I think I used a pico oscilloscope for my work a couple of years ago.... when I had to capture the very fast and intermittent 24V DC bus transients.... You can capture a lot of data... limited by the size of the laptop hard drive (on which the software runs...) It wasn't very expensive. You can set all sorts of triggers that can capture a lot of useful data / create a lot of useful events. The inputs can be configured for different ranges; digital inputs as well as analog inputs. I have Fluke 105B from my early (young) days when I was playing with audio/electronics in general. But I think you should give yourself a chance with pico scopes/meters... they are exactly what you need for the job at hand. Then, connect the inputs to desired points on your PCB while taking care of setting the inputs for digital or analog type, and a proper min&max range.
Is there a certain model I should look for? Perhaps something on eBay? I have an older Fluke 177 DMM that still works well. Just wondering if I can use it to perhaps probe something that may be bad? Thanks!
Excellent photos. Could this be a thermal design problem? Could it be a thermal protect trigger? How is pressure applied to the aluminum plate in order to maintain contact with all the heat sinks as the system heats and cools? Are there springs with the screws to apply even pressure? Were any of the mounting screws loose? After months or years, the contact between the plate and heat sinks degrades and thermal transfer becomes inadequate. From what I have read, these amps generate a fair amount of heat. They may have been able to get away with this thermal design with the lower power amps, but not this one. The 10% failure rate experience of Apollon is crazy.
I think I would try to just clean off the old thermal paste and keratherm. Reapply fresh paste to the heat sinks, remount the plate and see if it lasts longer until standby. You could also place the aluminum plate on a cold surface and see if that helps. If it is a thermal issue, the intervals from power to standby should decrease as things heat up. Turn the amp off and let it completely cool down, the interval should be the longest.
I think I would try to just clean off the old thermal paste and keratherm. Reapply fresh paste to the heat sinks, remount the plate and see if it lasts longer until standby. You could also place the aluminum plate on a cold surface and see if that helps. If it is a thermal issue, the intervals from power to standby should decrease as things heat up. Turn the amp off and let it completely cool down, the interval should be the longest.
Last edited:
They have blue keratherm and white adhesive thermal tape of some sort attached to the plate. They must have had problems with the plate maintaining contact with all of the heat sinks as it heats and cools. Or transformer vibrations are transmitted and amplified by the plate causing noise.
On CPUs, the heat sink is spring mounted to the CPU so that there is even pressure. Maybe springs on all of the mounting screws, then you could use just thermal paste.
On CPUs, the heat sink is spring mounted to the CPU so that there is even pressure. Maybe springs on all of the mounting screws, then you could use just thermal paste.
Last edited:
Possibly a few rubber washers between the screw head and plate. Like these.
10% failure in electronics design resulting in parts failure? I just don't see it. 10% failure after a few months due to a thermal design problem? Yes.
10% failure in electronics design resulting in parts failure? I just don't see it. 10% failure after a few months due to a thermal design problem? Yes.
Last edited:
There was definitely a noise problem, because they have a foam rubber block on the plate that presses on the bottom of the power supply transformer through the hole in the PCB. They also have dampening material on the plate under the power supply area. These are problems that don't show up in the CAD process. They obviously had some challenges based on the resulting Rube Goldberg design. The main noise problem is the PCB mounted transformer that is vibrating the PCB. If the transformer was chassis mounted with wire leads, they could have better mechanically isolated the transformer mounting from the PCB.
Last edited:
I can try and answer a few of your questions. As far as thermal design and the thermal protect trigger, that's a little Greek to me..lol. The bottom heat sink plate attaches to the pcb module by 18 machine screws. There are no rubber washers nor springs involved. When I first started loosening the screws, they were not difficult to remove. It is possible that some could have been tighter than others. I had to use a T-10 torx to get them off. I am not sure if that is the proper size, but it did work. I fairly sure that it could be a metric size. I may try and run up to the hardware store and see if I can find the proper size so when I do put the screws back, I can use my cordless Ryobi screw driver that can be adjusted to torque screws for a more consistent pressure? That way I will know that all of the screws should be evenly tightened. I did want to add that when I first mounted the module to the bottom of the chassis, I was told that I didn't need any thermal paste!? Like I said, I never had a problem with the amp module for years until perhaps 6 months ago?
