I’ve got an HK 870 and love it. Was my first stand alone amp bought in the early 80’s. The board connectors went bad and had my local audio shop rebuild it which included some of the power transistors (late 80’s). Serviced my father for several years powering Magnepan LRS 4 ohm speaker with no problems. Then it made it back to me.
Nice job. When you are poking around at 29:00, it looks like the PCB flexes around the heat sinks the transistors are screwed to. Bad design, it means that any forces that cause PCB flex will be transmitted directly into those solder joints (which were broken).
I believe a lot of the hk stuff in that era used direct coupled low negative feedback stages. So, the only limiting factor for frequency would be the transistors chosen for each stage. Some of them even had ultra wideband printed right on the front.
Another good design assembled poorly. Will make someone happy now. I actually bake sinks with transistors installed at 100°C in the oven and torque them down while still hot, compounds get runny and fill the gaps, while length contraction ensures they stay tight at high operating temperature. Bit tricky, but works great, especially with class A and high bias AB fet outputs.
Nice to see you back, this is a interesting specimen. 1983-1986............Not HK's best years for QC. Added a film cap to every service cap, hmmmmmmmmmm. Have every right to complain about that 1/2 ass white compound Goop. Never ever seen that light Blue stuff you purchased, very $$$$$$$$ Hell yes, those are big holes and little plastic support. Thanks for pointing out the thickness of the legs on those Burgundy colored Caps.
I used highest thermal conductivity compound with Peltier coolers. It is very effective, but after e while it was impossible do remove coolers from the heat dissipator. The ceramic in the compound was like cement. I had to break the coolers.
I think the thermal expansion of the long heatsink is ripping the joints apart in tension .On conventional heat sink arangements the legs flex sideways. I would have bent a cammel hump into the legs between the transistor and the solder pad to allow a bit of give to save the joint. But otherwise great job.
When I saw the two sets of big capacitors I was thinking it might be class G bias with a twin split power supply rail. I know Kenwood, Carver, and Hitachi did those designs. Have to see if Hifiengine has a schematic on this amp, would like to see the roadmap on this one. The emitter resistors may have gone up in value due to the heat and abuse, that might account for the drop in wattage. Then again, it may just be the way it is. Thanks for the ride along on this one.
If I remember correctly, you will get maximum heat dissipation in the output transistors on an analog amp at about 40% of rated output, because at that point, you have the highest product of current and voltage across the transistors. It should really cook then.
@@ancienttech4636 True, just wanted to clarify, that since analogue amp definition encompasses a wide variety of topologies and classes, the figure may be anywhere from no load to 80%
From where I'm sitting the problem with the transistors not sitting flat on the heatsink is due to the heatsink not being flat. There are signs that the heatsink is very slightly distorted, I would have been tempted to try to correct that so you don't need so much heatsink compound to fill the gap. It's amazing that it lasted long enough to dry out the old compound, those transistors must have been very robust but not indestructible !
The aluminum extrusion was flat. The old transistors had a slight dish to them which prevented them from making good contact. Not sure why or how they got dished
@@TrevorsBenchI suspect there might have been too much paste applied and/or the fasteners were over-torqued; or the paste that was applied ran out from near the fasteners when the amp heated up and/or aged. Simply put, the paste/heatsink arrangement didn’t age well.
Harmon Kardon used to market their amps (even the smaller ones) as having high current capacity which would probably account for the number of caps in the circuit design.
I like eating the thermal paste ! Reported you to Dig-Key LOL Without speaker protection you risk getting DC at your speakers,/ This could have been the cause of the transistors blowing as they took out the speaker (s) The shipping cost to the US makes it kind of problem
You can't replace those output transistors, , you are going to just gerry-rig the unit to work correctly which in my opinion is not fixing it correctly The manufacture never intended for that unit to work in 50 years , theyr knew SMD would replace hold through components I worked in electronic manufacturing years ago, and all that electronics is 100 percent oblate by today's standards Why did you choose that transistor as a replacement ? This is constructive criticism
Off course this all depends on the parameters of the replaced transistors. Lot's of older n.l.a transistors have very good modern replacements. If someone does their homework right, there is really no problem in doing so.
@@nicodenhaak3961 The original trasistor is still made, I mean the ONsemi company still makes the technically same identically spec to the original ones, not sure why he used other ones and what advantage that would have. I dont understand why he does not replace the resistors that are known to fail, the unit will fail very soon and will take some more transistors out with the next defect. Then you can start from scratch.
@@TrevorsBench Jerry Rigging because the OEM choose that transistor types for a reason as the final stage I know about transistor substation their was a lot of books in the eighties about cross referencing But 40 0r 50 years later that kind of thing is obsolete because the semiconductor manufactures discontinued so many types making it much harder to cross reference.
