Hi,
I'm currently working on a Pioneer A-70 Integrated amp. One of the power supply circuits uses 2 JFETs (2SK373-GR). How does one go about matching a Current Regulating Diode to the JFET? The 2SK373 has an Idss of 2.6mA min and 6.5mA max.
Here is what the circuit looks like:
One of the JFETs failed and it has been suggested that a CRD like the Semitek E-202 (https://www.mouser.com/ProductDetail/954-E-202) can take it's place.
I'm currently working on a Pioneer A-70 Integrated amp. One of the power supply circuits uses 2 JFETs (2SK373-GR). How does one go about matching a Current Regulating Diode to the JFET? The 2SK373 has an Idss of 2.6mA min and 6.5mA max.
Here is what the circuit looks like:
One of the JFETs failed and it has been suggested that a CRD like the Semitek E-202 (https://www.mouser.com/ProductDetail/954-E-202) can take it's place.
Thanks to the good information about voltages on this schematic, we can do some quick math. If the pass transistor has a gain of about 100 and we want the base current to be 10% of the Zener current, then we need about 3mA from the current source. Two Zeners have to add up to about 62V so 0.25 Watts/31V gives a max Zener current of about 8mA. 2mA is a bit low. I make the range 3mA to 8mA, which is why an Idss of 2.6mA to 6.5mA works fine. You could just use a resistor about 5.6K or 4.7K which would be more reliable but maybe a bit noisier. The FET is only dropping 17V but know that most JFETs are only good for 25 to 40V. Note the CRD is not much better. Don't get it backwards!!!
Last edited:
This particular OEM Toshiba JFET 2SK373-GR (which are almost impossible to find here) has rating for Vgds of -100V, and an Ig of 10mA
Here is the product chart for Semitek CRDs:
The E-202 is in the 2mA and below range.
Assuming I would go the CRD route for replacement, Since 2mA is a bit low, would the E-352 be more appropriate?
Yes, and that brings up another question I have about these, if one is to replace the FET with a CRD, which end of the diode is equivalent to the drain of the FET? The Anode goes to the higher voltage?
Ah, yes. If you are rcs16 on another forum, than many thanks. That doesn't mean I didn't value your answer - I did. Infact, that is the way I soldered them in last night. I have yet to power it on. As a famous wood worker once said "Measure twice, cut once." Being new to electronics DIY, I'm simply seeking as much info as I can get.
Last edited:
need help too with 2sk373-GR at Q904 for onkyo integra 817xd amp. Did your CRD workout to be perfect replacement? Wouldnt it be better if a newer FET be selected that can accomplish the task. output voltages (at 22.8) from FET are different from mine than yours why?
In short, it has not yet....and it might be the case where it will not.
I am not an electronics expert, but that N-Channel JFET is probably being used to regulate current(I) to the collector of Q901 not Voltage (V) as I understand it.
The 2SK373-GR has an Idss of 2.6-6.5 mA - so I guess that would be nominally 4.55mA?
Someone with experience perhaps can do the math to pick out a CLD: https://www.mouser.com/datasheet/2/362/P22-23-CRD-1729293.pdf
In my application the E-352 are not doing the job, they are for some reason burning up. Perhaps over voltage, or over current - I don't know yet.
I have some E-562s that I may try....
My un-educated advise is to try to find some OEM 2SK373s first from a very trusted source. I have not been able to find a newer FET that matches the specs.
I am not an electronics expert, but that N-Channel JFET is probably being used to regulate current(I) to the collector of Q901 not Voltage (V) as I understand it.
The 2SK373-GR has an Idss of 2.6-6.5 mA - so I guess that would be nominally 4.55mA?
Someone with experience perhaps can do the math to pick out a CLD: https://www.mouser.com/datasheet/2/362/P22-23-CRD-1729293.pdf
In my application the E-352 are not doing the job, they are for some reason burning up. Perhaps over voltage, or over current - I don't know yet.
I have some E-562s that I may try....
My un-educated advise is to try to find some OEM 2SK373s first from a very trusted source. I have not been able to find a newer FET that matches the specs.
neither am I am expert or electronics pro, but you can see the correct voltages are shown in schematics, after 2sk373 mine has 22.8 and yours 63.6. And these 22.8v are being fed into Base of 2SD1408-Y transistor (Q901) not the collector. collector is getting 47v in my schematic. I think you are matching transistor symbol with pins on transistor, the center leg in symbol is actually Base (Pin 1 in case of 2sd1408) and the right leg is collector which is Center pin on transistor.
So your amplifier is being wasted just because of jfet?
So your amplifier is being wasted just because of jfet?
