DC Heater Options

[TenderTendon]

I have successfully built 3 amps in the past, but they all have simply been a result of following instructions. My next build is going to more complex and I'd like to learn as I go. By the time I'm done, I hope to have a good understanding of every component function, along with how they interact with other components in the circuit and how to optimize for a desired outcome. I am also learning how to use LTSpice, as I feel it's a great tool to help those with weak math, like myself.
  My next build is a Soldono SLO-100. I'm have a couple decisions to make before I start and have a question about DC filament supplies. In the schematic I will be following (attached), there is an option to supply the first 2 pre-amp tube filaments with DC. Seeing how this is a very high gain amp, I feel it's worth pursuing. The PCB I am using already has a DC filament supply section, so I see no reason not to utilize it. I have discovered through LTSpice, that with the transformer I will be using (Hammond 373HX), additional capacitance and tweaking the dropping resistor, that I can power all 5 of the 12AX7 filaments with the same voltage and approximately half the ripple. If my Spice models are wrong, let me know and I'll chalk it up as a lesson learned. If the results are accurate, are there any reasons to NOT power all of the AX7's with DC? It would surely simplify wiring and theoretically reduce noise. I am hoping the answer is "Yes, it will work great and it's the best option", but I feel a lesson is about to begin. I welcome both though...

[sluckey]

I suggest following the solo schematic and only run V1 and V2 on dc. Notice that the solo has the ac heaters elevated? This is helpful because the tonestack driver cathode follower (V4B) has a high dc voltage on the cathode. Elevating the filaments to a dc voltage prevents exceeding the max cathode/heater voltage.

[TenderTendon]

Yes, I noticed the elevated heaters, but I thought it was used to minimize hum. I didn't know about the cathode/heater voltage limit. Is this voltage limit in place because of the close proximity of the two? To prevent possible arcing?

[sluckey]

Yes, I noticed the elevated heaters, but I thought it was used to minimize hum. I didn't know about the cathode/heater voltage limit. Is this voltage limit in place because of the close proximity of the two? To prevent possible arcing?

Yes and yes.

Look at page 2 of the linked pdf...

      http://www.tubebooks.org/tubedata/HB-3/Receiving-Type_Industrial_Tubes/6681_12AX7.PDF

[TenderTendon]

Thank you. I guess the question now is if I should use elevated AC on all 5, or run DC on the first 2? I assume the only way to find out would be to try it both ways, measuring the noise of each on a scope. I know the answer to this question could be quite subjective and vary from one designer to another, but given the option, which would be preferable for the input stages of a very high gain amp? This is assuming that both options are readily available and no additional time or cost is involved doing it either way.

[sluckey]

I suggest dc heaters for V1 and V2 only.

[TenderTendon]

OK, I decided to heat V1 and V2 with DC and the rest with elevated AC as recommended. I have been looking at that schematic for hours and I just can't figure out how it can work. My attempts at modelling the heater circuit in Spice, as shown on the schematic never come out as expected either. Here is my concern, or maybe just my confusion... If you look at the elevated AC heater portion of the schematic and the optional DC heater portion, what prevents the elevation DC voltage (+40v or so) from passing through the bridge rectifier and over-volting C41 and possibly others?

[Rontone]

http://www.valvewizard.co.uk/heater.html

Check out the elevation section and last section on DC heaters, about sharing the same 6.3v line with AC and DC and the rectifiers ground reference

The current draw goes up on the DC powered heaters too,

[TenderTendon]

Thanks, but I have read that before. It does not address what I am noticing. I believe there is a mistake on the schematic.

[sluckey]

Show me a current path that concerns you.

[TenderTendon]

Show me a current path that concerns you.

+40 or so volts from elevation voltage divider (C36), through transformer windings, through bridge rectifier U4, through dropping resistor R73, hitting filter cap C41, which is grounded.

[sluckey]

DC voltage current cannot flow through a capacitor.

[TenderTendon]

DC voltage cannot flow through a capacitor.

I don't understand what you are saying here. What I am trying to explain is that C41 is going to have over 40 volts applied to it. It is only a 16v cap. Spice simulation confirms this.

[sluckey]

I suppose your only option is to use AC to heat all tubes.

[TenderTendon]

I suppose your only option is to use AC to heat all tubes.

Do you agree that there is a mistake on the schematic, or am I losing my mind? I don't understand why using AC for all tubes is the only option. It seems that this can be fixed by simply re-working the DC filter section and removing the ground reference that is on C41. It seems that no one has actually looked at the schematic..

