Power Supplies And All That

Lately I had noticed that the power supply exhaust fan on my desktop HP would make noises like a coffee grinder occasionally. So when it started to get really annoying I figured I'd order a power supply, swap it out and then repair the old one.

No surprises here. The fan in the original was pretty noisy even after a dose of 3 in 1 oil, so I ordered another fan, cost $5. It arrived and it was the same make and model as the one I removed.

A couple of observations are in order. To get a Liteon power supply like the box came with and which you see pictured above you will cost you about a hundred bucks. So I ordered a cheapie "Replace Power" power supply from an online vendor, cost $22.79 and it was a drop in installation. The part number is RP-MPS3-420W, and it replaces the Liteon PS-5301-08HA.

Comparing the outputs reveals some differences. I'll put the Liteon's ratings in parentheses. +3.3v 11a (+3.3v 18a), +5v 12a (+5v 25a) +12v1 8a (+12v 19a) +12v2 8a (none) -12v 0.3a (-12v 0.8a) and +5vsb 2a (+5vsb 2a).

In most areas the Liteon has better capacity to deliver current on its individual legs but is limited to 300w combined.

There are a couple points of interest. The Liteon is much heavier and I believe that it is because of the massive extruded aluminum heat sinks. It is also thermostatically controlled and thus the fan does not run continuously as it does in the Replace Power unit. This is a nice feature because the exhaust fan rarely runs and most of the time doesn't bother me. It also has a small LED that indicates when power's on and it slowly fades as power is bled off after shutdown. The Replace Power does not have this feature.

The Replace Power also does not throw off a lot of RF noise, either. My last cheapie power supply for the Dell desktop that this replaced made a whistling noise that meant when it was running I couldn't listen to AM radio anywhere in the house. It's as good as the original in that respect.

Although they're both rated at 300w, the Liteon power supply seems to be more robust, and I'll either reinstall it or save it for another desktop if one should come my way as seems likely.

The Little Amp That Couldn't: The Egnater Vengeance

An Egnater Vengeance 60/100 head arrived here in the middle of a raft of other rocket science, dark matter, higgs boson amp projects last month so it took a little while to get to it.

The story given was that the owner while playing a gig and watching his power monitor saw a huge power surge at a particular local venue that will not be named. His Egnater died, laid down on the job, quit, gave up the ghost and generally took a dump right then and there. There are dark rumors and mutterings that this was not the first time this happened in this particular place although I haven't seen the proof of it.

If there's anything to take away here, when you've got an amp with no output, measure the voltages fer chrissakes. Often it will tell the story. On the average eight pin octal based tube such as everyone's favorites the 6L6GC, the 6V6, the EL34, and the 6550 you can measure pins 3-plate voltage, 4-screen voltage, and 5-negative bias voltage.

Missing voltages and no output has been the story around here the last couple of months. First there was the Sunn 100S with a defunct choke, then there was the Music Man 2100-65 with an open choke, an open filter capacitor and an open capacitor on the board I sussed out with my tone sniffer. Then there was the Vox AC15 HW with no output, a Marshall MA50 with no output, and a few others with various and sundry ailments. On the other hand it has given me an opportunity to put my scope to good use and take baby steps along the long learning curve to scope competency.


In this particular Egnater, there was no screen voltage present. A little work with the schematic located the correct voltage on the standby switch, and it dived into the power supply circuit board bit didn't come back out. A little more work revealed that one of the 100uf/500v radial snap in capacitors had shorted out, and two of the three 10w ceramic resistors were open.

Then the fun started. Getting the power supply board out to gain access to the back side required: removing the voltage selector switch, the power switch, the standby switch and the foot pedal board, unbolting the power transformer to get enough slack in the leads to push them far enough away so as to pull the circuit board up, clear the nearest 10w resistor to the chassis, and then remove the offending components to replace them. I would say it took a good solid hour and a half to accomplish this task.

While inspecting my nifty silver solder joints I managed to burn a hole through an unopened 12 ounce bottle of Coke and made a mess all over everything.

Here you can see the offending bulged shorted capacitor; some of the mass of wire on top of the PS board; ; more wire spaghetti ; yet more wire spaghetti; the PS board emerging; more wire spaghetti; and the new Ohmite 10w resistors.

In addition I was trying to do a clean install of XP on my library computer because one of its critical files had disappeared overnight.

This morning I expect to finish the XP reinstall and start the burdensome and onerous task of installing a bajillion updates and service packs.

Then, and only then, I may try the registry hack that will convince Microsoft that this old workstation is really a "point of sale" device and get more updates.

Of course the wireless business would not cooperate so things came to a screeching halt on the computer front. 

Somehow the usual abbreviation "P.O.S." seems appropriate.

UPDATE: I think that the long road of XP has come to the end. I knuckled under and bought a copy of Windows 7 and a new wireless adapter. These ought to arrive fairly soon and I can then see how well the box operates with a modern-ish operating system.

