I want to update the nx-Amp doc with any errors etc that have been found, and or any suggestions for tweaking. Please feed free to let me have your comments here.
Thanks
Thanks
Finally!!!! The boards arrived!!!!
They look small! And I ordered a really BIG(HUGE) 5U enclosure.... The BIG Amp one... It will look cool I think... A bit overkill 🙂 But I love overkill.
They look small! And I ordered a really BIG(HUGE) 5U enclosure.... The BIG Amp one... It will look cool I think... A bit overkill 🙂 But I love overkill.
PS: Should I wait for all the tweaks and other upgrades to be published? I don't have too much knowledge of electronics so I depend on the build instructions....
Cheers
Cheers
Sjoerd,
No, you can go ahead and build.
The tweeks (as far as I can tell from feedback received so far) are about improving the docs and making them easier to understand as use.
Yes, the boards are small!
I built 2 protos and have 1 final build I listen to regularly.
The thread will help you get your amps up and working. Good Luck!
No, you can go ahead and build.
The tweeks (as far as I can tell from feedback received so far) are about improving the docs and making them easier to understand as use.
Yes, the boards are small!
I built 2 protos and have 1 final build I listen to regularly.
The thread will help you get your amps up and working. Good Luck!
Last edited:
Hello,
I am going to build sx-amp, and after looking for parts I have 2 questions.
1. Can I use NJW0281/NJW0302 instead of NJW3281/NJW1302, cause I do not find where to buy NJW3281/NJW1302.
2. At online catalogs I find KSA1381ESTU and KSC3503DSTU, does that mean that these KSA1381 are only gain grade E, and KSC3503 only gain grade D? Or I can get any gain grade, so I should buy a handful and match them?
Regards,
Tomas
I am going to build sx-amp, and after looking for parts I have 2 questions.
1. Can I use NJW0281/NJW0302 instead of NJW3281/NJW1302, cause I do not find where to buy NJW3281/NJW1302.
2. At online catalogs I find KSA1381ESTU and KSC3503DSTU, does that mean that these KSA1381 are only gain grade E, and KSC3503 only gain grade D? Or I can get any gain grade, so I should buy a handful and match them?
Regards,
Tomas
1. Yes now NJW0281/NJW0302 are available instead of NJW3281/NJW1302 these are Obsolete now.
2.KSA1381 and KSC3503 with same gain grade is very difficult to get as Bonsai suggested you can use below ones
KSA1220A (this is a PNP)
KSC2690A (NPN)
2.KSA1381 and KSC3503 with same gain grade is very difficult to get as Bonsai suggested you can use below ones
KSA1220A (this is a PNP)
KSC2690A (NPN)
joshvi, a member has the E grade available for a good price.
http://www.diyaudio.com/forums/swap...2sc2911s-ksa1381e-ksc3503e-3.html#post4474208
http://www.diyaudio.com/forums/swap...2sc2911s-ksa1381e-ksc3503e-3.html#post4474208
Hi Guys
For rails less than 160V total, the KSA1220AYS - KSC2690AYS pair is the same gain range and inexpensive from Digikey and Mouser. Both have hfe>100.
I had to buy a huge quantity to get the 2SA1381 and 2SC3503 both in the same range, as both D. The pair above you can buy in single or whatever you need.
Have fun
For rails less than 160V total, the KSA1220AYS - KSC2690AYS pair is the same gain range and inexpensive from Digikey and Mouser. Both have hfe>100.
I had to buy a huge quantity to get the 2SA1381 and 2SC3503 both in the same range, as both D. The pair above you can buy in single or whatever you need.
Have fun
SX boards from Jims Audio arrived.
Will take some masurements and find what smd package that fits.
After checking and correcting the PSU+Prot board with ±35Vdc, I powered it up with ±49Vdc. Many of the resistors on this PCB run very hot.
I will keep these values for the duration of the tests still to follow, but I will change them later.
Even at ±35Vdc many resistors run very warm. The currents chosen for the LEDs are quite high and could easily be lowered without affecting circuit performance.
