Running DC heaters at the same time as elevating AC heaters

[xxmicrocutsxx]

Hi. I'm hoping someone can help me get my head around something that, to be perfectly honest, is making my brain melt!

I have a PRS MT15 amp which has quite a lot of hum. This is a very high gain circuit which uses DC Heaters for both V1 and V1 (first dedicated lead channel tube). While I've been changing out some CF resistors for MF (10k input grid, 2w MF for plates etc - everything in the amp, except for the 1ohm bias resisters were CF), to try and tame some hiss, I figured I'd chase out the circuit a bit. Seems as though they've taken the Mesa approach to DC heaters and put in two faux center taps - 330ohm resistors from the +/-, after DC rectification, to ground as well as two 330ohm resistors from the AC heater lines to ground - each being on either side of the board. It's exactly what Merlin points out as being wrong about the Dual Rec and Carvin Legacy amps. I'm guessing the 330ohm resistors are because they are also used the relay/led circuit so better to buy in bulk?

Anyway - I'm going to lift out the two 330ohm resistors from the DC Heater circuit, post rectification, so that only the AC side of the heaters is referenced to ground (the DC side can get it's reference from there through the rectifier). But, I've been thinking about elevating the remaining AC faux center tap, probably to around 40VDC. But here's where my confusion sets in - the Electrolytic caps in the DC heater circuit (two 10000uF) are rated at 16v each. They are in parallel across the DC lines acting as ripple filters. If I elevate the heaters and they reference 40VDC, will the 40VDC be across the 16v rated caps in the DC heater circuit? I've attached a schematic from a SLO100 clone build where elevated heaters are used along with an optional DC heater circuit. You can see the caps in the DC circuit are rated at 16v. It was brought to the designers attention that the two 4700uF 16v caps that reference ground in the DC heater circuit were a problem when utilising both elevated heaters and DC heaters. One of the solutions brought forward was to replace them with another single 4700uF 16v cap in parallel , like the others, for extra filtering, and get rid of the ground there.

Was the 'problem' that as soon as those 16v caps had a ground reference they would see the elevated voltage across then and exceed their rating? Why are they the 'problem' and not the other two 16v caps in parallel? I just don't understand how elevating AC heaters to 40VDC and having 16v caps in the DC circuit can play nicely together? I would much rather understand exactly why and implement it that way than blindly trust that it does without any knowledge. Is it due to that circuit 'floating' and not having a ground reference that makes the 16v caps OK to be used? Or because the elevated reference voltage doesn't actually carry any current on to the heater lines - therefore the DC measure across them is actually the elevated reference and not DC travelling along the heater wires themselves?

Apologies for such a long post! Hopefully I'm just missing something obvious and the explanation is a whole lot shorter, and more interesting, than this monologue!

I've also included a grab of the mesa DC circuit where you can see two taps from the heater line. Thanks for any help and advice!

[trobbins]

That 'optional' dc heater portion in the Solo 100 schematic does seem a bit confusing.

The CT of the heater winding sits at elevated dc. If the 'option' was installed then C39 and C40 would sit nominally at +/- a dc level from the elevated dc, except for the influence of the C41/42 gnd point. Imho, C41 would charge towards nominal elevated dc level but leakage through C41 would become noticeable as that dc elevation exceeded circa 18-20V, with leakage constrained by the 100k in the elevation divider. That leakage current would be quite low, and C42 would similarly see leakage from being reverse biased.

Is the 'GND-FS' node floating, or somehow linked to 'GND'. If it was floating then perhaps GND-FS is the intended node for C41/42, but then one would expect GND-FS to have some nominal resistive connection to ground/chassis.

Note that dc elevating the heaters, and dc powering the heaters, aim to suppress different causes of hum, and the benefit of dc powering heaters may be alternately achievable from using a humdinger pot (eg. in lieu of the heater CT connecting to the dc elevated divider).

[xxmicrocutsxx]

Thanks a lot for your reply!

So that tallies up to my thinking. In that schematic, C39 and C40 would sit at the elevated DC level. Another 'solution' put forward to that circuit was to just remove C39, C40 and that ground altogether. That theory is backed up by a schematic I found that was provide by the owner at amoclones.com a number of years ago (attached). You can see three 4700uF 25v caps in parallel in the DC circuit, no ground and with the AC elevated to 50VDC.

When looking at that, once again, my mind is confused as to how 25v caps can be used when there's 50+VDC being referenced.

The circuit in my MT-15 amp doesn't have elevated heaters right now and doesn't incorporate any ground from the caps in the DC circuit like that Solo 100 does (just the faux tap ground). The main thing I guess I want to confirm is that if I implement a very similar scheme to the one I've attached to this reply, will those caps be OK and why will they be OK?

I think the 'GRD-FS' node connects to a chassis ground in the layout for the Solo 100.

Since the DC heaters are already installed from the factory in my amp, I will keep them. it's just figuring out what I can do with the AC heaters to improve them without blowing caps.

[Stevem]

Is this the mesa dc supply circuit for V1 you speak of?

What help might it be to add these two caps to the heater supply ?

[xxmicrocutsxx]

The Mesa example was just to illustrate what the DC heater supply in my amp resembles - along with the doubling up of faux center tap, no-no'd by Merlin. My main concern it trying to ascertain, as per the schematic I attached to my second post, how you can elevate the AC heaters to 50VDC and at the same time use paralleled 16v/25v electrolytic caps in the DC heater circuit.

I think @trobbins was on to something with the cap leakage though. In an updated SOLO 100 schematic you can see a note saying to drop the C41/C42 caps, which go to ground, to 100nf/100v. I assume this is because the voltage that's charging and leaking to ground would exceed the original caps voltage ratings? What i don't understand, however, is the electrical science behind it - especially given some schematics show two caps and a ground reference in the DC heater supply, whilst others (including inside the MT-15), don't have any direct ground reference for the paralleled caps....... two ways to skin a cat and I'm looking for the instructions of how to make the knife.

Sorry, I don't think I'm being much help in my replies, but I appreciate you taking the time to reply!

[trobbins]

I think the leakage current, even with the e-caps either exceeding their voltage rating, or reverse biased, would have been too low to cause short-term damage. Certainly changing the e-caps to 100nF/100V would avoid any cap stress.

Those two caps to gnd provide some transient current path arising from the diode rectification process. If you have a chance then use UF400x for the bridge diodes to minimise any transients from taking an alternate path through the valve heater-cathode interface.

[danman]

@xxmicrocutsxx...I also had a lot of hum with my MT15. I found this post on another forum and after replacing the v6 effects loop tube with a 12au7 and the v3 tube with a 5751, the amp is now much quieter and actually sounds better.
https://sevenstring.org/threads/prs-mt1 ... um.336431/