Now that the thermistor mystery is (hopefully) solved it’s time for replacement transistors for the faulty 2N3055/MJ2955.
I’ve read loads of threads about alternative transistors and the MJ15003/MJ15004 combo comes up quite a lot. There doesn’t seem to be much follow up comments form people who have used them.
Anyone have any comments on the use of these for a 3020i? I realise these aren't a direct replacement and voltages will be different and heatsink may be inadequate...
By now you’ve maybe guessed I’m new to all this but every day is a school day
I’ve read loads of threads about alternative transistors and the MJ15003/MJ15004 combo comes up quite a lot. There doesn’t seem to be much follow up comments form people who have used them.
Anyone have any comments on the use of these for a 3020i? I realise these aren't a direct replacement and voltages will be different and heatsink may be inadequate...
By now you’ve maybe guessed I’m new to all this but every day is a school day
The MJ15003/04 pair is a more than adequate, even beefier, replacement for MJ2955/2N3055. Never tried this in a NAD 3020, though, but my experiences with them in the EQUA amplifier by Elektor are fine.
Heatsinking won't present any problems afterwards. As Rth j/c is lower for the MJ15xxx's, their dies will stay cooler at the same dissipated power and on the same sinks.
Best regards!
Heatsinking won't present any problems afterwards. As Rth j/c is lower for the MJ15xxx's, their dies will stay cooler at the same dissipated power and on the same sinks.
Best regards!
0.22 is a typical value, 0.33 or 0.47 play safer. 3 watt rating is plenty for an amp like this.
As with any 'mod', you are changing the original design and so you could find that the bias generator would perhaps overcompensate for heating (meaning the bias current falls when hot) which is perhaps no bad thing tbh.
If the original outputs have failed then I would also consider replacing the two drivers and also the VAS (voltage amplifier) that feeds the drivers. MJE340/350 come to mind given the supply voltages used.
As with any 'mod', you are changing the original design and so you could find that the bias generator would perhaps overcompensate for heating (meaning the bias current falls when hot) which is perhaps no bad thing tbh.
If the original outputs have failed then I would also consider replacing the two drivers and also the VAS (voltage amplifier) that feeds the drivers. MJE340/350 come to mind given the supply voltages used.
Ha, ha! This is growing arms and legs now - it's going to turn in to a complete overhaul!
Also read about adding base resistors. I'd rather just see if it fires up at all before going any further...
p.s resistors just fell through the door - they're huge compared to what I've used in the past
Also read about adding base resistors. I'd rather just see if it fires up at all before going any further...
p.s resistors just fell through the door - they're huge compared to what I've used in the past
Last edited:
This is the problem when you start to modify stuff. Fitting incorrect sized parts or having parts with wires extended from them can look awful (imo ).
These are the style I would use:
RESISTOR, WW 3W 5% 0R22 - WELWYN | CPC
The original 2N3055/2955 outputs cost buttons and same too the MJE340/350
Why not see how stable things are first without altering it all. Just fit new drivers, outputs and VAS.
).These are the style I would use:
RESISTOR, WW 3W 5% 0R22 - WELWYN | CPC
The original 2N3055/2955 outputs cost buttons and same too the MJE340/350
Why not see how stable things are first without altering it all. Just fit new drivers, outputs and VAS.
I haven't got the diagram in front of me but the drivers are the pair that directly feed the outputs (2SD669 and 2SB649 from memory) and the VAS was the BD139 feeding those.
The 2SD/2SB types are known to fail in peculiarly intermittent ways and could be the cause of the original failures. Same for the BD139.
VAS is the 'voltage amplifier stage' and is typically the single transistor that feeds the driver pair.
Also make sure you get semiconductors from approved sources and not well known auction sites as there are so many fakes around. CPC (resistor link above) are part of Farnell and do free delivery on orders over £5
You would need the emitter resistors even if you installed new 2955/3055's. The reason is because the original devices used Hometaxial construction... and you could "get away" without the emitter resistors in those. Newer Expitaxial transistors will have thermal runaway issues without the emitter resistors in place.
The On-Semi "G" suffix actually refers to "Green" or lead-free tinning of the leads and other manufacturing materials. This doesn't make any difference to the numbers or performance. I think you have to look more closely at what the columns of Hfe numbers to see what they actually mean - Hfe is a really a very broad range of gain depending on the vagaries of the manufacturing process and particularly, at what current the device is being tested. At maximum current rating, gain is always much less (the minimum figures quoted) than at levels of say, 1A or below.
As to substituting these output semis, I think it's a waste of time and money. The 3020 is only a 2 x 20W 8 Ohm amplifier and the 2955/3055 pair is already overkill. If this were a 30-45W 8 Ohm amplifier, yes, I'd think more seriously about it. Also, don't lose sight of the fact that much of the desired sound quality of these old amplifiers is actually down to the distortion generated by the particular types of semis, rather than any refinement of the design or superior parts quality.
Unless you have some experience with redesigning and testing amplifier circuits, I suggest you keep the refit as simple and direct as possible, with the least amount of distractions in multiple areas of the circuit. Modifying circuits is not simple or guaranteed to work so minimising the number of changes will give you better chances of success in fewer errors, costs and cussing along the way. As experience accumulates, you can always revisit the possibilities that still interest you.
As to substituting these output semis, I think it's a waste of time and money. The 3020 is only a 2 x 20W 8 Ohm amplifier and the 2955/3055 pair is already overkill. If this were a 30-45W 8 Ohm amplifier, yes, I'd think more seriously about it. Also, don't lose sight of the fact that much of the desired sound quality of these old amplifiers is actually down to the distortion generated by the particular types of semis, rather than any refinement of the design or superior parts quality.
Unless you have some experience with redesigning and testing amplifier circuits, I suggest you keep the refit as simple and direct as possible, with the least amount of distractions in multiple areas of the circuit. Modifying circuits is not simple or guaranteed to work so minimising the number of changes will give you better chances of success in fewer errors, costs and cussing along the way. As experience accumulates, you can always revisit the possibilities that still interest you.
there is a lot of truth in that.I hear what you are saying and agree. I have very little electronics experience so it's just a small project/challenge I've set to see if I can get it working. If I could find new old stock that wasn't too expensive that's exactly what I'd be using as it would make my task a lot easier.
I chose the outputs as they were the ones I read most about and I totally don't want to start changing circuit design or it'll never work!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.