Now I don't have to decide if I replace the last transistors that were original. I'm forced to.
I will try some bd139/mje340 or even 2SC4793.
The original 2sc1384 had a Ft of 200MHz.
BD139 is not mentioned in the datasheet.
MJE340 would be 10-20MHz.
2SC4793 Ft is 100Mhz.
Should I take this into account when choosing?
I will try some bd139/mje340 or even 2SC4793.
The original 2sc1384 had a Ft of 200MHz.
BD139 is not mentioned in the datasheet.
MJE340 would be 10-20MHz.
2SC4793 Ft is 100Mhz.
Should I take this into account when choosing?
I can't give an answer to that because ultimately it is how they perform when used under the operating conditions in the amp. Drivers and outputs on these old amps do tend to be very critical. You will need to make sure the bias is OK, that it is stable and that it actually adjusts within range in the first place.
You may need to alter/modify the circuit to get it to perform correctly... only way to know is by trying.
I will get 3 types and do some testing. I think I'll connect the board to a breadboarded power supply and get the signal with wires from the amp.
It's pretty hard to solder the wires and secure the pcb/heatsink assembly into the amplifier case.
The 2sc1384 with 200MHz ft had 500pF to keep oscillation at bay.
I presume with a slower type I can decrease that.
It's pretty hard to solder the wires and secure the pcb/heatsink assembly into the amplifier case.
The 2sc1384 with 200MHz ft had 500pF to keep oscillation at bay.
I presume with a slower type I can decrease that.
I'd like to know this before I try to install different transistors.
Last night I loaded the output (of the other amplifier board) with a 8 ohm resistor, and used different testing frequencies and measured H710 22ohm cathode resistor (R747).
Maximum at full output was 45mA, at any frequency.
Is there reason to suspect that a 100mA part would not be adequate in that spot? I'm trying to broaden my options so I have alternatives if I run into problems.
Last night I loaded the output (of the other amplifier board) with a 8 ohm resistor, and used different testing frequencies and measured H710 22ohm cathode resistor (R747).
Maximum at full output was 45mA, at any frequency.
Is there reason to suspect that a 100mA part would not be adequate in that spot? I'm trying to broaden my options so I have alternatives if I run into problems.
Try KSC2690. 500mA 1.2W, 150 MHz, 20pF Cob. Similar to the old Panasonic 2SC1567 which I used for years and years to replace every unobtainuim driver I came across. You're probably looking at changing to TO-126's in any case at this point, unless you want to risk getting a fake.
The current thru that resistor is essentially the output stage base current. 5A divided by hFE of 50 is about 100mA. But what you want is a driver with good gain at 100mA, not one with a max rating of 100mA. Big difference.
The current thru that resistor is essentially the output stage base current. 5A divided by hFE of 50 is about 100mA. But what you want is a driver with good gain at 100mA, not one with a max rating of 100mA. Big difference.
Couldn't get to that store today but managed to pick up:
bd139/bd140 (1.5A, unknown fT/cob)
mje340/mje350 (500mA, 20Mhz stated from some sources and 20pF cob maybe)
2sd669/2sb649 (1.5A, 140Mhz, 14pF cob)
bc327/bc337 (800mA, 210MHz, 15pF cob)
2n5551/2n5401 (600mA, 100MHz, 6pF cob as they are smaller die)
What would be most recommended to begin with?
I guess the smaller to92 would be up to task as it's 600/800mA rated but would go for best option.
On paper bc337/bc327 would be the closest.
bd139/bd140 (1.5A, unknown fT/cob)
mje340/mje350 (500mA, 20Mhz stated from some sources and 20pF cob maybe)
2sd669/2sb649 (1.5A, 140Mhz, 14pF cob)
bc327/bc337 (800mA, 210MHz, 15pF cob)
2n5551/2n5401 (600mA, 100MHz, 6pF cob as they are smaller die)
What would be most recommended to begin with?
I guess the smaller to92 would be up to task as it's 600/800mA rated but would go for best option.
