FourT6and2 wrote: ↑Wed Mar 26, 2025 5:54 pm
Fixed it! Bad cap in tone stack was causing thin/harsh sound. 470pF is the way to go.
FourT6and2 wrote: ↑Wed Mar 26, 2025 5:54 pmI did learn that Hiwatt does plan to implement the 22uF on a switch for future builds, as an option for people looking for more low end. That value does sound good now that the OT is wired properly, but it does disable the presence control.
That's not making sense to me. A very large cap will ground out all the NFB, and that might be useful, but NFB generally results in lower gain and more bandwidth, so more bass.
FourT6and2 wrote: ↑Wed Mar 26, 2025 5:54 pmI did learn that Hiwatt does plan to implement the 22uF on a switch for future builds, as an option for people looking for more low end. That value does sound good now that the OT is wired properly, but it does disable the presence control.
That's not making sense to me. A very large cap will ground out all the NFB, and that might be useful, but NFB generally results in lower gain and more bandwidth, so more bass.
What's not making sense? Yes, a 22uF cap here = more bass. If you hardwire the 0.01uF and then put the 22uF on a switch in parallel, you get a "depth" switch of sorts, at the cost of losing presence control functionality.
martin manning wrote: ↑Wed Mar 26, 2025 8:06 pm
A large cap will ground out (nearly) all of the NFB, increasing gain but reducing bandwidth. That's why the amp was stable with the primaries swapped.
Correct. Again... what's not making sense here?
Hiwatt added the 22uF to give the amp a low-end boost. I can think of better ways to implement a "depth" control, but this is how they chose to do it I guess.
martin manning wrote: ↑Wed Mar 26, 2025 8:27 pm
Where is this 22uF in the schematic? I was under the impression that it was positioned like a presence cap.
Ahhhh!
No, not a presence cap. It's after the presence circuit. Take a look at this DR504 layout from Ceriatone, for example. Note the 0.01uF cap in the bottom left corner. The one connecting the 470R to ground, after the presence circuit. This is the cap in question. You'll find it in most DR504 schematics.
This one? That is the presence cap. You have the speaker voltage feeding a 10k/2k2 divider, with that cap and the 470R in series paralleled with the 2k2. A large cap there will put the 470 in parallel with the 2k2, making the divider 10k/387, cutting the NFB by a factor of 5.7.
You do not have the required permissions to view the files attached to this post.
martin manning wrote: ↑Wed Mar 26, 2025 9:18 pm
This one? That is the presence cap. You have the speaker voltage feeding a 10k/2k2 divider, with that cap and the 470R in series paralleled with the 2k2. A large cap there will put the 470 in parallel with the 2k2, making the divider 10k/387, cutting the NFB by a factor of 5.7.
That schematic doesn't appear to match what I'm looking at. In this amp, the 2K2 is before the 470R. NFB is 100K/8ohm. The resistor providing ground ref. to the presence pot is 100R, not 470R. And there are a few other differences as well. This isn't a stock Hiwatt circuit. I can PM you the schematic if you'd like
The various HW schematics I have looked are all basically doing the same thing with respect to the GNFB and presence. PM or post yours, I'm curious to see if you have something different.
martin manning wrote: ↑Thu Mar 27, 2025 10:52 am
The various HW schematics I have looked are all basically doing the same thing with respect to the GNFB and presence. PM or post yours, I'm curious to see if you have something different.
It's the same circuit. The differences are the 100k vs. 10k or 22k NFB resistor, which reduces the NFB voltage fraction ~5-10x, and switching to the 8-ohm tap, which reduces the FB voltage another 1.4x. Thre is also the 100R vs. 470R from the Presence pot wiper to ground, but that isn't a big deal IMO.
martin manning wrote: ↑Thu Mar 27, 2025 5:37 pm
It's the same circuit. The differences are the 100k vs. 10k or 22k NFB resistor, which reduces the NFB voltage fraction ~5-10x, and switching to the 8-ohm tap, which reduces the FB voltage another 1.4x. Thre is also the 100R vs. 470R from the Presence pot wiper to ground, but that isn't a big deal IMO.
Similar yeah, I guess I wasn't thinking of the 0.01uF as "the" presence cap for the presence control, because it isn't. It's a hardwired low-pass after the presence circuit. So I view them as two independent filters. From AmpBooks:
"To the right of the presence control is a 470Ω resistor and a 0.01μF capacitor in series. They form a shunt impedance for a fixed low-pass filter for the feedback signal, thereby reducing negative feedback for high frequencies. This fixed treble boost combines with the presence control's variable treble boost to create the control dynamics that Hiwatt founder Dave Reeves wanted to achieve."
I'm probably going to put that 22uF on a switch because it does increase the "girth" of the amp quite a bit and would be nice to have. Just need to decide how to best implement. Either wire a switch to simply parallel the 22uF over the 0.01uF or place both caps on either side of a SPDT on-on and switch them out individually.
Ok, I stand corrected, the 10n is one of two presence caps, with the other one being the 100n connected to the presence pot. The 2k2 PI tail and the 10n presence cap are in parallel, and the presence pot adds or removes another 100n. The 10n and the 2k2 have a knee at around 7 kHz, and adding another 100n drops it down to around 650 Hz. The pot controls how much reduction in FB voltage occurs over that range of frequencies, and the range slides up in frequency as presence is turned up (it's a shelving filter). The pot also places 1n to ground at the plate of a preceding stage as the presence goes towards full off, reducing the HF content coming in.
martin manning wrote: ↑Thu Mar 27, 2025 8:09 pm
Ok, I stand corrected, the 10n is one of two presence caps, with the other one being the 100n connected to the presence pot. The 2k2 PI tail and the 10n presence cap are in parallel, and the presence pot adds or removes another 100n. The 10n and the 2k2 have a knee at around 7 kHz, and adding another 100n drops it down to around 650 Hz. The pot controls how much reduction in FB voltage occurs over that range of frequencies, and the range slides up in frequency as presence is turned up (it's a shelving filter). The pot also places 1n to ground at the plate of a preceding stage as the presence goes towards full off, reducing the HF content coming in.
Next step is to decide how to best implement this cap on a switch so it doesn't pop. Something like this, perhaps?
Low Pass Switch.jpg
You do not have the required permissions to view the files attached to this post.
Sent you a PM.