I have two Tascam MH-8 headphone amplifiers. One of them has some problems I would like to address. Generally, it is working but there are some issues.
The unit powers up and there is sound, aside for a couple channels with broken pots that I need to replace. The primary issue is a large section of the main board that looks cooked under 2 rows of 32 transistors each and a row of 32 resistors.
I'm assuming the problem is the transistors and not the resistors (180Ω 5% 3W). Is this a safe assumption?
The first row of 32 transistors alternates between C1815 and A1015, there are 16 of each. From what I understand, these are complimentary transistors.
The second row of 32 alternates between C2236 and A966, again, these are complimentary.
The other MH-8 has a much higher serial number to compare to. The main board on this unit is not burnt and 3 of the 4 different transistor types have different part numbers on this later unit.
All A1015's have been replaced with A1266 GR
All C2236's have been replaced with C3205 YK 743 KTC?
All A966's have been replaced with A1273 YK 805
After doing some research, all of these are legitimate replacements for their predecessors. The C3205 and A1273 are in a different package than the A1015 and C2236 which are in a TO-226 package. I'm not sure what it is. I can't find anything like it in any package charts I have looked at. Does this even matter?
Also, the newer unit has the exact same 3 watt resistors, same value and tolerance, this is largely why I am discounting them as the source of the problem.
While the unit does work, obviously something isn't right given the blackened board and it appears that Tascam realized this and substituted some of the transistors in newer units. While I'm replacing the broken pots I thought I would replace all or most of these transistors. Would it be a good idea to replace all of the A1015, C2236 and A966 devices with A1266, C3205 and A1273 respectively as is the case with the newer MH-8? Would it also be a good idea to replace the C1815's with new C1815's since they've probably been through hell?
One other difference between the 2 units in this area of the main board is another row of 16 resistors. On the "bad" unit, these are 1/8 watt 2.7kΩ 5% and on the good unit, they are 1/8th watt 2.7kΩ 2%. See R032 & R044 in both pics above. Should I change these from 5% to 2%?
Finally, if this is a good plan, how would I choose between KT and 2S variants of all of these transistors since apparently either is acceptable? I don't really understand the difference. For example, 2SC3205 vs. KTC3205.
I would appreciate any input on this
The unit powers up and there is sound, aside for a couple channels with broken pots that I need to replace. The primary issue is a large section of the main board that looks cooked under 2 rows of 32 transistors each and a row of 32 resistors.
I'm assuming the problem is the transistors and not the resistors (180Ω 5% 3W). Is this a safe assumption?
The first row of 32 transistors alternates between C1815 and A1015, there are 16 of each. From what I understand, these are complimentary transistors.
The second row of 32 alternates between C2236 and A966, again, these are complimentary.
The other MH-8 has a much higher serial number to compare to. The main board on this unit is not burnt and 3 of the 4 different transistor types have different part numbers on this later unit.
All A1015's have been replaced with A1266 GR
All C2236's have been replaced with C3205 YK 743 KTC?
All A966's have been replaced with A1273 YK 805
After doing some research, all of these are legitimate replacements for their predecessors. The C3205 and A1273 are in a different package than the A1015 and C2236 which are in a TO-226 package. I'm not sure what it is. I can't find anything like it in any package charts I have looked at. Does this even matter?
Also, the newer unit has the exact same 3 watt resistors, same value and tolerance, this is largely why I am discounting them as the source of the problem.
While the unit does work, obviously something isn't right given the blackened board and it appears that Tascam realized this and substituted some of the transistors in newer units. While I'm replacing the broken pots I thought I would replace all or most of these transistors. Would it be a good idea to replace all of the A1015, C2236 and A966 devices with A1266, C3205 and A1273 respectively as is the case with the newer MH-8? Would it also be a good idea to replace the C1815's with new C1815's since they've probably been through hell?
One other difference between the 2 units in this area of the main board is another row of 16 resistors. On the "bad" unit, these are 1/8 watt 2.7kΩ 5% and on the good unit, they are 1/8th watt 2.7kΩ 2%. See R032 & R044 in both pics above. Should I change these from 5% to 2%?
