5E3 Solid State Rec Sag Resistor?

[ncusack]

Hey im thinking about building a 5E3 and I have a few questions about the power section.

1. Can you achieve a relative similar tone using Solid State rectification and a sag resistor? and what value resistor would you use?
2. If you do use a solid state rec what value power transformer would you use to get the right B+?
3. Should the Power filter section be changed in any way for slolid state rec ie. Larger resevoir cap?

Thanks for your thoughts,

Neill

[fyl]

Can you achieve a relative similar tone using Solid State rectification and a sag resistor? and what value resistor would you use?

Yes. You need to drop app. 40V @ 100mA, which translates into a 400R. Nearest E12/E24 value is 390R. Use a big bad R, say a 25W chassis mounted model, even if power dissipation will only be 3W9 @ 100mA.

If you do use a solid state rec what value power transformer would you use to get the right B+?

Shoot for 370V *loaded* plate voltage. Plus 40V dropped => 410V DC after the recto. The actual AC value at the secondary depends on the type of recto you plan to use as well as the PT impedance ratio but, typically, 320V AC thru a bridge recto will get you quite close to 370V loaded.

Should the Power filter section be changed in any way for slolid state rec ie. Larger resevoir cap?

You may play with cap values but anything above 47µ for C1 will make the amp too stiff. I'd use 33µ max, with 22µ for the other nodes.

[ncusack]

That looks like some solid information. so what would you say about using that 390R sag in conjunction with a full wave diode rectifier on a hammond 270EX 275-0-275 @ 125mA. The 6.3 line is rated for 4A which should handle the tubes no problem. From my research full wave should give 1.4 x 275 which is 385VDC. That may also be a tab higher since i think that transformer is actually 115V primary so wall voltage could give up to 420VDC.

For that reservoir cap value would that be 33uf before any of the normal filtering giving 4 power section caps or would you just make the first cap in that network 33uf and make the other 2 22uf?

Thanks for the help on this one.

[tubeswell]

A 40V drop from 385VDC on the B+ will be in a good ballpark for a 5E3 with a 250-270R cathode resistor. Remember its not the current rating of the secondary winding that's going to cook your sag resistor, its the current that is drawn through the power rail by the (tube) load. The PT should always be rated equal to or better than the total current draw, but you base your resistor power rating on at least 2 x* the total current draw, and not (necessarily) on what the PT is capable of handling. Anyway, that's how I look at it**.

* (but the more - the merrier)
** not that that should mean anything other than 'others may see it differently'

[ncusack]

So tubeswell would you be suggesting that I use the current drawn from all of the tubes like i think its something like 70ma for 6v6 in push pull and like another 1.2ma each for the preamps. then use that value to calculate the proper wattage for that 390R sag resistor? These are numbers i've seen on a few data sheets by no means do I think my numbers are right here. Im only starting to get into the math behind this stuff.

[tubeswell]

Yep, that's a good basis for an overall upper limit ballpark B+ current draw for a 2 x 6V6 amp

[MWJB]

A 5E3 might run anywhere from 365-385vdc with a 5Y3, later models can even break 400v...feel free to use a 275-0-275VAC PT but your final voltages will be way down on a 5E3. That 1.4 multiplier rarely comes in right for cathode biased amps which have high idle currents.

Even a 5E3 with a 5Y3 rectifier will take 100uf at the first filter cap, it will still sound like a 5E3 (it will still sag due to your big B+ dropping resistor, or due to the 5Y3 in the original design, but it will hum less & hold together better), as you are changing the design anyway to accomodate a SS rectifier & B+ dropping resistor, it seems silly to worry about "changing the sound".

The sag in a tube rectifier can be emulated by a power supply resistor, but it does not sound quite the same, the sag in a tube rectifier is more dynamic in nature.

Why not use a BF Deluxe style PT, fit a 330ohm cathode resistor then you can switch between a SS plug in and any of the 5V tube rectifiers?

[fyl]

Even a 5E3 with a 5Y3 rectifier will take 100uf at the first filter cap

Not for long, but it will.

[MWJB]

Fyl, when you say "not for long", how long do you need it to last, a decade, a century, an eon?...With a 100uf first filter a 5Y3 will outlast the filter caps & power tubes...probably outlast you! :-)

[fyl]

I don't think so. The 5Y3GT can only accept 2.5A peak plate current for 200 ms. Using a typical amp, what is the inrush current when the cap charges for the first time? How long does it take to reach nominal voltage?

