When you get to PSU BOM include separate fuse and holder for each transformer. A fuse big enough for the power transformer may not blow if there is a fault in the aux transformer(s).
At which current level ?
Suppose Vgs of an average device is 4V at your chosen quiescent current level, then 3% off makes a difference of 120mV.
Without a source resistor, a jump of 120mV would make a change of drain current in the order of 100-150mA (see Vgs-Id graph)
The source resistor provides feedback, the current through it induces a voltage difference that lowers Vgs : lower Vgs, lower drain current.
At drain current levels of around 0.5A, transconductance is rather low, 1 to 1.5 ( see Id-Gfs graph)
In effect, the drain current difference is reduced to 80-100mA. (wet finger estimate)
The higher the Gfs number, the lower the feedback effect, reason for using higher value source resistors with high transconductance devices.
0.5A or 0.6A drain current also makes a 20% difference in dissipation.
Fortunately the heatsink provides thermal feedback by lifting the temperature of the output devices up by an identical number.
(a reason for higher thermal stability of class A amps)
With a heatsink at 50C and mica insulators, temperature difference of the dies would be in the order of 4 degrees C.
Means that 3% spread in Vgs leads to 5-6% difference of die temperature.
Unfortunately, it does not stop there.
Going from 25C to 150C, at a Vgs of 4.5, drain current goes up from 1A to ~2.5A, makes 12mA/C.
4 degrees temperature difference of the die would make Id go up again by 48mA.
Fortunately transconductance goes down, if die temperature goes up.
A lower transconductance level makes the feedback of the source resistor more effective.
Still, all in all, a Vgs spread of 3% leads to a much higher Id and dissipation difference.
A tire that spins 10% faster than the other not only wears out faster, it also leads to lower road handling.
Oh, missed the psu section. I think you want MUR2020RGOS-ND unless the PSU board has changed to accept dual rectifiers. I bought them at least 5 years ago.
The other thing with my PSU boards is that they were designed for a diode with pinouts opposite the MUR2020RGOS. My heatsinks have an open spot near the bottom which allowed me to reverse the mounting of the MUR2020 since I'd already assembled them and was checking when I noticed the issue. You can also mount the caps opposite indicated polarity and reverse the output connections unless you are using 4 pin caps. I've found appropriate diodes but can't remember the part number. Jim?
If you put a jumper in R3/4 you will eliminate your R in the CRC. Use another 0R47 there.
Buy a few values for R0 if you haven't got a parts bin going yet. My guess is 4K75 will be too bright, but YMMV. Suitable alternative values may be 10K, 20K. Smaller changes are likely to be barely noticeable.
Jacco's comments are important. If your devices are matched at lower currents and you go with them, pick the best matched source resistors in triplets from yout supply and then measure the voltage across them in the amp - look for tightly matched voltage indicating closely matched devices.
The other thing with my PSU boards is that they were designed for a diode with pinouts opposite the MUR2020RGOS. My heatsinks have an open spot near the bottom which allowed me to reverse the mounting of the MUR2020 since I'd already assembled them and was checking when I noticed the issue. You can also mount the caps opposite indicated polarity and reverse the output connections unless you are using 4 pin caps. I've found appropriate diodes but can't remember the part number. Jim?
If you put a jumper in R3/4 you will eliminate your R in the CRC. Use another 0R47 there.
Buy a few values for R0 if you haven't got a parts bin going yet. My guess is 4K75 will be too bright, but YMMV. Suitable alternative values may be 10K, 20K. Smaller changes are likely to be barely noticeable.
Jacco's comments are important. If your devices are matched at lower currents and you go with them, pick the best matched source resistors in triplets from yout supply and then measure the voltage across them in the amp - look for tightly matched voltage indicating closely matched devices.
That's right, the board only accept single Diodes, I've corrected my BOM accordingly.
I checked the traces on the pcb, and you're right, if I trust the MUR2020RGOS-ND datasheet, cathode will lead to ground and anode to capa positive, however, the sinks in the BOM seems to have this little slot on the bottom witch will let me bend the diodes legs to mount them in the right polarity (if the legs are long enough).
But if you find out the reference of a compatible diode, I will gladly change the BOM.
-For R0 (on the PSU), I've added 10K and 20K 3w resistor, I'll try them whem I'll be in the testing phase.
