Hi,
I saw years ago power amps with the 4 2sk135's & 4 j50's without the source resistors.
Can I do this with 4 2sk1058's & 4 2sj162's ? I just tried it using one 1058 & one j162 & it sounded Ok.
thanks.
I saw years ago power amps with the 4 2sk135's & 4 j50's without the source resistors.
Can I do this with 4 2sk1058's & 4 2sj162's ? I just tried it using one 1058 & one j162 & it sounded Ok.
thanks.
You can only do that with lateral MOSFET's.
No source resistors will cause thermal runaway in vertical FETs.
You will also get an imbalance with vertical mosfets with source resistors.
No source resistors will cause thermal runaway in vertical FETs.
You will also get an imbalance with vertical mosfets with source resistors.
Last edited:
Thanks,
But I wonder if anybody has tried using 4 pairs of 2sk1058/j162 without the source resistors?
I just built the classic 1977 NS/Hitachi using a pair but my friend wants to go to 100W, by using 4 pairs, but my PCB design didn't anticipate that!
I used 2n5402, 2sd669/b645 ( even mje340/350 here & it works fine!) instead of the recommended 2sa872, 2d756/b716 which are hard to find where I am.
& I wonder why amateurs aren't starting to build with this simple but very reliable topology.
thanks again.
But I wonder if anybody has tried using 4 pairs of 2sk1058/j162 without the source resistors?
I just built the classic 1977 NS/Hitachi using a pair but my friend wants to go to 100W, by using 4 pairs, but my PCB design didn't anticipate that!
I used 2n5402, 2sd669/b645 ( even mje340/350 here & it works fine!) instead of the recommended 2sa872, 2d756/b716 which are hard to find where I am.
& I wonder why amateurs aren't starting to build with this simple but very reliable topology.
thanks again.
I used matched triplets of K1058 without source resistors, 1,6A bias per each transistor. The only problem, one need to by much more pcs for selection, I would say three-four times more than needed amount.
Its wise to always use source resistors, even with laterals. minimum of 0.1ohm, but no more than 0.47ohm or you start eating into the damping factor.
I have five lateral mosfet ampliiers that I use for PA. So when they are used they are used hard!
The biggest has 8 pairs of Hitachi 2sk135 / 2sj50 per side. C-audio SR707
This amp did fail but was seriously over voltage, and had been used as a 4 ohm bridge producing 2.4kW.
http://www.diyaudio.com/forums/solid-state/150675-maximum-voltage-mosfet-output-stage.html
The next biggest has 4 pairs of 2sk1058/ 2sj162 per side. Rauch DVT 250s. 550W /4ohm per channel.
None of my amps have source resistors, and they are (now) reliable!
All of them are based on the hitachi app note circuit, but have an added driver stage for the fets.
Lateral mosfet amps are not prone to thermal runaway, and the mosfets are quite robust.
I know from experience that when the cooling fan has failed during a gig, and the amp has completely overheated, that there is no thermal runaway. The amplifier limits itself and sounds fuzzy but it does not die! 😱
The logic is quite simple - The device that is drawing the most current heats up, and turns itself down balancing the load. I would recommend matching your output devices to minimise this. Usually if I am rebuilding (recapping & tweeks mostly) I measure all the fets and group them to minimise the imbalance.🙂
I cant see you will have any problems with 2 pairs per side.
The biggest has 8 pairs of Hitachi 2sk135 / 2sj50 per side. C-audio SR707
This amp did fail but was seriously over voltage, and had been used as a 4 ohm bridge producing 2.4kW.
http://www.diyaudio.com/forums/solid-state/150675-maximum-voltage-mosfet-output-stage.html
The next biggest has 4 pairs of 2sk1058/ 2sj162 per side. Rauch DVT 250s. 550W /4ohm per channel.
None of my amps have source resistors, and they are (now) reliable!
All of them are based on the hitachi app note circuit, but have an added driver stage for the fets.
Lateral mosfet amps are not prone to thermal runaway, and the mosfets are quite robust.
I know from experience that when the cooling fan has failed during a gig, and the amp has completely overheated, that there is no thermal runaway. The amplifier limits itself and sounds fuzzy but it does not die! 😱
The logic is quite simple - The device that is drawing the most current heats up, and turns itself down balancing the load. I would recommend matching your output devices to minimise this. Usually if I am rebuilding (recapping & tweeks mostly) I measure all the fets and group them to minimise the imbalance.🙂
I cant see you will have any problems with 2 pairs per side.
agreed.
I bought 16pairs and got seven very good quads from that. The 8th quad had 3closely matched and the fourth was a few mVgs from the three.
Installed in the amplifier that poor quad causes the low Vgs device to run at about 10% more Id, generating ~10% more Pd. Yes, that device runs hotter and that is with 0r1 source resistors. I have not tried shorting out the Rs to see how bad the imbalance gets.
Good matching cannot perform miracles, but it helps even out the currents and the heats.
No matching and/or no source resistors is asking for trouble.
BTW, I matched the source resistors into quads as well.
Last edited:
Thanks guys!
I guess I have to find a way to put those source resistors cuz my PCB design doesn't accommodate them.
Did you guys used the recommended 2sd576 & 2sb716, & the 2sa782? I can't find them.
I guess I have to find a way to put those source resistors cuz my PCB design doesn't accommodate them.
Did you guys used the recommended 2sd576 & 2sb716, & the 2sa782? I can't find them.
- Status
- Not open for further replies.