UPDATE: I got the Egnater running this morning and on the scope it was putting out a good 90w clean, so I proceeded to put the head in the cabinet and to give it a last road test before it went home. The volume had dropped down to nothing, about 2 or 3 w. What now?

I took it back to the shop and opened it up again and this is what I found. I think I'm pulling the captive nut and spring clip and losing them somewhere. The screw had jammed right into the footswitch board. I'm guessing that the hole in the chassis and the mating hole in the cabinet were mislocated.

Laney AOR 50

A few weeks ago I saw this Laney combo on Craigslist not too far from here and by the time I jumped on it, a friend of mine had already snagged it for a very modest price. Well, he emails the other day and says it wasn't him, too heavy, and would I like it for the price he paid?

Of course I said yes. I'd been getting the itch for something new and I'd speculated about acquiring the Peavey Prowler that's on CL here at this very moment. Good thing I saved my dough.

Laney was said to be a downmarket Marshall clone, and that may be somewhat truthful as the schematic bears a certain resemblance to the JCM 800, with a few pull boost features thrown in. Reverb and a nice Fane speaker finishes out this amp that was built in 1986.

The transformers were from makers I'd never heard of on UK built amps. The output transformer was made by Skot, and the power transformer and choke were made by Ariel. Skot is still in business and they seem legitimate enough but no trace of Ariel thus far. Either way they've got to be better than the awful Dagnall crap that Marshall was using. The reverb is reasonably nice but not overly drippy, with an Accutronics tank in the bottom of the cabinet.

I had it on the bench today for some much needed maintenance and inspection. Cleaning and douching the pots and replacing both input jacks took care of business and a new set of Valve Art EL34s properly biased and a pair of new 12AX7s finished things up. The plate voltage is relatively modest at a measured 412v, and bias was accomplished with a nice pot located on the circuit board ala Marshall at 38 ma each. I do not know whether there is a connection between Marshall and Laney but I think it is more than likely.

Schematics are available readily from The Tube Store for free on their website.

It sounds good, and you can do some interesting Telecaster plonking easily enough. I'm not a high gain guy but I may never need it. The background noise level is just a little hissy with the master volume dimed but tolerable if you back it down a little.

These are pretty cheap and you can have a whole lotta fun so if you find one grab it quick.

Biasing the Albion TGT-50

An Albion TGT50 head came my way for some routine service and it occurred to me that it would be worth checking the output tube bias to make sure it was reasonably close to what it should be.

Of course I couldn't find any service information readily to hand. So I opened the small hatch on the rear of the chassis and saw what looked pretty close to something I've seen on Marshalls and on a Randall head I used to own-a pair of potentiometers and three pins. I then removed the chassis for a general lookover and it seemed like a pretty conventional 50w channel switching head. Nothing was remarkable except that the tube sockets are chassis mounted and wired to the board.

A nice touch, and one that makes service people happy, if you followed my adventures with hard wiring a Blackstar 100w head recently because the board had burned up.

After using my bias probes and measuring the plate voltage at about 470v DC I adjusted the pots and got the results I needed. I then went back to the original setup, set my multimeter to millivolts and measured the left and right pins with the ground probe to the center pin.

The results were quite similar which led me to believe that the designer of this amp, Steve Grindrod,  used the same setup that Marshall used-not surprisingly when you figure that he worked for Marshall for a while. I figure I was reading across a 1 ohm resistor in the cathode string for each power tube.

Or, of course, you can take the better path and check bias with a pair of probes and a couple of dedicated multimeters. It'll give you a chance to look around inside while you're visiting.

One Day In 1970 Or, How We Fixed Your Telephone

I was having a discussion at the Other Place with a few fellows on the subject of fixing electronics, namely guitar amplifiers, sans a digital multimeter or a scope. So this was rolling around inside my head and maybe it was the story in the AARP bulletin today on the subject of memory loss that had me casting about for something to prove that I'm not dead yet.

So anyway, back in 1970 I was working for New Jersey Bell Telephone Company, better known as Ma Bell. Mother was very picky about a lot of things and when she'd settled on something she liked, nothing else would ever do and no off the shelf stuff either. Mother loved MacMillan Ring Free motor oil and that's all we ever used. Macmillan's only claim to fame was that they'd sponsored the Skoda racing team in the fifties.  The supply crib was well stocked with goodies like Ma Bell bandaids and aspirin, spray paint (olive drab, safety yellow, and black) and other interesting things. They had the best damn first aid kits you ever saw, complete with ampules of antiseptic. They were packed ten to a box with a glass ampule inside a plastic tube with a cap on one end. Pull off the cap, crush the ampule, squeeze the plastic tube and anoint thy self, brother....where was I? Oh, I remember.

When there weren't any big projects going on the repair crew would be sent out to clear defects. What that meant was, there was a defect on a particular pair of wires that the installers and frame men couldn't fix, so they'd bypass the pair (spare pairs, doncha know?) and sooner or later the splicing gang would have it.