[Rontone]

I'm not sure but isn't the 6.3v 'riding' the DC elevation, so its alternating between +36.85v and +43.15v with +40v as its 'zero' reference when it has been elevated, so is that rectifier only seeing 6.3v of swing?

According to Merlins site and book, the reference, whether standard ground or elevated is used should be on one side of that heater rectifier not both, that " the other side will received its reference through the rectifier."

So possibly that is a mistake on the schematic and the ground between C41 and C42 could be left off

[TenderTendon]

So possibly that is a mistake on the schematic and the ground between C41 and C42 could be left off

Thanks for taking the time to reply, but I was hoping to find out if is a mistake, not just a possible mistake. Because if it is, it's also a safety hazard (unless hitting 16v caps with 40-50 volts by design is OK). If it's a mistake, the designer of the schematic and the PCB should be notified. I am not qualified to make that decision. I am here to consult with those who are.

[glass54]

Throw my hat in the ring. 
I would question the use of C41 and C42 in their present form (Reply #12 and #13). All the serious filtering is done by C39 and C40.
However if C41 and C42 were 100nF/100V rated I would be happy. This would (couple)/decouple DC Heaters to ground should any spikes/noise etc occur.
My thoughts (without the use of Spice or consulting Merlin).
Regards
Mirek

[TenderTendon]

Throw my hat in the ring. 
I would question the use of C41 and C42 in their present form (Reply #12 and #13). All the serious filtering is done by C39 and C40.
However if C41 and C42 were 100nF/100V rated I would be happy. This would (couple)/decouple DC Heaters to ground should any spikes/noise etc occur.
My thoughts (without the use of Spice or consulting Merlin).
Regards
Mirek

Thanks for those thoughts. I've already abandoned the ground reference after the rectifier and re-purposed C41 & C42 in parallel, as an additional node with another 1 ohm dropping resistor. This gives me slightly over 6vdc to V1 and V2, with less than 15mv of ripple. By bumping up the values of C39-42, I can get the ripple below 5mv, but don't know if there would be any benefit from that. All of these figures are based on Spice modelling. The PCB won't be here for a couple more days. I do understand that these DC rails will be floating in space. I'm not sure if I will have do deal with that or not.

[acheld]

I'd ask Andrea Antonello (the board designer) to explain . . . 

He is a) very experienced, and b) takes pride in his boards, c) knows his stuff, and d)  late to answer emails at times.   You can also post on his company's forum, and he will answer.  His boards are very high quality in general, though I have no experience with this one.

All that said, I'm not clear why C41/42 are referenced to ground at all, since the center tap is referenced to ground already (albeit through that 100K resistor).   

[TenderTendon]

I have already sent him an email. Thanks!

[glass54]

+1 acheld
Would it not make more sense to ref C41/42 at R60/C36 (and junction of R59)?
Keen to see if TT get an explanatory email. We may learn something that we haven't tried before.
Regards
Mirek

[sluckey]

Do you agree that there is a mistake on the schematic, or am I losing my mind? I don't understand why using AC for all tubes is the only option. It seems that this can be fixed by simply re-working the DC filter section and removing the ground reference that is on C41. It seems that no one has actually looked at the schematic..

I had overlooked the voltage rating of C41. Looks like you have a valid concern. I believe that C41/C42 are intended to provide an AC ground reference for the DC heater supply for filament hum reduction.

Since this circuit is listed as optional I wonder if it has ever been built and tested?

[acheld]

Since this circuit is listed as optional I wonder if it has ever been built and tested?

The good thing about his PCB's is that they allow for "options" in the build; the bad thing is that is that you really have to go through the schematic and know exactly what you are doing -- and avoid pitfalls like this. 

I built one of his Dumble clones awhile back, and it really helped me "start" to understand what goes on in those circuits.  Great learning exercise, and the amp (after being neutered into a 6V6 machine) has turned into my daily driver!

If I were building this, I'd delete C42 and the associated connection to ground, and take the negative lead of C41 over to the negative rail.  I'd hope the heater CT reference to ground through R60 would be ok . . . 

In fairness to the designer, he does call out the heater center tap issue, but it's not as clear as it might have been.

[TenderTendon]

I had overlooked the voltage rating of C41. Looks like you have a valid concern. I believe that C41/C42 are intended to provide an AC ground reference for the DC heater supply for filament hum reduction.

Since this circuit is listed as optional I wonder if it has ever been built and tested?

My thoughts exactly. The PCB and schematic appear to have been released very recently, so I may be the first to build this, with or without the DC option. I can't find one single discussion about this PCB in any of the forums.