On paper bc337/bc327 would be the closest.
Remember that link to a thread I posted a while back. The guy with the amp that we were helping him attempt to rebuild. And how despite all our best efforts with modern replacements it just wouldn't bias up or be stable.
You are going to have to dive in here and see what works best. Personally I would be tempted to try the MJE340/350 if they are modern versions such as from CDIL.
Do I need a load to test for oscillation?
I have a variac, so I want to use that to couple with the regulated supply I made the other day to offer the correct DC voltage so I can observe the bias and the rest.
Also can I inject signal from my laptop headphone jack into the output stage directly?
Not having a load would allow me to use the small power supply (LM317 has a whimpy heatsink).
I have a variac, so I want to use that to couple with the regulated supply I made the other day to offer the correct DC voltage so I can observe the bias and the rest.
Also can I inject signal from my laptop headphone jack into the output stage directly?
Not having a load would allow me to use the small power supply (LM317 has a whimpy heatsink).
Yes I read that thread. Ended on a sad note
I hope that my work will help others if they are stuck with new production parts. Hope I sort this out as well.
Right now I'm not in the "let's tinker" mood anymore as I've finished this project once and I just want to listen to the amp. I really loved the way it sounded.
I put the board back into the amp, thought it's faster.
I installed 2SD669/2SB649 and I think it is oscillating. I added 470pF on 2SD669 instead of 500pF as it was on the 2sc1384.
I measured bias and I saw that the DC varied constantly so I scoped it (H001 emitter) and saw this.
In DC there was 100mA or so and I could adjust it but it was very sensitive and was droping fast. At min I still have a sawtooth waveform, but less in amplitude. Increasing the bias increases the sawtooth amplitude only, not frequency.
The frequency is 100Hz
I installed 2SD669/2SB649 and I think it is oscillating. I added 470pF on 2SD669 instead of 500pF as it was on the 2sc1384.
I measured bias and I saw that the DC varied constantly so I scoped it (H001 emitter) and saw this.
In DC there was 100mA or so and I could adjust it but it was very sensitive and was droping fast. At min I still have a sawtooth waveform, but less in amplitude. Increasing the bias increases the sawtooth amplitude only, not frequency.
The frequency is 100Hz
Attachments
Last edited:
Bd139/bd140 with 220pF oscillates as well.
Managed to burn another output transistor from the other channel on the same board. Forgot I had the amp set in bridge. The 0.5ohm emitter resistor of the burnt transistor is out as well. Good thing I didn't throw away the old resistors as they come in handy now while testing
Tomorrow I want to move 2sc1384/2sa684 from the other channel to the one I was dealing with so I'm sure it works and nothing else is bad, and instead I'll try mje340/mje350 pair. Is 220pF good? There are 2 extra resistors around the driver, that don't show up in the schematic. I'll try and add them to the schematic.
If this fails as well I'll move to to92 devices like 2n5551 bc337 pairs.
Also will extend the board with wires so I can work faster. Maybe I will try and tinker with the capacitor value.
The worst thing actually is burning mje15032. Cost adds up :/ I'm up to 5 so far. Switching to tip41 is not an option as I know the mje15032 is stable in this amp.
Managed to burn another output transistor from the other channel on the same board. Forgot I had the amp set in bridge. The 0.5ohm emitter resistor of the burnt transistor is out as well. Good thing I didn't throw away the old resistors as they come in handy now while testing
Tomorrow I want to move 2sc1384/2sa684 from the other channel to the one I was dealing with so I'm sure it works and nothing else is bad, and instead I'll try mje340/mje350 pair. Is 220pF good? There are 2 extra resistors around the driver, that don't show up in the schematic. I'll try and add them to the schematic.
If this fails as well I'll move to to92 devices like 2n5551 bc337 pairs.
Also will extend the board with wires so I can work faster. Maybe I will try and tinker with the capacitor value.
The worst thing actually is burning mje15032. Cost adds up :/ I'm up to 5 so far. Switching to tip41 is not an option as I know the mje15032 is stable in this amp.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.