powering off sequence - amp , preamp , source
broken signal chain (and RCA is first breaking GND connection, leaving floating hot) is often resulting either in burst or oscillation or both , it doesn't matter did you heard it or not
even with much less sensitive (not fragile, sensitive in meaning of responsive) amps , you simply don''t do that
it's the same as you trying to seat in your car while your lovely wifie is driving it , at least 30mph
go figure
from same reasons , you have powering up sequence - source , preamp , amp ......... and powering off sequence - amp , preamp , source
broken signal chain (and RCA is first breaking GND connection, leaving floating hot) is often resulting either in burst or oscillation or both , it doesn't matter did you heard it or not
Last edited:
It makes a perfect sense especially your note about connecting hot end of RCA first.
I'm wondering whether protection transistors Q5, Q6 from F5 would help to prevent from such failures.
What do you think about fusing DC rails. I searched this thread and it looks like nobody is doing this... Why?
(a) beacuse it's useless as even the fastest fuse is too slow to prevent mosfets
(b) it has some drawbacks
(c) both (a) and (b)
I'm wondering whether protection transistors Q5, Q6 from F5 would help to prevent from such failures.
What do you think about fusing DC rails. I searched this thread and it looks like nobody is doing this... Why?
(a) beacuse it's useless as even the fastest fuse is too slow to prevent mosfets
(b) it has some drawbacks
(c) both (a) and (b)
After thinking about my amplifier issues for a LONG time, I have an idea that might help.
I wired my F5T with the output boards for each polarity in parallel.
So, the +'ve pair is in parallel and the -'ve pair are in parallel for each channel.
Does this mean that the anti oscillation capacitor on the output boards ( normally one per polarity) should be duplicated?
Meaning parallel output boards should have one capacitor per output board????
I really don't know if this would be helpful/not. Could anyone comment please? I'd like to get this finished!
I wired my F5T with the output boards for each polarity in parallel.
So, the +'ve pair is in parallel and the -'ve pair are in parallel for each channel.
Does this mean that the anti oscillation capacitor on the output boards ( normally one per polarity) should be duplicated?
Meaning parallel output boards should have one capacitor per output board????
I really don't know if this would be helpful/not. Could anyone comment please? I'd like to get this finished!
Last edited:
to Maciej #5933
Hello Maciej,
as Zen Mod wrote it in #5932 - never connect/disconnect any cables as long
your amp is in on state!
That is a `golden rule`!
I think no cap or fuse (are they fast enough?) will protect you from damage to your amp if you brake that `golden rule`.
And if you switch an amp off - wait a short while till you turn it on again.
Also if you use a soft start or thermistor in your AC input to transformer.
I hope your speakers survived....
Greets
Dirk
Hello Maciej,
as Zen Mod wrote it in #5932 - never connect/disconnect any cables as long
your amp is in on state!
That is a `golden rule`!
I think no cap or fuse (are they fast enough?) will protect you from damage to your amp if you brake that `golden rule`.
And if you switch an amp off - wait a short while till you turn it on again.
Also if you use a soft start or thermistor in your AC input to transformer.
I hope your speakers survived....
Greets
Dirk
Yup. Learning golden rules costs and gold is precious
My amp is back up-and-running. I replaced mosfets, added caps and fuses - I agree that thease measures may not protect from issues like this, but at least help limiting the damage.
I had dummy load connected for measurements when failure happened, so no harm was made. I would be mad if my Quadral Vulkan suffered a damage... TL is a rare design nowadays.
My amp is back up-and-running. I replaced mosfets, added caps and fuses - I agree that thease measures may not protect from issues like this, but at least help limiting the damage.
I had dummy load connected for measurements when failure happened, so no harm was made. I would be mad if my Quadral Vulkan suffered a damage... TL is a rare design nowadays.
I'm going to build F5-T monoblocks (<=100W class A - each). I have a couple of questions in terms of parts supply.
- I would like to use bridge rectifiers instead of diodes - is VS-GBPC3510A good enough or should i buy something else ? VS-GBPC3510A Vishay Semiconductors | Mouser Polska
- Mosfets - is tubestore on ebay reliable source ? There is no specified producer - it might be Vishay or IR. Someone mentioned before that IR and Vishay are same. I would like to buy 2 sets 8+8 - 8 PIECES MATCHED IRFP240 IRFP9240 MOSFET FOR PASS F5 TURBO AMPLIFIER | eBay - is it good idea or should i buy those somewhere else ?
- 2SC4793/2SA1837 - I found original toshiba factory matched pairs in retro audio store, but of course used. Is there any problem with that ? These are not cheap so i would like to have them if it's really good idea only. If not - can you provide me place where i can order new ones ?
- Also i would like to go with 24+24 toroidal transformer 800VA (per monoblock) - is it enough ? Or better would be to buy something bigger like 1000VA/higher voltage ?
Thanks
- I would like to use bridge rectifiers instead of diodes - is VS-GBPC3510A good enough or should i buy something else ? VS-GBPC3510A Vishay Semiconductors | Mouser Polska
- Mosfets - is tubestore on ebay reliable source ? There is no specified producer - it might be Vishay or IR. Someone mentioned before that IR and Vishay are same. I would like to buy 2 sets 8+8 - 8 PIECES MATCHED IRFP240 IRFP9240 MOSFET FOR PASS F5 TURBO AMPLIFIER | eBay - is it good idea or should i buy those somewhere else ?
- 2SC4793/2SA1837 - I found original toshiba factory matched pairs in retro audio store, but of course used. Is there any problem with that ? These are not cheap so i would like to have them if it's really good idea only. If not - can you provide me place where i can order new ones ?
- Also i would like to go with 24+24 toroidal transformer 800VA (per monoblock) - is it enough ? Or better would be to buy something bigger like 1000VA/higher voltage ?
Thanks
Last edited:
bridges are OK
dunno for that fleabay store , maybe someone had experience
for cascodes you don't need actual types (Papa is certainly having them in tons, so why not using them) - you can use BD139/BD140 there without problems , taking care of possible pinout differences (didn't checked)
dunno for that fleabay store , maybe someone had experience
for cascodes you don't need actual types (Papa is certainly having them in tons, so why not using them) - you can use BD139/BD140 there without problems , taking care of possible pinout differences (didn't checked)