Been buying from this place before. Good store but seems they don’t have 1943/5200 trannies.
I saw Digikey and Mouser has. Will be ordering from there i think.
Thx anyway
The junctions are in fact the diode drop over each transistor.
So pre drivers, drivers and outputs are six drops in total.
I would have aimed for the larger output transistors. 1943/5200 may have a tough job at +/- 75 volts and low impedance.
/Figge
So pre drivers, drivers and outputs are six drops in total.
I would have aimed for the larger output transistors. 1943/5200 may have a tough job at +/- 75 volts and low impedance.
/Figge
where Q320 IS LISTED, Q329 is ment. Typo , misprint or scanerror.
Uneven numbers are the left channel.
Elaborate (left ch, top2btm, "//" is parallel):
Q325 - (Q329 - Q337//Q341)//(Q333 - Q345//Q349): 3 junctions
(Q351//Q347 - Q335)//(Q343//Q339 - Q331) - Q327: 3 junctions
3+3=6
Total bias voltage is some ≥ 4Vdc expected. (Calculation of bias net needed.)
Regulated +/-82Vdc is for front stages, neglectable.
I am wrong! The Pd 54W total is ok, but not the classcalc.
Say every branch 50mA, 4 branches per channel. Yields 200mAdc
Swing can be from +200 mA to -200 mA, 400mA toptop total.
Iac is approx 1/3 of toptop value (sinuswave for simplicity), 133 mA.
PclassA = Iexp2 * Zload -> Zload = 8Ω to start with.
(0.133)exp2 * 8 = 0.14W
You can do the bridge mode (twice the toptop value) calculation yourself.
But in the end it proofs to yank up the idle to a max permitted value to prevent nasty switching of bjt's. (Despite more or less succesfull solutions.)
I'd prefer for this reason only classA amps (Iset ≥ 1A); think Hiraga or Pass.
As a rule of thumb, as long as I can keep my hand on the central heating, it's fine. I do not know the NAD designer nor the claim 'n proof.
Good. It is a solid design, Krell-like. If I had to live with an AB-amp, such was the topology.
I'm currently & still developping a class A amp around these 5200-1943 bjt's in an alternative bridged Hiraga with 4-5A idle per side, some 25WA@8. Pdiss is 300W -- not very green...
Last edited:
When you are already inside the amp, its a good idea to increase the heatsinking of the series elements for the +- 82v powersupply to save the pcb and the nearby capacitors cause it gets more than usual hot.
Quoting myself is not customary, but I'm wrong again...
I haven't done these calc's lately, so after posting I didn't sleep well until I woke up screaming what I'm quoting now (not customary either).
Vset TP301-303 = 8mV +/- 2mV; Re = 0.22Ω
Iidle = 8mV/0.22= 36mA
4 branches yields 144mA, say 0.15A
Top value before other side switches off (lineair Vbe curve for simplicity) is 0.3A. AC current is (sinus for simplicity) 0.7 * 0.3 = 0.21A
Pa = Iexp2 * ZL (8Ω) = (0.21)exp2 * 8 = 0.35W
This is the same for bridge mode, because it only counts when bjt's switch of running in B mode. In bridge there is a larger voltage swing, but bjt-switching-off is at the same currents.
To prevent to much consecutive posts as were it wirings on social platforms, the bias calc is as follows:
Q323: 2 * Vbe = 1.3V approx, continuous value
Q321: (680+220+100)/(220+100)*Vbe = 1000/320*0.7 = 2.2V min
Q321: (680+220)/(220)*Vbe = 900/220*0.7 = 2.9V max
Total Vbias range Q323 + Q321 = 3.5 -- 4.2 Vdc (expected was 4V)
.
Thankx for all info and calculations. Quite much tech stuff. Have to dig into it. Math is not my strong suit.
the main powersupply is +-72/76v. I mean the regulator +-82v more precise the two transistors on a heatsink in the bottom left side of the cabinet 2sa1837 and2sc4793. They need better heatsinking.🙂
When I talk central heating, I'm mentioning the grill (aka heat sink) with the end bjt's.
There on the unregulated +/- 75Vdc rails.
pdul talks about the small heat sinks for the regulated +/-82V rails: Q701 & Q702.
There on the unregulated +/- 75Vdc rails.
pdul talks about the small heat sinks for the regulated +/-82V rails: Q701 & Q702.
Ok, any idea where to find a bigger heatsink? Thinking also rebuilding the whole heater. I have already components for that here except grill.
Regulators for lower current sections (preamp, regulated VAS, etc) often get put on heat sinks that are barely good enough for the dissipation, and get run that hot 24/7. Hence, all the discolored PCBs and oxidized solder joints you often see in vintage gear. If there is a place to put a bigger heat sink on it, by all means do it.
Have checked fleabay and Aliexpress and don’t find any heatsink that supports 2 trannies in that size. Can only be higher. Packed with components around. Thinking of extending original sink with some copper fins/plates.
Whatabout the main psu caps, they seem a bit on the small side considering ”big amp”... footprint of them is 30x52mm so any bigger wont fit, for example 5600uf/100v is 35mm
Whatabout the main psu caps, they seem a bit on the small side considering ”big amp”... footprint of them is 30x52mm so any bigger wont fit, for example 5600uf/100v is 35mm
The +/-17v heatsinks are quite small they get hot as heck including 2pcs 1k/1w fusible resistors.
Also the other heater have 4x100R /1w fusible too and you can boil egg on those.
A1493/C3788 that are sitting pushed down on each channel doesn’t have a heatsink at all and can’t barely touch them with my fingers.
What was Nad thinking? The S200 looks the same
Also the other heater have 4x100R /1w fusible too and you can boil egg on those.
A1493/C3788 that are sitting pushed down on each channel doesn’t have a heatsink at all and can’t barely touch them with my fingers.
What was Nad thinking? The S200 looks the same