either follow @steveu and use the resistor, let him calculate the resistor value for you. Or you can order 2sk373 from https://www.aliexpress.com/item/1005004001185912.html?spm=a2g0o.order_detail.0.0.3527f19cu1JDxU the supplier is pretty confident and has guaranteed it to be original japan. or wait till my order arrives and will see if it works.
and have you performed recap of the amplifier, if not do it, for safety but also from my experience getting the original sound back. i have replace all electrolytics except 2 15000uf capacitors. I used ukw capacitors and few ukz. the most improvement in sound i got was in instrument separation.
Voltage drop of R903 58.4-47 =11.4
Current through R903 11.4/330 =.03445A
Voltage drop from jfet 47-22.8v =24.2
Resistor to replace jfet 24.2/.03445 =702, Isnt it right?
Most of the 0.03445A goes to the Q901 collector, not the JFET. The JFET current is the Zener current plus the Q901 base current. The Zener current can not be more than 11mA (0.011A) or you burn them out (assuming 1/4W Zener) and the base current will be the collector current divided by the transistor Beta, which will be about 100. The base current should be less than the Zener current but the Idss spec for 2SK373 min is 0.0006A, (0.6mA) not much more than transistor base current (about 0.3mA). The 2SK373 max Idss is 6.5mA so typical is probably about 3.5mA (0.0035A). A suitable resistor is then 24.2/0.003445 = 7K. Min resistor value would be 24.2/6.5mA= 3.7K. Max resistor value would be 24.2/0.6mA=40K.
I tried 8.8k, 6.6K also 2.2k, 2.2k resistor from drain to source/gate was giving 1-2mv, resistor was getting hot. I have ordered zener diodes should arrive after 2 days, then i will replace both zener and R503 resistor, then i will try 7K resistor and let you know. 2sk373 will arrive after 1 month so as your recommendation i will try to workaround it with resistors.
There are a lot of possible alternatives. An LM7824 could replace the whole circuit except that the input voltage is a bit too high, but you could fix that with a larger value for R903. Or use an LM317 but you have the same issue unless you get the high voltage version, and you have to calculate the feedback resistors. I accidently blew a Zener in a similar circuit once but a couple resistors and a transistor and I used the negative regulator as a reference to "follow" the positive supply. An LM7812 or LM7805 can also be wired like a LM317 to get an "adjustable" voltage.
High voltage LM317
High voltage LM317
@steveu congratulation you have done it, amp came out of protection, i am using 6.8k ohm resistor in place of 2sk373 fet, but i will have to wait for original parts, my replacement clone of transistor 2sd526 npn is getting more hot than the original pnp transistor, replaced zener is performing but voltage is lower at 21.5mv. for transistor how high in frequency can I go, the original was 8mhz. and can I go for emitter-base voltage Vebo of 6v, original was 5v.
as a backup got harman kardon hk3270, which was supposed to be working, but i found it otherwise after receiving it. it turns on but goes to protection after 2 seconds.
IMHO the Semitech CRDs are only useful for voltage drop of around 10~20V continuous, except for those < 1mA.
On top of that, their rated voltage of 100V is for pulsed wave and not continuous.
And their max rated power is 300mW.
For reliability, I would never want to exceed 50mW continuous.
In the circuit you showed in post #1, the CCS is used to bias 2x Zener diodes // a cap.
That means the actual current value is not super critical, due to the Zener's low impedance.
Also it only sees the full 79V briefly at the start up when the cap is not fully charged.
At steady state, it see a differential voltage of just 15V, which is near the max impedance part of the CRD operating curve.
If I were you, I would just use E352 and not worry too much about hitting a specific current value.
Cheers,
Patrick
On top of that, their rated voltage of 100V is for pulsed wave and not continuous.
And their max rated power is 300mW.
For reliability, I would never want to exceed 50mW continuous.
In the circuit you showed in post #1, the CCS is used to bias 2x Zener diodes // a cap.
That means the actual current value is not super critical, due to the Zener's low impedance.
Also it only sees the full 79V briefly at the start up when the cap is not fully charged.
At steady state, it see a differential voltage of just 15V, which is near the max impedance part of the CRD operating curve.
If I were you, I would just use E352 and not worry too much about hitting a specific current value.
Cheers,
Patrick
"2sk373-gr arrived from AliExpress, working fine for last 2 days."
So how much voltage (Vds) did you test them with, larger than 70V ?
Patrick
So how much voltage (Vds) did you test them with, larger than 70V ?
Patrick
Errm....you did say you were ordering components from Aliexpress, right? I'd have thought most DIYs would by now have realised that the correlation between Chinese or most other substitute/generic semis and original types, is unreliable to say the least. I would definitely not try their products out on a paying customer so I wouldn't suggest it to other DIYs either. The more reliable path would be the resistors, as suggested by steveu and that at least looks sensible.
Unfortunately, reviving old gear isn't as cheap and simple as matching up catalog part numbers and ordering from any parts distributor, especially since they are obsolete and hard to find types. That's precisely the sort targeted by retail platform and boutique parts sellers. Buyers are just suckers, trusting enough to spend a few dollars and wait some weeks, only to find there are still problems and more failures, whilst PCB damage is becoming an issue too.
Unfortunately, reviving old gear isn't as cheap and simple as matching up catalog part numbers and ordering from any parts distributor, especially since they are obsolete and hard to find types. That's precisely the sort targeted by retail platform and boutique parts sellers. Buyers are just suckers, trusting enough to spend a few dollars and wait some weeks, only to find there are still problems and more failures, whilst PCB damage is becoming an issue too.
I have not found a CRD will last more then a few seconds on the mains. They seem fine when the amp is connected to a dim bulb tester but as soon as I remove the DBT, the CRD on the positive side starts smoking at the solder joint. I have tried the 202s, 352s, and the 562 to no avail. I have completely rebuilt that section - new caps, new rectifier, and new Zeners and I still get the same results. The amp comes out of protection and several seconds later the smoke starts. The 202s latest the longest, I actually played several minutes of music and powered the amp down. Next time I powered the amp up, they were toast a few seconds later..
Doesn't both of those CRDs see a constant +- 79V?
The way they are oriented, the cathode of the CRD on the positive sees +79V and the anode of the other see -79V.
Ok, well, I think I can chalk this up to a bad solder joint. I unsoldered that leg of the CRD cleaned it up, got rid of extra flux and reflowed that joint.
Powered the unit up, and viola - out of protection and no smoke!
And it works!
I've been using it for the last couple of days. That E-352 seems to be one for replacing that 2SK373.
I must say...this Pioneer A-70 is a really nice sounding amp. It was definitely worth the effort. It's a very nice sounding amp with crisp highs and nice deep lows.
It's currently driving a couple of old Bose book shelf speakers playing Collage - Moonshine.
Powered the unit up, and viola - out of protection and no smoke!
And it works!
I've been using it for the last couple of days. That E-352 seems to be one for replacing that 2SK373.
I must say...this Pioneer A-70 is a really nice sounding amp. It was definitely worth the effort. It's a very nice sounding amp with crisp highs and nice deep lows.
It's currently driving a couple of old Bose book shelf speakers playing Collage - Moonshine.
In my opinion modern amplifiers are not worth the price if compared with a clean vintage amplifier. The build quality that i have on my integra can only be available now on $3000 and above amplifier, I got this for $150, did a recap $100, still two large power capacitors are remaining those would be $100. in total $350.