[MWJB]

I know so.

[loudthud]

I ran a simulation with Duncan's PSUD (version 2.0.1) using the transformer data I got from a vintage 5E3, YMMV, this transformer has a pretty high resistance.

Primary resistance: 7.9 ohms.

Secondary resistance: 374 ohms end to end, 179.9 ohms on one side,
194.6 ohms on the other side. I split the difference and put in 187 ohms.

Secondary voltage unloaded: 775V end to end with 121.3V on primary.

PSUD called this to be 387.5V with lumped impedance of 267.6 ohm.

With a 5Y3-G rectifier, 100uF (2 ohm) filter and 100mA constant current load, the first cycle peak current was 1.2069A. Changing the ESR of the filter cap to 0.1 ohm, the peak current increased slightly to 1.2119A.

The reason I ran this simulation is I wanted to see how the calculation of the sag resistor matched simulator results. Because the sag resistor is before the first filter cap, you can't just use Ohm's Law to calculate the value. But much to my surprise, it was pretty close. Using a 5Y3 model, 16uF (2ohm) cap the peak DC out was 388.27V. Using 1N4007 diodes and a 330 ohm resistor the peak DC out was 386.92V. Using 390 ohms the DC was 375.3V. The high impedance of the transformer must dominate the rectifier as a sag source.

[ivey]

LC input with didod tube or Solid State

It is not recommended to run any thing higher than 60uf in to a LC filter PSU. 60uf or less is considered good, with tube diodes. Information can be found in RCA tube manual 1963.

Solid State is a different matter. You can load up to 100uf. I use photo caps because they are design to handle high voltage and current. Plus they store the charges real good and funnel it out a great deal faster.

If you use a full wave rec., use 0.01 caps across them. The diodes coming into the USA today, could cause you problems.

Ivey

[MWJB]

You can use 100uf with any of the commonly used 5V tube rectifiers (5Y3/5U4/5V4/5AR4) remember to connect the cap to the hot side of the standby, not the cold side like tweed schems.

Data sheets are typically conservative in many respects, most popular tube amps, even those from the late 50's contradict the data sheets.

[barfoden]

That looks like some solid information. so what would you say about using that 390R sag in conjunction with a full wave diode rectifier on a hammond 270EX 275-0-275 @ 125mA. The 6.3 line is rated for 4A which should handle the tubes no problem. From my research full wave should give 1.4 x 275 which is 385VDC. That may also be a tab higher since i think that transformer is actually 115V primary so wall voltage could give up to 420VDC.

For that reservoir cap value would that be 33uf before any of the normal filtering giving 4 power section caps or would you just make the first cap in that network 33uf and make the other 2 22uf?

Thanks for the help on this one.

The current is high enough at 125mA for a 5e3 with a small sag resistor but your plate voltage will be lower.. The 1.41 x 275 will probably give you a +B of around 390V unloaded. As you have a lot of current capacity from the PT your loaded +B will not be that low, and it will probably drop about 25 volts to around 365V +B. Depending on the OT resistance the plate voltage will be around 350-355V.
Maybe you do not need a sag resistor. The 5e3 is cathode biased which have a bit of natural sag and compression to it. Besides that a hot cathode biased amp will not sag as much as a fixed bias amp with a 5y3 as the idle current is probably allready around 90% dissipation and biased towards class A. The 5e3 has a large 5k screen dropping resistor. This resistor will induce some compression on hard transients when current will be pulled through it and voltage will drop.

Another amp builder here on the forum built a budget 5e3 with diode rectifier and a 260V/100mA power transformer, classictone 5e3 8k OT and no sag resistor. He got around a +B of 340V and a plate voltage of 328V. He used a 270 ohm bias resistor and increase the preamp and PI voltage by exchanging the 22k resistor to a 10k resistor. HE kept the 4.7K screen dropping resistor. He is very satisfied with the outcome, so my suggestion is to build the amp first without a sag resistor. If you are not satisifed add a sag resistor and compensate a bit by lowering the 5k screen resistor and 22k resistor to get voltage closer to the schematic values.