The process consisted of going to the frame (usually Somerville), locating the pair, and blasting it with the breakdown set, which was a zapper that would send a jolt down the wire and maybe dry out the fault, but usually ended up fusing the pair at the site of the problem. So we'd read the resistance which would tell us how many miles away it was, and attach a tone generator to the pair. We knew from the print what bundle it was in in the cable.

Off we'd go in the truck watching the odometer and when we got close, a pole with a probe was stuck out the window. The probe was connected to a small amplifier and when we reached the site of the defect the tone would quit. We'd do a little more walking around until we had it located pretty close and then we'd put the bucket up to the cable and open it up to determine the final location. A slack puller on the wire rope could pull enough slack into it that we could dig around inside the cable with a probe until we found the pair. This took a little bit of doing since a 2,200 pair cable is not an easy thing to wrestle with.

Once the pair was located we'd fix it with a splice, put everything back together and install a splice case. We'd wait for a little bit until the air pressure came up, check the case with soap solution for leaks and then head to the diner for coffee.

Tweed Super On The Cheap

While trolling my usual places on fleabay (the "Why pay more?" store), I ran across a trio of partially built tweed amps that consisted of Weber chasses, pots, knobs, and Mercury Magnetics iron but not much else. There were a 5E7 Pro, a 5F6A Bassman, and a high power tweed twin.

Since I'd already built a couple of 5E7 Bandmaster kits I opted to snag the 5E5 Pro and build it as a 5F4 Super. All three of these amps are pretty much the same with some very minor differences.

Once the partially constructed chassis arrived I made an eyelet board for it and loaded it with F&T electrolytics, Mojo Dijon capacitors and mostly metal film resistors. At about the same time I bargained with Dirk Newton for a pair of Weber Ferromax 10 inch speakers and contracted with J.D. Newell for a lacquered tweed cabinet.

Looking through my stash I came up with a trio of new old stock Cinch shielded 9 pin sockets for the preamp tubes and a trio of new-ish Chinese ceramic top loading sockets. For glassware I used new old stock RCA preamp tubes, blackplate power tubes, and a Mullard 5AR4 rectifier tube.

The chassis went together easily as this was my third one, and everything else went together when the cabinet arrived, and it is a beauty. I mostly wired it with 22 gauge pre tinned cloth pushback wire from Stewart MacDonald.

There's a lot of controversy about  what sort of wire and gauge to use for general hookup and filament service but I think it is a relatively simple matter to determine what the current load is and wire accordingly. For instance the total current draw for the preamp and power tubes is about three amps total once the initial inrush is over. At that point we see that 21 awg chassis mounted wire maximum load is 9 amps so that should be well above the margin of safety. Some folks think that they have to use 18 gauge single strand for the filament wiring and there's simply no need for anything that heavy.



Here's the breakdown on prices which includes shipping:

$375.00 chassis
$260.00 cabinet
$  75.00 speakers

Glassware and other components were from stock

Peavey weirdness now morphing into Blunt Instruments

UPDATE: I think I've found an appropriate name for the amps I build, I checked TESS and nobody's sitting on the trademark, so henceforth and forevermore, anything I build will be yclept a "Blunt Instrument". You heard it here first.



Last week as I considered the state my monster amp project was in I decided to plunge for power and output  transformers for the Super Bee.

There are two potential sources for this class of iron, and a possibility of a third. There is, of course, the ubiquitous offerings from Mercury Magnetics about which more anon, or the offerings from Mojotone which are manufactured by Heyboer. If Classic Tone made these transformers I would have shopped there but alas! they do not.

The net total was $117.42 for the Mojo 753 power transformer, and $109.70 for the Mojo 756 output transformer-which came with multiple taps by the way. Shipping was another $20 or so. Mercury would have run me $425 plus shipping so it was a no brainer.  I suppose I could have used a Twin Reverb power transformer which I have as well as a Twin Reverb output transformer and then had to come up with a 5v filament transformer to run a rectifier tube which would have saved some dough but would not be close to the original setup. This amp has to look and sound like something Fender built. If I wanted to hunt til I'm old and gray I could possibly have turned up an original p/n 7993 Triad power transformer and a 45268 output transformer but that is not a likely proposition. I could die waiting.

The transformers were installed with blue Loctite on the screws. I've seen too many get loosened up with age.

I've selected a new old stock Mullard GZ34  for rectifier duties and RCA black plate 6L6GCs for power section duties. I think Uncle Boris over at Sovtek can do the  rest. There is also a Dynaco socket mod for protecting the rectifier tube which I plan on incorporating- to power 4 6L6GCs is asking a lot of a 5AR4.

With the board built, all I have to do now is build a bias board and a cap board, install some controls and start stitching. The biggest unknown is fitment and speaker selection. That will come later.



On the top you can see my layout. This was possible only because a friend owns a high power tweed twin and I was able to lay out the components in the original way. By comparison there is a shot of a recently built tweed twin replica and the output transformer and cake pan are reversed.