-For the pairing of the IRFP240, I honestly don't know at witch VGS they where paired, I think they probably been paired under 25V rails as the pack is meant to build an Aleph 30 Full Set of Matched Semiconductors for Pass Aleph 30 Class A Power Amplifier | eBay
I've chosen this solution because I was a little afraid to engage the pairing myself (lack of knowledge, and the feeling that it will cost me more to buy a full pack of 240 and 9610 to find pairs) but I probably should have taken the risk, my bad, however, it will probably do with what i'd get...or not😱
So I'll buy a full pack of 50 0R47 3W resistors to be able to find 4 ideal triplets for the source resistors.
-For over-current protection I already have some 3.15A 5x20 slow blow fuses for the primary (600VA under 230V) and 15A 5x20 slow blow fuses for the secondaries (24V, 12.5A) with corresponding holder (5A chassis type holder and 2x 20A on wire type holders)
I checked the traces on the pcb, and you're right, if I trust the MUR2020RGOS-ND datasheet, cathode will lead to ground and anode to capa positive, however, the sinks in the BOM seems to have this little slot on the bottom witch will let me bend the diodes legs to mount them in the right polarity (if the legs are long enough).
But if you find out the reference of a compatible diode, I will gladly change the BOM.
-For R0 (on the PSU), I've added 10K and 20K 3w resistor, I'll try them whem I'll be in the testing phase.
-For the pairing of the IRFP240, I honestly don't know at witch VGS they where paired, I think they probably been paired under 25V rails as the pack is meant to build an Aleph 30 Full Set of Matched Semiconductors for Pass Aleph 30 Class A Power Amplifier | eBay
I've chosen this solution because I was a little afraid to engage the pairing myself (lack of knowledge, and the feeling that it will cost me more to buy a full pack of 240 and 9610 to find pairs) but I probably should have taken the risk, my bad, however, it will probably do with what i'd get...or not😱
So I'll buy a full pack of 50 0R47 3W resistors to be able to find 4 ideal triplets for the source resistors.
-For over-current protection I already have some 3.15A 5x20 slow blow fuses for the primary (600VA under 230V) and 15A 5x20 slow blow fuses for the secondaries (24V, 12.5A) with corresponding holder (5A chassis type holder and 2x 20A on wire type holders)
R0 doesn't need to be 3W. The space is tiny, with small diameter leads. You'll have just under 20V rectified, less 5V for the diode drop and you're at 0.02W in R0. 1/8W is overkill.
If you haven't already purchased IRFP240s look up member h_a, he has them matched and will tell you the matching conditions. Do real IRFP240s have shiny leads? My (old) Fairchild parts are in TO-3P packages and have flat dull leads. Current VIshays are in TO-247. With all the horror stories I'd be VERY leery of a Hong Kong seller. Matcing yourself aid the learning process...
If you haven't already purchased IRFP240s look up member h_a, he has them matched and will tell you the matching conditions. Do real IRFP240s have shiny leads? My (old) Fairchild parts are in TO-3P packages and have flat dull leads. Current VIshays are in TO-247. With all the horror stories I'd be VERY leery of a Hong Kong seller. Matcing yourself aid the learning process...
Last edited:
You're probably right 
even if my Wife is Chinese (and man I love her) I know about all the counterfeits you can find from HK sources, I was hastly in this case.
The fets have not been shipped yet (although I already paid for them 2 weeks ago), I don't know if I can still ask for an order cancelling and refund.....
even if my Wife is Chinese (and man I love her) I know about all the counterfeits you can find from HK sources, I was hastly in this case.The fets have not been shipped yet (although I already paid for them 2 weeks ago), I don't know if I can still ask for an order cancelling and refund.....
I posted on the h_a thread.
I think that even if my current order cannot be canceled i'll still take two sixlets of irfp240 from h_a (if it's possible)
For the PSU diodes (the MUR2020 are not available), I think this model should work:
MUR1560GOS-ND
Also I think can substitue the 1N4004DICT-ND diode of the psu with the 1N4007 I already have on hand right?
I think that even if my current order cannot be canceled i'll still take two sixlets of irfp240 from h_a (if it's possible)
For the PSU diodes (the MUR2020 are not available), I think this model should work:
MUR1560GOS-ND
Also I think can substitue the 1N4004DICT-ND diode of the psu with the 1N4007 I already have on hand right?
Last edited:
MUR1560 is fine.
Yes, 1N4004 and 1N4007 are interchangeable in this application, differing basically in PIV rating.
Yes, 1N4004 and 1N4007 are interchangeable in this application, differing basically in PIV rating.
