There is, here are the IRF540 gates with bias CCW.
Ch1
Ch2
Ch3
Ch4
When I turn up the bias it causes a higher frequency oscillation. If I turn it slightly higher it smooths out, and if I turn it CCW it oscillates at a higher amplitude as I approach fully CCW. This is Ch 3 for instance:
Bias fully CCW:
Bias turned up about "normal"
Bias turned back fully CCW
It seems to do this on all channels, with some of them returning to their idle state at fully CCW
Ch1
Ch2
Ch3
Ch4
When I turn up the bias it causes a higher frequency oscillation. If I turn it slightly higher it smooths out, and if I turn it CCW it oscillates at a higher amplitude as I approach fully CCW. This is Ch 3 for instance:
Bias fully CCW:
Bias turned up about "normal"
Bias turned back fully CCW
It seems to do this on all channels, with some of them returning to their idle state at fully CCW
I looked at the diagram again. Why Rx39?
Until you resolve the problem of the oscillation, you're going to continue having this problem.
How distorted is the output signal with the bias fully CCW?
Until you resolve the problem of the oscillation, you're going to continue having this problem.
How distorted is the output signal with the bias fully CCW?
Just grasping at straws and noticed the differences, thought it might lead somewhere.
What should I drive into it to look at the output? It plays audio without noticable distortion.
What should I drive into it to look at the output? It plays audio without noticable distortion.
Input source:
Output with bias fully CCW:
Output with bias around normal position:
Output with bias turned back fully CCW:
Output with bias fully CCW:
Output with bias around normal position:
Output with bias turned back fully CCW:
I haven't - I didn't think I had any suitable replacements on hand (short of some 0.01uF, but that seemed kind of high), but it looks like I could steal some 5pF caps from a trashed 800.2 board.
I also simply don't have the understanding to see why that would be a viable solution when I have another amp that doesn't do this, using the same value parts? Unless you think the parts are bad?
I also simply don't have the understanding to see why that would be a viable solution when I have another amp that doesn't do this, using the same value parts? Unless you think the parts are bad?
Are the amps the same revision number?
4.7 to 5 wouldn't make a difference but you could try paralleling them.
4.7 to 5 wouldn't make a difference but you could try paralleling them.
No, not the same revision. Working one is a 'C', this problematic one is an 'E'. They have different manufacturer components, but I haven't found any difference in face values on anything
Last edited:
The solution on this one may mean taking quite a bit of time and several failures to find something that works.
The A06 > A42 mentioned earlier may be the solution.
The A06 > A42 mentioned earlier may be the solution.
Could I lift some components to sufficiently isolate a given channel to more easily see where the issue is occuring?
In order I lifted R316, R416, R216. Still had oscillation & the bias oddities on Ch 1, so I lifted R116 and replaced R216...oscillation remained on Ch2, so that didn't help eliminate any one channel.
Hey Perry,
The best I could find on hand were 270pf smd caps. I switched out C113 & C115 with those. I am able to bias the channel more normally now, and output on the scope seems to be comparable to the other 400a4 I have.
The best I could find on hand were 270pf smd caps. I switched out C113 & C115 with those. I am able to bias the channel more normally now, and output on the scope seems to be comparable to the other 400a4 I have.
With that much of a change, it may have limited the frequency response of the amp. You need to run a 20Hz-20kHz sweep to confirm that the frequency response is still OK. It doesn't need to be more than enough signal than it takes to get a volt of output.
Clamp the FETs down before doing this.
Clamp the FETs down before doing this.
Seems to be okay through the sweep. I can order some other lower values to try - what would you recommend?
Still bothers me that it does this while the other amp doesn't. I also previously switched out the 3 bias transistors on this channel without change, not too surprisingly.
Still bothers me that it does this while the other amp doesn't. I also previously switched out the 3 bias transistors on this channel without change, not too surprisingly.
I would recommend ordering the next higher 3, 4, 5... values, at least 10 of each if you do any kind of repair work, regularly and start by installing them (lowest value first) to see the minimum needed and use the value just above the absolute minimum that works.
Post the results if you do this. As far as I know, you're the only one to resolve this problem. Others may be interested.
Post the results if you do this. As far as I know, you're the only one to resolve this problem. Others may be interested.
Mouser has them available in every tenth of a pF - what would be a good range? My tester supposedly is accurate to 1% down to 20pF, and unclear below that. When I test the parts I pulled they both show around 8pF (supposed to be 4.7pF)
I don't know what your budget is or how much time you are going to spend on this. Looking at the document attached, the E6 series of values would probably be the minimum. I'd suggest buying values up to the value you're currently using since you know those work.
The original caps are RF part CDS-047CAAA, 4.7PF 50 VOLT 5% NPO 0805. Never heard of E-series before you mentioned it, but wouldn't 5% put it into E24? Or are you saying more along the lines of using the E6 20% tolerance range steps, but with 5% tolerance parts? I am trying to balance having the necessary parts vs a lot of inventory I'll never use. I pretty much only work on this style amp.