I counted 10 compensation capacitors in the MKII, which is 9 more than can be designed in a straightforward manner. I am amazed that Kenwood got it to work.
Ed
Ed
And there are caps with 3p and 4p values. I would not trust those small caps for compensation as the internal capacitance of the transistors can easily outweigh those caps.
It sounds like I really need to find the original outputs for this thing... at the very least. Other transistors appear to have been changed on the amp boards as well.
Last edited:
Here's some great info from Kenwood expert EchoWars over on AudioKarma.
https://audiokarma.org/forums/index.php?threads/kenwood-l-07m-ii-issues.97091/
These definitely sound like a nightmare.
https://audiokarma.org/forums/index.php?threads/kenwood-l-07m-ii-issues.97091/
These definitely sound like a nightmare.
Last edited:
As long as you don’t load the speaker terminals, it should be safe to force the output relay into or out of conduction for convenience in testing the two feedback states. Shorting base of Qk7 to ground will prevent the relay from energizing. Shorting its collector to ground will force the relay to close. Also safe to keep base shorted and short/remove collector short to control the relay.
According to the EchoWars thread I linked, these regularly oscillate themselves to death. Those low value compensation caps go out of spec due to heat and should be replaced with silver mica as those are the most stable. Whoever worked on this before used silver mica for some of those caps but not all of them. It seems to explain what happened with the second amp (that I originally stated seemed to be okay).
It worked fine until it warmed up, then went nuts. It had to warm up enough for that cap value to shift by just enough...
It sounds like they can be rebuilt with the much slower onsemi outputs but won't meet original spec. Not sure if that shortcoming will be audible.
Even the Sankens are far slower than the original transistors they used in the II version.
So my next step would be to replace the rest of the compensation caps he listed with silver mica, check everything else he mentioned, and hope for the best.
Sounds like a safe assumption it is oscillating.
It worked fine until it warmed up, then went nuts. It had to warm up enough for that cap value to shift by just enough...
It sounds like they can be rebuilt with the much slower onsemi outputs but won't meet original spec. Not sure if that shortcoming will be audible.
Even the Sankens are far slower than the original transistors they used in the II version.
So my next step would be to replace the rest of the compensation caps he listed with silver mica, check everything else he mentioned, and hope for the best.
Sounds like a safe assumption it is oscillating.
I could use much faster outputs if I could find a way to fit TO-3P parts.
These adapters exist, but they wouldn't fit as the positions for the TO-3 transistors are recessed. I'd have to have some risers machined.
https://hoppesbrain.com/product/to3-to-to3p-adapters/
These adapters exist, but they wouldn't fit as the positions for the TO-3 transistors are recessed. I'd have to have some risers machined.
https://hoppesbrain.com/product/to3-to-to3p-adapters/
You should ask restorer John over at ASR about this: https://www.audiosciencereview.com/forum/index.php?members/restorer-john.1308/
It's possible to belt-sand off the over-hanging parts of the TO-3P so it fits the profile.
Will that decrease the device power handling at all?
I thought of that but wasn't sure I could do it without taking too much material off.
If I can make that work I can use 30Mhz parts instead of 4Mhz parts. Not quite the 70 of the originals but a lot closer...
I see you are the vendor. It is cool that you made these.
I thought of that but wasn't sure I could do it without taking too much material off.
If I can make that work I can use 30Mhz parts instead of 4Mhz parts. Not quite the 70 of the originals but a lot closer...
I see you are the vendor. It is cool that you made these.
In the video I posted in #13, it shows how to use TO3 transistors in the L07M II...........
@claudio Nice! I should have looked more closely at that.
I see you used the MLCCs in place of the "black flag" capacitors. How are those for temp stability etc compared to the silver mica parts EchoWars recommended in the thread I linked in post #24.
I'll have to look again to see if there are new TO-3P that are 70Mhz. The 2SC5200/compliment I often use are 30Mhz. But I might have enough TO-3P Sanken pulls around here from AVR units to make a full matched set.