Finally, if this is a good plan, how would I choose between KT and 2S variants of all of these transistors since apparently either is acceptable? I don't really understand the difference. For example, 2SC3205 vs. KTC3205.
I would appreciate any input on this
Last edited:
So how hot is it running?
This is not new, is it? (That phenolic board is, uh, 'old school'.) If it has been run 24/7 for 30 years, it won't give actual trouble for another decade, leaving lots of time to think. Turn-off except when heads are in cans? Add fan?
This is not new, is it? (That phenolic board is, uh, 'old school'.) If it has been run 24/7 for 30 years, it won't give actual trouble for another decade, leaving lots of time to think. Turn-off except when heads are in cans? Add fan?
I'm not the original owner so I don't know the full history. I haven't been able to get it very hot yet. Just sitting on the bench for hours it stays fairly cool. Maybe in a rack with all 8 channels working and low headphone impedance it might really start to sizzle.
That's a good question about the age. This model is still in production but obviously they have made some revisions. I don't know if newer boards look any different. I have a few other Tascam devices from the last decade and they have the same style boards. It might be typical Tascam. I repaired a Tascam reel to reel from the 70's a while ago that looked exactly the same. Who knows.
I did some digging and even the transistors on the newer unit are now obsolete but they can still be had on ebay etc. There are current replacements that are still available but I really don't know if you can simply swap out an old transistor for a newer equivalent. I know you need to be sure the pin configurations are the same but I don't know if there are other considerations that need to be taken into account other than comparing the main specs on the datasheets. That's mainly why I posted here. I was hoping someone far more knowledgeable might say, "just swap A for B and X for Y and you'll be good". Something like that. Lol.
The unit actually sounds good. I really can't discern any audible difference with the other unit.
I'm currently doing a bunch of electronics projects and making sure Digikey & Mouser stay afloat so I thought it might be a fun/educational project to update the transistors on general principle. Apparently people aren't chomping at the bit to chime in here. Maybe there is a lot more involved than I am thinking. Maybe it would be more work than it's worth.
That's a good question about the age. This model is still in production but obviously they have made some revisions. I don't know if newer boards look any different. I have a few other Tascam devices from the last decade and they have the same style boards. It might be typical Tascam. I repaired a Tascam reel to reel from the 70's a while ago that looked exactly the same. Who knows.
I did some digging and even the transistors on the newer unit are now obsolete but they can still be had on ebay etc. There are current replacements that are still available but I really don't know if you can simply swap out an old transistor for a newer equivalent. I know you need to be sure the pin configurations are the same but I don't know if there are other considerations that need to be taken into account other than comparing the main specs on the datasheets. That's mainly why I posted here. I was hoping someone far more knowledgeable might say, "just swap A for B and X for Y and you'll be good". Something like that. Lol.
The unit actually sounds good. I really can't discern any audible difference with the other unit.
I'm currently doing a bunch of electronics projects and making sure Digikey & Mouser stay afloat so I thought it might be a fun/educational project to update the transistors on general principle. Apparently people aren't chomping at the bit to chime in here. Maybe there is a lot more involved than I am thinking. Maybe it would be more work than it's worth.
Last edited:
Little plastic-encapsulated transistors like these are usually limited to less than half a watt of dissipation before they overcook to death. If the engineer who designed the circuit knew what he/she was doing, he/she would have limited power per transistor to half that (< 300 mW) to improve reliability.
Meantime, those 32 resistors are rated at 3 watts apiece. This doesn't mean they're actually dissipating 3 W apiece - but bigger resistors cost more, so we can guess the designer specified 3W resistors because the power dissipation in each resistor was high enough to require that rating.
If this guess is anywhere close to correct, it means each of those resistors is dissipating much more power than the adjacent transistors. Maybe six or eight times as much.
So my guess is the heat-damage to the circuit board came from hot resistors, not hot transistors.
