Tube DIY Asylum

Do It Yourself (DIY) paradise for tube and SET project builders.

Return to Tube DIY Asylum


Message Sort: Post Order or Asylum Reverse Threaded

Refurbing old PA amp, need help with PSU

69.125.143.12

Posted on May 7, 2020 at 17:49:03
GRH
Audiophile

Posts: 444
Location: CT
Joined: March 22, 2004









For no good reason I am trying to bring an old Sears Silvertone PA amp back to life. If I get it working I will then convert it to a guitar amp for my stepson.

Based on style and dates codes of 803 and 807 on the old can caps I believe this probably dates to 1948 or 49. It uses one 7b4 for each of the two Mic inputs and a 7n7 for the phono, summing, and tone control. A second 7n7 is the final amp and phase splitter. The rectifier is a 6U4, and the outputs are 6l6. The back panel identifies the tubes none of which have letter suffixes so I have to assume it was designed for the original version of the 6u4 and 6l6's. The transformers appear to be OK via a cursory ohm check. I think the output transformer must have been replaced at some point based on the poor soldering and only two clearly non-original screws holding it down. I haven't been able to locate a schematic so I am forced to reverse engineer it and hence my problem.

I have attached a photo of a sketch of the PSU. It seems pretty straightforward except for three issues I'm hoping to get help with:
1st.) The shunt resistor has no value lettering on it anymore and both power resistors are blown. Assuming a design point of 360V Plate and 270V screen based on original 6l6 data then I believe the shunt needs to be around 10K ohms. That current draw plus the screen and preamp current would get the 90V drop across the coil and 2250 ohm series resistor. Am I in the ballpark?
2nd.) and more confusing to me. after disconnecting the high voltage outputs and with no tubes in I measured 400-0-400V. That was with 117V in from a variac. It will be even higher with todays line voltage. Won't that high an input into the 6u4 and then into 40uF result in something close to 500V rather than something closer to the 360V max at the plates?
3rd.) I have a twin 50uF can cap. If I put a CL-80 (47ohm cold) between the 6U4 and the50uF cap will I be OK?

Thanks in advance. You guys always help me out and I really appreciate it.
Gary

 

Hide full thread outline!
    ...
RE: Refurbing old PA amp, need help with PSU, posted on May 7, 2020 at 20:39:01
Eli Duttman
Audiophile

Posts: 10455
Location: Monroe Township, NJ
Joined: March 31, 2000
The rectifier is a 5U4 and either of NOS GE or current production ElectroHarmonix (EH) 5U4GBs will be fine. Steer clear of Chinese and Russian ST bottle 5U4 "equivalents", as they are utter trash.

As the PSU filter is cap. I/P, that bleeder resistor only serves to discharge the filter caps., after the unit is turned off. Select a rational replacement value and stop fretting.

5881 equivalent Russian 6Π3C-E (6p3s-e) stock as the O/P tubes will give you a margin of increased toughness over either original metal envelope 6L6s or electrically equivalent ST bottle 6L6Gs. No more worries over today's higher average AC mains voltage.

Put a CL-140 inrush current limiter between the 5U4 and the 1st 50 μF. cap., to protect against turn on arcing. A CL-80 might not fully warm up under the conditions present in your unit.


Eli D.

 

RE: Refurbing old PA amp, need help with PSU, posted on May 7, 2020 at 21:03:46
trobbins
Audiophile

Posts: 189
Location: Melbourne
Joined: March 26, 2012
Restoring old PA amps can be fun, but also tedious if all the wire insulation is poor, and many resistors have drifted too much, and any electrolytic needs replacing.

It would be a rarity if the amp hadn't been modified/repaired, and yes output transformers were often replaced (whether failed, or due to incompatible speaker impedances).

Some advice relating to PA amp restoration in link:

 

RE: Refurbing old PA amp, need help with PSU, posted on May 7, 2020 at 21:44:51
Eli Duttman
Audiophile

Posts: 10455
Location: Monroe Township, NJ
Joined: March 31, 2000
Yes, that old cloth wire insulation deteriorates. If flexibility is not a significant concern, that cloth can be consolidated with clear nail lacquer and then protected with polyolefin heatshrink tubing.


Eli D.

 

My rationale for "Shunt" vs "Bleeder", posted on May 8, 2020 at 06:04:38
GRH
Audiophile

Posts: 444
Location: CT
Joined: March 22, 2004
Thanks Eli, as always you are most helpful.

There are two reasons I thought the shunt is part of the regulator. First, the old part is a wirewound of at least 10Watts. I don't think a bleeder would need to be a low enough value to need that kind of dissipation. Second, per the data sheet the two screens together only draw 5ma quiescent and 15ma full power. Add 5ma for the four preamp tubes and the draw is 10-20 ma, a 100% variation. Even 20ma won't drop much voltage over the 2400 ohm series resistance. I figured the shunt also might be there to reduce variation in the screen voltage as power varies. Am I overthinking this and still being too much of a worry wart? Perhaps the two WW's are not original and the 2250 should be higher. I've seen 5000 in other 5u4 feeding 6l6 schematics.