I think you should replace the JFET with a ~4.7K resistors. You can't get a resistor backwards and the Zener will be ~safe. (They are probably shorted by now). The difference in noise will be almost impossible to detect.
what would be shorted by now? also can you give calculation how you came up with 4.7k value referring to the schematics.
Reversing a CRD or similar mistake will instantly destroy the Zener(s), most likely as a shorted condition. 4.7K was calculated from the original circuit, ie the voltage drop divided by a current in the middle of the useful range. The circuit in post #10 has only one Zener and the current is not as obvious. For that circuit, the current will be about (58.4-47)/330/(min Q901 beta of about 100) *1/10% base current = 3.55mA; (47-22.8)/3.55mA= 6.8K. R903(330) may be damaged as well.
I certainly wont claim to know that equation. R903 was always running very hot, but then I was doing recap and could have made a mistake on a bypass resistor on Q901. and the result amp in protect, no voltage on emitter of Q901 and after trying to diagnose further R903 was gone and eventually Q904 also failed, I have replaced R903, have placed order for 2sk373-gr and also 2sd1408-y. what else should I order, knowing that would be helpful as it will save a month or so as that is what AliExpress takes to deliver.
ok now for understanding, what is decreasing voltage from 47 to 22.8? 2sk373? if yes, then why it has different voltages in pioneer a-70 schematic?
ok now for understanding, what is decreasing voltage from 47 to 22.8? 2sk373? if yes, then why it has different voltages in pioneer a-70 schematic?
Last edited:
It's not very complicated. R903 drops 58.4 to 47 volts so we know the current in that resistor is (58.4-47=11.4)/330=34.55mA. Most of that goes to Q901. Q901 will have a Beta of about 100 so the base current will be about 34.5mA/100 = 0.35mA but for good regulation, that should be say 10% of the Zener current, so the total JFET current will be about 10x that, ie 3.5mA. The JFET conducts Idss over a wide range of voltage, but we know the JFET voltage should be 47-22.8=24.2V. We read from the 2SK373 JFET data sheet that Idss is between 0.6mA and 6.5mA, so 3.5mA is probably typical. Note 0.6mA is more than the 0.35mA base current and 6.5mA is OK for a 250mW Zener, ie 0.0065x22.8=148.2mW. But a resistor chosen to replace the JFET would be 24.2V/3.5mA=6914, ie 6.8K. The resistor will pass a bit of ripple noise, but you have a capacitor C903 and the Zener D904 filtering out that noise.
The Zenner will conduct whatever current it has to hold the voltage down to (in this case) 22.8V, but it is limited to typically 250mW, so more than 0.25/22.8=11mA will destroy the Zener. So, if the JFET is shorted or reversed, you get too much current and the Zener is toast.
The Zenner will conduct whatever current it has to hold the voltage down to (in this case) 22.8V, but it is limited to typically 250mW, so more than 0.25/22.8=11mA will destroy the Zener. So, if the JFET is shorted or reversed, you get too much current and the Zener is toast.
one of the things that I became aware of is that the those CRDs have a current drop-off as voltage increases. The datasheet (yes, I actually read it and studied the I/V diagram 🙂 ) shows that the current (I) has a fairly significant drop-off as Voltage (V) increases. As the datasheet indicates, those CRD are rated for 100 volts, but the peak limiting current is produced at just 10V. So, for diode E-352 at 10V the current is 3.5ma. At 63V the current passed through the diode ( as the below graph shows) is much less then 3.5mA and is around 2.5mA. So it seems to me that I should probably be using the E-562 which has a current of around 3.5ma at 63V. The E-352s are not holding up in the circuit, they are failing - perhaps because they are not a good fit as the graph might suggest or because I still have something else awry. Fortunately, the Zeners are still good - oh and they add up to 61V (a 32 and a 29)
When I put the 352 in place (in the right direction this time - that is another story) the Amp came out of protection and al looked good at zero input and zero volume. I let it sit there and warm up for a while. Powered it down, buttoned it up, and attached a CD player, powered it back up, and viola, it actually worked. I left it at low volume for probably 10 minutes or so. Had to do something else so I powered it off. Came back several minutes later, powered it back on and POOF - smoke - the CRDs were toast.
When I put the 352 in place (in the right direction this time - that is another story) the Amp came out of protection and al looked good at zero input and zero volume. I let it sit there and warm up for a while. Powered it down, buttoned it up, and attached a CD player, powered it back up, and viola, it actually worked. I left it at low volume for probably 10 minutes or so. Had to do something else so I powered it off. Came back several minutes later, powered it back on and POOF - smoke - the CRDs were toast.
Last edited:
too much current during startup being allowed from CRD? my FET comes after resistor R903, your is directly connected to rail?