Rectification question

[ampgeek]

I am chasing sluggish voltages in my homegrown Dlite. The amp sounds OK but I would like to tweak the voltages closer to the range of accepted values.

I am using a Bnote 325-0-325 PT with 115VAC at the outlet and am measuring 320VAC on each leg of the diode string. My first B+ node measures 420VDC with all tubes removed and the FET simulation resistor lifted from ground.

With all tubes loaded (5881's at 40'ish ma and 12AX7's up front) and the FET resistor connected I am only getting ~395VDC at the output tube plates.

It "feels" like I am experiencing excessive voltage drop across the diode chain. I have looked closely for any small shorts to ground in the power supply/filtering sections that might be loading them down but have found none.

Does that diagnosis sound right? Is there any way to evaluate diodes in-situ for "plugging"? Are there any other areas that I should be looking into?

TIA,
Dave O.

[heisthl]

Somethin' be wrong there. Open the string before the screen node and measure again. Make sure you get a good reading on the initial B+ (if not remove B+ filter(s) and cliplead in a 500v cap and/or disconnect OT). If opening the string gets you back in range go down the string, disconnecting and measuring until you find the problem.

[Structo]

I am using the Brown Note PT with similar rating and my B+1 ( right side of choke resistor ) is 430 v to the power tubes.
Hope that helps.

[Structo]

Also what does your house current voltage measure?
That can throw it off a few volts.

[ampgeek]

Thank you gentlemen!

I am running 115VAC at the power switch. Guesstimating that I should be in the 430 to 450 VDC on the power tube plates.

I disconnected the string at the standby switch and measure 288VDC in the appropriate places when switched on and off (e.g., nothing funky going on there).

Should the suggested cap go to ground like a filter?

In an attempt to increase my theortical knowledge...Do diodes need to be trickling off AC and/or passing some threshold quantity of current to work properly?

Again, many thanks,
Dave O.

[heisthl]

You need a filter (even 20uf will do) to get an accurate measurement of DC after the diodes. Yes that would be a cap from B+ to ground.

[Funkalicousgroove]

You should be getting a MAX loaded plate voltage of 413vdc, a 5881 (other than sovtek which is NOT a true 5881) is a 25W tube and should be biased around 36ma.

Diodes are cheap, I'd replace them all as a first measure, making sure to check all your solder connections and wire along the way.

What kind of filter caps are you using on the 1st b+ node? caps with VERY high ESR would act this way and cause funky voltage as well-

[jdh]

Connecting a voltmeter will draw current through the diodes and let them rectify, but heisthl is right: you need a capacitor to measure the peak voltage the supply normally charges up to. Without the cap the meter sees pulsating DC, and will measure somewhere between average (63.6% of peak) and rms (70.7% of peak).

[ampgeek]

Great stuff as usual guys. Thanks again!

I slapped in a 20uf/500V Atom just upstream of the standby switch, disconnected the first filter (2 X 100uf/350 in series with equalizing resistors) and broke the B+ chain just after the choke simulation resistor. FET simulation resistor is still lifted and all tubes are removed.

Measuring at the end of the diode chain, I get the following:

In standby, 430VDC
Out of standby gives 430 VDC (OT is still connected)
Re-connected B+ chain (e.g., 2nd, 3rd, 4th and 5th filters in), 427VDC
Re-connected first filter drops to 420VDC
All tubes in and 5881's at 35ma gives 410VDC and high 390's VDC at tube plates

I replaced all of the diodes, re-flowed/closely inspected all joints in the PS/Filter zone and repeated the above with similar results.

All PS filters are Xicon 140-XAL-**** electrolytic type.

What do you think? Filter caps leaking excessive DC? Struggling PT HV winding?

TIA,
Dave O.

[heisthl]

Doesn't read like you have a problem other than a transformer that's way under spec.

[dogears]

Put in a beefier PT. Shoot for 440V on tubes. You'll thank me!

[kleinm]

Could a bad filter cap be causing this problem??

I had strange things going on in my PS, although not similar to your experience. The voltage would randomly drop and touching a lead with my DMM regained voltage.

The problem turned out to be a faulty electrolytic capacitor. Recapping the entire board solved the problem. You're getting good voltage after the rectification, so I'd rule that out.

Have you tried retubing your amp and remeasuring voltages? BTW, I'd heed Funk's warning. The RCA published spec (Class AB1) for 5881 max plate voltage is 400VDC, with max plate dissipation at 26W. If you like the sound of the 5881, you might try the TAD 6L6GC-WGC, it has slightly similar tonal characteristics to a WGC (5881). However, it is a true 30W 6L6.

I was fond of them for quite sometime, but I'm becoming fond of the JAN Phillips 7581A. I luckily had two pairs laying around from my BiValve era. I'd like to give the SED 6L6GC another shot, too.

[ampgeek]

Yep...I reackoned that the "problem" was largely due to a not quite up to snuff PT. As H suggests, probably just the nature of the set-up.

The amp does sound really good over a large range of tones. It just doesn't quite nail that sweet OD tone that is characteristic of the D clips here regarded as "right on".

No matter, this build was meant to be a D trial anyway and it really is a nice sounding amp that I wouldn't be embarrased to play anywhere. My eye is toward stuffing a JCM800 and D style circuit into an empty Ampeg V4 head cab just crying to be used for something!

I have tried MANY different types of output tubes along the way. I "inherited" a boatload of NOS tubes from a lab that was being shutdown. The most common power pentodes included many Tungsol 5881s, a couple of RCA Blackplate 6L6, RCA 5881s, 6V6s and GE 6L6s. I included a dual biasing mod so slapping in just about any pair and balancing them out is a breeze.

Call me crazy....but perhaps the best sounding tube to my ear in this circuit has been Sovetek 5881 WXT. A servo tube I think. Go figure!

Just for the record: What type of filter caps are being used in the "benchmark" builds that ya'll are doing? I know that there are much higher quality caps out there but the various benefit discussions that I have read are ambiguous. I noticed that the BNote deluxe Dlite included Elna Blackgates(?) which made me ponder an upgrade to my build. All in the name of "science" and experience (he says to his wonderful and understanding wife!!!)!

Thanks for your input and help,
Dave O.

[Structo]

I believe the E caps in the deluxe D'Lite's are F & T German made.

[tele_player]

Structo - that's correct.