Hi everybody 😉
I was quite busy those last two days as I learnt that I will be a Dad for the first time😀😀😀😀😀
I'll have to rush the building a little, cause I won't have that much time in a few months 😉
Well, I will buy the irfp240/irf9610 from h_a (witch is a great dude
), I'm just waiting to know if he can supply the ztx450 too.
I have thought about C2 (witch is the one cap on the signal path) and I'm wondering if a SILMIC2 220uf 35v (12.5mm so will fit perfectly here) would work here (I know that my rails are a little high, but the caps is still rated a little higher) C1 and C3 should remain Nichicon FG 50V
Also, as I will get perfectly matched fets from h_a, do I still need to buy more 0R47 resistors for source res than required?
I love you all!!!! (gosh it's better than drugs 😀 )
I was quite busy those last two days as I learnt that I will be a Dad for the first time😀😀😀😀😀
I'll have to rush the building a little, cause I won't have that much time in a few months 😉
Well, I will buy the irfp240/irf9610 from h_a (witch is a great dude
), I'm just waiting to know if he can supply the ztx450 too.I have thought about C2 (witch is the one cap on the signal path) and I'm wondering if a SILMIC2 220uf 35v (12.5mm so will fit perfectly here) would work here (I know that my rails are a little high, but the caps is still rated a little higher) C1 and C3 should remain Nichicon FG 50V
Also, as I will get perfectly matched fets from h_a, do I still need to buy more 0R47 resistors for source res than required?
I love you all!!!! (gosh it's better than drugs 😀 )
Last edited:
congrats !
Silmic is good
if you are going to use Panasonic 3W-ers , no need for tighter matching
loving us is easy .......... if not most clever thing to do
Silmic is good
if you are going to use Panasonic 3W-ers , no need for tighter matching
loving us is easy .......... if not most clever thing to do
Thank you 😉
for the Panasonic resistor, do you mean those ERX-3SJR47 I can't find any ERS serie.
Those are 5% tolerance, isn't it a little too much? The vishay bc in my bom are 1%, although I would gladly take the ERX as they are far cheaper 😀
....clever?? who, where?? 😱 😱 😛
for the Panasonic resistor, do you mean those ERX-3SJR47 I can't find any ERS serie.
Those are 5% tolerance, isn't it a little too much? The vishay bc in my bom are 1%, although I would gladly take the ERX as they are far cheaper 😀
loving us is easy .......... if not most clever thing to do
....clever?? who, where?? 😱 😱 😛
Last edited:
Congratulations on your impending fatherhood. Buckle up! It's a journey I'm glad I took, and even more glad to see them as self supporting adults. 😎
I'd only mentioned resistor matching in case of poorly matched IRFPs, which won't be an issue buying from h_a.
Don't sweat the ZTXs. Any GP NPN with the same order pinout is ok - EBC or CBE, just watch the orientation. MPSA18, MPSA42, BC546, BC550, etc.
You're going to end up with 30-32V rails, IIRC. For that 35V rated caps are sort of close, but probably OK.
I'd only mentioned resistor matching in case of poorly matched IRFPs, which won't be an issue buying from h_a.
Don't sweat the ZTXs. Any GP NPN with the same order pinout is ok - EBC or CBE, just watch the orientation. MPSA18, MPSA42, BC546, BC550, etc.
You're going to end up with 30-32V rails, IIRC. For that 35V rated caps are sort of close, but probably OK.
Last edited:
ARF!!
who cares for a comforter when you can get a sparkling metal film resistor 
Thank you BobEllis, I hope my journey's as good as your's seems to be
Congratulations! In a few months you can spend your sleepless nights
studying schematics. 🙂
Others have commented on ZTX450 being not critical. My recollection is
that the Aleph 3 used MPSA18 in the same position, for example.
For the 0.47 ohm resistor, just get a bag of 100 (big price drop). That
part shows up so often that they're unlikely to go to waste. (You can
never stop at one amp. 🙂 )
And welcome to the world of fearless amplifier builders...
studying schematics. 🙂
Others have commented on ZTX450 being not critical. My recollection is
that the Aleph 3 used MPSA18 in the same position, for example.
For the 0.47 ohm resistor, just get a bag of 100 (big price drop). That
part shows up so often that they're unlikely to go to waste. (You can
never stop at one amp. 🙂 )
And welcome to the world of fearless amplifier builders...
- Status
- Not open for further replies.