It does appear that now when the amp is first turned on, it does take a couple of minutes before it starts going into standby, then it will continue once the amp has been on. So I'd have to say it seems like it is a thermal problem. I also tried measuring the transistors to see if I could get consistent numbers on my DMM, and I did find one transistor to be a little off. Not sure if that can help you determine anything?
It does appear that now when the amp is first turned on, it does take a couple of minutes before it starts going into standby, then it will continue once the amp has been on. So I'd have to say it seems like it is a thermal problem. I also tried measuring the transistors to see if I could get consistent numbers on my DMM, and I did find one transistor to be a little off. Not sure if that can help you determine anything?
Thermal compound is know to degrade over time. There are a lot of thermal barriers between the plate and the heat sinks that can potentially go bad.
I think it would be easy to solve the thermal issue. Just reattach the plate with thermal paste. If that stops the standby issue, great. If it doesn't, then more troubleshooting (as described above) is needed.
I think the plate attached with only thermal paste will create a noise problem from the transformer vibrating. The DIY solution to that is to remove the transformer and maybe even pot it. Figure out some way of attaching it to the PCB with adequate dampening, and solder jumper wires from the PCB to the transformer leads.
It's a little hard to tell from pictures, but it looks like the transformer of the 2000AS2 is attached differently to the PCB.
I think it would be easy to solve the thermal issue. Just reattach the plate with thermal paste. If that stops the standby issue, great. If it doesn't, then more troubleshooting (as described above) is needed.
I think the plate attached with only thermal paste will create a noise problem from the transformer vibrating. The DIY solution to that is to remove the transformer and maybe even pot it. Figure out some way of attaching it to the PCB with adequate dampening, and solder jumper wires from the PCB to the transformer leads.
It's a little hard to tell from pictures, but it looks like the transformer of the 2000AS2 is attached differently to the PCB.
Last edited:
OK, so I have some Ceramique 2 tri-linear ceramic thermal compound, would that be ok to use? I take it I should remove that blue keratherm as well?
One last thing: Should I also use the thermal paste between the chassis and heatsink/module? Thanks!
One last thing: Should I also use the thermal paste between the chassis and heatsink/module? Thanks!
Yes. I would take it down to the white tape, which I guess (as Extreme_Boky pointed out) is providing electrical isolation (under the keratherm). At first, I don't think you need paste between the chassis (case) and amp. That's what you mean, right?
Last edited:
Yes correct. I didn't put any before and it worked fine, so I'll just do the same when I put the module back in the chassis. I'll report back after I do this and tell you what it is doing. Thanks again!
OK just finished up and it came on for maybe a minute, then started going crazy into standby, ON, standby, ON, and repeating. So I turned it OFF. Just seems impossible to get that warm/hot that fast, as my room is cool. So back to the drawing board..lol
I can't imagine that you did anything to make it worse. If it's a thermal issue, a fan blowing on the amp might help. If you did get an oscilloscope, the wait for data points won't be long.
The DIYA scope these days is REW and a USB audio interface. You need an attenuator to keep the voltages within the limits of the audio interface. REW gives you trigger and storage capabilities.
Maybe the best solution is to have someone that can fix the issue as it may be cheaper than buying the extra gear to trouble-shoot? Like I said, I know a little bit, but never had any experience using an oscilloscope. If you believe you can find the problem and willing to fix it for a fee, I can ship the module to you if you wanted?
I appreciate that! Perhaps you or someone here may know somebody that has perhaps worked on these 1200as2 amps? I don't really have anything local around my area that can do this. I'd be willing to invest a little to get an oscilloscope like you mentioned, but I would need some assistance.
REW is the free software. A lot of articles here on using it. You need a two channel USB audio interface like Focusrite, Steinberg or Behringer. There are a lot of them out there, and you can find them used, open box or discounted. I have used it for FFT and distortion measurements. REW has scope capabilities with trigger and storage. Oscilloscopes are also getting cheaper.