JCM800 build - Filter cap inrush current proposal, maybe

[pjd3]

Hi all,

OK, so here comes a JCM800 2204 build and at first, the rumination is, "To standby or not standby".

My first 2 amp builds I put in a standby and have no reason to regret it.

For this 2204 build, I'm going stock for starters on the Power supply with a couple of 3 amp/1000v diodes into a pair of 50uF caps that are wired in parallel. Thats what I'm seeing, right? So it looks like we are looking at 100uF ready for some current at startup. Am I wasting precious mental energy thinking that this is alot of inrush current for a pair of Rubycon 47uF 500volt caps?

Anyhow, lets say that it is alot of inrush current for 2 of those caps at start up. I was discussing this with a fellow tech at my job who is also a musician tube amp user that used to work repairing tube amps. He believed it was feasible to choose a resistor value to put in series after the 2 diodes on the way to the first filter cap pair, that would decrease the "charge rate" of the cap to a certain extent. Then, you simply actuate a normal switch that would bypass the resistor. I had no reason to argue that this could be means of making life a little easier for when the amp is initially turned on, and very possibly extend the life of well, filter caps and maybe even make life better for the diodes and anything else that is serving to feed the caps that would initially look like a closed circuit for a window of time upon start up.

Does anyone refute the wisdom of this approach? Am I missing any aspects of consideration that would make this a bad idea, or at best, not very effective? My best thinking tells me that the right choice of resistor value could offer a much nicer start up event for caps n things wait a few seconds, and toggle a switch to bypass the resistor for normal operation.

Thanks for mulling this over with me. It just seems like a relatively large capacitance to hit at start up.
Best,
Phil D.

[pdf64]

Merlin’s ‘trickle current’ resistor http://www.valvewizard.co.uk/standby.html

[BrownIsound]

I was contemplating repurposing the standby switch as a “tube rectifier”/ silicone diode setting with resistor(s) and having the other half of the dpdt switch adjust the bias with a resistor to keep the bias current the same in both modes. The “tube rectifier“ setting giving you a slower inrush.

This is for a superlead fwiw.

[pdf64]

Without additional series resistor/s, a superlead capable PT would probably have too low supply impedance for a valve rectifier.
The cap value limit is probably suggested on the basis of that being sufficient to prevent excessive anode current.
Bear in mind that despite their CT, typical superlead PTs aren’t intended for 2 phase rectifiers as per GZ34, rather FWB types.

[pjd3]

pdf64, Thank you for that link.

It lines up perfectly with my conversation about an approach and also adds the benefit of only exposing AC to the switch.

That should be fine to take the edge off the inrush current to all components from transformer to 1st filter cap and for all I know, even beyond that.

Think I will definitely install that!

Thank you, very good!

Best,
Phil D.

[rangdipkin]

Any love for a pair of NTC thermistors right off the AC power inlet? I've been shotgunning a pair of CL-80 or thereabouts right off my IEC socket lugs, with an MOV between the two on the input side. You can watch the pilot light bloom to life if you size 'em just so!

[pjd3]

Another thing I like about a post diode "in rush" resistor/switch/diode is that theoretically it may give the negative bias the opportunity to come up full before the power tube plate voltages reach a significant voltage. I'm not sure if it always works like that in practical terms but, being that the bias is taken immediately off the HT before all the "standby/inrush" components is a nice thought.

Best,
Phil D.

[pdf64]

Any love for a pair of NTC thermistors right off the AC power inlet? I've been shotgunning a pair of CL-80 or thereabouts right off my IEC socket lugs, with an MOV between the two on the input side. You can watch the pilot light bloom to life if you size 'em just so!

I like to use an NTC on the live in of my amps, something like a CL60, to limit that initial surge. A pair of CL80s may be a bit much though, tending to reduce the mains voltage at full power output, and hence reduce the max power output.

[rangdipkin]

Agreed, if yer gonna pair em up they gotta be sized appropriately for the current draw

[pjd3]

I brought an idea from Valve Wizard to a few of the top EE's at my company (Zoll Medical) and they thought it should take care of some things adequately, particularly, having a resistor bypassing the "standby switch" followed by a diode just before the first 100uF filter cap.

the proposed benefits according to Valve Wizard would be:
1. A 47K ohm resistor in parallel or bypassing the standby switch to slow inrush current making life much better for PT, 2 rectification diodes and filter cap
2. A diode also placed after the "standby switch" keeping the switch in a pulsed DC environment as opposed to a more steady DC environment making life better for the switch (switches don't deal with as much arcing and thus corrosion of the contacts).
3. The 47K ohm resistor enables a current to continue in the tubes (trickle) to minimize cathode surface breakdown and to thwart interface resistance.

Well, if this were even half true then its certainly worth my time and money to install a few more common and cheap components. What I don't know is how much of a "mute" action the switch will enable with the 47K bypass "trickle" resistor but, thats OK - I have volume pots on my amps and guitars. No problem.,

Thanks for hanging in with me on this topic. At least for now, this all seems to bring some resolve to an old rumination over the whole truth or myth of the "Standby switch" debacle.

Best,
Phil D.

[pjd3]

Oh, I should have also mentioned that the Valve Wizard circuit also includes a 1 Meg ohm resistor from the downstream end of the switch to ground, to keep the switch referenced to ground.

With the inclusion of the 47K bypass trickle resistor I really don't understand the function of the 1 Meg ohm resistor, as this is usually relegated to components that are floating and benefit from a large resistance to ground. The 47K ohm resistor is always providing a closed electrical condition but I'm probably just missing something in my head.

thanks,

Phil D

[pjd3]

So as far as I can tell, this is looking like a good thing.

I went to Valve Wizard and configured the 2 diode rectifier scheme as they showed, with the switch following the 2 diodes plus 47K ohms of resistance bypassing the "standby" switch. I followed all that up with another diode for keeping the switch in a pulsed DC environment. The only thing I haven't put in (yet) is the 1 Meg resistor from the switch to ground.

Although I haven't calced out any time constant values for this, it seems just from watching the voltage rise across the 100uF of first filter cap that this is an effective solution to the in-rush current. With the "standby" switch open and upon hitting on the power switch, the voltage across the 1st filter cap rises to about 40vdc in one second, 120vdc in two seconds, around 170vdc in three seconds, etc... so that is looking like quite a substantial slow down of current build into the cap. At that point of course, you close the "standby" switch and the HV DCV instantaneously rises to its full 430vdc. This is in an unloaded condition however, there are no tubes or even more power node resistor/caps following the first filter cap. But, I believe we get a fair picture of how the 47K ohm "trickle" resistor across the switch is going to allow you a decent curtailment of what might be an aggressive in-rush I.

Thank again for steering me towards what looks to be an effective remedy for the real or imagined traumatic in-rush current with 100uF as the first filter cap in this JCM800 2204 build.
Best,
Phil D.