[TenderTendon]

  I've taxed my brain way too much, trying to figure out the best way to combine elevated AC and DC using the same windings. None of them seem to look good. No matter what I try (in Spice), with or without any ground reference after the rectifier, there is always very dirty +42vdc (the AC heat elevation) riding on both of the DC rails (referenced to ground). I'm guessing the rectifier diodes are chopping up the elevation voltage on every reversal of the 6.3vac wave passing through. Adding a ground (or elevation voltage) reference after the rectifier just makes things worse in one form or another.
  I think I'm learning a lot from this, but I'm nowhere near capable of figuring out a solution. The best solution my 60 year old brain can come up with is a small 1:1 isolation transformer just before the rectifier. That may be a terrible idea and not work at all. Like I said, I have MUCH to learn. I'm sure you all have better ideas. I'm just going to wait to hear back from Andrea. Tracking shows the PCB is in New Jersey, so I should have it in a couple days. I have all the components here, so I'll be able to wire up the heater circuit on the bench and to some actual testing.

[sluckey]

This guy posts over on The Amp Garage under the name frusciante89. Here's the post when he first announced this pcb in December 2022. He was last active on Tue Feb 07, 2023 10:26 am.

     https://ampgarage.com/forum/viewtopic.php?f=2&t=35893

I'm a bit cautious about buying V1.0 of most anything.   

[frusciante89]

Hello everyone! I’m “the guy”, Andrea.
As you might have seen, the slo100 is a relatively recent board. I have designed over the span of three months last summer with the goal of including all the most common mods variations, including DC heaters, AC elevation and relay/ldrs. At the moment I am travelling for work leave and I’ll be back in around a week: I don’t recall any particular errors/mistakes in the section you are mentioning, but the board and the documentation is relatively new, hence a typo might be present.
I will go through the schematic and layout once again once I’m back - in the meantime, I’ll try to reply here if needed (I’m in Asia and have very limited connection)
Andrea Antonello 

[frusciante89]

This guy posts over on The Amp Garage under the name frusciante89. Here's the post when he first announced this pcb in December 2022. He was last active on Tue Feb 07, 2023 10:26 am.

     https://ampgarage.com/forum/viewtopic.php?f=2&t=35893

I'm a bit cautious about buying V1.0 of most anything.   

I’d normally agree with you on not buying things that are V1.0, but just for a bit of context, I have 7 boards in my portfolio and all of them worked at the first iteration :)
The fact that some of the boards (like the Dumble) are at a higher version is simply due to aesthetic changes (silkscreen font, pcb colour, etc.)

[TenderTendon]

I will go through the schematic and layout once again once I’m back - in the meantime, I’ll try to reply here if needed (I’m in Asia and have very limited connection)
Andrea Antonello

Thanks for jumping in Andrea. Have a safe trip. I (possibly all of us) look forward to hearing your thoughts when you return.

[aaelectronics]

Hello everyone! I finally had some time to look at the schematic in the airport lounge :)

As a clarification, the AC elevation and DC heaters were originally intended as two mutually exclusive mods, inspired from successful builds of SLO100 clones (from personal experience and from people at sloconeforums.com). I think nothing prevents from implementing them together (AC elevation AND DC heaters).
However, if that's the case, what TT is pointing out is indeed correct: when implementing both AC elevation and DC heaters, C41 and C42 must be treated with care.
Worth noting that the negative pad of C41 and the positive pad of C42 are connected to the PCB ground pour.

Hence, I can think of three options (chime in if you see more):
- skip C41 and C42 altogether
- (to provide extra filtering) keep C41 and C42, keep the reference to ground, but change values to 100nF @100V
- (to provide extra filtering) skip the reference to ground and connect a single axial cap between the positive pad of C41 and the negative pad of C42. Anything in the range of 1000-4700uF@16V should be fine.

let me know if the above makes sense!

PS: I have a couple of people who have already successfully completed the build, I will share some pics of the shared builds as soon as they share them with me

[TenderTendon]

- skip C41 and C42 altogether
- (to provide extra filtering) keep C41 and C42, keep the reference to ground, but change values to 100nF @100V
- (to provide extra filtering) skip the reference to ground and connect a single axial cap between the positive pad of C41 and the negative pad of C42. Anything in the range of 1000-4700uF@16V should be fine.

Thanks.

[aaelectronics]

- skip C41 and C42 altogether
- (to provide extra filtering) keep C41 and C42, keep the reference to ground, but change values to 100nF @100V
- (to provide extra filtering) skip the reference to ground and connect a single axial cap between the positive pad of C41 and the negative pad of C42. Anything in the range of 1000-4700uF@16V should be fine.

Thanks.

Hello TenderTendon, were you able to sort this out? Let us know :)
Cheers from London,
Andrea