You could measure DC voltage across one of those resistors, and calculate actual quiescent dissipation, if you like. P = V*V/R
That seems perfectly logical. But as this is an old unit, could there have been some other cause? Did someone apply too high a power supply voltage long ago, cooking the board? Was the device mounted in a location with no airflow?
We don't know, but these are plausible explanations. So I don't think we can discount the resistors as the cause of the excess heat.
I don't think that is the right conclusion. Many transistor types rapidly become obsolete, and production stops. Tascam may have switched to a newer transistor simply because the old part(s) was/were no longer being manufactured.
My two cents: this is a Very Bad Idea (TM). Don't fix what isn't broken...chances are, you'll create more problems than you solve.
IMO, blanket parts replacement is a bad idea - again, why fix what isn't broken?
Most likely it will make no difference at all. As manufacturing methods have improved over the years, tighter-tolerance resistors have become cheap, and manufacturers have no reason not to use them.
Any properly designed transistor circuit will not be that fussy about tolerance values. Transistors themselves have enormous tolerances, and a good transistor circuit is designed to work properly in spite of these huge variations.
Dunno if you want my input - but my feeling is, if the circuit works now, and there are no obvious problems, don't fix what isn't broken. Just enjoy your "new" headphone amplifier. (And the money you save on unnecessary repairs can go into something that will actually make a difference - maybe a new set of strings for your bass guitar, for instance. 😀 )
-Gnobuddy
Wow, that was a phenomenal response. Thanks for taking the time to write it.
This is exactly what I was hoping for, someone who knows infinitely more than I do to shed light on the situation and set me in the right direction.
Everything you said makes a lot of sense.
I did a little reading on quiescent dissipation and did the calculation for several of the 3W resistors. The voltage drop is 4.8V (14V going in and 9.2 coming out). The current is 27 mA. If I understand it correctly the quiescent dissipation is 128 mW. If that's correct, that's a lot less than I would have expected.
Out of curiosity, I measured the temps of the resistors and the adjacent transistors. The resistors are around 125F and the transistors are about 135F. With the top on while the unit is working things probably get hotter. I'm not sure if this means anything, just throwing it out there. The fact that it still works perfectly probably means this is inconsequential.
Very good input. I really appreciate it. I took your advice and bought some strings today.
Thanks
This is exactly what I was hoping for, someone who knows infinitely more than I do to shed light on the situation and set me in the right direction.
Everything you said makes a lot of sense.
I did a little reading on quiescent dissipation and did the calculation for several of the 3W resistors. The voltage drop is 4.8V (14V going in and 9.2 coming out). The current is 27 mA. If I understand it correctly the quiescent dissipation is 128 mW. If that's correct, that's a lot less than I would have expected.
Out of curiosity, I measured the temps of the resistors and the adjacent transistors. The resistors are around 125F and the transistors are about 135F. With the top on while the unit is working things probably get hotter. I'm not sure if this means anything, just throwing it out there. The fact that it still works perfectly probably means this is inconsequential.
Very good input. I really appreciate it. I took your advice and bought some strings today.
Thanks
Thank you, my friend. I really appreciate that warm response.
Wonderful! Making music is so much more important than faffing about with transistors, don't you think? And I say this as a person who's spent an enormous chunk of my life faffing about with transistors. I thought it was the most important thing in the world, for many years.
I'm a musician, too. Limited though my own musical talents might be, I've surely created more happiness in the world by making music, than I ever have with any number of transistors. It took me a long time to see that, too.
I think that kindness, love, and music are the best things that human nature has to offer. (And I spent most of my life not having the faintest idea that this was the truth.)
You are a musician. Evolution gave you a huge gift, the ability to create music. And that ability is also a gift to those who enjoy your music, even if they don't realize it consciously.
So by all means, more music-making, and less transistor-soldering! 😀
(I still faff about with transistors, but almost exclusively in the pursuit of more music-making - pretty much all my electronics projects have something to do with making music.)
Gnobuddy