The first Power ON indicating LEDs are fed with 5k6. = 5.9mA @ ±35Vdc and = 8.4mA @ ±49Vdc
I like to use an added Zener for these to indicate OFF, or very dim, if the supply voltage falls well below normal.
Adding a 20V 400mW Zener in series reduces the 5.9mA to 2.3mA @ ±35Vdc and the LED is sufficiently bright to act as an effective indicator.
I will add a 33V 400mW Zener for my nominal ±49Vdc supply. That reduces the current to 2.5mA
This massively reduces the heat in the two 5k6 resistors. at a worst case ±51Vdc the resistor current becomes 2.9mA and dissipation is down to 47mW. Without the added Zener this worst case is 432mW The 33V Zener operates at <24% of rating.
If you don't like the idea of using Zeners, then I suggest you adopt higher resistance values for the LED current limiters.
I will keep these values for the duration of the tests still to follow, but I will change them later.
Even at ±35Vdc many resistors run very warm. The currents chosen for the LEDs are quite high and could easily be lowered without affecting circuit performance.
The first Power ON indicating LEDs are fed with 5k6. = 5.9mA @ ±35Vdc and = 8.4mA @ ±49Vdc
I like to use an added Zener for these to indicate OFF, or very dim, if the supply voltage falls well below normal.
Adding a 20V 400mW Zener in series reduces the 5.9mA to 2.3mA @ ±35Vdc and the LED is sufficiently bright to act as an effective indicator.
I will add a 33V 400mW Zener for my nominal ±49Vdc supply. That reduces the current to 2.5mA
This massively reduces the heat in the two 5k6 resistors. at a worst case ±51Vdc the resistor current becomes 2.9mA and dissipation is down to 47mW. Without the added Zener this worst case is 432mW The 33V Zener operates at <24% of rating.
If you don't like the idea of using Zeners, then I suggest you adopt higher resistance values for the LED current limiters.
Last edited:
I ran the PSU+AMP pcbs last night.
My 2sa1220y/c2960y have a quiescent operating hFE of 284 (Ib=0.13mA, Ic =37mA)
The outputs are operating with hFE=141 for the NPN (upper) and 161 for the PNP (lower).
I bought over a hundred of the a1381/c3503 in D and E grades. I got 1pr closely matched for hFE and 1pr close enough. Just enough for a stereo NX using the 1220/2690 as drivers.
2sa1381e hFE varies from 120 to 146
ksa1381e varies from 125 to 138
2sc3503e varies from 139 to 158
ksc3503e varies from 156 to 176
ksc3503d varies from 82 to 97
That is over £47 for a yield of four (2pairs) of To126 transistors.
AndrewT;4516994 That is over £47 for a yield of four (2pairs) of To126 transistors.[/QUOTE said:
/U.
Warm means to me feeling warm when I touch them.
Hot means hot when I touch them and that is probably a touch hotter than very warm.
Too hot means they burn my finger within a couple of seconds when I touch them. Very subjective, but that works for me.
Zeners are a very cheap solution that works if one wants an indicator that turns OFF when PSU voltage drops substantially.
Otherwise just use my last suggestion.
Are you going to respond to my private comments/observations on U1?
Hot means hot when I touch them and that is probably a touch hotter than very warm.
Too hot means they burn my finger within a couple of seconds when I touch them. Very subjective, but that works for me.
Zeners are a very cheap solution that works if one wants an indicator that turns OFF when PSU voltage drops substantially.
Otherwise just use my last suggestion.
Are you going to respond to my private comments/observations on U1?
Very warm is ok. Resistors run 'very warm' in millions of end products all over the world.
However, if you don't feel comfortable about very warm resistors, up the resistor to 12 or 15k.
However, if you don't feel comfortable about very warm resistors, up the resistor to 12 or 15k.
Last edited:
It's unfortunate that these devices no longer come in matched gsin ranges. I was very lucky a few years ago and got a few matched gain pairs.
It's clear that as a community we are going to have to find some viable, low Cob alternatives. NXP do a SOT223 device BF67x (IIRC) that may offer a general solution - but that would require a change to the board layout.