Bad work this time, very bad. You did not change any of the resistors on the vertical board??? I saw so many broken, there is literature and articles about them over the whole internet. I just switched it on for testing a few times and they blew, blew some transistors nearby. This amp could have also been thrown in the bin, if you use it like that, it will kill itself anyway, waste of time for a bad job that will maybe hold a few more months. Sad for the unit, I like those units, but this is not a lasting repair, I would not suggest ANYONE to buy this or run it like this. This is a receipe for desaster, the knowledge to work on 870 is clearly not there. And the fuse cover must be spread apart, and not pushed together, dont ask me how I know but I tried a bit too hard the wrong way, very flimsy board that is. Also the instruction is not very good, because the four metal covers dont fit with the bigger transistors that he used, so not sure if he left them out or bent them or what, not very clear video and unclear in many places. He said in the end he did not do it, how can you make an instruction video where half of the work is not done. Sad, could have been a proper video for 870 repair and maintenance, the crusty paste is something I also came accross, but if you go after that video, you can leave it as it is anyway, will not make any difference on the result, the 870 will fry itself and that was it. Then it is better not invest work and let it die than to make a bad job and let it die anyway. No care taken here.
Help me out here. I'm trying to understand why you would want to trash my work or this amp. Is it to prevent a possible sale? Are you trying to provide some kind of service for the other viewers?
I’ve got an HK 870 and love it. Was my first stand alone amp bought in the early 80’s. The board connectors went bad and had my local audio shop rebuild it which included some of the power transistors (late 80’s). Serviced my father for several years powering Magnepan LRS 4 ohm speaker with no problems. Then it made it back to me.
Good old Harman Kardon wave soldering and corrosive glue. Nice work as always Trevor.
Awesome job, Trevor.
170 kHz at 100W?! That thing is a longwave transmitter. 😄
Always loved the look of the plain box style amps.
Nice job.
When you are poking around at 29:00, it looks like the PCB flexes around the heat sinks the transistors are screwed to. Bad design, it means that any forces that cause PCB flex will be transmitted directly into those solder joints (which were broken).
Beefy amp today, weird comments. I like the copper braid that you laid in there for the grounds, that's a nice touch.
Great video! It was nice that there were some good replacement transistors available. Should give someone lots of years of service now.
I believe a lot of the hk stuff in that era used direct coupled low negative feedback stages. So, the only limiting factor for frequency would be the transistors chosen for each stage. Some of them even had ultra wideband printed right on the front.
Another good design assembled poorly. Will make someone happy now. I actually bake sinks with transistors installed at 100°C in the oven and torque them down while still hot, compounds get runny and fill the gaps, while length contraction ensures they stay tight at high operating temperature. Bit tricky, but works great, especially with class A and high bias AB fet outputs.
Nice to see you back, this is a interesting specimen. 1983-1986............Not HK's best years for QC. Added a film cap to every service cap, hmmmmmmmmmm. Have every right to complain about that 1/2 ass white compound Goop. Never ever seen that light Blue stuff you purchased, very $$$$$$$$ Hell yes, those are big holes and little plastic support. Thanks for pointing out the thickness of the legs on those Burgundy colored Caps.
Great work! I would have made an offer but the shipping would kill me!! All the best,
Thanks Trevor enjoyed the video, all the best
Do you still have the harmon kardon hk 870 that you rebuilt for sale? I'm interested.
Sold?
I used highest thermal conductivity compound with Peltier coolers. It is very effective, but after e while it was impossible do remove coolers from the heat dissipator. The ceramic in the compound was like cement. I had to break the coolers.
It was probably thermal adhesive. You can buy paste that will set up and harden with heat
I think the thermal expansion of the long heatsink is ripping the joints apart in tension .On conventional heat sink arangements the legs flex sideways. I would have bent a cammel hump into the legs between the transistor and the solder pad to allow a bit of give to save the joint. But otherwise great job.
Never saw that style of transistors before.
When I saw the two sets of big capacitors I was thinking it might be class G bias with a twin split power supply rail. I know Kenwood, Carver, and Hitachi did those designs. Have to see if Hifiengine has a schematic on this amp, would like to see the roadmap on this one. The emitter resistors may have gone up in value due to the heat and abuse, that might account for the drop in wattage. Then again, it may just be the way it is. Thanks for the ride along on this one.
The emitter resistors all checked fine. They get tested because of the shorted transistors.
@@TrevorsBench 👍
If I remember correctly, you will get maximum heat dissipation in the output transistors on an analog amp at about 40% of rated output, because at that point, you have the highest product of current and voltage across the transistors. It should really cook then.
@@ancienttech4636 Depends on bias and output stage topology, Douglas Self published detailed results in his Self on Audio compendium.
@@paulb4661 You're right, but this is either a true or quasi-complementary output design, so it's more than likely some form of Class AB.