I have UOS metal 6l6's and glass 6l6G's, and NOS 6l6WGB's. I wanted to stick with the UOS for this application. I'm mainly doing it for amusement and don't want to spend much or waste valuable tubes.
G

 

Thanks, posted on May 8, 2020 at 07:10:56
GRH
Audiophile

Posts: 444
Location: CT
Joined: March 22, 2004
Thanks for the great link. All of the safety pointers are already part of my standard process but there was a ton of useful ideas if I go forward with the guitar aspect of this.
Gary

 

RE: My rationale for "Shunt" vs "Bleeder", posted on May 8, 2020 at 12:31:18
Eli Duttman
Audiophile

Posts: 10455
Location: Monroe Township, NJ
Joined: March 31, 2000
In the future, we'll have to pay New Sensor's price for Reflektor made 6Π3C-E stock. This is what EBone is currently showing for the Russian 5881s and the $ seem "reasonable".


Eli D.

 

RE: Refurbing old PA amp, need help with PSU, posted on May 8, 2020 at 15:40:45
trobbins
Audiophile

Posts: 189
Location: Melbourne
Joined: March 26, 2012
A lot of vintage amps had internal rail loads - mainly to sustain choke current to above nominal critical conduction level, which provides a better regulation range, and reduces stress on the filter capacitors (which were used to a bare minimum if at all).

Nowadays, capacitor voltage rating is cheap (even when using series connected to get the peak voltage rating) and regulation is not such a big thing, so initially I'd suggest making sure all the caps (including coupling caps) can withstand the max B+ without valves inserted (as the 5U4's will pass current well before any amp valves). You do still need some loading, to avoid capacitor voltage lingering after power turn off.

 

Still am concerned, posted on May 8, 2020 at 18:09:38
GRH
Audiophile

Posts: 444
Location: CT
Joined: March 22, 2004
I certainly can stiffen the screen regulation with more capacitance (my sketch was wrong, only 20uF) and I plan to do that and I can easily add something like 50K 2W as a bleeder.. My concern is I don't see how I get the screen voltage down to 270V without heftier shunt load. If my B+ is well above 400V and I'm only drawing 20ma through 2400ohms then that will only be a 48V drop so both the plate and screen are going to be well above the 360V/270V spec for early 6l6's. Eli does not seem to think it is an issue.
G

 

RE: Still am concerned, posted on May 9, 2020 at 19:31:02
trobbins
Audiophile

Posts: 189
Location: Melbourne
Joined: March 26, 2012
Regulated screen voltage can be one means to minimising distortion, and datasheets will present regulated rail data to show what distotion can be achieved.

In PA and guitar land, screens are almost never regulated, and effectively rely on screen voltage sag to swing the screen voltage down sufficiently to suppress the rise in screen current under heavy loading, so as not to over-heat the screen. Screen voltage sag may markedly change the over-drive response of the output stage, and so is a 'tone' adjustment used by amp makers to suit a certain style or sound from the amp, and that includes any filtering 'support' to maintain a higher screen voltage under certain signal conditions.

Many PA amps just used a simple resistor dropper to feed the screens off the main B+, and often the NOS valves could cope with seemingly very high idle screen voltages. So it is not a total no-no to allow high screen voltage levels at idle, but you need to appreciate how that voltage will swing down and by how much, as full output and over-drive conditions are approached, and whether all generic valves can cope with the screen voltage for starters.

 

Got it working, posted on May 11, 2020 at 19:49:55
GRH
Audiophile

Posts: 444
Location: CT
Joined: March 22, 2004
I modeled it in PSUD with guesstimated load currents and rebuilt it with 4K in series and 12K shunt because thats what I had on hand. Bringing it up on a variac I got pretty good numbers of 390V on plates and 275V on the screens and if I subtract cathode bias those numbers are even lower. Also like you said the screen voltage comes down as output is increased so I'm comfortable now. I'll either lower the 4k a bit or raise the 12K a bit the next time I order any parts but it's fine for now.

Thanks for your help and reassurance. I wasn't worried about distortion and having regulated screens or achieving max power, I just didn't want to destroy any 6l6's.

The unit seems to work fine. I have some 60HZ from the first 7n7 preamp that I have to track down but otherwise all good.
Gary

 

RE: Got it working, posted on May 11, 2020 at 22:13:03
trobbins
Audiophile

Posts: 189
Location: Melbourne
Joined: March 26, 2012
Sounds like you are through the first phase, with a working amp.

Minimising hum can be a fickle process - there are some faultfinding guidelines in link that may help if you haven't been down that path before.

There are so many options as to how to go from here - at least with an old PA amp, making changes can be cheap and easy and don't have to meet some aesthetic benchmark or diminish the value of some expensive amp.

 

Page processed in 0.025 seconds.