I did replace the main filter caps. I used the blue Chemicon caps. Screw terminals. I screwed a terminal onto them with solid copper wire coming out. I pounded the wire flat so it would fit through the board without modification. I'll post pictures.
I see you used the MLCCs in place of the "black flag" capacitors. How are those for temp stability etc compared to the silver mica parts EchoWars recommended in the thread I linked in post #24.
I'll have to look again to see if there are new TO-3P that are 70Mhz. The 2SC5200/compliment I often use are 30Mhz. But I might have enough TO-3P Sanken pulls around here from AVR units to make a full matched set.
I did replace the main filter caps. I used the blue Chemicon caps. Screw terminals. I screwed a terminal onto them with solid copper wire coming out. I pounded the wire flat so it would fit through the board without modification. I'll post pictures.
Nooo, I didn't do that video: Prime Hifi did.
About using ceramic instead of silver mica, just use NP0 type: no temp variation and no microphonicity.
Another possibility is to get the originals from UTsource: they have them used. For the complementary you have to inquire them. Might try them if the cost is not too much.
About using ceramic instead of silver mica, just use NP0 type: no temp variation and no microphonicity.
Another possibility is to get the originals from UTsource: they have them used. For the complementary you have to inquire them. Might try them if the cost is not too much.
No problem shaving them off a bit. It's still more mass of copper heat-spreader than a regular TO-3 has, and the copper of the transistor's heat-spreader sandwiched with the adapter is 3.2mm thick; it should actually move heat slightly better than a TO-3.
The issue with simply bolting TO-3P's to a TO-3 heatsink, is that the copper contact patch of a TO-3P heat-spreader is only 1/3 the size of a TO-3. Copper conducts heat twice as fast as aluminum, so the more the heat is spread out in copper before it gets to the heatsink, the faster that heat is transferred to the aluminum.
Last edited:
I talked to the client and will explain better to him on the phone tomorrow. But he's not too concerned with costs or timeframe, so this will go ahead with the best solution and I can take the time to get it right.
I haven't studied the Kenwood circuit, but most amps require output transistors with Vce and Vbe at least twice the rail voltage, because they are pulled negative while the other side is conducting.
Have you decided on transistors to try with this beast?
Have you decided on transistors to try with this beast?
@Phloodpants I wasn't sure what parts I was going to use but I just got my hands on some Sanken 2SC2837/2SA1186. Same voltage/current/power specs as original. fT of 60Mhz. 10Mhz short of the originals. Hope that's close enough.
When using a lower Ft output device it needs more high frequency compensation. Same applies to the drivers. 100Mhz drivers and 60Mhz outputs sounds fine to me. I would still replace the compensation caps with higher value ones.
A bias of 600mV? There should be 12mV over those 0.47ohm resistors. Is Df9 still on the heatsink?
Cheers and good luck
Ruben
Edit: typos
A bias of 600mV? There should be 12mV over those 0.47ohm resistors. Is Df9 still on the heatsink?
Cheers and good luck
Ruben
Edit: typos
Last edited:
If I have to rebalance this crazy amp I'll never get it right...
The original compensation caps were the infamous "black flag" ones which fail and kill these amps. In the thread I referenced a few posts up AK legend and Kenwood expert Echowars changed them to silver mica caps since they supposedly offer the best stability. And those low value caps need to be ultra stable. All but 2 of the caps he mentions have already been changed to silver mica. The other two are some other sort of ceramic or MLCC. I guess I'll replace those with the silver mica too to give the best chance.
The original compensation caps were the infamous "black flag" ones which fail and kill these amps. In the thread I referenced a few posts up AK legend and Kenwood expert Echowars changed them to silver mica caps since they supposedly offer the best stability. And those low value caps need to be ultra stable. All but 2 of the caps he mentions have already been changed to silver mica. The other two are some other sort of ceramic or MLCC. I guess I'll replace those with the silver mica too to give the best chance.