@@ancienttech4636 True, just wanted to clarify, that since analogue amp definition encompasses a wide variety of topologies and classes, the figure may be anywhere from no load to 80%
From where I'm sitting the problem with the transistors not sitting flat on the heatsink is due to the heatsink not being flat. There are signs that the heatsink is very slightly distorted, I would have been tempted to try to correct that so you don't need so much heatsink compound to fill the gap.
It's amazing that it lasted long enough to dry out the old compound, those transistors must have been very robust but not indestructible !
The aluminum extrusion was flat. The old transistors had a slight dish to them which prevented them from making good contact. Not sure why or how they got dished
@@TrevorsBenchI suspect there might have been too much paste applied and/or the fasteners were over-torqued; or the paste that was applied ran out from near the fasteners when the amp heated up and/or aged. Simply put, the paste/heatsink arrangement didn’t age well.
Harmon Kardon used to market their amps (even the smaller ones) as having high current capacity which would probably account for the number of caps in the circuit design.
This amp is likely 4 ohm stable. I never checked at 4 ohms but it's built for the extra current
Where did you buy the transistors from?
LCSC
Guess it fun to do. Myself I would not waste my time on it. Life is short and so many things to do.
for dry thermal paste use alchool . any type will do . and it will come off with a cloth
In this case alcohol wouldn't touch it, Only the brass brush and tetrachloroethylene would remove it
Aha, the oldfashion tri. Not easy available here anymore
What the heck happened last Friday I was worried about you eh !
It's the summer schedule
I like eating the thermal paste ! Reported you to Dig-Key LOL Without speaker protection you risk getting DC at your speakers,/ This could have been the cause of the transistors blowing as they took out the speaker (s) The shipping cost to the US makes it kind of problem
I cannot believe the output device soldering method. Tack-soldering??!? Really? Doom by design.
Great design, poor execution.
You can't replace those output transistors, , you are going to just gerry-rig the unit to work correctly which in my opinion is not fixing it correctly The manufacture never intended for that unit to work in 50 years , theyr knew SMD would replace hold through components I worked in electronic manufacturing years ago, and all that electronics is 100 percent oblate by today's standards Why did you choose that transistor as a replacement ? This is constructive criticism
Off course this all depends on the parameters of the replaced transistors. Lot's of older n.l.a transistors have very good modern replacements. If someone does their homework right, there is really no problem in doing so.
@@nicodenhaak3961 The original trasistor is still made, I mean the ONsemi company still makes the technically same identically spec to the original ones, not sure why he used other ones and what advantage that would have. I dont understand why he does not replace the resistors that are known to fail, the unit will fail very soon and will take some more transistors out with the next defect. Then you can start from scratch.
@@stefanegger
As i understand it, he checked the resistors, and they where in factoryspec.
HK units are great, but not always easy to get right.
I picked those transistors because they are the best fit for this application. Since when do people call substituting transistors 'jerry-rigging"?
@@TrevorsBench Jerry Rigging because the OEM choose that transistor types for a reason as the final stage I know about transistor substation their was a lot of books in the eighties about cross referencing But 40 0r 50 years later that kind of thing is obsolete because the semiconductor manufactures discontinued so many types making it much harder to cross reference.
Bad work this time, very bad. You did not change any of the resistors on the vertical board??? I saw so many broken, there is literature and articles about them over the whole internet. I just switched it on for testing a few times and they blew, blew some transistors nearby. This amp could have also been thrown in the bin, if you use it like that, it will kill itself anyway, waste of time for a bad job that will maybe hold a few more months. Sad for the unit, I like those units, but this is not a lasting repair, I would not suggest ANYONE to buy this or run it like this. This is a receipe for desaster, the knowledge to work on 870 is clearly not there. And the fuse cover must be spread apart, and not pushed together, dont ask me how I know but I tried a bit too hard the wrong way, very flimsy board that is.
Also the instruction is not very good, because the four metal covers dont fit with the bigger transistors that he used, so not sure if he left them out or bent them or what, not very clear video and unclear in many places. He said in the end he did not do it, how can you make an instruction video where half of the work is not done.
Sad, could have been a proper video for 870 repair and maintenance, the crusty paste is something I also came accross, but if you go after that video, you can leave it as it is anyway, will not make any difference on the result, the 870 will fry itself and that was it. Then it is better not invest work and let it die than to make a bad job and let it die anyway. No care taken here.
Cry a bit more. Boo hoo 😢😢😢😢😢😢😢😢😢😢😢😢😢😢😢😢
Help me out here. I'm trying to understand why you would want to trash my work or this amp.
Is it to prevent a possible sale? Are you trying to provide some kind of service for the other viewers?
@@TrevorsBench No he's trying to be an arsehole and is doing very well. Ignore the twat.